• Open Access

An intellectual property sharing initiative in agricultural biotechnology: development of broadly accessible technologies for plant transformation

Authors


(Tel 530 752 1411; fax 530 752 2278; email abbennett@ucdavis.edu)

Summary

The Public Intellectual Property Resource for Agriculture (PIPRA) was founded in 2004 by the Rockefeller Foundation in response to concerns that public investments in agricultural biotechnology benefiting developing countries were facing delays, high transaction costs and lack of access to important technologies due to intellectual property right (IPR) issues. From its inception, PIPRA has worked broadly to support a wide range of research in the public sector, in specialty and minor acreage crops as well as crops important to food security in developing countries. In this paper, we review PIPRA’s work, discussing the failures, successes, and lessons learned during its years of operation. To address public sector’s limited freedom-to-operate, or legal access to third-party rights, in the area of plant transformation, we describe PIPRA’s patent ‘pool’ approach to develop open-access technologies for plant transformation which consolidate patent and tangible property rights in marker-free vector systems. The plant transformation system has been licensed and deployed for both commercial and humanitarian applications in the United States (US) and Africa, respectively.

Introduction

Scientific advances in agriculture have been historically treated as public goods. Universities and other public-sector institutions were leaders in developing improved crop varieties that were transferred to farms through cooperative extension services in the US or equivalent organizations internationally (Conway and Toenniessen). However, this model has changed rapidly in the last few decades due to greater utilization of formal intellectual property (IP) protection of agricultural advances by the public sector, as well as the development of a research-intensive private sector that now makes major contributions in enhancing the productivity of US agriculture (Kowalski et al., 2002). The growth in patents related to agricultural biotechnology, in particular, has been on the rise since about 1980, and both private companies and public research institutions, most notably the land grant universities, have contributed to the increasing use of formal IP protection to support the translation of basic research into products. Several changes in the legal and policy framework greatly expanded the possibility of patenting and licensing biotechnology inventions over the last 20 years. For example, in 1980, the Bayh-Dole Act was passed which encouraged US universities to patent their innovations and license them to private sector companies to encourage their commercial use (Bennett and Boettiger, 2009).

For scientists, the relevance of IP during the research and development process may vary by jurisdiction depending on the research-use laws in each country. Research-use or safe harbour exemption refers to the legal right to use a proprietary technology in noncommercial research. The scope and degree of clarity in the law with regard to research-use (or safe harbour) defence is highly debated, often ill-defined and inconsistent across jurisdictions (McBratney et al., 2004). As a result, academics often assume the use of patented technologies in exploratory research is immune from infringement. The research exemption tends to be more liberal in some parts of the world, like Europe, where research-use exemptions are often embedded in the law. In Canada, statute provides an exemption for reasonable use during development and regulatory approval. Research-use defence still remains uncertain for Australian researchers because of the lack of case law or statutory exemption. In contrast, in the US, the landmark Madey v. Duke University case practically invalidated any sense of a safe harbour exemption for university researchers, although in practice US Universities are rarely a target of suing cases for patent infringement. Regardless of the jurisdiction in which scientists develop their research; the use of proprietary technologies for commercial purposes is clearly considered patent infringement in all jurisdictions.

All research institutions, whether private or public, now face an increasingly complex IP landscape that can influence the development of genetically engineered (GE) crops. This is particularly true of so-called enabling technologies—the research tools such as Agrobacterium-mediated transformation, selectable markers and promoters which are necessary to produce transgenic plants. The fragmented ownership of IPRs across multiple public and private sector owners produces situations where few single institutions can provide a complete set of IP rights to ensure freedom-to-operate (FTO) with any given technology, giving rise to the development of a so-called anticommons (Heller and Eisenberg, 1998). This situation is particularly true for subsistence and specialty crops, the historically important work of public-sector research, where only isolated efforts have been made to assemble complete IP portfolios applications in these fields (Wright, 1998; Conway and Toenniessen, 1999). A prominent example of the complexity resulting from fragmented IP ownership is the case of Golden Rice in which over 40 patents or contractual obligations associated with Material Transfer Agreements (MTA) represented potential constraints for its commercial development (Kryder et al., 2000). While this situation at first appeared intractable, the Golden Rice was manageable, although the transaction costs were high. The serious difficulties faced by the public research institutions in assembling necessary IP rights to support FTO of a project have less to do with blocking patents and more to do with misinformation, high transaction costs and a lack of an up-front IPR strategy to support the translation of the research to applications. Public sector researchers often lack access to IPR information, affordable due diligence services, strategic IPR management and resources available for negotiating rights, when necessary.

The IP landscape in agricultural biotechnology has always been dynamic, but this is particularly true today. The increasingly rapid discovery of new genes through genome sequencing is contributing to the continued growth in new patent applications and at the same time, patents that arose at the dawn of the biotechnology revolution are now expiring and the technologies entering the public domain. Although there have been arguments that the proliferation of IPRs in agricultural research has constrained new developments, the opposite is more likely to be correct. In addition, a recent paper argues persuasively that strong IP regimes in developing countries will be critical to support the agricultural innovations that will be needed to feed the growing global population (Grimes et al., 2011).

In response to the growing significance and complexity of IP and regulatory issues in agricultural research, and, in particular, in plant biotechnology, there have been a number of public sector initiatives including the Center for Application of Molecular Biology to International Agriculture (CAMBIA), the Public Interest Intellectual Property Advisors (PIIPA), the African Agricultural Technology Foundation (AATF), the International Service for the Acquisition of Agri-biotech Applications (ISAAA), the Specialty Crops Regulatory Assistance (SCRA) and the Public Intellectual Property Resource for Agriculture (PIPRA). PIPRA was founded in 2004 by the Rockefeller Foundation, and its primary motivation was to address the concern that public investments in agricultural biotechnology targeted to developing countries were facing delays due to IP and technology transfer issues (Atkinson et al., 2003; Delmer et al., 2003). From its inception, PIPRA has worked broadly to support development of specialty and minor acreage crops by public and private sectors as well as crops important to food security in developing countries. In this paper, we present PIPRA as a case study of the public sector’s efforts in navigating the landscape of IPRs by providing access to due diligence legal information, assistance in negotiating licenses, and in developing specific research tools designed to lower IPR transaction costs.

Here, we review PIPRA’s evolution and the development of information and analysis, educational and laboratory research platforms to support the development of agricultural biotechnology. PIPRA has occupied a unique position, straddling the disciplines of law and science, and operating at the interface of public and private sectors. It has a wealth of insights to offer from years of practical experience that should be used to inform practices and policies in public sector institutions, private sector interactions with the public sector, and national policy debates.

Information, analysis and educational outreach

A critical element in addressing IP issues in agricultural research is the ability to access and interpret IP information, a skill that is not particularly well developed within public research institutions. The basic information is embodied in patent databases, which can be widely accessed although more detailed information is embedded within each researcher’s laboratory and includes MTAs and so-called shrink-wrap licenses that accompany many commercial research reagents. One of PIPRA’s major objectives has been to support public research institutions as well as funding agencies and foundations in gathering the information and analysis to determine whether IP issues will constrain or facilitate the translation of basic research to applications. This has been particularly important for research sponsors whose primary goal is to support the development of innovations with clear applications to improve agricultural production for either commercial or humanitarian purposes.

To improve IP information access, PIPRA created a public institution patent clearinghouse, an online, publicly accessible database that consolidated thousands of patents owned by universities and public sector research organizations. The interface provided direct links to view the patents and, for many institutions, a link to the corresponding licensing officer to obtain IPR status. However, for reasons of institutional confidentiality, the PIPRA patent clearinghouse did not provide access to licensing information and the database itself was limited to public research institutions. These two limitations were problematic and, as a result, the clearinghouse did not provide added value relative to the diversity of existing online patent databases (Thangaraj et al., 2007). The PIPRA patent clearinghouse represented one model of a ‘patent commons’ (Bennett and Boettiger, 2009) and, as a result, PIPRA became a key resource for consultation on a wide variety of patent commons models across agriculture, energy and health. It is worth drawing the distinction between an informational or clearinghouse commons and a patent pool or commons of patents that are licensed under common licensing terms. PIPRA’s experience in developing its patent clearinghouse provided the basis to link together a network of over 60 public research institutions and to utilize this network as a basis for assembling a patent pool to support the development of enabling technologies for plant transformation.

Access to IP information is just a first step, and the ability to analyse IP information and IP landscapes is also a critical. For example, many companies routinely analyse the FTO of a research project or product, assessing whether making, using or selling it is likely to infringe existing patents or other types of IPRs. The resulting information contributes to a larger assessment that may involve a range of options: identifying in-licensing targets, considering the substitution of technologies, deciding to ignore the potential infringement, investing in work-around technologies or perhaps deciding to abandon the project all together. Increasingly, public research institutions or foundations who sponsor research in public research institutions are asking the same questions. The Bill and Melinda Gates Foundation is one such organization who requires, as a condition of grant support, that institutions conduct a complete assessment of the IP landscape of the project and to identify and resolve any issues that might impede the commercial application of the research in its intended markets. On many occasions, PIPRA has conducted these IP assessments as a service to the Foundation. It is important to note that IP rights are granted by individual countries and only apply in the geography in which they are granted. In the case of the Bill and Melinda Gates Foundation supported projects, the research is typically carried out in the United States, often using components imported from other developed countries, the product is intended to be produced in a developing country and ultimately to be sold in the same or other developing countries. As you can be imagined, the analysis of the IP landscape needs to be conducted in each geography to adequately assess the potential for IP hold-ups in deploying the research results. It is interesting to note that even when significant IP issues have been identified in projects targeted towards agricultural development in developing countries, they have been easily resolved through the use of royalty-free nonassertion or license agreements from the IP owners. Although many of these IP donations are not publicized, there are a few well-known recent examples of agricultural biotechnology companies donating technology for crop development targeted to alleviate poverty in Africa (AATF, 2009; Monsanto, 2011). These recent examples signal a changing environment where IP is less of a constraint especially for research targeted to parts of the world where there is little potential for significant commercial markets, but there is still a high social value for the anticipated research results. Nevertheless, it is important to examine the IP landscape of a project at its inception, to plan the research in ways that minimize IP issues and then to address and resolve the remaining issues before the project starts.

In addition to having access to IP information and analysis, there is a real need for public research institutions to develop the capacity to analyse and evaluate that information in the context of a translational research project. As a consequence, education and capacity building primarily in developing countries is one of the most significant services PIPRA offers. Among its contributions, PIPRA, along with the Centre for Management of Intellectual Property in Health Research and Development, published a two-volume IP Handbook entitled ‘Intellectual Property Management in Health and Agricultural Innovations’ (Krattiger et al., 2007). Given the demand, editions were translated into Spanish in 2010 (Anguita et al., 2010). Printed and publically available electronic versions of the Spanish and English versions are available from http://www.pipra.org. An educational curriculum has been developed to accompany the Handbook, and courses have been delivered in Asia, Africa and Latin America as well as an in-depth course taught at PIPRA’s headquarters in Davis, California.

Public Intellectual Property Resource for Agriculture is an ongoing experiment by the public nonprofit sector to develop a framework for IP access and sharing globally and at the root of access, and sharing is having the information and analytical capability to clearly assess the IP landscape. The experience to date has been that facilitating access to IP information, assessing IP landscapes and training personnel provide the greatest value to public research institutions. This is especially true in developing countries who are often just beginning to identify their own potential for generating valuable IP assets and recognizing the need to evaluate IP belonging to others. PIPRA has established an international presence through its network that extends to over 60 universities and research institutions in 17 countries. In 2009, PIPRA and the Foundation for Agricultural Innovation in Chile developed a joint programme (FIA-PIPRA) aimed at catalysing the innovation process in Chile and South America (http://pipra.fia.cl). In the future, we anticipate that Latin America may be one of the most rapid adopters and developers of GE crops and that there will be a strong need to support the emergence of stronger IP regimes in this region.

Plant transformation enabling technology with FTO

Navigating the patent ‘thicket’ in plant transformation

Public Intellectual Property Resource for Agriculture was challenged by the Rockefeller Foundation to address the limited FTO (Atkinson et al., 2003), or legal rights to use third-party IPRs (Fenton et al., 2007), for the enabling technologies required to make any new GE crops. These enabling technologies include a range of technologies, such as transformation methods, selectable markers and promoters. In the early 2000s, when PIPRA was established, IPRs to key enabling technologies were controlled by a few major agricultural biotechnology companies. In transformation methods, the public sector invented and patented some of the cornerstone technologies but then restricted its own access to these technologies by granting exclusive commercial licenses (Graff et al., 2003). In the area of selectable markers and promoters, some of the most widely used technologies were patented by the private sector (Graff et al., 2003; Dunwell, 2005). Lack of clarity of which IPRs are relevant and access to the entire package of enabling technologies directed biotechnology investment to focus on certain crops, traits and institutions. Investment was deterred particularly in small acreage crops for which the cost for access to enabling technology IPRs and regulatory compliance was (and is) not economically viable (Pray and Naseem, 2005). Obtaining FTO also increased the transaction costs for the deployment of crops developed for humanitarian or nonprofit purposes as was the case for Golden Rice (Potrykus, 2001). In essence, IP and regulatory hurdles were, and continue to be, excluding the public sector from transferring their scientific discoveries to meet the agricultural and food demands of the global challenges that lie ahead.

Navigating the myriad patents in Agrobacteria-mediated transformation is not a simple process, as described in a thorough review (Nottenburg and Rodríguez, 2008). In view of this patent thicket, it was compelling for PIPRA to attempt to develop plant transformation tools that minimized, where possible, third-party IPRs. PIPRA’s undertaking was to make broadly available a plant transformation enabling toolkit which consists of a suite of plant transformation DNA vectors that were designed with considerations of technical, regulatory and IPR issues (Box A). Technical functionality criteria should ensure that the transformation vectors and technologies are as convenient and easy to use as technologies that are routinely used by the research community today. To address regulatory and public acceptance issues, the toolkit should have features that address the current or potential regulatory framework – for example, features that allow for ‘marker-free’ transformation and for the incorporation of all-plant-derived constructs. Public acceptance of GE crops is reported as high as 80% when the plant contains only plant DNA and only 20% if the transgene is from nonplant sources (Lusk and Sullivan, 2002). Thus, PIPRA’s transformation vectors were designed so that the resulting plant will, to the extent possible, be compatible with technologies that allow for all-plant-derived constructs, such as the incorporation of plant-derived T-DNA borders (Rommens, 2004). Finally, with regard to IP FTO, the enabling toolkit should, to the extent possible, be free of IP restrictions or be freely available for research and humanitarian uses and affirmatively available for commercial applications under prenegotiated terms.

The Rockefeller Foundation sponsored the up-front design and validation costs of the suite of enabling technologies for plant transformation. The resulting pPIPRA vectors consolidate access to IPRs under prenegotiated licenses, thereby lowering transaction costs and encouraging a shift in the capacity of universities and public sector organizations to adopt a transformation platform that is compatible with the translation of plant biotechnology research to commercial applications. PIPRA developed a design process that incorporated input from an extensive patent attorney network that provided legal information and opinions, from regulatory consultants, and from scientists and that resulted in a strategy to develop plant transformation technologies addressing the technical, regulatory and IP considerations. PIPRA focused on the suite of technologies included in transformation vectors; FTO for the use of the vectors with specific traits, transformation methods, germplasm, etc. would need to be re-evaluated for each specific application.

First, a thorough and comprehensive review of the patent landscape around plant transformation identified areas with limited FTO (i.e. selectable markers). In areas that represented IP ‘road blocks’, PIPRA identified scientific comparable technologies that offered greater FTO, because they were not patented (and in the public domain) or patented by a PIPRA-member university and available for a nonexclusive license option (i.e. promoters with FTO). In addition to IPRs, the owners of tangible property used in the laboratories, like DNA vectors or other biological materials, can stipulate how the materials are used by third parties. To avoid research-use only restrictions often embedded in MTAs, PIPRA synthesized de novo a large number of the vector components, including the vector backbone. PIPRA’s laboratory then tested and validated the research and marker-free plant transformation platforms in model plants (Box A).

Patent bottleneck for plant selectable markers ‘uncorks’

In 2004, when PIPRA started the quest to develop an enabling toolkit for plant transformation, the field of plant selectable markers was an IP minefield. Finding technical replacements for other enabling tools, like promoters, was not difficult because we were able to identify many technologies that were in the public domain (not patented) or available for licensing. However, for plant selectable markers, there were only a few options of technically viable systems, and IPRs protecting the three most used selectable markers (Miki and McHugh, 2004) created a bottleneck. And, the technology is critical for genetic engineering because of the inherent low transformation efficiency, of about 2% to 20% in monocots and dicots, respectively (Opabode, 2006; Tzfira and Citovsky, 2006).

Notably, although there are over fifty plant selectable marker systems used in plant genetic engineering, only three were used in over 90% of the scientific publications: antibiotic resistance to kanamycin (nptII, neomycin phosphotransferase II) and hygromycin (aphIV, hygromycin phosphotrasferase), and III herbicide tolerance to phosphinothricin, the active ingredient in BASTA (Bayer CropScience, Hawthorn, Victoria, Australia) (bar or pat, phosphinothricin N-acetyltransferase).

Bayer CropScience (formerly Aventis CropScience) holds patents in different countries related to phosphinocitricin-based herbicide selection. Selected family members in the US include US5561236, US5646024 and US7112665, expiring no later than 2014 and US5648477 expiring in 2023.

Novartis Ag (formerly owned by Eli Lilly & Company) owns an international portfolio including two patents in the US covering the use of the antibiotic hygromycin-based plant selection, US6048730 and US5668298, with predicted expiration date of 16 Sept. 2014.

Public Intellectual Property Resource for Agriculture’s navigation of the IP landscape focused on the preferred kanamycin resistance system. The bacterial nptII gene is by far the most commonly employed selectable marker in transgenic crop research and US field trials, and all commercially released crops had employed either nptII or bar or pat (Miki and McHugh, 2004; Petersen et al., 2005). The prevalent use of nptII in commercial crops is attributed in part to the selection system’s effectiveness and deregulated status in the US and numerous other international government agencies. The selectable marker (nptII) is listed by the United States Food and Drug Administration as generally recognized as safe (GRAS).

Monsanto Co. held an extensive and international patent family for antibiotic-based and kanamycin resistance systems (Figure 1). In particular, US6174724 was a key IP asset that granted Monsanto broad, exclusive rights to use any antibiotic resistance gene, under any plant expression promoter, to select any plant cell. A member of this patent included US 5034332 with narrower protection limited to particular promoters to drive expression of the antibiotic resistance gene.

Figure 1.

 Patent Family Tree for Antibiotic and Kanamycin-based Plant Selection. Patent filing applications are shown in gray background and patents are shown with white background.

Public Intellectual Property Resource for Agriculture also kept a close watch over Monsanto’s US patent application 08/127,100 because of its potential, if awarded, to have restricted access to this technology for an additional 17 years. This application filing remained under prosecution for almost 16 years.

Because of the limiting FTO available in this technology space at that time, PIPRA sought alternative selectable markers and initiated license negotiations with the Universities of Kentucky and Tennessee to incorporate the AtDEF1 and AtDEF2 (Dirk et al., 2001, 2002) or Atwbc19 (Mentewab and Stewart, 2005) to the enabling toolkit. However, research testing showed these alternative systems were not as widely applicable as the nptII gene.

In 2008, the patent landscape of antibiotic-based plant selectable markers was significantly altered with the expiration of Monsanto’s US 6174724 and US 5034322 on July 23rd. And after almost two decades of prosecution, the US application 08/127,100 was abandoned on 24 Nov. 2008.

At that time, with the exception of two US patents, members of Monsanto’s international family tree were expired (Figure 1). The two US patents contain claims for the use of specific promoters, CaMV 35S and 19S, to drive any chimeric gene, including the nptII gene. The remaining two patents in the US, US5352605 and US5530196, recently expired on 4 Oct. 2011.

Considering this remarkable change in the IP landscape, PIPRA designed and developed plant transformation DNA plasmids with maximum FTO using the nptII gene as selection system driven by a promoter different than the CaMV 35S and 19S. As described later, PIPRA chose the University of California’s FMV34S promoter because of its legal access and comparable expression profile as the CaMV35S (Sanger et al., 1990). Consideration of the evolving patent landscape allowed PIPRA to develop a FTO path within the heavily patented field of plant selectable markers. Similar strategies can be applied to other key agricultural areas.

Promoters with FTO to support new agricultural crops

Transcription regulatory elements that control the expression of desirable traits or selectable markers in specific plant or tissue organs and developmental stages are critical when developing biotechnology products. Despite the number of promoters described in scientific literature, to date, viral-derived promoters, like the Cauliflower Mosaic Virus CaMV35S and its derivatives, particularly the enhanced CaMV35S (Odell et al., 1985; Kay et al., 1987), are often preferred because of their ability to drive high and constitutive expression of genes in monocotyledonous or dicotyledonous plants. Commonly used plant binary vectors utilize the CaMV35S promoter to drive plant selection markers, reporter genes and genes of interest. However, use of CaMV35S is restricted by numerous overlapping patents and is often distributed in vectors with limiting MTAs (see section above and Dunwell, 2005). Mining scientific and legal information, PIPRA developed a database with over 800 promoters with a wide range of tissue and expression profiles. From this resource, we identified the 34S promoter from the Figwort mosaic virus M3 strain (FMV 34S) that confers constitutive gene expression. The FMV34S (M3 strain) promoter is the subject of US Patent 6051753, which expires on 18 April 2017. Although the patent is co-assigned to Calgene (now Monsanto Company) & University of California, the University solely manages and grants TP and IP rights (Box A).

Several reports have shown the FMV34S and CaMV35S exhibit analogous expression profiles and that enhanced versions of both promoters increase expression in the model plant, tobacco (Kay et al., 1987; Sanger et al., 1990; Maiti et al., 1997). To test whether the FMV34S is a technically suitable substitute to the patented promoter, PIPRA’s laboratory characterized the promoter expression of FMV34S in Arabidopsis and tomato (A.B. Bennett et al., University of California Davis, unpublished data). As schematically represented in Figure 2a, chimeric constructs were created in which full-length or enhanced versions of CaMV35S and FMV34S promoters drove the expression of the reporter gene uidA (GUS) encoding the β-glucuronidase enzyme. There was no significant increase in transcription activity between the full-length and enhanced FMV34S promoter constructs in tomato or Arabidopsis (Figure 2b–d). In tomato, the enhanced CaMV35S promoter had an increased expression of 30% and 16% versus the full-length CaMV35S, in leaves and fruit, respectively (Figure 2c,d). The FMV34S promoter showed about 3.5-fold higher expression level in leaves than in fruit (Figure 2c,d).

Figure 2.

 Characterization of Figwort mosaic virus, 34S Promoter Expression (a) To assess promoter expression, pCAMBIA1381Z-derived binary vectors containing chimeric promoter-gene uidA (GUS) constructs were generated. The chimerical constructs consist of full-length (34S::GUS pPIPRA10, 35S::GUS pPIPRA47) or enhanced versions of the FMV34S and CaMV35S (e34::GUS pPIPRA15, e35::GUS pPIPRA41) promoters fused to the GUS reporter gene and nopaline synthase 3′UTR (NOS). A promoter-less construct (PL) was included for control purposes. FMV34S promoter consists of a fusion of nucleotides 718–1617 and 371–414 from NCBI M59930 and X16673, respectively. The enhanced FMV34S promoter consists of an enhancer region, nucleotides 1236–1567, NCBI M59930, fused to the full length promoter. The CaMV35S constructs were used as described by Kay; 35S consists of nucleotides −343 to +9 and e35S nucleotides −343 to −90 fused to −343 to +9 (Kay et al., 1987). Promoter expression was evaluated visually by histochemical detection and semi-quantitatively by enzyme assays as previously described (Jefferson, 1987) in Arabidopsis (b); tomato leaves (c) and immature green fruit (d) from transgenic plants. Fruit were sliced transversely and stained for GUS activity (d). Ten independent tomato lines and five independent Arabidopsis lines per construct were analyzed. One representative picture of tissue from each transgenic line is shown. The bars show the average of GUS activities for all transgenic lines of each construct ± standard error. Unpublished data by A.B. Bennett et al., University of California.

The data obtained from our promoter analyses led us to conclude that the FMV34S drives significant expression, particularly in leaf, and that the enhancer duplication used in this study was not necessary to increase the FMV34S M3 strain promoter expression. In addition, the FMV34S promoter was effectively used to drive gene expression in other important crops such as rice, alfalfa and lettuce, supporting its use as efficient promoter in both monocots and dicots (A.B. Bennett et al. and K.J. Bradford et al., University of California Davis, unpublished data).

Consolidating patent rights

Public Intellectual Property Resource for Agriculture explored the feasibility of consolidating patent rights to key enabling biotechnologies and offering nonexclusive licenses in a convenient one-stop shop (Atkinson et al., 2003; Delmer et al., 2003). The goal of consolidating or ‘pooling’ IPRs was to lower transaction costs for adopters of the enabling technologies and to have a transparent mechanism with predetermined costs to translate fundamental research projects to commercial applications. University technology managers were in agreement with this IPR pooling model. The main stipulations for the ‘pool’ of technologies were that IPRs would be licensed to the pool under nonexclusive terms, in exchange for a reasonable fee for commercial use and under royalty-free terms for humanitarian use (defined geographically) (Bennett, 2007). Technology providers could also license their individual technologies outside of the pool, so long as terms were nonexclusive and would not limit use of the technology within the pool. For many managers, the PIPRA ‘pool’ was welcome since it was perceived as an additional marketing vehicle. Negotiations mirrored the legal discussions that have occurred in the development of patent pools in other industries. For example, valuation of the technology with respect to other components in the pool and reconciling differences in patent life of different components were points of significant discussion. In the end, technology providers agreed that all component technologies would be valued equally, as the patents, collectively, were more valuable than any individual patent and any commercial revenues would be shared among the technology providers in proportion to the number of technology components they provided. To decrease transaction costs, technology providers strongly supported the administration of the patent pool through a single entity which had proscribed permissions to sublicense third-party patent rights. Since PIPRA is not a legal entity, the University of California, serving as the host institution, manages and licenses the technology pool on behalf of PIPRA-member universities.

During the evolution of this undertaking, the number of proprietary technologies necessary to achieve FTO was greatly reduced, mainly due to the expiration of patent rights corresponding to ‘roadblocks’ in the patent landscape. This was the case for selectable markers used in the transformation process. Currently, the only patents rights in the simplified toolkit are for the promoters, and since the vector system is modular, the number of patented technologies can range between zero and two, depending on the number of proprietary promoters utilized (Box A).

Public Intellectual Property Resource for Agriculture selected a viral and bacterial constitutive promoter to drive the selectable marker gene and the gene of interest (GOI) because of their ability to drive expression across both monocotyledonous and dicotyledonous plant species. In the case of FMV34S, patent rights can be licensed from the University of California. For the MAS promoter, the IPRs were developed at Purdue University but exclusively licensed to a company that is supportive of PIPRA’s mission and is facilitating IPR access on a case-by-case basis. Biological or tangible rights to the FMV34S promoter and PIPRA’s plant transformation vectors are solely managed by the University of California and can be obtained under royalty-free terms for research or humanitarian applications or on a nonexclusive, fee basis for commercial purposes.

In the end, the strategic design of the enabling toolkit was successful in demonstrating that achieving a high degree of FTO with enabling transformation technologies is possible. Determining what IP is relevant, ownership and accessing rights to certain technologies is a major issue in translational research. PIPRA’s expertise and resources in navigating and facilitating access to agricultural technologies continue to allow the provision of an important service to minimize IP restrictions. This can be achieved by early consideration of IP issues, either by utilizing preassembled packages, such as PIPRA’s vector patent pool, or by addressing FTO through assessment of applicable patents and MTAs on an continuing basis throughout the life of a research project.

Public Intellectual Property Resource for Agriculture, through two independent public–private partnerships, is currently validating the transformation systems in research and development pipelines for both commercial and humanitarian applications. The AATF was recently granted royalty-free access to PIPRA’s enabling, marker-free, technologies and Arcadia Biosciences’ traits to develop nitrogen-use efficient and salt-tolerant African rice varieties (AATF, 2009). A private seed company has also been granted a commercial license to develop GE crops with potentially broad commercial applications.

Conclusion

Public Intellectual Property Resource for Agriculture was created in 2004 to address IP issues in agriculture. Its timely creation addressed one of the important barriers to the public sector’s engagement in translational agricultural research. IP issues are important but only when coupled with the adoption of a comprehensive translational research paradigm for public agricultural research. PIPRA has sustained its focus and expertise in addressing IP issues but has also taken on a broader mandate is addressing other barriers in translational agricultural research including regulatory requirements and the development of strategies to bring research project results to scale. Perhaps most importantly, PIPRA has developed a significant education and outreach programme targeted at increasing the capacity of developing-country researchers and technology managers to address IP and translational research barriers in their own unique institutional environments.

The public agricultural research sector continues to be important in developed countries but to a diminishing extent, relative to private sector R&D. The opposite is true in developing countries where the public sector is the sole player in agricultural R&D and must play the major role in addressing the looming issues of food security. PIPRA has expanded its range of services and set its sights on contributing in a variety of ways to ensure that the public sector can continue to fulfil its historic role in research, education and agricultural food security.

Box A. PIPRA’s enabling packages: plant transformation DNA vectors

Enabling transformation technologies require a platform of fundamental components, including a gene delivery ‘vehicle’, selectable markers and promoters assembled in convenient vectors. When compared with other vector systems currently in use, with primarily static vector components, pPIPRA vectors are modular and flexible for customization and efficient use in plant transformation.

Single T-DNA binary vectors are the predominant option when creating GE crops. The 1T-DNA construct facilitates selection of genetically modified cells by co-integration of the selection marker and GOI. Considering the legal information of NPTII and scientific preference for this marker, we opted to use this plant selectable marker. We designed PIPRA’s 1T-DNA for research applications in mind, testing of new trait genes. However, the vector may also be used to engineer crops with commercial application. The safety of NPTII protein has been confirmed by numerous international regulatory agencies and the World Health Organization. Furthermore, the selectable protein has been used in a variety of crops approved for commercial use including corn, potato, tomato, cotton, flax, chicory, cotton and oilseed rape (Miki and McHugh, 2004).

Public Intellectual Property Resource for Agriculture’s 1T-DNA vectors allow cloning of the GOI through a two-step processes, via a shuttle vector or directly into the plant binary vector. In the two-step cloning strategy, the GOI is first introduced into the multiple cloning site (MCS) of the FMV34S-shuttle vector, pPIPRA522 GenBank JF811681 (Figure 3a). This shuttle vector contains the constitutive promoter, FMV34S and pea RuBisCo E9 3′UTR. In a subsequent cloning step, the GOI cassette can be transferred from the pPIPRA522 shuttle vector into the pPIPRA560 plant binary vector, GenBank JF811682 (Figure 3b). For this subcloning step, the GOI cassette may be excised from the shuttle vector, with the enzyme recognition sites PacI (blunt) or KasI (Klenow-filled) and ligated into the unique PacI site of pPIPRA560. Customization of the pPIPRA522 shuttle vector is possible by replacing the promoter and 3′UTR components using the flanking rare-cutting restriction endonuclease recognition sites (Figure 3a, ClaI and AsiSI). The second cloning strategy is a one-step process, in which the GOI may be cloned directly into MCS of the pPIPRA561 plant transformation plasmid, GenBank JF811683. With the exception of EagI, the MCS of pPIPRA561 is identical to that in pPIPRA522 (Figure 3b).

Figure 3.

 Public Intellectual Property Resource for Agriculture Plant Transformation 1TDNA, Co-Transformation and Transposase-based Marker Free Systems. (a) pPIPRA 522 shuttle vector containing a GOI cassette with a MCS. Circular arrows indicate promoter and 3′UTR components that may be substituted using the flanking restriction sites. (b) pPIPRA binary plant transformation vectors backbone include bacteria origin of replication, ColE1 for E. coli and pVS1 for Agrobacterium/Rhizobium, and a kanamycin bacterial antibiotic resistance marker. 1TDNA includes pPIPRA 560 binary vector with a PacI unique restriction site to clone the GOI cassette, from pPIPRA522 shuttle vector. pPIPRA561 is a binary vector that contains the FMV34S promoter and 3′UTR and a MCS where the GOI can be cloned. Plasmids pPIPRA522 and pPIPRA560 share the same MCS (a), with the exception of EagI, which is only unique in pPIPRA522. NCBI accession numbers are shown in grey. The design of the co-transformation plasmids includes separate plasmids to deliver the GOI and selection (NPTII) and segregation markers (CodA). The transposase-based vector system includes the transposase enzyme, GOI cassette flanked by left and right transposon recognition sites (LTR, RTR), and selectable (NPTII) and segregation markers (CodA). Integration of isopentenyl transferase cassette is integrated in the backbone of the marker-free vectors.

Anticipating industry standards and consumer preference, PIPRA is finalizing the validation of two transformation platforms designed for the production of marker-free plants using cotransformation or transposon-mediated DNA-excision strategies (Figure 3b). The cotransformation system includes the two separate plasmids and separate T-DNA for the trait and selectable marker genes. There are numerous patents in this space, most bearing on the delivery of two T-DNAs in a single Agrobacteria strain. We employ an approach, previously utilized in the literature, in which T-DNA’s are delivered separately.

The transposon-mediated systems uses the maize Ac/Ds, previously used to generate marker-free tomato plants (Yoder et al., 1988; Goldsbrough and Yoder, 1993). The transposon system may be particularly favourable for plant species recalcitrant to transformation because numerous unique insertion lines may be generated from a single primary transformed plant. With this strategy, unnecessary DNA (e.g. backbone vector DNA), transposon and other T-DNA elements can be segregated to generate a final plant that only contains the trait of interest. Patents covering the Ac/Ds system to generate marker-free events were awarded to the University of California, US5225341 and US5792924. Both patents expired in 2010; however, given the complexity of this multicomponent vector, TP rights for PIPRA’s Ac/Ds-transposon vector can be granted through the University of California.

One of the broadest patents in dicot transformation, US6051757, was granted to Washington University and licensed to Syngenta. The IP is anticipated to expire 18 April 2017 and pertains to methods of using an Agrobacteria strain that does not contain a functional cytokinin gene (Nottenburg and Rodríguez, 2008). PIPRA’s modular design allows for the integration of isopentenyl transferase (IPT) cytokinin gene in the plant transformation vectors backbone. While this IP consideration may only be applicable when transforming dicot plant cells, the IPT gene also serves as an efficient marker to detect undesirable backbone integration events (Rommens, 2004) and thus is useful for transformation in general.

To minimize potential TP issues due to MTA, pPIPRA T-DNA binary vectors, pPIPRA560 and pPIPRA561 include de novo synthesized bacterial origins of replication (ColE1 for E. coli and pVS1 for propagation in Agrobacterium or Rhizobium) and the kanamycin-resistant bacteria selection marker aadA for maintenance in bacterial cells.

Given the modular design of PIPRA’s vectors, it is possible to develop the transformation vector with nonpatented components. Current versions of the vectors are offered with promoter elements that are proprietary but for which rights can be obtained. In the case of FMV34S promoter, IP and biological rights to the FMV34S promoter included in our vectors as well as pPIPRA’s plant transformation vectors are solely managed by the University of California and can be obtained under royalty-free terms for research or humanitarian applications through a shrink-wrap license (http://www.pipra.org) or on a nonexclusively, fee basis for commercial purposes.

Acknowledgements

Public Intellectual Property Resource for Agriculture was established with funding from the Rockefeller and McKnight Foundations and has also been supported by grants from the US Department of Energy, the US Patent and Trademark Office, the Bill and Melinda Gates Foundation, the Pierce’s Disease/Glassy Wing Sharpshooter Board, the UC Discovery Grant Program, the Sasakawa Peace Foundation, the Global Alliance for Livestock Veterinary Medicine, The African Agricultural Technology Foundation, the Fundación para Innovación Agraria, and USAID-HortCRSP Horticulture Collaborative Research Support Program. Special thanks to the numerous law firms and attorneys that provide pro-bono services to PIPRA.

Ancillary