October 20-21, 2010
Purpose and Workshop Overview The purpose of the NIH Common Fund Workshop: Renewable Protein Capture Reagents, October 20-21, 2010, Hilton Washington DC/Rockville Hotel & Executive Meeting Center was to gather information on the state of the art in technology for production of a comprehensive public resource of affinity reagents for the human proteome (http://nihroadmap.nih.gov/proteincapture/).
The NIH Common Fund is exploring the technical and practical ability to produce a comprehensive resource of renewable affinity reagents against the human proteome. As a beginning, in 2010 and 2011, common fund initiatives will solicit production of affinity reagents against human transcription factors. Although monoclonal antibodies are currently a mature methodology with clear downstream applications already developed, they have drawbacks that may be a practical barrier to creating a more comprehensive set of protein capture reagents. For example, the high cost of development and the quantity/type of downstream validations that are applied to ensure wider utility of the reagents.
In addition to monoclonal antibodies, there is a range of alternative technologies such as recombinant antibodies, engineered proteins, and aptamers, which are at various stages of development, and present encouraging opportunities towards generating a comprehensive set of protein capture reagents. Given the combined considerations of cost and range of end use, it may be that multiple approaches will be needed to generate a comprehensive resource. Central questions remain in what are the most promising approaches for the generation of a complete set of protein affinity capture reagents for wide use by the research community; what are the limitations and advantages of each approach; what type of affinity reagent can be productively encouraged to move forward at a "production level" what are timelines; what is the best way to stimulate each; what are the advantages or disadvantages of "waiting" until one or more of these are ready for production?
The concept of a "comprehensive" set of capture reagents begs the question of how many human proteins should be considered to represent the proteome in a way that is meaningful for biomedical research. How will the resource address the issues of splice variants and post-translational modifications? Clearly, these considerations could have a large effect of the ultimate cost in addition to the biological and technical issues involved in producing “comprehensive” reagents. The intent of the workshop was to solicit practical advice about regarding these issues and how to balance costs and utility in a framework with reasonable milestones for a comprehensive resource.
The NIH is aware that there are other similar efforts, at various scales and with various aims. The workshop participants (see attached participant list) were asked to provide feedback on how other similar and parallel efforts should be integrated or considered in a federal resource. In addition, participants were asked to provide discussions of the balance of intellectual property constraints and resource openness to the scientific community in the context private and public approaches.
The Workshop agenda organized the proceedings into plenary and breakout sessions. The opening session consisted of presentations in plenary session (see attached agenda) which introduced general discussions in topical areas of: (a) Current Technologies, (b) Alternative/Future and (c) Coordination and Resource Distribution. Presenters offered discussions and a range of opinions based on special knowledge and interests which included existing widely applied polyclonal and monoclonal immunoglobulin technologies, developing recombinant engineered protein affinity technologies, aptamers, and advanced chemical methodologies. These discussions served as starting content for the break-out sessions which divided the participants into three topics: (1) What are the prime considerations in developing a scalable protein capture resource now? (2) Technologies for the near future: What is needed for significantly improving throughput and reducing cost within 3 to 5 years, and (3) Strategic issues: Integrating with other worldwide efforts, maximizing resource accessibility and downstream utility.
- The consensus of the workshop was that there are least two approaches ready to scale up, to some level, the production of renewable affinity reagents: traditional monoclonal and recombinant antibodies. The Human Protein Atlas, lead by Dr. Mathias Uhlen, has put most of its effort in the production of polyclonal antibodies and recently reached the goal of producing reagents for about 10,000 proteins. However, a small fraction of these are monoclonal. The Structural Genomics Consortium, lead by Dr. Aled Edwards, is another significant effort that was recently started and it aims at the production of monoclonal antibodies. The existence of these parallel efforts will provide opportunities for coordination to avoid duplication and cooperation on targets, standards and dissemination as the NIH effort is initiated.
- Workshop participants thought that reagent quality was paramount and were unanimous in urging the extensive experimental standardization of protein capture reagents including the determination of affinity (Kd) and standard performance validation in each of the several important areas of application including ELISA, immunofluroescence, immuniprecipitation, etc. This extensive and more costly characterization raised concern that the projected budget for the production effort of about $5-7K per reagent ($10M over five years for reagents for 1000-2000 human transcription factors) would not be adequate to support the necessary effort including validation, startup equipment purchases and incremental technical development through the project. Reagents for the production of antigen targets for the effort may be already available through other open and publically funded initiatives such as those mentioned above or the Arizona State plasmid repository (http://dnasu.asu.edu/DNASU/Home.jsp) where many clones including those generated within the NIGMS Protein Structure have been deposited (http://psimr.asu.edu/).
- The workshop participants highlighted several technologies including aptamers and alternate recombinant display approaches which offered prospects of scalable development over 3-5 years. More focused technical developments and enhancements of production pipeline operations might also be worthwhile investment. These opportunities justified potential plans for an investment in several awards totaling $15M over three years.
- Participants in the workshop supported the availability of protein capture reagents usable in standard assays at a minimal cost without intellectual property constraints and with nonexclusive licensing arrangements which would facilitate a combined profit/ non-profit distribution scheme. Under such an arrangement, private sector partners would be able to add value through multiple modes of distribution or through products or services which leverage basic reagent production.
Resume of Break-out Discussions
- i. What are the prime considerations in developing a resource now
- ii. Technologies for the near future: a survey
- iii. Strategic issues: Integrating with other worldwide efforts; maximizing resource accessibility and downstream utility
i. What are the prime considerations in developing a resource now? The discussion focused on considerations relating to the direct and immediate development of a protein capture reagent resource. The consensus among meeting participants was that based on renewability, cost, throughput, consistency of efficiency and applicability in the hands of many researchers that monoclonal antibody and recombinant binder technologies are ready for scaling for production. The program would envision cost control and the use of the reagents by academic or private sector researchers should be unconstrained by burdensome intellectual property restrictions. Whether in a centralized or distributed format, the effort should be managed from beginning to end, from antigen selection to product catalogue and the process be as transparent and integrated as possible. The research community should recognize the NIH supported resource as a uniquely useful and reliable brand based on performance and reinforced by a carefully chosen name, term and logo. In the case of a monoclonal antibody reagent resource careful attention to quality assurance and quality control will be the essential for the success of the program. Information on the quality and characterization data of the affinity binder reagents should be easily and readily available. Characterization of the reagents by the primary producer and also by independent core testing facilities should be organized in a single data document together with clear "how to use" methodology included or offered in easily accessed secondary sources, as done in focused public efforts to develop and distribute antibodies of interest to specific research communities (e.g. http://antibodies.cancer.gov). In addition to the initial characterization of the antigen a standard set of criteria for the characterization such as those already been adopted elsewhere (e.g. MIAPAR, http://www.nature.com/nbt/journal/v28/n7/full/nbt0710-650.html) should be applied. Performance within the several broad areas of application of such reagents including enzyme-linked immunofuorescence (ELISA), Western blotting, immunohistochemistry and immunoprecipitation should be specified. Where the possible application in more specialized assays, such as ChIP-seq or flow-cytometry, should be included in the product specification data sheet. While recombinant binders have so far seen fewer downstream applications it was thought that they represent a clearly viable approach for scalable developments. Given the availability of antigens for development a choice based on the relative merits of these two approaches which revolve around issues of availability and costs of screening technologies and other practical matters should be left to the process of application and peer review of responses to a formal funding opportunity announcement. Responses to such an announcement should be judged by standards which include the clear goals and milestones for quality, performance and cost with improvements in efficiency and output over the life of the effort. Distribution of the affinity reagent products of this effort should be centralized and stable over the long term with the Developmental Studies Hybridoma Bank (DSHB, http://dshb.biology.uiowa.edu/) or the Mammalian Gene Collection (MGC, http://mgc.nci.nih.gov/) considered useful models.
ii. Technologies for the near future: a survey focused on a discussion of technologies which might be capable of large scale production in the near future (3-5 years). There was a consensus that single-chain variable fragment fusion proteins and other immunoglobulin display derivatives were ready for testing during that timeframe. Perhaps also aptamers might be considered in this case. Other technologies including combinatorially engineered scaffolds such as Affybodies and fibronectins, imprinted polymers and chemical reagents were considered but not discussed extensively. There was the sense that these technologies might not be at the same level of development as the recombinant antibodies but might have good value in specific applications. Combinatorially engineered protein scaffolds as alternatives to the immunoglobulin fold were considered valuable, particularly in the context of well adapted production platforms such as yeast or phage. As with these approaches, every candidate technology should have demonstrated initial proof of concept. Discussants agreed that whatever class of affinity reagent chosen that convincing proof of these technologies would require demonstration of sensitivity and specificity according to common standards varying with the assay (e.g. ELISA, immunoblot, immunohistochemistry, immunoprecipitation, etc.). Initial performance baseline should be set at minimum to the level of existing technologies. The ultimate proof of utility at pilot scale would require deploying the technology in a monitored pipeline. This scheme should be established according to preselected milestones and operable determine of costs, quality and throughput throughout the process. The product of developments should be judged quantitatively by affinity (Kd), sensitivity, specificity, cost and more generally based on renewability, throughput and scalability of the process, ease of use and utility tied to satisfactory performance in user’s hands and generalizability of the reagent technology to other protein classes including post-translationally modified proteins. High affinities in the nanomolar range have often been coupled with “high quality” reagents. Although a clear cut off here can’t be determined, because it is dependent on the application, the minimum is that all reagents are well characterized by their affinities and specificities and validated for their intended application. It should be mandatory that release of the quality controlled product reagents at the end of the process should be combined with the fully documented validation experiments documenting the quality of the reagent. Additionally, data, methods and technology related to the reagents should be broadly disseminated through publications, presentations and the web. Successful applicants for these scale-up projects should include plans for the distribution of the target antigens employed during the development and the capture reagents which are produced by this pilot activity. These projects should be funded and administered as cooperative agreements to facilitate the production and ultimate distribution of the products. Monitoring of progress by NIH program staff should include a mid-course review with the specific inclusion of adjustments of budget and scope within the individual projects in order to maximize the impact of the overall initiative. Among the potential approaches there is still a great deal to be learned about which may be the most attractive for further focused effort. The program should take an opportunity to select a range of approaches with the object of capturing the best of potentially scalable technologies. Longer term research and development efforts that can not show scalability to proteome scale within 3-5 years should be outside the scope of the present effort.
iii. Strategic issues: Integrating with other efforts, maximizing resource accessibility and downstream utility focused on the integration of the proposed NIH resource with ongoing and planned activities planned in Europe (http://www.proteinatlas.org/, and http://www.proteomebinders.org/) and Asia and on the promotion of resource dissemination and community-wide access and impact. The NIH aims to create a community resource leveraged through consensus and coordination with other efforts minimizing costs and maximizing timely impact. Division by agreement among the cooperating partners of the target human proteome by protein class, systematic collections, biological sub-proteome [sic] or end use of the reagent would be an advantage. This will also provide the opportunity to prioritize targets during the effort to focus the effort on the most valuable targets at the various stages of the effort. This coordination may provide for overlap where this might be an advantage, for example where some reagent technologies are more suitable for different classes of protein target, but avoiding wasteful redundancy. However, it should be noted that the need for multiple reagents for the same target has been recognized and that at this stage some overlap was identified as beneficial. The establishment of a common approach to intellectual property addressing the interests of stakeholders and consistent with laws and regulations in participating countries will be essential. The governing principle should remain minimal intellectual property constraint, so that these reagents can be as widely used as possible, and that creative scientists are enabled to perform experiments and to devise creative downstream applications for the reagents. Similarly, a mixture of public resources and private sector entities would be an advantage for dissemination of data, methods and materials. Examples of such public-private partnerships in several public sector research efforts suggest the practicality of this approach (http://mgc.nci.nih.gov/). All the products from the sponsored effort would be available to users at minimal cost and with a minimum of complication and formality through a comprehensive public resource. Nonexclusive licensing of technology developed by the research network to private sector entities would provide an alternative source of reagents and also derived technology. Participants advised NIH staff to take strong measures to ensure this, including obtaining a Determination of Exceptional Circumstance. The effort and product of the NIH supported effort should be identified to the research community as a uniquely useful and reliable brand. A related name, term and logo should be adopted to facilitate marketing and advertisement. One or several web-based portals should provide a ready source for information about the reagents and resources available for distribution (see for example http://www.antibodypedia.org, http://dshb.biology.uiowa.edu/ and http://www.sbkb.org/). The methods and technology developed should be compiled and offered for dissemination for academic research purposes and with appropriate licensing for commercial use. This information portal should provide results from validation testing performed prior to the release of the reagent and also provide a forum for ongoing evaluation and reporting of the performance of the materials by the users. The information portal(s) should provide the community with an opportunity to provide input on the operation and management of the activity and to make requests for the preparation of certain reagents which are of particular value and interest in their individual research endeavors. Acknowledgement of the use of the reagents should be required in publications by customers. The web resource should provide pointers to publications produced by the research network and also to those by resource users where the reagents were used in the work.
The importance of establishing a rational priority among protein target pursued within the production effort (i.e. human transcription factors) and targets within other classes of proteins which might be selected by pilot studies in earlier developmental stages. Care should be taken to distinguish targets which are ready for production and reagent preparation from those which require extra effort. The early establishment of pipeline operations requires attention to availability of appropriate target materials for operation.