What happens when 80 scientists and NIH officers put their heads together to figure out how to use proteomics technologies to understand the brain? As it turns out, they don’t agree on much, but they highlight the diversity of approaches it takes to study such a complex system.
At the workshop entitled “Proteomics in the Neurosciences” held by the National Institute of Neurological Disorders and Stroke in Washington D.C. last week, international proteomics researchers, bioinformaticians, and neuroscientists presented their research, discussed the needs for neuroproteomics resources and technologies, and made recommendations on how NINDS should invest its funds in this area.
The meeting was organized by Danilo Tagle, program director in neurogenetics at NINDS, along with fellow NINDS program director Randall Stewart; Seth Grant, a professor of molecular neuroscience at the University of Edinburgh; and Samir Hanash, president of HUPO and professor of pediatrics at the University of Michigan, and had a variety of participants, mostly from academia and NIH institutes.
Opinions differed widely on how proteomics should tackle the brain. While some attendees argued in favor of large-scale projects, such as assembling “parts lists” or protein profiles in defined tissues or cell types, or studying protein-protein interactions on a global scale, others prefer studying and validating distinct functional complexes, one at a time. For example, Grant presented his analysis of the NMDA receptor complex, which regulates many activities of a nerve cell — so far, he has identified 168 components.
Christian von Mering from EMBL in Heidelberg, on the other hand, showed a comparative analysis of large-scale protein-protein interaction datasets obtained by different methods. “Even small-scale experiments have an error rate, and only the comparison of large datasets helps find true complexes,” commented Edward Marcotte from the University of Texas in Austin, who gave a talk on integrating data from different sources.
However, participants seemed to agree largely on the need for certain resources. These include antibodies specific to neuroproteins that work under a variety of assay conditions, and well-defined clone sets for protein expression. Antibody-type chips, according to various participants, would be useful for expression profiling, validating biomarkers, and certain functional studies. The complexity of the nervous system warrants new methods to separate different cell types, and the importance of membrane proteins as neuronal receptors means that proteomics technologies need to be adapted to study these.
Finally, a number of participants expressed a need for establishing databases in order to make neuroproteomics data widely accessible. Data from the collaborative Human Brain Proteome Project in Germany, for example (see ProteoMonitor 02-11-02), is only available to the participating researchers at present. Unlike the Human Genome Project, which made sequence data available via central databases, “in proteomics, we have none of this,” Henning Hermjakob from the European Bioinformatics Institute told ProteoMonitor. “We either need to go to dozens of sites, or we don’t get the data at all.”
So what is the role of HUPO in all of this? According to Hanash, this and future workshops could lead to a “HUPO-promoted” neuroproteomics initiative next year. Similar to HUPO’s plasma and liver proteomics initiatives, he envisions a pilot and an execution phase for such a project, as well as a standard target. This target could be, for example, a specific brain region, that would be analyzed by different investigators with “a fleet of technologies.”
One size doesn’t fit all
But the idea of a standardized model faced some opposition from participants. “Not every ship in this fleet may head in the same direction,” said Richard Caprioli, a mass spectrometrist from Vanderbilt University who presented his MALDI tissue imaging technology at the meeting (see ProteoMonitor 07-01-02). Since each technology has its own strengths and weaknesses, they might not all be appropriate to analyze the same model, he said. And while the meeting was dominated by researchers focusing on mammalian models, Mike Snyder, a yeast geneticist from Yale University, and others pointed out the value of studying the nervous systems of other model organisms, for example flies and worms.
The final session, during which participants made recommendations for research priorities, collaborations, and funding strategies, was not open for unrestricted reporting. However, based on an informal survey of a number of participants by ProteoMonitor prior to the session, views differed widely on how NINDS should best fund neuroproteomics. While some participants favor large proteomics centers, focusing on distinct diseases or models, others hope for additional funding for individual investigators. Yet others would like to see international collaborations facilitated.
Based on recommendations from the workshop, which Tagle and Grant plan to publish as part of a meeting report in a scientific journal, NINDS might start new initiatives in neuroproteomics, and might request applications as early as next year, Tagle told ProteoMonitor. However, such initiatives are subject to a decision by an internal committee, he said, and neither the level of funding has been decided nor whether funding should go to large centers — similar to the way the National Heart Lung and Blood Institute’s proteomics initiative awarded contracts to ten proteomics centers across the US this fall — or to individual investigators.
But the workshop was also aimed at determining overlapping interests in neuroproteomics between those NIH institutes that focus on specific neurological disorders. “We would like to define areas of focus for neuroscience institutes vs. NIH as a whole,” said Robert Baughman, associate director for technology development at NINDS. Not surprisingly, about half of the workshop participants — not including the speakers — came from NIH institutes — besides NINDS, the National Institute of Mental Health, the National Institute on Drug Abuse, the National Institute on Aging, the National Eye Institute, the National Institute on Deafness and Other Communication Disorders, the National Institute on Alcohol Abuse and Alcoholism, the National Cancer Institute, and the National Center for Research Resources.
Is an NIH Neuroproteomics Network Next?
These institutes might form a trans-NIH neuroproteomics working group, said Tagle. He could imagine an international collaboration springing out of this effort, similar to the Human Genome Project. Such an effort might tie in with projects in other countries, he said, for example a European Human Brain Proteomics consortium that is currently seeking €40 million from the European Union’s 6th Framework Program and plans to focus on neurodegenerative disorders.