AT A GLANCE
Name: John C. Norvell
Position: Director of the NIGMS research training program and of the NIGMS Protein Structure Initiative
Prior Experience: Protein Crystallographer at Brookhaven National Laboratory and National Institutes of Health
Structural biology, structural genomics, structural proteomics — by whatever name researchers refer to solving protein structures, to John Norvell, head of the structural genomics program at the National Institute of General Medical Sciences, an institute of the National Institutes of Health, the field fits quite handily into proteomics. In fact, the evolving name of the field mirrors the technological progress that has ushered protein structure determination into the realm of big biology.
When Norvell first started solving protein structures 30 years ago,the process required months of work to clone, express, and purify the protein, not to mention the time required to perform x-ray crystallography or nuclear magnetic resonance experiments in an attempt to solve the structures. Furthermore, biologists had not yet fully realized the power of tunable x-rays from synchrotron sources — traditionally the realm of physicists — to probe protein structure.
“As I recall, in the beginning [synchrotron radiation] didn’t seem that useful for biologists,” Norvell said. “Most biologists would rather have stayed home; only when it was demonstrated that data was much better” did biologists start to take notice.
Norvell himself studied magnetism as a graduate student in physics at Yale University before being drawn into structural biology while a postdoc at the University of Wisconsin. “I’d always had this interest in biology, and I had toyed with the idea of changing in grad school, but was torn by my interests in biology and physics and there weren’t many joint programs then.” After completing his postdoc, Norvell moved to Brookhaven National Laboratory, and then to NIH, where he solved protein structures as part of an intramural research program.
Four or five years ago, when it became apparent that the reams of genomic data pouring out of sequencing centers, combined with advances in protein expression, purification, and crystallization methods might enable a large-scale effort to determine protein structures, Norvell turned to NIGMS Director Marvin Cassman for help in obtaining a slice of the expanding NIH budget.
But first, Norvell and other members of structural biology held a series of workshops and meetings to discuss whether such a project was technologically feasible, and if so, how to efficiently allocate resources for such a broad project. In spite of the skeptics, Norvell and his colleagues came up with a plan for covering all of the protein structure “space.” Rather than solve the structure of every protein encoded in the genome, the researchers made plans to choose one protein for each of the estimated 20,000 to 40,000 families of protein structures.
Two years later, Norvell found himself in charge of a $40 million-a-year program that funds nine structural genomics research centers, each led by principal investigators at academic institutions such as the Rockefeller University and the University of Washington. The centers have a five-year mandate to develop high-throughput methods for selecting targets, producing the proteins, and collecting x-ray and NMR data, Norvell said, and each center is aiming to solve proteins at a rate of 100 to 200 structures a year by the end of 2004. When consortium members solve a structure, they place the data in the Protein Data Bank, curated by the Research Collaboratory for Structural Bioinformatics.
But the structural genomics initiative is still officially a pilot project, and Norvell admits that the member centers are purposely steering clear of the types of proteins especially difficult to crystallize or solve, such as membrane proteins and protein complexes. “The idea is to try to set it up, and see what the bottlenecks are,” he said. At the end of the five years, Norvell and his advisory committee at NIGMS will have to decide whether to continue the initiative in its current form, or create some other organizational structure for coordinating structural genomics experiments.
However the public structural genomics initiative evolves, Norvell maintains that the project will continue to serve both academic and industry researchers. “At the end of [five years] we’ll have produced something like 10,000 unique non-redundant protein structures,” he said, “and hopefully it’ll be a quick way to get from protein sequence to structure to function.” — JSM