National Institute of
General Medical Sciences

Four large-scale centers, all established during the PSI pilot phase, constructed structural genomics pipelines for the production and structural determination of proteins in a high-throughput operation. Over the 5-year period of PSI-2, these centers coordinated their efforts to generate more than 4,000 structures of protein families, and thus meet the program’s goals of increasing structural coverage of sequenced genes and providing benefits for the entire biomedical community. The centers developed new technologies to improve the pipeline and to increase success rates and lower costs. Each center also had an individual structural genomics research project, as described below.

 

During the PSI pilot phase, the JCSG developed a scalable, high-throughput structural genomics pipeline. The design of the JCSG pipeline was based on the development and integration of automated approaches for every experimental and computational step in the structural genomics process. During PSI-2, the JCSG contributed to the overall goal of maximizing structural coverage of protein families with no structural representation and continued to develop and disseminate innovative new technologies for structural biology. The consortium is determining structures of a large number of high-value targets, including structurally challenging proteins, such as membrane proteins and protein complexes. The JCSG consortium theme was the “central machinery of life” — proteins that are conserved in all kingdoms of life.

Principal investigator: Ian Wilson, Scripps Research Institute

Participating institutions: The Scripps Research Institute, the University of California at San Diego, the Burnham Institute, Stanford Synchrotron Radiation Laboratory and the Genomics Institute of the Novartis Research Foundation

Web site: http://www.jcsg.org

 

During the PSI pilot phase, this center automated synchrotron-based protein structure determination and solved many unique structures. In PSI-2, the multi-institutional consortium rapidly determined the structures of large numbers of strategically selected proteins using x-ray crystallography both to provide structural coverage of major protein superfamilies and to elucidate the entire protein folding space. Targets include proteins from pathogens and higher organisms. The efficient structure determination pipeline included the classification of genomic sequences, cloning and expression of genes and gene fragments, purification and crystallization of proteins, analysis of structures for folds and functions and homology modeling of related structures.

Principal investigator: Andrzej Joachimiak, Argonne National Laboratory

Participating institutions: Argonne National Laboratory, European Bioinformatics Institute, University College London and the University of Toronto

Web site: https://www.anl.gov/bio/midwest-center-for-structural-genomics​

 

During the pilot phase, this academic consortium determined more than 200 structures of both prokaryotic and eukaryotic proteins, including human proteins, using X-ray crystallography and solution-state NMR spectroscopy. It also developed several new technologies for eukaryotic protein production and protein structure analysis. In PSI-2, the consortium solved both prokaryotic and eukaryotic structural representatives from the major domain families constituting the eukaryotic proteome. Its biological-focus project targeted networks of co-functioning proteins (BioNets) involved in developmental biology and cancer.
Principal investigator: Gaetano Montelione, Rutgers University
Participating institutions: Columbia University, Hauptman Woodward Medical Research Institute, Ontario Cancer Institute at the University of Toronto, Pacific Northwest National Laboratories, Rutgers University, Robert Wood Johnson Medical School at the UMDNJ, Weill Medical School at Cornell University and Yale University

Web site: http://www.n​esg.org

 

This collaborative group of industry, academic and government labs participated in the PSI pilot phase. They developed a fully-integrated, high-throughput center for all the components of structural genomics, including an industrial protein production robotic pipeline. In parallel to automated protein production in bacterial cells, the center used production in insect cells for problematic eukaryotic proteins. During PSI-2, the consortium’s individual project focused on new targets, principally protein phosphatases and multidomain eukaryotic proteins.

Principal investigator: Stephen K. Burley, Eli Lilly and Company, San Diego, Calif.

Participating institutions: Eli Lilly and Company; Albert Einstein College of Medicine; Brookhaven National Laboratory; Case Western Reserve University; Columbia University; and University of California, San Francisco

Six specialized centers developed innovative methods, approaches and technologies for producing and determining the structures of proteins that traditionally have been difficult to study, including small protein complexes, membrane proteins and proteins from higher organisms. These efforts were to overcome bottlenecks in the structural genomics pipeline and approach high-throughput operation. The specialized centers also expected to determine a significant number of these challenging protein structures.

 

This center focused on the development, operation and deployment of novel approaches in miniaturization, integration and automation with an aim toward lowering the overall cost of gene to structure for all researchers. Primary initiatives included whole gene synthesis for the production of optimized protein constructs and parallel protein purification with integrated small-scale biophysical characterization, micro-capillary based crystallization technologies, and the use of a mini-synchrotron.

Principal investigator: Lance J. Stewart, Emerald Biostructures

Participating institutions: Emerald Biostructures; The Scripps Research Institute; University of Chicago; Lyncean Technologies, Inc.; and Micronics, Inc.

 

During the PSI pilot phase, this center focused on structures from eukaryotic genomes, initially from the plant Arabidopsis thaliana and subsequently from human, mouse, zebra fish and others. As a specialized center, the focus was on the determination of eukaryotic protein structures, especially human proteins related to disease or cell differentiation and proteins from families represented only in eukaryotes. The center uses two forms of protein production: expression of targets in E. coli and automated cell-free production using wheat germ extracts. The group uses X-ray crystallography and solution-state NMR spectroscopy to solve structures.   

Principal investigator: John L. Markley, University of Wisconsin, Madison
Participating institutions: University of Wisconsin, Madison; Medical College of Wisconsin

Web site: http://www.uwstructuralgenomics.org

 

The center refined several steps in the structure determination pipeline, from sample preparation to the collection of crystal diffraction data. The researchers’ efforts will center on protein expression, purification and the crystallization of small protein complexes and transmembrane proteins (all from yeast). Researchers made and purified antibodies that target transmembrane proteins and crystallize the protein-antibody complexes. They worked on cryopreservation of protein crystals and are improving automation and analysis of crystallization screening results.

Principal investigator: George T. De Titta, Hauptman-Woodward Medical Research Institute

Participating institutions: Hauptman-Woodward Medical Research Institute,

University of Rochester, Cornell University, Stanford Synchrotron Light

Source, University of Pittsburgh, University of Washington at Seattle and the University of Toronto

Web site: http://www.chtsb.org

 

This specialized center determined structures of membrane proteins derived from archaea, bacteria and humans. Genes were selected using bioinformatics to optimize coverage of classes of membrane proteins. Investigators used four expression systems, including MISTIC (Membrane-Integrating Sequence for Translation of Integral membrane protein Constructs) systems, which allow expression of eukaryotic proteins in E. coli and subsequent targeting of the proteins to the bacterial membrane. Structures were determined by detergent-based crystallization followed by X-ray structure determination at the Advanced light source (ALS), 2-D crystallization in lipid bilayers, single particle electron imaging and NMR. Protein function was assayed.

Principal investigator: Robert M. Stroud, University of California, San Francisco

Participating institutions: University of California, San Francisco;

The Salk Institute; Advanced Light Source (ALS) Lawrence Berkeley National Laboratory; University of California, Los Angeles; and University of California, Davis

Web site: http://csmp.ucsf.edu

 

The investigators addressed problems related to the key steps of the production of soluble proteins and crystallization. Emphasis was on protein solubility and crystallization of small protein complexes, soluble domains of membrane proteins and eukaryotic proteins. This center developed chaperone-assisted crystallography and engineered proteins for crystallization by mutating their surfaces and adding fusion proteins.

Principal investigator: Thomas C. Terwilliger, Los Alamos National Laboratory

Participating institutions: Los Alamos National Laboratory; University of Chicago; University of Virginia; University of California, Los Angeles; Lawrence Livermore National Laboratory; and Lawrence Berkeley National Laboratory

The center’s goal was to speed structure determination for membrane proteins, starting with bioinformatics analysis of sequence. The investigators also addressed bottlenecks in the expression, purification and structure determination of membrane proteins. The center used automated cloning and expression, high-throughput expression screening, detergents to help solubilize folded proteins and large-scale purification and labeling of proteins for NMR and X-ray crystallography. Proteins were expressed in insect and mammalian cells as well as E. coli.

Principal investigator: Wayne A. Hendrickson, New York Structural Biology Center

Participating institutions: Columbia University, Albert Einstein College of Medicine, New York University, The Rockefeller University, SUNY Buffalo, University of Medicine and Dentistry of New Jersey and Yale University

Two homology modeling centers developed innovative computational methods for reliably predicting the three-dimensional structures of proteins, bringing the PSI closer to its long-range goal of making it easy to determine the shapes of proteins from genetic sequence information. The centers used PSI structures solved experimentally to help develop computational methods and verify model accuracy.

 

Joint Center for Molecular Modeling
Collaborating scientists used mathematical tools to build software programs that couldaccurately predict the structures of proteins distantly related to ones with known structures. They also developed novel ways to extract rules and trends of how structures change and evolve, enabling protein structure prediction from sequence. Future applications could include predicting, for example, how a human protein differs from its counterpart in mice or bacteria. The modeling tools were made available to researchers worldwide through open-source databases.
Principal investigator: Adam Godzik, Burnham Institute for Medical Research
Participating institutions: University of California, San Diego
Web site: http://jcmm.burnham.org/

 

New Methods for High-Resolution Comparative Modeling
This center brought together a team of investigators in biophysics, mathematics, statistics and computer science to improve the modeling accuracy of proteins both closely and distantly related to proteins with known structures. The team developed modeling methods for a wide range of proteins, including cancer proteins and ones that could serve as drug targets for new cancer treatments. To explore how modeling can enhance experimental structure determination, the center modeled some structures for molecular replacement as they’re being solved by the PSI research centers. 
Principal investigator: Roland Dunbrack, Fox Chase Cancer Center
Participating institutions: University of Washington, Washington University in St. Louis, and the University of California, Berkeley
Web site: http://dunbrack.fccc.edu/nmhrcm/

Two centers were established to centralize resources developed by the PSI centers and increase the scientific community’s access to them.

 

PSI:Biology Materials Repository
Established in September 2006, the PSI:Biology Materials Repository (PSI:Biology MR) worked to collect, sequence, annotate and store the more than 90,000 plasmid clones already generated by the PSI research centers. By the end of this PSI phase, nearly 30,000 clones were available for searching and ordering at http://dnasu.asu.edu . The PSI:Biology MR moved to the Biodesign Institute at Arizona State University.
Principal investigator: Joshua LaBaer, Arizona State University
Web site: http://psimr.asu.edu/
Learn more about the PSI:Biology MR in Nu​cleic Acids Research​

PSI:Biology Structural Biology Knowledgebase
Launched in February 2008, the PSI:Biology Structural Biology Knowledgebase made all the products generated by the PSI centers available to the greater scientific community. This Web portal enables scientists to search for the best available information on protein structures, their biological function and their experimental determination. The PSI SBKB, which offers a free tutorial on using the knowledgebase, merges these features with external resources, making it possible for researchers to access a wealth of information from one site.
Web site: http://sbkb.org/