Recent advancemens in protein separation, mass spectrometry and bioinformatics have enabled the simultaneous analysis of a large number of proteins in cells and tissues. The interaction of these proteins, especially those that respond to specific challenges, are rapidly identified by ingenious screening techniques such as the phage expression or yeast two hybrid systems. These studies, collectively termed proteomics, have greatly enriched our knowledge of cellular physiology and pathology in normal and diseased tissues. A systematic identification of molecular interfaces between interacting proteins in natural cellular environment represents the next major challenge toward a molecular understanding of the complex and dynamic cellular events. Such a profile of molecular interfaces would also be invaluable as a tool in drug development by targeting crucial molecular interfaces. We are initiating such a proteomics project and focusing on the application of chemical cross-linking approach toward a systematic and global identification of protein-protein interfaces in complex systems such as the contractile machinery of muscle cells, the cytoskeletal filaments and organelles of nonmuscle cells. First, high-resolution gel electrophoresis and fractionation techniques are used to identify and index expressed proteins in skeletal and heart muscle tissues. Second, the neighboring relationship is defined by chemical crosslinking with a wide range of reagents. Third, the molecular interfaces of neighboring proteins are identified by isolating the crosslinked complexes, cleaving the proteins and isolating crosslinked peptides for protein sequencing with conventional chemical methods or by mass spectrometry. Fourth, the peptides at the interfaces will be synthesized and their interaction characterized further by capillary electrophoresis and biosensor techniques. A protein sequence database of molecular interfaces of proteins with known and unknown identities will be established to complement and enhance the utility of genome and proteome databases for basic and clinical research. We are evaluating the feasibility and integration potential of each steps of this multifaceted project. In particular, we have investigated the choice of crosslinking agents and the identification of extent, the stoichiometry of crosslinked complexes to define nearest neighbors. As a first step, contractile machinery, myofibrils from vertebrate skeletal muscle (rabbit psoas) and a novel tubular muscle from Midshipman fish sonic muscle were crosslinked by a zerolength crosslinker. The crosslinked complexes were identified by the gel mobility and Western blot of the complexes. Further efforts are made to purify selected complexes, followed by digestion, purification and sequencing and mass spectrometric analysis. Our work so far has clearly indicated the needs to employee or develop crosslinking agents with spectroscopic and mass tags to facilitate the identification and separation of crosslinked peptides. As a second method, protein interface is identified by screening synthetic peptides. The feasibility is demonstrated by the identification of nebulin/calmodulin interface by capillary electrophoresis of clmodulin with a set of nebulin peptides. The data are being compared with direct crosslinking and mass spectrometry. We conclude that, for our purposes, electrospray mass spectrometry is inferior to laser assisted matrix desorption mass spectrometry. The latter provides simple and interpretable spectra that are sufficiently accurate to calculate the composition of the complexes. - proteomics, protein interface, mass spectrometry, capillary electrophoresis, protein sequencing, crosslinking, genomics

Agency
National Institute of Health (NIH)
Institute
National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS)
Type
Intramural Research (Z01)
Project #
1Z01AR041120-02
Application #
6289046
Study Section
Special Emphasis Panel (LPB)
Project Start
Project End
Budget Start
Budget End
Support Year
2
Fiscal Year
1999
Total Cost
Indirect Cost
Name
National Institute of Arthritis and Musculoskeletal and Skin Diseases
Department
Type
DUNS #
City
State
Country
United States
Zip Code