Prof. C Forbes Dewey, Jr

Professor Dewey's group uses microscopy to probe the effects of mechanical forces on living cells. The targets of interest, endothelial cells, protect the artery wall from the inflammatory reactions that result in atherosclerosis. The cells have a mechanical structure that depends on a network of polymerized actin molecules. Fluorescent dyes are used as markers to tag individual molecules of actin, and we look at how they move and diffuse through the cell. This can tell us how much of the actin is polymerized and how much is in monomeric form. Application of a fluid shear force comparable to those that blood flow applies to these cells causes a major transient decrease in polymerization, and we believe that this is correlated with the observed realignment of the cells in the flow direction. Additional measurements of cell motility and intercellular gap junction proteins is extending our understanding of the interaction process. We are also undertaking major computational modeling of actin polymerization in collaboration with colleagues at the University of Rochester. The ability to peer inside individual cells with modern fluorescence techniques has powered a profound change in our quantitative understanding of cellular biological function. Imaging research also includes the reconstruction of 3-D images from stereo pairs taken using scanning electron microscopy (SEM). These images allow us to see the interaction of individual structural features within the cells at the molecular level. We currently use antibodies tagged with gold particles to visualize the specific conformation and interaction between different proteins, for example between polymerized actin filaments and the actin binding protein filamin-A.

A second major program within the lab is the development of robust information systems for biological data. The goal is to develop a common technology that will capture all major experimental biological data types in a single ontological framework. So far, we have shown this with gel electrophoresis, microarrays, and fluorescence-activated cell sorting, and are currently implementing mass spectrometry and optical microscopy. We've shown that these data can all be stored and searched from within a single database or from collaborating (federated) databases. The methods of federation are novel and have significant potential for application. A second area of research is the development of new algorithms for extracting structured information from unstructured documents. Additional information on these results will be available in the Fall of 2003.