This program project addresses fundamental questions concerning the relationship between genome sequence, protein structure, and function. The present period of the project has focused on structural studies of 'hypothetical' proteins from bacteria. We will now address one of the most surprising outcomes from genome scale studies of higher Eukaryotes: in the last two years it has become clear that in excess of 30% of human genes are expressed in at least one alternatively spliced form. Our thesis is that alternative splicing represents a fundamentally different mechanism for the generation of functional diversity in higher Eukaryotes, but that its utility is moderated by the constraints of maintaining structural integrity. On the one hand, reuse of exons has advantages similar to the reuse of software modules - common sub-function is preserved. On the other, protein substructures are not inherently modular - random combinations of exons would only very rarely lead to viable, folded structures.
We aim to thoroughly explore the interplay between these two principles - what kinds of structural change can be accommodated, and what kinds of functional diversity result? In addition to providing an understanding of the relationship between alternative splicing, structure and function, the study will illuminate aspects of protein evolution - under what circumstances can substructures be recombined to generate modified folds? What kinds of conformational change accompany such recombination? Do frame shifts result in folded, completely different structures? Finally, many diseases are associated with alternative splicing. Are the corresponding structures usually non-viable structurally, or do they form alternative structures with pathological functions? To answer these questions, we will determine and analyze a series of carefully chosen structures (50 to 100 over 5 years). Proteins for study will be chosen on the basis of three considerations: likely impact of the splice forms on protein structure, functional role of the splicing, and relevance to disease.
Showing the most recent 10 out of 52 publications
