September 9, Thu 2010
2:00 pm, MRB 200 Conference Room
Dr. A. Keith Dunker
Department of Biochemistry and Molecular Biology, Indiana University School of Medicine
Bioinformatics and Intrinsically Disordered Proteins (IDPs)
The standard paradigm for proteins is amino acid sequence â†’ 3D structure â†’ biological function. Fischerâ€™s 1894 lock-and-key hypothesis for an enzyme-substrate interaction and Erlichâ€™s 1897 use of the lock-and-key model to explain antibody-antigen interactions were key steps in the development of this paradigm. By the late 1930s, however, it was clear that not all antibodies obeyed the Fischer-Erlich lock-and-key molecular recognition hypothesis, but rather many antibodies were each able to bind to several, very differently shaped antigens. In 1940 Linus Pauling proposed that such multi-specific antibodies were initially unfolded but used the different antigens as templates so the same sequence would fold into different structures upon interaction with the different antigens. According to Paulingâ€™s idea, protein functions such as multi-specificity required lack of prior folding. By the mid-1990s more than 100 proteins of many types were shown to use lack of fixed structure, or â€œintrinsic disorderâ€ﾝ for biological function, yet none of these â€œintrinsically disordered proteinsâ€ﾝ (IDPs) has been discussed in any of the major biochemistry textbooks, even though many of their biological functions are extremely important. Starting in the mid-1990s, we began to apply the concepts and methods of bioinformatics to these IDPs. We formulated the hypothesis that the amino acid sequences of IDPs code for their absence of fixed 3D structure, thus we reasoned that it should be possible to predict presence or absence of 3D structure from amino acid sequence. The focus of this seminar will be on four items: 1. the development of predictors of structure and disorder from amino acid sequence; 2. the use of such predictors to develop the new paradigm of amino acid sequence â†’ intrinsically disordered protein ensemble â†’ biological function; 3. the use of such predictors to estimate the commonness of IDPs across the various types of living cells; and 4. the use of such predictors to understand which protein functions are associated with prior structure and which functions are associated with prior lack of structure. The study of IDPs is now one of the fastest growing areas of experimental protein science; we argue that this important new field owes its existence to the findings from bioinformatics that these proteins share common amino acid sequence features.