Enzyme specificity is thought to arise from differential stabilization of the transition state structure of a reaction for different substrates. Computational models of transition state stabilization are based on the precise three-dimensional structure of the enzyme active site. Protein design methods for predicting the effects of enzyme sequence changes on active site structure, and thus on activity and specificity, have been developed, but typically assume that the structure of the polypeptide backbone of the enzyme will remain fixed. In this thesis, I demonstrate that an algorithm, which explicitly accounts for backbone flexibility, can be used to identify mutations predicted to alter enzyme specificity. Furthermore, these predictions can be validated experimentally. In particular, mutants of the enzyme human guanine deaminase with increased activity towards ammelide were identified using such an algorithm. Biochemical and structural characterization revealed the high accuracy of the predictions made with this computational method.An Agilent 1200 Series HPLC (Agilent, Santa Clara, CA) was used with a Zorbax NH2 Analytical Column, 4.6 x 250 mm, 5 um (Agilent). Isocratic elution with 22% 5 mM NaP04 pH 6.0, 78% acetonitrile was performed at 1 mL/min, andanbsp;...
|Title||:||Alteration of Enzyme Specificity by Computational Loop Remodeling and Design|
|Publisher||:||ProQuest - 2008|