UCSF

Structural information on nonnative states of proteins, including folding intermediates and folding and unfolding transition states is crucial for understanding folding and unfolding mechanisms. Kinetically stable proteins such as α-Lytic protease (αLP) combine a very high barrier to unfolding with extraordinary unfolding cooperativity to uncouple their native state from the unfolded states. This unusual energetic landscape results in a remarkable resistance to proteolytic destruction and thus is crucial for αLP's biological function. The uncoupling of native and unfolded states makes the differences between the native state and the transition state most relevant for protein function and thus, the height of the unfolding barrier rather than the thermodynamic stability is the relevant metric to investigate.

In order to characterize the structural determinants of protein unfolding in αLP, I used salt bridges to probe for structural rearrangements and cooperative contributions in the unfolding transition. From these studies, I have identified protein regions that are frequently unfolded in the transition state, as well as a mechanism to couple cooperative contributions from distant regions in the protein. This analysis led to an energetic dissection of cooperativity allowing for a quantitative assessment of this property for the first time. Also, two other findings, support the previous findings of an extremely rigid αLP native state: The extremely low native state pK_a values for the salt bridge carboxylates point to a significantly dampened native state dynamics strengthening the salt bridges. In addition, experiments investigating the role of the disulfide bridges in αLP unfolding found an extreme insensitivity of the unfolding barrier to reducing conditions, suggesting that these disulfide bridges are also protected by rigid native state dynamics and that protein unfolding is all or nothing.

αLP folding transition involves and even higher barrier than that for the unfolding transition. Uncatalyzed, a molten-globule like intermediate converts extremely slowly to mature protease. Understanding this transition with structural detail has been the motivation to develop a structure determination method for nonnative states of proteins. Nonnative proteins provide extreme challenges for structure determination; a multiplicity of structural states and aggregation at high enough concentrations to name a couple. The structure determination method I developed involves the use of cross-linkers to identify distance constraints in the protein structure via mass spectrometry. While the structure determination method was not sensitive enough to provide the level of structural detail I aimed to obtain, findings facilitated by precursor ion scanning during the development of this method are crucial to improve the detection limit for future studies.

Last, I studied the substrate length dependent k_cat effect for αLP substrate catalysis. Crystal structures of αLP with varying lengths of boronic acid inhibitors that are thought to mimic one of the tetrahedral intermediates in catalysis were compared in terms of active site protein dynamics. Through the analysis of anisotropic B-factors, a correlation was found with increased thermal motion in catalytic atoms and decreased thermal motion in the substrate binding pocket, and increasing inhibitor length. The latter with the higher overall disorder may provide an explanation for the k_cat effect with an increased order in catalytic transition states in the presence of longer substrates.