D3SC: Nonlinear Optical Analysis of Proteins in Glasses

点击翻译按钮获取中文摘要

With support from the Chemical Measurement and Imaging Program in the Division of Chemistry, Professor Garth Simpson and his group at Purdue University are devising sophisticated new optical methods and machine learning-based data analysis tools capable of probing the changes of structure and flexibility that occur when proteins transition from solutions to glassy matrices. Some proteins retain function in amorphous glasses, even in the absence of liquid water. This effect enables some living creatures to survive even below the freezing point of water. Glassy matrices can also support long-term storage and delivery of therapeutic proteins - another context in which it is important to understand the properties of proteins in these matrices. The methods being developed by the Simpson group are designed to improve our fundamental understanding of the nature of protein/matrix interactions upon removal of water. Dr. Simpson is communicating the principles behind his research to broad audiences through short instructional videos and through industrial interactions in an NSF Industry/University Collaborative Research Center. The Simpson group is working on three innovative approaches to probing the protein and sugar structural and dynamic changes that accompany the lyophilization-induced transition of sugars from co-solutes in aqueous environments to protein-stabilizing matrices. First, they are developing new Fourier transform methods for substantially improving the information content in fluorescence recovery after photobleaching (FRAP) in order to enable rapid assessment of mobility/relaxation within low volumes (~100 nL) of both aqueous solutions and lyophilized materials. Second, they are devising means of atomic-level mapping of structural evolution within proteins upon water removal using a one-of-a-kind instrument designed and built in collaboration with Argonne National Laboratory. The approach combines nonlinear optical imaging, synchrotron X-ray diffraction, and unique data-handling algorithms based on non-negative matrix factorization. Third, a novel physics-encoded artificial neural network (PENN) is being designed, trained, and tested as a means of gaining improved understanding of the physical origins of observed structure and mobility changes. The long-term aim is to develop experimental/computation methods that can improve fundamental understanding and provide capabilities for predicting the dynamic transformations in proteins within glassy matrices.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.