Investigating the Structure of Macromolecular Protein Assemblies Using Ultrafast Spectroscopy Techniques

Infrared (IR) spectroscopy has been an invaluable tool for structural and biophysical analysis of proteins due to its ability to capture structural changes and quantify the amount of secondary structural components within a protein in in vitro conditions. Advances in nonlinear IR spectroscopic has further solidified its role in protein biophysical analysis. Topics in this dissertation include discussion and development of various IR spectroscopic techniques useful for studying protein structure within large, supramolecular protein structures, and application of these techniques to systems such as phase separating proteins and a novel fibril structure.

Methods of collecting spectral information through microscopy such as hyperspectral microscopy are hindered at IR wavelengths because protein macromolecular structures are typically at or smaller than the mid-IR light used to characterize protein structure. Here we examine the use of dielectric microspheres to provide resolution enhancement beyond the resolution limit using IR light. Furthermore, we demonstrate the ability to collect hyperspectral images of samples through the microspheres.

Next, we use IR spectroscopy to investigate droplet phase separation, a phenomenon where a solution spontaneously separates into two distinct liquids, which is involved in some biological processes and has been implicated in several neurodegenerative diseases, namely as a method for nucleation in amyloid formation. We use both Fourier Transform IR (FTIR) spectroscopy and Two-Dimensional IR (2DIR) spectroscopy to show that protein folding related secondary structural is associated with liquid phase separation.

In the final sections, we use 2DIR to study a new cross α-fibril structure formed by amyloid forming proteins involved in Staphylococcus aureus toxicity and pathogenicity. 2DIR spectroscopy was used to confirm the existence of the new fibril structure in vitro and showed evidence of polymorphisms within the fibers. Furthermore, we employ spectroscopic simulations to report on new peak patterns in the 2DIR spectrum which distinguish the fiber structures from the monomeric form. Finally, we evaluate our model and show that coupling between α-helices in the extended fibril structure leads to new excitonic states. These states contribute unique signals to the 2IDR spectrum which may allow future detection and identification of these fibers in new samples.