Methods development and applications for ultrafast protein crystallography
File(s)
Author(s)
Fadini, Alisia
Type
Thesis
Abstract
Time-resolved crystallography follows the structural rearrangements that underlie protein functionality in real space and time. X-ray free electron lasers (XFELs) enable the temporal resolution of these experiments to extend to the ultrafast regime. Through time resolved serial femtosecond crystallography (TR-SFX), atomic motions in proteins on timescales that are relevant for chemical bond isomerization and breaking, or for electron movement, can now be visualized.
My thesis employs these properties of time-resolved crystallography and explores methodological developments that can improve on its current capabilities, with the focus on understanding protein photochemistry: a complete study of photon-induced protein dynamics requires the ability to confidently identify the formation of transient species and photoproducts in protein electron density starting at ultrafast timescales.
The first results presented characterize mapping of time to energy in protein diffraction patterns simulated with an energy-chirped polychromatic XFEL pulse. TR-SFX is next employed to resolve a photoisomerization pathway, the hula twist, in a fluorescent protein; tracking the reaction from femtoseconds to microseconds yields the first experimental structural evidence of a hula twist mechanism in a protein on its femtosecond-to-picosecond timescale and insight on how chromophore isomerization and twisting lead to secondary protein structure rearrangements. New tools are developed to identify time-evolving species in electron densities. Gaussianity statistics, cross-validation, and denoising procedures find crystallographic map generation parameters that maximize signal and reveal new structural details in existing data. Finally, orientation-dependent merging of TR-SFX intensities is proposed: birefringent decomposition of an optical laser field used for photoexcitation leads to an effective power density titration across protein crystal orientations. This can be analyzed to understand the effect of high intensity photoexcitation in recorded signals. Populations modeling is implemented for a model system to assess the magnitude of the birefringent effect and determine the data collection conditions that would enable this type of analysis.
My thesis employs these properties of time-resolved crystallography and explores methodological developments that can improve on its current capabilities, with the focus on understanding protein photochemistry: a complete study of photon-induced protein dynamics requires the ability to confidently identify the formation of transient species and photoproducts in protein electron density starting at ultrafast timescales.
The first results presented characterize mapping of time to energy in protein diffraction patterns simulated with an energy-chirped polychromatic XFEL pulse. TR-SFX is next employed to resolve a photoisomerization pathway, the hula twist, in a fluorescent protein; tracking the reaction from femtoseconds to microseconds yields the first experimental structural evidence of a hula twist mechanism in a protein on its femtosecond-to-picosecond timescale and insight on how chromophore isomerization and twisting lead to secondary protein structure rearrangements. New tools are developed to identify time-evolving species in electron densities. Gaussianity statistics, cross-validation, and denoising procedures find crystallographic map generation parameters that maximize signal and reveal new structural details in existing data. Finally, orientation-dependent merging of TR-SFX intensities is proposed: birefringent decomposition of an optical laser field used for photoexcitation leads to an effective power density titration across protein crystal orientations. This can be analyzed to understand the effect of high intensity photoexcitation in recorded signals. Populations modeling is implemented for a model system to assess the magnitude of the birefringent effect and determine the data collection conditions that would enable this type of analysis.
Version
Open Access
Date Issued
2023-04
Date Awarded
2023-07
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
van Thor, Jasper
Sponsor
Imperial College London
Grant Number
No. BB/P00752X/1
Publisher Department
Life Sciences
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)