Frequency-swept Excitation in Distance Measurements by EPR: Spin 1/2 systems

Abstract

In the toolbox of structural biology, pulsed Electron Paramagnetic Resonance (EPR) experiments have proven their worth in determining distance distributions in the nanometer range. For systems without longrange order, such information can be difficult to obtain by other methods. The distance range accessible by EPR depends on the surroundings of the electron spins. In biologically relevant environments, the distance range is typically limited to 1.5-5 nm, and often particularly short in lipids. Sensitivity is often the limiting factor in accessing distances of sufficient length. Recent advances in technology provide access to amplitude and frequency modulation for EPR. Application of such pulses has for example already led to large sensitivity gains in distance measurements of GdIII–GdIII spin pairs. In such systems with spin > 1/2, the larger transition moment simplifies spin state transfers. However data analysis can be complicated by high-spin effects. This study instead focuses on spin 1/2 systems. The overarching objective of this work was to explore the use of frequency-swept pulses in distance measurements. The developed methodologies were subsequently used in application projects and put in the context of other distance determination methods. In a first scenario, the detection pulses of a ’pump–probe’ experiment were left as monochromatic pulses while the pump band was frequency-swept. These pulses were employed in a variant of the Double Electron Electron Resonance (DEER) experiment with increased sensitivity. Dynamical decoupling by optimal timing of refocusing reduces losses due to interactions with the environment in the sequence called 5-pulse DEER. Remarkably, we found that the increase in sensitivity holds true for various spin environments, including biologically relevant ones. The frequency-swept pump band enabled suppression of a known artefact contribution. Additional artefacts were found and identified as being caused by overlap of the excitation bands. Tuning inversion efficiency and spectral separation of the frequency-swept pump pulse allowed for finding an optimal pulse setup with respect to sensitivity and purity. A data processing algorithm for removing a remaining artefact contribution based on shifting the artefact with respect to the main signal was rigorously tested; before the method, using the optimal pulse setup, was applied to biologically relevant systems. In a second scenario, frequency-swept excitation of both the pump and the probe band was explored. These experiments were performed on a spin 1/2 system with a broad spectrum: Cu(II). The gain by the use of frequency-swept pulses in the individual bands was assessed independently as well as in combination. Interestingly, we found the sensitivity of dipolar traces with frequency-swept observer pulses to be lower than would be predicted by the enhanced echo intensity. Optimized frequency-swept pump pulses yielded similar sensitivities as a different experiment in which the pumped spin is flipped by relaxation. This finding is rather remarkable, seeing that frequency-swept excitation of Cu(II) can compete with the ’infinite bandwidth’ of relaxation. In the third scenario, frequency-swept excitation was employed for one, unified band exciting both pumped and probed spins in an experiment abbreviated as SIFTER. This approach enabled suppression of orientation selection in systems with correlated geometries, an effect which can confound distance determination. We assured reliable determination of distances by performing SIFTER on a series of ’molecular rulers’. Furthermore, the decay characteristics of SIFTER were found to be prolonged to a similar extent as those of 5-pulse DEER. The prolonged dipolar evolution time allowed to estimate a distance of ca. 7 nm, which is, to the best of our knowledge, the longest distance so far measured by EPR between spins in a protonated lipid environment. In the last part of the thesis, applications of the developed methodologies are presented. First, the ability of SIFTER to suppress orientation selection is exploited to verify distance information in stiff compounds intended to study Förster Resonance Energy Transfer (FRET): The EPR data provided the basis for modeling molecular rulers with nitroxide or fluorescence labels. These predictions allowed comparison of the accuracy of distance determination by the two methods as well as to provide experimental proof for the seminal theory by Förster with respect to orientation-dependence. Finally, 5-pulse DEER is applied for structure determination of the protein Pin1. The prolonged distance range allowed for characterization of interdomain distances, which provide valuable complementary information to short-range NMR restraints for this system.