Optimal Control Methods for Biomolecular NMR Spectroscopy
by David Joseph
Date of Examination:2024-12-10
Date of issue:2025-04-04
Advisor:Prof. Dr. Christian Griesinger
Referee:Prof. Dr. Christian Griesinger
Referee:Prof. Dr. Marina Bennati
Persistent Address:
http://dx.doi.org/10.53846/goediss-11195
Files in this item
Name:diss_DaJo_for_publication_final.pdf
Size:13.7Mb
Format:PDF
The following license files are associated with this item:
Abstract
English
Bio-molecular NMR studies are usually limited by the sensitivity and resolution, especially for concentration limited samples such as IDPs, in-cell samples and metabolomics studies, etc. Hence it is essential to use a strong magnet such as the recently made available 1.2 GHz (28.2 T) ultrahigh-field magnet to improve the S/N as well as the resolution of the spectra acquired on these samples. To achieve uniform manipulation of the spins dispersed in such a large frequency range broadband pulses are required. This means the commonly used hard pulses would require a strong B1 field and a high power source to generate these pulses. Hence, tolerance to such high power is a stringent criterion that needs to be satisfied by probe designs and is a limiting factor on the dimension of the CryoProbes coils, which are popular in solution bio-NMR studies due to their increased S/N ratio. At present this limitation is visible at the ultrahigh field magnets and for the above reason, the commercial manufacturer of 1.2 GHz spectrometers (Bruker) provides only a 3 mm CryoProbe. This is a serious limitation for studying concentration limited samples where measuring on a 5 mm CryoProbe with a larger sample volume would have provided an additional improvement in the S/N ratio. In this work, we have designed a set of low power optimal control pulses for broadband universal rotation of 1H and 15N nuclei (24 times less power on 1H, 3 times lower power on 15N) using the optimal control module in open source software Spinach. A new approach to designing optimal control based low power band selective pulses is introduced (with power of either 4 / 7 times or 2 / 3.5 times lower power on 13C). These pulses were then utilized to construct low power biomolecular NMR sequences. In all the OC-pulse sequances constructed we observed a performance improvement and they were highly tolerant to B1 miscalibration and hence are user friendly and can be fully automated at the spectrometer. We further show that these low power optimal control bio-NMR experiments can enable large volume measurements (or 5 mm CryoProbes) at the 1.2 GHz magnets and would be useful at the new generation of ultrahigh-field magnets we expect to see in the future. We also have developed a library of optimal control bio-NMR experiments to be used at these ultrahigh-field magnets. This thesis further explores the ability of multiband optimal control pulses to compensate for the Bloch-Siegert shift. The Bloch-Siegert shift is the perturbation of the true resonance frequency of a nucleus due to an off-resonance B1 field. This is common when an off-resonance decoupling pulse is used for homonuclear decoupling, such as during the carbon indirect dimension of 3D experiments, or for direct decoupling in carbon detection experiments. We show that multiband optimal control pulses have an inherent ability to compensate for the phase error induced by the Bloch-Siegert shift. This allows us to replace the commonly used three-pulse and delay compensation scheme with a shorter one-pulse scheme. Compensation for B1 inhomogeneity for these pulses provides an additional 11% improvement in signal-to-noise ratio. Furthermore, we show that to compensate for the frequency shift due to Bloch-Siegert shift, it is necessary to have two counterrotating B1 fields from the decoupling bands. Such an optimal control pulse is used to develop a direct decoupling scheme called Optimal control BAnd-Selective Homonuclear Decoupling (O-BASHD). This allows us to achieve up to 83% improvement in the signal-to-noise ratio for carbon detection experiments, saving up to 6 times the experimental time compared to experiments using a virtual decoupling scheme. In the last part of the thesis, we develop method for obtaining reliable intermolecular contact information from biological complexes. The study of interactions between biological macromolecules is important for understanding the functional mechanism of complexes such as DNA-protein, RNA-protein and protein-protein. The set of filtered/edited NOESY experiments are crucial to obtain intermolecular NOEs and thus information about the atomic sites that interact to form the above complexes. The reliability of structural models of these complexes thus depends crucially on our ability to record filtered-NOESY spectra without spurious NOE peaks. This depends on the purging capability of the filtering scheme used in the pulse sequence. Adiabatic pulses with matched sweeping to suppress 1H-13C J-coupling have been proven useful in improving the filtering schemes. These pulses work by sweeping at a rate that is tuned to the linearity of the 1H-13C J-coupling and the carbon chemical shifts. Thus, different sweep rates are required to study RNA/DNA-protein and protein-protein complexes and they do not compensate for the J-coupling over the entire 200 ppm range of the 13C spins. In this work, we design a robust J-coupling compensated optimal control pulse element with a large compensation range of 10-300 Hz J-coupling and 200 ppm 13C chemical shift range (60 kHz at 1.2 GHz) that can be used for filtered NOESY experiments. More importantly, the filter efficiency does not rely on the linearity between the chemical shift and the magnitude of the J-coupling, which makes this approach superior to previous implementations with adiabatic pulses. These pulses are significantly shorter (270 μs duration) than the adiabatic pulses (1-2 ms duration) and thus suffers little from relaxation losses, which would benefit the study of larger complexes. We observe, improved filtering when the J-coupling compensated optimal control pulse element was used for acquiring filtered NOESY spectrum on a Calmodulin/Munc13-1 complex.
Keywords: NMR; optimal control; cryoprobe; 1.2 GHz spectrometers; Bloch-Siegert shift; filtered/edited NOESY; bio-molecular NMR
