Abstract: Among the diverse NMR techniques, the investigation of spin relaxation stands out as a well-developed method for probing molecular motions. Relaxation in NMR is a result of stochastic fluctuations in local magnetic fields induced by thermal motions of molecules, with relaxation rates dependent on the external magnetic field strength, commonly known as nuclear magnetic relaxation dispersion (NMRD). The measurement and analysis of NMRD curves, forming the field of NMR relaxometry, provide a means to extract both dynamical and structural information for various molecular systems, with a list of applications in fundamental and practical studies across disciplines such as biology, food science, and soil science [1,2]. Notably, recent work by Wang and co-workers demonstrated the effectiveness of combining NMR relaxometry with high-resolution detection in studying metabolite-protein interactions in biological fluids [3]. One challenge in NMR relaxometry is the limited timescale of nuclear spin relaxation, typically spanning several seconds, restricting studies to faster processes. Overcoming this limitation involves exploiting the symmetric properties of a specific spin state, such as the singlet state of a pair of protons, which corresponds to zero total spin. Storing initial magnetization as a population imbalance between singlet and triplet states, known as singlet order (SO), has shown promise in overcoming fast relaxation limits, as demonstrated in previous research on ligand-protein interactions [4]. In our investigation, we analyze the field dependence of singlet order relaxation time in a small molecule (alanine-glycine, Ala-Gly) engaged in weak binding with a protein (human serum albumin, HSA). We establish a connection between the observed NMRD data and the parameters of the ligand-protein interaction by fitting it with a theoretical model grounded on established analytical expressions for the field dependence of 1 and . Furthermore, our findings reveal that the contrast, indicating the relative change in relaxation time in the presence of the protein, offered by surpasses that provided by 1 across all field ranges. Methods The application of NMR relaxometry to investigate ligand-protein interactions relies on the sensitivity of spin relaxation to molecular motion timescales. Typically, a protein's slower overall rotation compared to a ligand's results in faster relaxation of nuclei in the ligand-protein complex at low magnetic fields. When binding is reversible, the measured relaxation time is an average over both bound and free forms of the ligand, with the bound form often undetectable (Fig. 1). For most types of motions associated with biomolecules, the dispersion of relaxation times occurs in the range of magnetic fields around 0.1 T. A common technique to measure relaxation at such fields exploits the stray magnetic field of a superconducting NMR magnet. In our study we use a custom-build experimental setup based on 700 MHz Bruker Avance HD NMR spectrometer, with a stepper motor placed at the top of the spectrometer bore. The motor drives a plastic rail with a carriage for a standard 5 mm NMR sample tube. This allows shuttling the sample in the range of fields from 4 mT to 16.4 T with transfer time shorter than 0.4 s. Figure 1. Schematic process of the weak ligand-protein interaction with the indication of the 1 and relaxation mechanisms, accounted in the fitting model (top). Typical effect of the addition of protein on the 1 and magnetic field dependences (bottom). Results The experiments yielded a collection of relaxation kinetics measured across a range of magnetic fields, specifically probing ten field values logarithmically distributed from 4 mT to 16 T. To extract the relaxation times from the kinetics, a least square fitting procedure with a single exponential function was conducted. We obtained NMRD curves for a sample containing only the ligand molecule and for samples containing 50, 100, 200 and 300 M of added HSA. The fitting results exhibit strong conformity with the experimental data for both Ala-Gly (Fig. 2.A). The 1 NMRD curves, observed in the presence of protein, manifest the expected behavior, reaching their maximum at high fields and experiencing a significant decline as the magnetic field decreases. Consistent with the model, an increase in protein concentration leads to a faster relaxation at low fields. The shape of the NMRD curves presents a more complex pattern (Fig. 2.B). Upon decreasing the field from 16 T to 1 T, the lifetime of the singlet order notably extends as it aligns with the eigenstate of the methylene protons, where the difference in their Larmor frequencies approaches their J-coupling. Subsequently, as the magnetic field further diminishes, the rapid relaxation of the bound form becomes evident, leading to a reduction in the observed singlet order lifetime. The reported procedure not only confirms the occurrence of ligand-protein interactions but also offers qualitative insights into this process. However, given the significant disparity between relaxation times of magnetization and long-lived spin order, visualizing the binding effect on NMRD curves is best achieved by considering the "relaxation contrast" = ( 1, ⁰ − 1, ᴴˢᴬ)/ 1, ⁰ , indicating the relative change in relaxation time upon protein addition. Analysis of relaxation contrast as a function of magnetic field reveals that NMRD curves obtained with the singlet order exhibit greater sensitivity to ligand-protein binding compared to 1 NMRD curves. Thus, employing the singlet order results in approximately a 10% enhancement in relaxation contrast for various concentrations of HSA (Fig. 2.C). This increased relaxation contrast enables the detection and quantification of the binding process with lower concentrations of the protein. This work is supported by the Russian Science Foundation (Grant #23-73-10103).

Cite: Kozinenko V.P. , Kiryutin A.S. , Yurkovskaya A.V.
Probing Weak Ligand-Protein Binding Using Long-Lived Spin Order Relaxometry
In compilation 21th International School-Conference MAGNETIC RESONANCE AND ITS APPLICATIONS Proceedings. – Издательство ВВМ., 2024. – C.109-11.

Dates:

Published online:

Mar 1, 2024

Identifiers: No identifiers

Citing: Пока нет цитирований