TL;DR
This paper identifies and explains "bump" artifacts in solid-state NMR relaxation dispersion experiments, which previously confounded the study of protein dynamics. By demonstrating that these bumps arise from a specific proton-mediated recoupling phenomenon (MIRROR), the authors show that applying low-power proton decoupling during the experiment effectively suppresses these artifacts, enabling accurate measurements of microsecond protein motions even in fully protonated samples.
What Problem Does This Paper Address?
Understanding the dynamic motions of proteins on the microsecond to millisecond (μs-ms) timescale is crucial for deciphering their biological functions, such as enzymatic activity and ligand binding. Bloch-McConnell Relaxation Dispersion (BMRD) NMR spectroscopy is a powerful tool for probing these dynamics. While well-established in solution, applying BMRD to solid protein samples presents significant challenges due to strong dipolar couplings between nuclear spins, particularly between ¹⁵N and ¹H. These couplings complicate the spin dynamics and can mask or mimic the very motions researchers aim to study. Historically, high levels of deuteration and fast magic-angle spinning (MAS) were thought to sufficiently suppress these unwanted effects, allowing for quantitative analysis. However, researchers have observed unexpected "bump"-like features in ¹⁵N R₁ρ BMRD profiles, even under these optimized conditions. These artifacts distort the relaxation dispersion curves, making accurate quantification of protein dynamics impossible and leading to misinterpretations. This paper aims to identify the origin of these "bump" artifacts and provide a method to eliminate them.
Methods
The researchers employed a combination of experimental solid-state NMR techniques and computational simulations.
- Experimental NMR: They performed ¹⁵N R₁ρ BMRD experiments on protein samples (ubiquitin and TET2) under various conditions: different magnetic field strengths (600 and 700 MHz), MAS frequencies (55.56 and 100.0 kHz), and protonation levels (perdeuterated and fully protonated). A key methodological innovation was the adoption of an "one-point" measurement scheme, adapted from solution NMR, which significantly accelerates data acquisition by measuring intensity at a single relaxation delay for each spin-lock frequency. This allowed for a much larger number of spin-lock frequencies to be sampled, providing higher resolution dispersion profiles.
- Computational Simulations: Spin dynamics simulations were used to model the behavior of the spin system under different conditions, particularly to investigate the proposed MIRROR recoupling mechanism.
- Artifact Suppression: The core of their solution involved applying low-power continuous-wave (CW) ¹H decoupling simultaneously with the ¹⁵N spin-lock pulse. They systematically tested different decoupling strengths (8 kHz and 16 kHz CW) and compared the results with and without decoupling.
Datasets
The study utilized two protein systems:
- Ubiquitin: Both perdeuterated and fully protonated ubiquitin samples were used. These were prepared with specific isotopic labeling (¹³C/²H/¹⁵N) and crystallized with MPD.
- TET2: Perdeuterated TET2 samples were also used, prepared and packed as described in previous literature.
The paper does not introduce new benchmark datasets but rather uses these well-characterized protein systems to investigate and resolve an experimental artifact.
Key Insights and Ideas
- The "Bump" Artifact is Widespread: The paper demonstrates that these artifactual bumps are not isolated incidents but are commonly observed across different protein samples, research groups, and experimental setups, suggesting a fundamental underlying cause.
- MIRROR Recoupling is the Culprit: The authors definitively identify the origin of the bumps as a second-order three-spin Mixed Rotary and Rotational Resonance (MIRROR) recoupling condition. This occurs when the ¹⁵N spin-lock frequency matches the chemical shift difference between two coupled ¹H spins, leading to unwanted ¹H-¹H dipolar coupling recoupling and affecting the ¹⁵N relaxation rate.
- Protons are Essential: The artifacts are significantly more pronounced in protonated samples, strongly implicating ¹H spins in their generation.
- Decoupling is the Key to Suppression: Applying low-power CW ¹H decoupling during the ¹⁵N spin-lock pulse effectively eliminates the MIRROR recoupling and the associated bumps. This is a crucial breakthrough for solid-state BMRD.
- Quantitative Measurements are Now Possible in Protonated Samples: The ability to suppress artifacts with decoupling opens the door to accurate μs-ms dynamics measurements in fully protonated proteins at high MAS frequencies (≥100 kHz), a feat previously considered extremely challenging or impossible.
- Artifacts Offer Structural Clues: Interestingly, the "bump residual" (the difference between decoupled and non-decoupled profiles) can provide site-specific information about the local proton environment around a ¹⁵N spin, offering a potential new avenue for structural insights.
Key Results
- Artifact Identification: The paper systematically links the observed "bump" features to the MIRROR recoupling condition through experimental variation of magnetic field strength, MAS frequency, and protonation levels, supported by simulations.
- Effective Artifact Suppression: The application of 16 kHz CW ¹H decoupling during the ¹⁵N spin-lock pulse was shown to completely remove the artifactual bumps in both perdeuterated and fully protonated ubiquitin samples at 100 kHz MAS.
- Quantitative Dynamics in Protonated Proteins: The study demonstrates that accurate BMRD measurements, yielding comparable results to perdeuterated samples, are achievable in fully protonated proteins at 100 kHz MAS when ¹H decoupling is applied.
- Impact on Kinetic Parameters: Analysis shows that the presence of artifactual bumps can lead to significant misinterpretations of exchange rate constants (kex) and exchange amplitudes (Φex), highlighting the critical need for artifact removal for reliable quantification. For example, sites previously thought to show significant dispersion might now be identified as artifactual.
Limitations and Future Work
- Experimental Time: While the "one-point" measurement scheme significantly accelerates data acquisition compared to traditional multi-point methods, BMRD experiments, even with this improvement, can still be time-consuming, especially for collecting a large number of spin-lock frequencies.
- Model Dependence: The interpretation of BMRD data, even after artifact removal, relies on fitting to specific dynamic models (e.g., two-site exchange). The paper acknowledges that the use of a two-site model for evaluation does not imply that the exchange is necessarily a two-site process.
- Generalizability to Other Nuclei: The paper primarily focuses on ¹⁵N BMRD. While the authors speculate that similar artifacts might arise with other heteronuclei (¹³C, ²H), they suggest these would likely be less prevalent or occur at different frequencies due to differences in gyromagnetic ratios and chemical shift dispersion. Further investigation into these cases would be valuable.
- Future Work: The authors suggest that the "bump residual" itself, representing the artifactual contribution, could be used to gain site-specific structural and spectral insights into the local proton environment around a spin. This opens up a new avenue for utilizing these previously problematic signals. They also note that similar three-spin recoupling conditions might affect other types of relaxation dispersion experiments, such as NEar Rotary Resonance Dispersion (NERRD), suggesting further investigation is warranted.
Summary updated