the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Insights into Protein Dynamics from 15N-1H HSQC
Abstract. Protein dynamic information is customarily extracted from 15N NMR spin-relaxation experiments. These experiments can only be applied to (small) proteins that can be dissolved to high concentrations. However, most proteins of interest to the biochemical and biomedical community are large and relatively insoluble. These proteins often have functional conformational changes, and it is particularly regretful that these processes cannot be supplemented by dynamical information from NMR. We ask here whether (some) dynamic information can be obtained from the 1H line widths in 15N-1H HSQC spectra. Such spectra are widely available, also for larger proteins. We developed computer programs to predict amide proton line widths from (crystal) structures. We aim to answer the following basic questions: is the 1H linewidth of a HSQC cross peak smaller than average because its 1H nucleus has few dipolar neighbors, or because the resonance is motionally narrowed? Is a broad line broad because of conformational exchange, or because the 1H nucleus resides in a dense proton environment? We calibrate our programs by comparing computational and experimental results for GB1 (58 residues). We deduce that GB1 has low average 1HN order parameters (0.8), in broad agreement with what was found by others from 15N relaxation experiments (Idiyatullin et al., 2003). We apply the program to the BPTI crystal structure and compare the results with a 15N-1H HSQC spectrum of BPTI (56 residues) and identify a cluster of conformationally broadened 1HN resonances that belong to an area, for which millisecond dynamics has been previously reported from 15N relaxation data (Szyperski et al., J. Biomol. NMR 3, 151-164, 1993). We feel that our computational approach is useful to glean insights into the dynamical properties of larger biomolecules for which high-quality 15N relaxation data cannot be recorded.
This preprint has been withdrawn.
-
Withdrawal notice
This preprint has been withdrawn.
-
Preprint
(3098 KB)
-
Supplement
(667 KB)
-
This preprint has been withdrawn.
- Preprint
(3098 KB) - Metadata XML
-
Supplement
(667 KB) - BibTeX
- EndNote
Interactive discussion
Status: closed
-
RC1: 'Comment on mr-2021-30', Geoffrey Bodenhausen, 15 Apr 2021
The comment was uploaded in the form of a supplement: https://mr.copernicus.org/preprints/mr-2021-30/mr-2021-30-RC1-supplement.pdf
-
AC1: 'Reply on RC1', Erik Zuiderweg, 16 Apr 2021
I appreciate the comments of Dr. Bodenhausen, but I think he is too pessimistic about what can be learned from the HSQC. He states:
“The problems are summarized in the conclusions: “The theory of 1HN R2 for proteins is not iron-clad; issues such as “like” and “unlike” R2, “in-phase/antiphase” relaxation, “selective” and “unselective” R1 rates and cross-correlated R2 relaxation all play roles in these issues.” I cannot agree more. The main problem is that these issues cannot be resolved by establishing clear boundaries.”
Generally true of course, and dealing with R2 relaxation for all protons would be too difficult. But if one focusses just on the amide protons in a HSQC, I think some boundaries are automatic. Also, the value of the approach depends which precision of analysis / correlation with experiment one is comfortable with. More about that later.
It is clear that amides and aliphatics are always un-like. Amides themselves can be like if they have the same chemical shift within the linewidth. Since we know the structure, and shifts, we can take that into account in the calculations (it will only make a real difference when they are also close in space).
In-phase / anti-phase ratios are also knowable since we can calculate the scalar coupling– but I must admit that I only thought about the length of the acquisition time, and not about the R2 itself; this will make the influence of the anti-phase term only smaller.
Reviewer states about selective/non selective R1:
“The distinction between “selective” and “unselective” R1 rates of neighbouring scalar-coupled protons (that contribute to transverse relaxation of antiphase terms) depends on the degree of saturation, the breadth of the rf irradiation (the statement “hence the 5 kHz r.f. field “hits” those HA whereas the 500Hz r.f. field does not” does not leave any room for a grey area.)”
Reviewer is referring to the T1rho experiment. I think the r.f. argument is not relevant to the HSQC. In addition, an offset-dependent grey area for the T1rho falls, luckily, between the amide and aliphatic resonances. I am happy that reviewer seems otherwise content with the T1rho approach.
What may be an issue of concern, is what the state of the HA’s is during the data acquisition in the HSQC. Certainly I did not expect them to be uniformly in +Z, and I reasoned that they would become quicky saturated through J(0) with the other (saturated) aliphatics. Coming to think of it, it is likely better to run the HSQC with a purge pulse in the first INEPT, then everything not 15N-labeled will be saturated.
He also states:
it is tricky to extend consideration of cross-correlated fluctuations to a manifold of densely packed neighboring protons.
Sure, I actually started the entire project with this, because I thought it would be a real problem I could solve. I spent lots of calculations on this issue, but could show that the deviations caused by cross-correlations are not a major issue at this level of precision (Appendix).
Still, if one could be experimentally and computationally perfect, I do not believe that one will ever get a super correlation between the HSQC experiment and the calculations. Afterall, proteins are not static – the inter-proton vectors will fluctuate in length and direction. And small distance changes make for big relaxation changes. Despite that, I do get an overall agreement, with a RMSD of a ~ 2 Hz. And with that, one can define “outliers”, and the outliers are to my interest: they tell us something about the protein dynamics.
This is a philosophical argument: does one require real precision – as is the hallmark of Reviewer’s science, or is one content with progress over what was previously done (the latter being: hey, this HSQC line is narrow, it must belong to a dynamic area in the protein, and the converse). For me the progress is most important – and I show with the analyses of the HSQC of BPTI that one can indeed identify areas of conformational dynamics that correspond to what was found before with 15N relaxation methods. The latter swayed me towards accepting the imprecision, and to enjoy the progress. Reviewer does not comment on the BPTI part, which for me is the most important.
So, I hope that this may sway Dr. Bodenhausen, and that he could go along with publishing despite our philosophical differences about scientific progress (small changes of course to be made)
For me, I will continue with the T1rho, since that is the better defined experiment, and with perdeuterated proteins once I can make them again, for another publication – the timespan for the Festschrift and corona rules do not allow me to do that for this manuscript anymore.
Citation: https://doi.org/10.5194/mr-2021-30-AC1
-
AC1: 'Reply on RC1', Erik Zuiderweg, 16 Apr 2021
-
RC2: 'Comment on mr-2021-30', Anonymous Referee #2, 18 Apr 2021
The comment was uploaded in the form of a supplement: https://mr.copernicus.org/preprints/mr-2021-30/mr-2021-30-RC2-supplement.pdf
-
EC1: 'Comment on mr-2021-30', Jörg Matysik, 07 May 2021
Dear Eric,
Unfortunately, after consulting several editors, we cannot accept your paper for publication in MR.
Conceptually your idea to use sensitive 15N-1H HSQC spectra for dynamics analysis of proteins is very nice and could be very useful for a large number of biological applications. Thus, the software you developed for that would be a valuable contribution. However, the results still show a poor correlation between predicted (cf Fig. 4 and 6) and measured relaxation properties and are therefore not convincing yet. Therefore the dynamics analysis still remains very qualitative in the end. As various experts indicated, there can be several reasons for this, such as slow dynamics/exchange and indirect saturation transfer effect. Since good methods exist to study those, the current paper seems to be premature.
Whereas the current manuscript cannot be accepted, we encourage you to progress these studies, to investigate improvements via, e.g., MD simulations and to obtain more convincing data in particular. We think that this could lead to a balanced publication that gives credit to your software and its theoretical analysis. We know that you will be disappointed not to add an accepted manuscript to the Special Issue for Robert Kaptein. When you want, it is possible to leave your current manuscript as a contribution to MR Discussions and in that case it would still be connected to this SI, as indicated by Dr Bodenhausen. Since our task as guest editors for this SI will end, you can also enquire with the executive editors how this contribution could be merged into a full publication later on.
With kind regards
Jörg (Matysik)Citation: https://doi.org/10.5194/mr-2021-30-EC1 -
AC2: 'Reply on EC1', Erik Zuiderweg, 07 May 2021
Dear Jorg,
This is dissapointing, but I would have been surprised if you could have put the review of the Editor-in-Chief to the side. It is too bad that the perfect has once more become the enemy of the good. I would like to keep the manuscript posted, with all comments. In that way, others may still learn from the idea and see differences in scientific philosophy exposed.
Citation: https://doi.org/10.5194/mr-2021-30-AC2
-
AC2: 'Reply on EC1', Erik Zuiderweg, 07 May 2021
Interactive discussion
Status: closed
-
RC1: 'Comment on mr-2021-30', Geoffrey Bodenhausen, 15 Apr 2021
The comment was uploaded in the form of a supplement: https://mr.copernicus.org/preprints/mr-2021-30/mr-2021-30-RC1-supplement.pdf
-
AC1: 'Reply on RC1', Erik Zuiderweg, 16 Apr 2021
I appreciate the comments of Dr. Bodenhausen, but I think he is too pessimistic about what can be learned from the HSQC. He states:
“The problems are summarized in the conclusions: “The theory of 1HN R2 for proteins is not iron-clad; issues such as “like” and “unlike” R2, “in-phase/antiphase” relaxation, “selective” and “unselective” R1 rates and cross-correlated R2 relaxation all play roles in these issues.” I cannot agree more. The main problem is that these issues cannot be resolved by establishing clear boundaries.”
Generally true of course, and dealing with R2 relaxation for all protons would be too difficult. But if one focusses just on the amide protons in a HSQC, I think some boundaries are automatic. Also, the value of the approach depends which precision of analysis / correlation with experiment one is comfortable with. More about that later.
It is clear that amides and aliphatics are always un-like. Amides themselves can be like if they have the same chemical shift within the linewidth. Since we know the structure, and shifts, we can take that into account in the calculations (it will only make a real difference when they are also close in space).
In-phase / anti-phase ratios are also knowable since we can calculate the scalar coupling– but I must admit that I only thought about the length of the acquisition time, and not about the R2 itself; this will make the influence of the anti-phase term only smaller.
Reviewer states about selective/non selective R1:
“The distinction between “selective” and “unselective” R1 rates of neighbouring scalar-coupled protons (that contribute to transverse relaxation of antiphase terms) depends on the degree of saturation, the breadth of the rf irradiation (the statement “hence the 5 kHz r.f. field “hits” those HA whereas the 500Hz r.f. field does not” does not leave any room for a grey area.)”
Reviewer is referring to the T1rho experiment. I think the r.f. argument is not relevant to the HSQC. In addition, an offset-dependent grey area for the T1rho falls, luckily, between the amide and aliphatic resonances. I am happy that reviewer seems otherwise content with the T1rho approach.
What may be an issue of concern, is what the state of the HA’s is during the data acquisition in the HSQC. Certainly I did not expect them to be uniformly in +Z, and I reasoned that they would become quicky saturated through J(0) with the other (saturated) aliphatics. Coming to think of it, it is likely better to run the HSQC with a purge pulse in the first INEPT, then everything not 15N-labeled will be saturated.
He also states:
it is tricky to extend consideration of cross-correlated fluctuations to a manifold of densely packed neighboring protons.
Sure, I actually started the entire project with this, because I thought it would be a real problem I could solve. I spent lots of calculations on this issue, but could show that the deviations caused by cross-correlations are not a major issue at this level of precision (Appendix).
Still, if one could be experimentally and computationally perfect, I do not believe that one will ever get a super correlation between the HSQC experiment and the calculations. Afterall, proteins are not static – the inter-proton vectors will fluctuate in length and direction. And small distance changes make for big relaxation changes. Despite that, I do get an overall agreement, with a RMSD of a ~ 2 Hz. And with that, one can define “outliers”, and the outliers are to my interest: they tell us something about the protein dynamics.
This is a philosophical argument: does one require real precision – as is the hallmark of Reviewer’s science, or is one content with progress over what was previously done (the latter being: hey, this HSQC line is narrow, it must belong to a dynamic area in the protein, and the converse). For me the progress is most important – and I show with the analyses of the HSQC of BPTI that one can indeed identify areas of conformational dynamics that correspond to what was found before with 15N relaxation methods. The latter swayed me towards accepting the imprecision, and to enjoy the progress. Reviewer does not comment on the BPTI part, which for me is the most important.
So, I hope that this may sway Dr. Bodenhausen, and that he could go along with publishing despite our philosophical differences about scientific progress (small changes of course to be made)
For me, I will continue with the T1rho, since that is the better defined experiment, and with perdeuterated proteins once I can make them again, for another publication – the timespan for the Festschrift and corona rules do not allow me to do that for this manuscript anymore.
Citation: https://doi.org/10.5194/mr-2021-30-AC1
-
AC1: 'Reply on RC1', Erik Zuiderweg, 16 Apr 2021
-
RC2: 'Comment on mr-2021-30', Anonymous Referee #2, 18 Apr 2021
The comment was uploaded in the form of a supplement: https://mr.copernicus.org/preprints/mr-2021-30/mr-2021-30-RC2-supplement.pdf
-
EC1: 'Comment on mr-2021-30', Jörg Matysik, 07 May 2021
Dear Eric,
Unfortunately, after consulting several editors, we cannot accept your paper for publication in MR.
Conceptually your idea to use sensitive 15N-1H HSQC spectra for dynamics analysis of proteins is very nice and could be very useful for a large number of biological applications. Thus, the software you developed for that would be a valuable contribution. However, the results still show a poor correlation between predicted (cf Fig. 4 and 6) and measured relaxation properties and are therefore not convincing yet. Therefore the dynamics analysis still remains very qualitative in the end. As various experts indicated, there can be several reasons for this, such as slow dynamics/exchange and indirect saturation transfer effect. Since good methods exist to study those, the current paper seems to be premature.
Whereas the current manuscript cannot be accepted, we encourage you to progress these studies, to investigate improvements via, e.g., MD simulations and to obtain more convincing data in particular. We think that this could lead to a balanced publication that gives credit to your software and its theoretical analysis. We know that you will be disappointed not to add an accepted manuscript to the Special Issue for Robert Kaptein. When you want, it is possible to leave your current manuscript as a contribution to MR Discussions and in that case it would still be connected to this SI, as indicated by Dr Bodenhausen. Since our task as guest editors for this SI will end, you can also enquire with the executive editors how this contribution could be merged into a full publication later on.
With kind regards
Jörg (Matysik)Citation: https://doi.org/10.5194/mr-2021-30-EC1 -
AC2: 'Reply on EC1', Erik Zuiderweg, 07 May 2021
Dear Jorg,
This is dissapointing, but I would have been surprised if you could have put the review of the Editor-in-Chief to the side. It is too bad that the perfect has once more become the enemy of the good. I would like to keep the manuscript posted, with all comments. In that way, others may still learn from the idea and see differences in scientific philosophy exposed.
Citation: https://doi.org/10.5194/mr-2021-30-AC2
-
AC2: 'Reply on EC1', Erik Zuiderweg, 07 May 2021
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
709 | 689 | 90 | 1,488 | 156 | 76 | 64 |
- HTML: 709
- PDF: 689
- XML: 90
- Total: 1,488
- Supplement: 156
- BibTeX: 76
- EndNote: 64
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1
Erik R. P. Zuiderweg
This preprint has been withdrawn.
- Preprint
(3098 KB) - Metadata XML
-
Supplement
(667 KB) - BibTeX
- EndNote