the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
γ-effects identify preferentially populated rotamers of CH2F groups: side-chain conformations of fluorinated valine analogues in a protein
Abstract. Using cell-free protein synthesis, the protein G B1-domain (GB1) was prepared with uniform high-level substitution of valine by (2S,3S)-4-fluorovaline, (2S,3R)-4-fluorovaline, or 4,4'-difluorovaline. The 19F nuclear magnetic resonance (NMR) signals are distributed over a wide spectral range. The fluorinated samples maintain the relative 1H chemical shifts of the wild-type protein, opening a convenient route to assigning the 19F NMR signals. For the singly fluorinated residues, the 13C chemical shifts of the remaining CH3 group are subject to a γ-effect that depends on the population of different rotameric states of the CH2F group and correlates with 3JFC coupling constants. In addition, the preferentially populated rotamers are reflected by the γ-gauche effect on 19F chemical shifts, which correlates with 3JHF couplings. Some of the side-chain conformations determined by these restraints position the fluorine atom near a backbone carbonyl group, a non-intuitive finding that has previously been observed in the high-resolution crystal structure of a different protein. Through-space scalar 19F–19F couplings due to transient fluorine–fluorine contacts are observed between residues 39 and 54.
Competing interests: The authors declare no competing interests, except that the communicating author is an editor of Magnetic Resonance.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.- Preprint
(7901 KB) - Metadata XML
-
Supplement
(2662 KB) - BibTeX
- EndNote
Status: open (until 16 Oct 2025)
-
RC1: 'Comment on mr-2025-12', Geoffrey Bodenhausen, 21 Sep 2025
reply
-
AC1: 'Reply on RC1', Gottfried Otting, 25 Sep 2025
reply
Are deeply buried amino acids necessarily rigid?
Response:
Good question!
To decompose into 2 questions: (1) What drives the 19F relaxation of buried CH2F groups versus more solvent-exposed ones?
In Table S2, the deeply buried residue 54 shows the fastest transverse and slowest longitudinal relaxation rates. Conversely, the most highly solvent-exposed residue 21 shows the slowest transverse and faster (though not the fastest) longitudinal relaxation rates. This pattern suggests longer correlation times for residue 54 than 21, which would underpin both CSA and dipolar relaxation. As far as the relaxation is driven by dipolar interactions with 1H spins, the protons of the CH2F group dominate very much, as shown by the HOESY cross-peak intensities (Fig. 6). Compared with the 1H relaxation of CH3 groups, which are complicated by fast methyl rotation, the 19F relaxation of CH2F groups depends on slower movements that are more akin to those governing the 1H relaxation of long side chains such as those of lysine and arginine, where the correlation times get shorter for 1H spins nearer to the end of the side chain.
(2) Are amino acid side chains in the interior of a protein always less flexible than on the surface?
High-resolution X-ray and NMR structures usually report single conformations for buried amino acid side chains, whereas solvent-exposed side chains show much more variable conformations, as also evidenced by averaged three-bond coupling constants. There are, however, exceptions, as the packing in the core of a protein may leave room for multiple conformations.
Our observation of 19F-NMR line width correlating with depth of burial is experimental and qualitative. At this stage, we prefer not to try and interpret the relaxation rates quantitatively.
The expression “installed” seems a bit unfortunate. How about “incorporated”?
Response:
The terms are used interchangeably in the literature. We agree that “incorporated” is more adequate for several instances.
Why not write out “gamma-gauche effect” without using a Greek letter, for the convenience of data bases?
Response:
We prefer technology to serve humans rather than the other way around.
Are di-fluorinated amino acids also commercially available?
Response:
Analogues of amino acids with two coupled fluorine atoms are commercially available. When Honek and co-workers incorporated difluoromethionine into calmodulin, however, the 19F-NMR spectrum was quite complicated (J. Am. Chem. Soc. 1999, 121, 8475). A simpler spectrum would be obtained, if the amino acid can be genetically encoded. For example, 2,3-difluorophenylalanine (CAS number 236754-62-4) is inexpensive and genetic encoding systems have already been reported by Zhang et al., ACS Synth. Biol. 2023, 12, 2408. The amino acid may also be recognized by the polyspecific aminoacyl-tRNA synthetase selected for the incorporation of pentafluorophenylalanine (Qianzhu et al.,
ACS Sens. 2025, 10, 3152). In proteins, however, I would worry about CSA relaxation spoiling the party of long-lived states.
Citation: https://doi.org/10.5194/mr-2025-12-AC1
-
AC1: 'Reply on RC1', Gottfried Otting, 25 Sep 2025
reply
-
RC2: 'Comment on mr-2025-12', Geoffrey Bodenhausen, 27 Sep 2025
reply
Rarely has a naive question triggered such a rich reply !
Citation: https://doi.org/10.5194/mr-2025-12-RC2 -
RC3: 'Comment on mr-2025-12', Geoffrey Bodenhausen, 27 Sep 2025
reply
Rarely has a naive question triggered such a rich reply !
Citation: https://doi.org/10.5194/mr-2025-12-RC3 -
RC4: 'Comment on mr-2025-12', Anonymous Referee #2, 05 Oct 2025
reply
This reviewer concurs with the views of the first reviewer that this is a thorough study demonstrating an angular dependence of the gamma-gauche effect of 19F on 13C chemical shifts. The ability to observe CH2F rotameric populations without 19F NMR via the chemical shift of the opposite methyl group is intriguing.
The observation of unique 1H chemical shifts for the CH2F groups as shown in Figure S3 is interesting. The authors have quite obviously spent a significant amount of time considering the structural environments of the four valine groups in this study. Do they care to speculate further about why some show unique 1H chemical shifts (particularly GB1-2 valine 54), while others do not?
The authors note qualitative analysis of the 3J(HA-HB) couplings via analysis of COSY cross peak intensities. Is there any evidence that the intensity of those cross peaks varies between the WT, GB-1, GB-2, and GB1-d samples, which would suggest that fluorination is affecting the chi1 populations?
Minor comments:
- On line 314, there appears to be an incomplete sentence at the beginning of the paragraph starting with a lowercase “using”.
- Perhaps change “little destabilized” to “slightly destabilized” (as an adverb seems more appropriate here than an adjective) on line 538.
- The melting temperatures were described to be about 10 °C lower than WT. However, the data for WT was not shown in Figure S1, nor given in the cited reference (Tan et al., 2024). Could the WT melting temperature be given and the source of that information be clarified? Are the reported melting temperatures from reversible or irreversible unfolding?
Citation: https://doi.org/10.5194/mr-2025-12-RC4 -
AC2: 'Reply on RC4', Gottfried Otting, 07 Oct 2025
reply
Do they care to speculate further about why some show unique 1H chemical shifts (particularly GB1-2 valine 54), while others do not?
Response: In principle, the chemical shifts of diastereotopic protons should always be different. The exceptionally large chemical shift difference observed for the protons of the Cγ2H2 group of residue 54 can be explained by ring currents from the nearby indole side chain of Trp43. In the wild-type protein, the 1H-NMR signal of the γ2-methyl group of Val54 appears at -0.3 ppm. Based on the 3D structure of GB1 and the rotamer shown in Figure 13, larger ring currents are expected for the γ22- than for the γ23-proton. Interestingly, even the highly solvent-exposed residue 21 shows different chemical shifts for the protons of its Cγ1H2F group, together with a large 3JFC coupling constant in GB1-1 and a large 13C γ-effect (Table 1). We conclude that incomplete averaging by bond rotation is the rule rather than the exception for FVal residues.
The authors note qualitative analysis of the 3J(HA-HB) couplings via analysis of COSY cross peak intensities. Is there any evidence that the intensity of those cross peaks varies between the WT, GB-1, GB-2, and GB1-d samples, which would suggest that fluorination is affecting the chi1 populations?
Response: The signal-to-noise ratio in the DQF-COSY spectra was insufficient to draw more quantitative conclusions on the size of the 3J(HA-HB) couplings, especially as the spectra had been recorded without 19F decoupling. The conservation of 1H chemical shifts supports the notion of little change of χ1 angles. Specifically, the relative order of 1H chemical shifts of the FVal CH3 groups is maintained in GB1-1 and GB1-2 (Fig. 10).
Minor comments:
- On line 314, there appears to be an incomplete sentence at the beginning of the paragraph starting with a lowercase “using”.
Response: The missing part of the sentence is “To explore the full range of the γ-effect, DFT calculations were performed”. We apologize for the oversight.
- Perhaps change “little destabilized” to “slightly destabilized” (as an adverb seems more appropriate here than an adjective) on line 538.
Response: we’ll make the change as suggested.
- The melting temperatures were described to be about 10 °C lower than WT. However, the data for WT was not shown in Figure S1, nor given in the cited reference (Tan et al., 2024). Could the WT melting temperature be given and the source of that information be clarified? Are the reported melting temperatures from reversible or irreversible unfolding?
Response: thank you for the careful reading and pointing out the error! The correct reference is Tan et al., 2025. We observed reversible unfolding for the proteins GB1-1 and GB1-2. For GB1-d, no CD spectrum was recorded after heat denaturation and we don't have the original protein anymore. We believe that GB1-d also refolds reversibly as it survived NMR measurements for many days without precipitation. For the wild-type protein, the inflection point of the heat denaturation curve is at about 79 oC, which we will cite in the legend of Fig. S1.
Citation: https://doi.org/10.5194/mr-2025-12-AC2
-
AC2: 'Reply on RC4', Gottfried Otting, 07 Oct 2025
reply
-
RC5: 'Comment on mr-2025-12', Anonymous Referee #3, 07 Oct 2025
reply
The manuscript by Abdelkader et al. presents a detailed and insightful investigation into how fluorination of valine residues in the GB1 protein can be used to study side-chain conformations. The work is interesting, carefully executed, and suitable for publication in Magnetic Resonance after minor revision. I have several questions and suggestions that may help to clarify and strengthen the manuscript.
- Additional species in GB1-d (Fig. 2c):
Could you expand the discussion regarding the presence of additional species observed in Fig. 2c? For instance, why are these species not present in GB1-1 and GB1-2? You write, “Therefore, the low-intensity peaks in the 19F-NMR spectrum of GB1-d most likely originate from the species with one valine and three diFVal residues.” Does this imply that the absence of one diFVal residue causes a significant change in the chemical shifts of the other 19F-labeled valines? Additionally, could you specify which residue is missing, since the small number of low-intensity peaks suggests that not all possible combinations are present in solution? - Differences in 19F chemical shifts between variants:
Could you briefly discuss potential reasons why the 19F chemical shifts differ between GB1-1 and GB1-d, and between GB1-2 and GB1-d? This may become clearer later in the manuscript when you discuss how rotamers are affected by fluorine substitution, but at this earlier stage the distinction is not immediately obvious to the reader. - Formatting issue (page 17, line 314):
There appears to be a paragraph break and an unfinished sentence at this location. Please check and correct this formatting issue. - Fig. 12 and DFT calculations:
If I understand correctly, the DFT calculations do not indicate a strictly linear dependence of the FC coupling on the chemical shift, but rather a more complex relationship. Could you comment on this in the text, clarifying that the linear fit is used primarily to illustrate the presence of a correlation?
In addition, could you discuss the relevance of the DFT calculations performed on (2R)-1-fluoro-2-methylpropane(3-13C) to the conformations of valine residues in GB1? A brief justification of this model system would be helpful. - Fig. 13:
It might be useful to include the structures of the valine residues identified in GB1-d alongside those from GB1-1 and GB1-2, to facilitate direct comparison. - Clarification of 3JFC coupling statement:
The sentence “Assigning preferential rotamers in GB1-d is more difficult, as the diFVal residues contain no CH3 group, which makes 3JFC coupling measurements difficult” is not fully clear. Could you elaborate on how the absence of a CH3 group specifically complicates the 3JFC coupling measurements? - Potential applications to protein–ligand interactions:
It would be valuable to expand the discussion on how these fluorinated labels might be applied to studying protein–ligand interactions. Would such interactions be detectable as perturbations in the 19F chemical shifts? Do you expect these effects to be site-specific and sensitive to local changes, or rather global, given the apparent sensitivity of the 19F shifts to the overall protein structure? - Estimating energy differences between rotamers:
Based on your experimental data and the DFT calculations of 3JFC couplings, would it be possible to estimate the energy differences between the various rotamers?
Citation: https://doi.org/10.5194/mr-2025-12-RC5 -
AC3: 'Reply on RC5', Gottfried Otting, 09 Oct 2025
reply
- Additional species in GB1-d (Fig. 2c):
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
159 | 36 | 16 | 211 | 16 | 11 | 11 |
- HTML: 159
- PDF: 36
- XML: 16
- Total: 211
- Supplement: 16
- BibTeX: 11
- EndNote: 11
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1
This is a highly knowledgeable report on an exhaustive study of the effects of fluorine atoms substituted in methyl groups of valine residues in a protein. The authors have carefully investigated the effects of the bulky fluorine atom on the local conformations and overall protein structure. Heat denaturation monitored by circular dichroism indicates melting temperatures about 10 degrees lower than for the wild-type protein. There seems to be a slight bias for the fluorine atoms to sit near positively polarized carbons of carbonyl groups. The inverse correlation of 3JHF couplings with 19F chemical shifts is a manifestation of the gamma-gauche effect. “The broadest signals were observed for the most deeply buried residues, indicating that the peak heights are sensitive indicators of the side-chain mobilities”: why should deeply buried amino acids necessarily be rigid?
The expression “installed” seems a bit unfortunate. How about “incorporated”?
Why not write out “gamma-gauche effect” without using a Greek letter, for the convenience of data bases?
One of the curious observations is the existence of through-space scalar 19F-19F couplings due to transientfluorine-fluorine contacts between two remote residues, despite the polarity of the C-F bonds that may discourage direct 19F-19F contacts.
From the point of view of an ‘aficionado’ of long-lived states and coherences involving two or more 19F nuclei, it would be interesting to know if a similar study could be conducted with di-fluorinated methyl groups. Are di-fluorinated amino acids also commercially available?