Computational model of the full-length TSH receptor
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This manuscript presents the first molecular dynamics (MD) simulations of full-length membrane-bound Thyroid Stimulating Hormone Receptor (TSHR). The authors find that its linker region (LR) is disordered, contrasting previous models. While this is largely a solid study that would interest researchers working in computational modeling, thyroid hormone metabolism, and signaling, the rationale for the arbitrarily chosen starting model and unclear mechanistic relevance need to be clarified further.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Structural Biology and Molecular Biophysics (eLife)
- Reading list (BiophysicsColab)
Abstract
(GPCR)The receptor for TSH receptor (TSHR), a G protein coupled receptor (GPCR), is of particular interest as the primary antigen in autoimmune hyperthyroidism (Graves’ disease) caused by stimulating TSHR antibodies. To date, only one domain of the extracellular region of the TSHR has been crystallized. We have run a 1000 ns molecular dynamic simulation on a model of the entire TSHR generated by merging the extracellular region of the receptor, obtained using artificial intelligence, with our recent homology model of the transmembrane domain, embedded it in a lipid membrane and solvated it with water and counterions. The simulations showed that the structure of the transmembrane and leucine-rich domains were remarkably constant while the linker region (LR), known more commonly as the ‘hinge region,’ showed significant flexibility, forming several transient secondary structural elements. Furthermore, the relative orientation of the leucine-rich domain with the rest of the receptor was also seen to be variable. These data suggest that this LR is an intrinsically disordered protein. Furthermore, preliminary data simulating the full TSHR model complexed with its ligand (TSH) showed that (a) there is a strong affinity between the LR and TSH ligand and (b) the association of the LR and the TSH ligand reduces the structural fluctuations in the LR. This full-length model illustrates the importance of the LR in responding to ligand binding and lays the foundation for studies of pathologic TSHR autoantibodies complexed with the TSHR to give further insight into their interaction with the flexible LR.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
Weaknesses:
The methods section lacks sufficient detail, and arbitrary choices made in the simulation setup may have biased the results. The author's finding that the LR is disordered does not provide obvious mechanistic insights, and the simulations with the bound ligand are too preliminary to make solid conclusions. Although this manuscript is technically strong, the significance of the results is often unclear.
We did not make “arbitrary choices”. The set up choice (only one) which we made was guided by heuristics and its adequacy was amply confirmed by the robustness of the simulated system. We have emphasized this in the revision.
Reviewer #2 (Public Review):
Strengths:
- The authors have focused on the LR region of TSHR and perform rigorous MD simulations to identify its …
Author Response
Reviewer #1 (Public Review):
Weaknesses:
The methods section lacks sufficient detail, and arbitrary choices made in the simulation setup may have biased the results. The author's finding that the LR is disordered does not provide obvious mechanistic insights, and the simulations with the bound ligand are too preliminary to make solid conclusions. Although this manuscript is technically strong, the significance of the results is often unclear.
We did not make “arbitrary choices”. The set up choice (only one) which we made was guided by heuristics and its adequacy was amply confirmed by the robustness of the simulated system. We have emphasized this in the revision.
Reviewer #2 (Public Review):
Strengths:
- The authors have focused on the LR region of TSHR and perform rigorous MD simulations to identify its various conformers and tried to give a reasoning for this observation. The authors also showed the stability of LR increased in the presence of the ligand, TSH.
- The authors have done many simulations of the TMD helix bundle and meticulously tried to quantitate the differences by assessing the changes in helix length, radius and angles.
Weaknesses:
- Although the focus of the paper was the full model of the TSHR, the authors have broken down the whole protein into smaller sequences and have done separate simulations, and discussed the result. The whole picture of the TSHR is not clear. For example in Figure 5, the various confirmation (and secondary structures) of only LR is shown at different times. For the TMD helix bundle, separate tables have been shown, focusing only on TMD.
The whole picture of the TSHR is shown on Figure 1. The reason the TMD is not shown in Figure 5 is because there is little variation in the TMD (as analyzed in detail in Tables 1-3) and not including it allowed us to show more detail in the ectodomain.
- The authors have analyzed the cysteines in the LR doing simulation, showing the propensity for various pairs of disulfide formation. However, the authors have not further discussed this point. Can this information be used to better guide the modelling process?
We have added a statement in the Discussion section suggesting that the closeness of these cysteines during the simulation indicates that they indeed should form disulfide bonds. Furthermore, their separation when TSH was introduced indicated the likely role of these disulfide bonds in signal transduction.
- Based on the data in this manuscript, the authors claim that the LR domain makes significant contact with the TSH ligand. However if one refers to the crystal structure of FSH ligand with the ectodomain of its receptor (pdb: 4AY9) , the corresponding loop for LR is missing, directing to the point that either the interaction between this loop (LR) and ligand is either very weak or there is no interaction.
While our results on the TSH-TSHR complex are still preliminary we pointed out in the revision that (a) the TSH-LR contact we see involves the part of the LR that are missing in the FSH-FSHR structure (b) there is only 39.3% sequence identity between the LR of TSHR and FSHR and (c) the large fluctuations we see in the LR conformations suggests that it is very unlikely that the contacts seen are artifacts of the initial structure.
-
Evaluation Summary:
This manuscript presents the first molecular dynamics (MD) simulations of full-length membrane-bound Thyroid Stimulating Hormone Receptor (TSHR). The authors find that its linker region (LR) is disordered, contrasting previous models. While this is largely a solid study that would interest researchers working in computational modeling, thyroid hormone metabolism, and signaling, the rationale for the arbitrarily chosen starting model and unclear mechanistic relevance need to be clarified further.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
This publication shows a strong understanding and implementation of large-scale multiprotein MD simulations. It is the first application of MD simulations to full-length membrane-bound TSHR. The authors showed that the LR is intrinsically disordered, contrasting a previously published homology model. Some simulation results are supported by cryo-EM structures. Finally, it is significant that the inclusion of TSH in the binding site altered the dynamics of the LR region, supporting a hypothesis that the LR is involved in a signaling mechanism, though the authors acknowledge this result as preliminary.
Weaknesses:
The methods section lacks sufficient detail, and arbitrary choices made in the simulation setup may have biased the results. The author's finding that the LR is disordered does not provide obvious …Reviewer #1 (Public Review):
This publication shows a strong understanding and implementation of large-scale multiprotein MD simulations. It is the first application of MD simulations to full-length membrane-bound TSHR. The authors showed that the LR is intrinsically disordered, contrasting a previously published homology model. Some simulation results are supported by cryo-EM structures. Finally, it is significant that the inclusion of TSH in the binding site altered the dynamics of the LR region, supporting a hypothesis that the LR is involved in a signaling mechanism, though the authors acknowledge this result as preliminary.
Weaknesses:
The methods section lacks sufficient detail, and arbitrary choices made in the simulation setup may have biased the results. The author's finding that the LR is disordered does not provide obvious mechanistic insights, and the simulations with the bound ligand are too preliminary to make solid conclusions. Although this manuscript is technically strong, the significance of the results is often unclear. -
Reviewer #2 (Public Review):
The authors have built up from their previous work where they modelled the transmembrane region of TSHR protein and combined this result with the LRD domain obtained from Alphafold2 to get the full model of TSHR. Their final model is supported by the various MD simulations performed in this manuscript. The authors have also provided the most complete version of the model by supplementing with the DPPC bilayer, water, and ions to mimic the real transmembrane protein. Moreover, the main point that the authors make is that their model takes into consideration the large conformational movement of the LR region which was not taken into account earlier. Also based on the simulations, they presented and compared the models of the TSHR in complex with ligand (TSH), stimulating and blocking antibody.
Strengths:
1. …Reviewer #2 (Public Review):
The authors have built up from their previous work where they modelled the transmembrane region of TSHR protein and combined this result with the LRD domain obtained from Alphafold2 to get the full model of TSHR. Their final model is supported by the various MD simulations performed in this manuscript. The authors have also provided the most complete version of the model by supplementing with the DPPC bilayer, water, and ions to mimic the real transmembrane protein. Moreover, the main point that the authors make is that their model takes into consideration the large conformational movement of the LR region which was not taken into account earlier. Also based on the simulations, they presented and compared the models of the TSHR in complex with ligand (TSH), stimulating and blocking antibody.
Strengths:
1. The authors have focussed on the LR region of TSHR and perform rigorous MD simulations to identify its various conformers and tried to give a reasoning for this observation. The authors also showed the stability of LR increased in the presence of the ligand, TSH.2. The authors have done many simulations of the TMD helix bundle and meticulously tried to quantitate the differences by assessing the changes in helix length, radius and angles.
Weaknesses:
1. Although the focus of the paper was the full model of the TSHR, the authors have broken down the whole protein into smaller sequences and have done separate simulations, and discussed the result. The whole picture of the TSHR is not clear. For example in Figure 5, the various confirmation (and secondary structures) of only LR is shown at different times. For the TMD helix bundle, separate tables have been shown, focussing only on TMD.2. The authors have analyzed the cysteines in the LR doing simulation, showing the propensity for various pair of disulphide formation. However, the authors have not further discussed this point. Can this information be used to better guide the modelling process?
3. Based on the data in this manuscript, the authors claim that the LR domain makes significant contact with the TSH ligand. However if one refers to the crystal structure of FSH ligand with the ectodomain of its receptor (pdb: 4AY9) , the corresponding loop for LR is missing, directing to the point that either the interaction between this loop (LR) and ligand is either very weak or there is no interaction.
-
