Mapping Protein–Protein Interactions at Birth: Single-Particle Cryo-EM Analysis of a Ribosome–Nascent Globin Complex

Interactions between ribosome-bound nascent chains (RNCs) and ribosomal components are critical to elucidate the mechanism of cotranslational protein folding. Nascent protein–ribosome contacts within the ribosomal exit tunnel were previously assessed mostly in the presence of C-terminal stalling sequences, yet little is known about contacts taking place in the absence of these strongly interacting motifs. Further, there is nearly no information about ribosomal proteins (r-proteins) interacting with nascent chains within the outer surface of the ribosome. Here, we combine chemical cross-linking, single-particle cryo-EM, and fluorescence anisotropy decays to determine the structural features of ribosome-bound apomyoglobin (apoMb). Within the ribosomal exit tunnel core, interactions are similar to those identified in previous reports. However, once the RNC enters the tunnel vestibule, it becomes more dynamic and interacts with ribosomal RNA (rRNA) and the L23 r-protein. Remarkably, on the outer surface of the ribosome, RNCs interact mainly with a highly conserved nonpolar patch of the L23 r-protein. RNCs also comprise a compact and dynamic N-terminal region lacking contact with the ribosome. In all, apoMb traverses the ribosome and interacts with it via its C-terminal region, while N-terminal residues sample conformational space and form a compact subdomain before the entire nascent protein sequence departs from the ribosome.

In the structure, the authors iden�fy similar contacts to the ribosome that have been observed previously in the vicinity of the PTC and the exit tunnel core.What is new here are NC interac�ons with the ribosome exit tunnel ves�bule region and the surface of the ribosome near the end of the exit tunnel.Based on the size of crosslinked NC species and cryo-EM density, the authors propose a model in which a nonpolar or hydrophobic patch on ribosomal protein L23 might contribute to early steps of protein folding in a manner similar to a chaperone.This region of L23 is not an overlapping binding site with that for trigger factor (TF) but is a major site of interac�on with the signal recogni�on par�cle (SRP).Given the transient nature of ribosome-NC interac�ons, it is not yet clear whether compe��on between NC binding to this region of L23 and SRP would be relevant to protein folding trajectories for soluble protein domains.
Overall, the observa�on of surface interac�ons between the NC and L23 provide new structural insights, but the func�onal implica�ons of the interac�on and whether they are in some sense general are not clear.It is also not clear exactly what is new by adding the previously published fluorescence anisotropy data, as one of three possible models for NC folding is inconsistent with the new cryo-EM structural model independent of the fluorescence anisotropy data.The authors also don't delve into whether the nonpolar region of L23 observed to interact with the NC here is conserved across bacteria and/or across archaea and eukaryotes as well.For these reasons, the manuscript is probably beter suited to another journal such as Biochemistry in its present form.Some other issues the authors should consider addressing are given below.
1.The authors provide arguments for why the crosslinked species in Figure 1d likely involves L23 and L29.In their prior work (ref. 45, i.e.) they used western blots against L23 and L29.Why not use those approaches here?2. In Figures 2-4, it would be helpful to show the NC density in the absence of ribosome density, i.e. mask away the ribosome density.This would help the reader see more clearly how much density is available to be filled by the NC model.S3 is a litle confusing.Presumably the red density is very low, not high as indicated.Furthermore, a slice through the exit tunnel would be more helpful to see compared to showing the le� and right panel.The top panel is helpful for the surface interac�ons.lower-resolu�on NC density.The authors should consider an alterna�ve approach of genera�ng a family of models consistent with the density, perhaps using Roseta-based tools or others developed by Tom Terwilliger's group.A detailed list of our responses to the reviewers' comments is provided below.For added clarity, the modifica�ons implemented to the manuscript as a result of the comments by the reviewers are underlined.In addi�on, we also uploaded a marked-up copy of the revised manuscript, showing in red the changes that we implemented as a result of the reviewers' comments.

Suppor�ng figure
In response to Reviewer #1: This reviewer greatly appreciated the significance of our work.He/she had a construc�ve sugges�on on how to improve the quality of the manuscript.Only minor revisions were requested.Specific responses to the comment by this reviewer are listed below.
Comment #1: I find it difficult when the authors write: "Finally, frequency-domain fluorescence anisotropy decay 46 experiments were carried out, to identify N-terminal dynamic regions that were undetectable by sp-cryo-EM." and these measurements have not been carried out during the preparation of the current manuscript.This is misleading.It would be more scholarly to transparently state that the authors relate to an earlier publication -which they do.That needs to be clearly stated.
Response #1: We completely agree with this comment by the reviewer and we sincerely apologize about the ambiguous formula�on of the original wording in the manuscript.We have now modified the per�nent sentence of the manuscript as follows: "Finally, previously acquired 46 frequency-domain fluorescence anisotropy decay experiments were cri�cally evaluated, to help iden�fying N-terminal dynamic regions that were undetectable by sp-cryo-EM".We hope that the above updated wording sounds transparent, appropriate and fair.
In response to Reviewer #2: This reviewer had several construc�ve comments on how to improve the quality of the manuscript.Specific responses to the comments by this reviewer are listed below.
Comment #1: Overall, the observation of surface interactions between the NC and L23 provide new structural insights, but the functional implications of the interaction and whether they are in some sense general are not clear.
Response #1: Indeed, our work represents the first �me the interac�on between ribosomal protein L23 and a nascent protein chain outside the ves�bule of the ribosomal exit tunnel has ever been reported.While the func�onal role of this interac�on has admitedly not been elucidated yet, the fact that the interac�on site on L23 is highly hydrophobic in nature is en�rely unique.The highly nonpolar character of this interac�on strongly suggests that its role is related to cotransla�onal protein folding and/or cotransla�onal nascent chain solubility.While future studies will be devoted to unequivocally and directly assessing the func�onal significance of this interac�on, its mere existence and its clearly iden�fied nonpolar character represent a significant step forward is our knowledge of the ribosomal characteris�cs in the context of transla�on.
In support of the importance of this newly iden�fied interac�on, recent publica�ons listed below, showed that solubiliza�on of nascent chains is one of the major func�ons of the ribosome.Phys. Chem. B 124, 6488-6507 (2020).Therefore, it is highly sugges�ve -though not proven as of now --that the L23-nascent-chain interac�on iden�fied here, given its nonpolar character, may contribute to preserve the solubility of nascent-chains by interac�ng with their nonpolar regions.While this study does not prove the func�onal nature of the newly iden�fied RNC/L23 interac�on, our research moves an important step in the right direc�on by linking the nonpolar character of the newly discovered interac�on with the known nascentchain solubilizing ac�on of the ribosome.To take the reviewer's comment into account and to beter illustrate the above concepts, we adjusted and expanded the manuscript text in the sec�on �tled "The L23 r-protein may serve as a molecular chaperone".
Comment #2: The authors also don't delve into whether the nonpolar region of L23 observed to interact with the NC here is conserved across bacteria and/or across archaea and eukaryotes as well.
Response #2: We completely agree with the reviewer on the fact that Conserva�on Analysis of the novel interac�on site within the L23 r-protein was a notable missing aspect in our manuscript.To address this shortcoming, we carried out Conserva�on Analysis of the L23 ribosomal protein.The results are graphically illustrated in the following new por�ons of the manuscript: panel "e" of Figure 5, suppor�ng Figure S9, suppor�ng Table S4 and a new sec�on of the Results �tled "The nonpolar nascentchain interac�on site of L23 is highly conserved".A new Methods sec�on describing how the analysis was performed has also been added to the manuscript.Very interes�ngly, the new conserva�on analysis shows that the L23 RNC interac�on-site residues iden�fied in this work are highly conserved across a variety of species.This finding adds significance to the results presented in this work by showing the nonpolar L23 region interac�ng with RNCs is recurrently present in mul�ple organisms.There is no doubt that addi�on of the conserva�on analysis adds value to our work, and we are extremely grateful to the reviewer for the sugges�on to perform it.

Comment #3: [listed as comment 1 by the reviewer]
The authors provide arguments for why the crosslinked species in Figure 1d likely involves L23 and L29.In their prior work (ref. 45, i.e.) they used western blots against L23 and L29.Why not use those approaches here?
Response #3: Indeed, as men�oned by the reviewer, a Western blot (WB) approach could have also been adopted here.However, the presence of an interac�on of the apoMb RNC with the L23 ribosomal protein is already clearly established by the cryo-EM work presented here.Therefore, we believe that performing WB analysis is unnecessary in our work.In addi�on, the WB approach already showed RNC interac�ons with L23 in the case of an intrinsically disordered protein and a globin protein with similar structure to apoMb, in previously published studies.So, while certainly feasible, WB studies seem unnecessary and would not add any new informa�on.2-4, it would be helpful to show the NC density in the absence of ribosome density, i.e. mask away the ribosome density.This would help the reader see more clearly how much density is available to be filled by the NC model.

Comment #4: [listed as comment 2 by the reviewer] In Figures
Response #4: We thank the reviewer for this interes�ng comment.On the other hand, showing the nascent-chain alone would not reveal any informa�on that is not currently available in the present display.We tried, and the density visualiza�on ends up being very similar to that of the present display.For this reason, we elected to preserve the current display.On the other hand, we are certainly willing to display both versions (with and without the surrounding ribosome) if the reviewer insists on this point.S3 is a little confusing.Presumably the red density is very low, not high as indicated.Furthermore, a slice through the exit tunnel would be more helpful to see compared to showing the left and right panel.The top panel is helpful for the surface interactions.

Comment #5: [listed as comment 3 by the reviewer] Supporting figure
Response #5: This is an important point and we are extremely grateful to the reviewer for bringing this item up.The red-colored density in suppor�ng figure S3 does not correspond to a very high-resolu�on region but is, unfortunately, an ar�fact resul�ng from overfited noise.The presence of this noise is unavoidable due to threshold requirements to adequately visualize the nascent-chain density.There was no way that this ar�fact could be avoided and visualiza�on of ribosome slices actually produced worselooking results.Therefore, we have le� the image in suppor�ng figure s3 as is.However, to avoid incorrect interpreta�ons and to address the reviewer's comment, we added the following text to the legend of this figure: "Note that red-colored density does not correspond to a high-resolu�on region but is an ar�fact due to overfited noise.The presence of this noise is unavoidable due to threshold requirements to adequately visualize the nascent-chain density".We hope that this ac�on adequately takes care of the reviewer's concern.S5 is not really useful in the context of this paper.However supporting figures 6 and 7 could be.It is not clear, however, how well constrained the Ramachandran plot could be based on the lower-resolution NC density.The authors should consider an alternative approach of generating a family of models consistent with the density, perhaps using Rosetta-based tools or others developed by Tom Terwilliger's group.

Comment #6: [listed as comment 4 by the reviewer] Supporting figure
Response #6: This is another interes�ng sugges�on by the reviewer.We have indeed examined Rosetabased tools as suggested by the reviewer.Here are our resul�ng thoughts.According to Roseta's user guide, there are three different scenarios where Roseta could poten�ally be useful.The first op�on is using model refinement against a low-resolu�on structure.Unfortunately, op�on this requires an ini�al model, and while we have models of the ribosome, there are no models of the apoMb nascent chain tethered to the ribosome within the literature.The second scenario is the use of Roseta to rebuild missing por�ons of a model.Unfortunately, this alterna�ve op�on also requires the input of previously solved structures, which again, are not available within the literature.The third scenario, de novo building within a density using only a known sequence, would be the most valuable tool within Roseta for us to use.However, this tool requires near atomic resolu�on and the ability to fit known sequences within the density.Unfortunately, our data are not at a sufficiently high enough resolu�on to use this tool.As a reference, here is the source web site that we studied in detail, to provide the above answer: htps://www.rosetacommons.org/demos/latest/public/electron_density_structure_refinement/structure_re finement Regarding the Ramachandran map in suppor�ng figure S5, we believe that it is overall useful to retain it, and keep it as a suppor�ng figure (but of course not in the main manuscript).Our reasoning is behind this choice is that it seems useful to the readers, at least to some extent, to appreciate that, considering all dihedral angles of both RNC and hos�ng ribosome shown in this figure, there are very few dihedralangle viola�ons in our final structure.
In summary, we made our best efforts to address the reviewers' concerns and believe that the manuscript has substan�ally improved as a result of their comments.We sincerely hope that this revised version meets the requirements for publica�on in ACS Central Science.

Comments to the Author
In the revised manuscript by Masse et al., the authors now provide a conserva�on analysis of the nonpolar surface residues in L23 iden�fied in the cryoEM study interac�ng with the apoMb N-terminal region.This improves the manuscript, but the methods do need to be clearer.How did the authors choose the L23 sequences to include in the alignment and conserva�on analysis?Presumably there are thousands of such sequences from bacteria that could be used, not just the 20 or so listed in the methods?It would be helpful to know how widely and evenly distributed the sequences are across bacterial phylogeny.The authors should ensure this is the case and use enough sequences for their analysis to be meaningful.
In Figure 5e, the authors could spread out the sequences a bit in the horizontal direc�on to make the more readable and to u�lize the available space.

2.
On p. 18 the authors present data that are not shown.This should be discouraged, and the authors could put it in the supplement.The caveat due to EDC crosslinking make sense, and this can be noted while s�ll including the data in the manuscript.

3.
In Figure S3, the authors can reduce the significant figures from 4 to 2 in the resolu�on key.
Author  A detailed list of our responses to the reviewers' comments is provided below.For added clarity, the modifications implemented to the manuscript as a result of the comments by the reviewers are underlined.In addition, we also uploaded a marked-up copy of the revised manuscript, showing in red the changes that we implemented as a result of the reviewers' comments.
In response to Reviewer #2: This reviewer had a few additional helpful comments on how to further improve the quality of the manuscript.Specific responses to the comments by this reviewer are listed below.
Comment #1: In the revised manuscript by Masse et al., the authors now provide a conservation analysis of the nonpolar surface residues in L23 identified in the cryoEM study interacting with the apoMb N-terminal region.This improves the manuscript, but the methods do need to be clearer.How did the authors choose the L23 sequences to include in the alignment and conservation analysis?Presumably there are thousands of such sequences from bacteria that could be used, not just the 20 or so listed in the methods?It would be helpful to know how widely and evenly distributed the sequences are across bacterial phylogeny.The authors should ensure this is the case and use enough sequences for their analysis to be meaningful.
Response #1: We agree with this comment, and we have now provided a more comprehensive conservation analysis, including more organisms.Importantly, we have included more detailed written descriptions to justify the choice of the bacteria employed in the conservation analysis.The text accompanying the improved conservation analysis is provided both in the Results and Discussion (pages 16-17) and in the Materials and Methods (pages 25-26).In addition, we have generated improved and new Tables (supporting Tables 4 and 5) and Figures (supporting Figures S11 and S12) to better describe the choices made and provide additional details.Briefly, the gene encoding the L23 ribosomal protein is essential for cell viability in E. coli.Thus, we carried out the conservation analysis across all bacteria of a database known as DEG (Database of Essential Genes), which contains genomes of organisms that bear one or more essential genes.First, we focused on the bacteria in the DEG database where L23 serves as an essential protein.Next, we performed an additional new conservation analysis comparing E. coli L23 to all L23s in bacteria (within the DEG database) where L23 is non-essential.It is worth notig that, while the conservation analysis reported in this work is limited to the bacteria in DEG database, more comprehensive comparisons across a wider set of bacteria could and should also be carried out in the future.However, these comparisons are beyond the scope of the present study, which focuses on the RNC structure and dynamics in E. coli and on RNC interactions with the ribosomal surface.
Comment #2: In Figure 5e, the authors could spread out the sequences a bit in the horizontal direction to make the more readable and to utilize the available space.
Response #2: We are thankful to the reviewer for this comment, and we have updated the conservationanalysis diagrams in Figure 5e so that they are now horizontally spread as much as possible, to facilitate inspection and readability.
Comment #3: On p. 18 the authors present data that are not shown.This should be discouraged, and the authors could put it in the supplement.The caveat due to EDC crosslinking make sense, and this can be noted while still including the data in the manuscript.
Response #3: We are glad about this request by the reviewer, and we have now added supporting Figures S9 and S10, which include the lifetime and anisotropy results originally mentioned in the manuscript but not explicitly shown before.We have also cited these new figures and properly adjusted the text in the Results and Discussion (page 19).Finally, we have added a new section in the Materials and Methods, to properly describe how the fluorescence lifetime and anisotropy-decay experiments were carried out.
Comment #4: In Figure S3, the authors can reduce the significant figures from 4 to 2 in the resolution key.
Response #4: In response to this comment by the reviewer, we were happy to reduce the number of significant figures in the index bar of Figure S3 (Supporting Information) , as requested by the reviewer.

In response to the Formatting Updates requested by the ACS Central Science Editorial Office:
Comment #1: SI PG#S: The supporting information pages must be numbered consecutively, starting with page S1.
Response #1: The pages of the Supporting Information have now been appropriately renumbered, as requested.
Comment #2: SI AU EMAIL: Please include the email address of the corresponding author on the first the Supporting Information, with an asterisk next to their name in the author list.
Response #2: We have taken care of this requested update.
Comment #3: AU EMAIL: Please include the email address of the corresponding author on the first page of the manuscript, and the Supporting Information if submitted, with an asterisk next to their name in the author list.Please be sure to label "email." Response #3: We have taken care of this requested update.
In addition to addressing the above comments by the reviewers and the requested formatting changes, we have also corrected some typographical errors and made small adjustments to Figures 6 and 7 and the accompanying text, to further improve clarity and better summarize our results.
In summary, we made our best efforts to address the comments by reviewer #2, and believe that the manuscript has substantially improved as a result of our manuscript updates.We sincerely hope that this newly revised version meets the requirements for publication in ACS Central Science.
to submit a newly revised version of manuscript oc-2023-00777q.R1 titled "Mapping Protein-Protein Interactions at Birth: Single-Particle Cryo-EM Analysis of a Ribosome/Nascent-Globin Complex" co-authored by, Meranda Masse, Rachel B. Hutchinson, Christopher E. Morgan, Heather Allaman, Hongqing Guan, Edward Yu, and myself.This manuscript is to be considered as an Article for ACS Central Science.All authors have approved the submission of this 2 nd revision.
1. 'Complementary Role of Co-and Post-Transla�onal Events in De Novo Protein Biogenesis' J.
Please do not hesitate to contact me in case you have any ques�ons.
's Response to Peer Review Comments: Please do not hesitate to contact me if you have any questions.