Mechanical forces impair antigen discrimination by reducing differences in T‐cell receptor/peptide–MHC off‐rates

Abstract T cells use their T‐cell receptors (TCRs) to discriminate between lower‐affinity self and higher‐affinity foreign peptide major‐histocompatibility‐complexes (pMHCs) based on the TCR/pMHC off‐rate. It is now appreciated that T cells generate mechanical forces during this process but how force impacts the TCR/pMHC off‐rate remains debated. Here, we measured the effect of mechanical force on the off‐rate of multiple TCR/pMHC interactions. Unexpectedly, we found that lower‐affinity TCR/pMHCs with faster solution off‐rates were more resistant to mechanical force (weak slip or catch bonds) than higher‐affinity interactions (strong slip bonds). This was confirmed by molecular dynamics simulations. Consistent with these findings, we show that the best‐characterized catch bond, involving the OT‐I TCR, has a low affinity and an exceptionally fast solution off‐rate. Our findings imply that reducing forces on the TCR/pMHC interaction improves antigen discrimination, and we suggest a role for the adhesion receptors CD2 and LFA‐1 in force‐shielding the TCR/pMHC interaction.

bond (e.g. 1G4) and catch bond (e.g. OT-1) receptors in hand, and they are therefore in a position to test this prediction.
Referee #2: Detailed comments on the ms. titled "Mechanical forces impair antigen discrimination by reducing differences in T cell receptor off-rates" by Pettmann et al. (for submission to the authors) Pettmann et al. combine theoretical modelling with a laminar flow chamber (LFC) assay, molecular dynamics simulations as well as previously published T cell activation assays to assess the role of mechanical forces exerted on T cell antigen receptors (TCRs) bound to peptide-loaded MHC molecules (pMHCs) on T-cell antigen discrimination. Based on their experimental data, calculations and simulation results they conclude that low affinity TCR-pMHC interactions are, at least when compared to higher affinity TCR-pMHC interactions, less susceptible to disruption by mechanical forces. The authors then reason that antigen discrimination by T cells suffers from TCR-pMHC exerted forces -unlike what has previously been proposed -and benefits from cell-cell adhesion mediated for example by LFA-1 and CD2. While I find the chosen approach stimulating and some of the data shown peculiar, I am, as I will state in more detail below, not yet convinced by the evidence provided. My reservation is primarily based on my concerns regarding (i) the choice of force ranges applied, (ii) the power and validity of the laminar flow chamber assay and (iii) the conclusions drawn from the resulting experiments.
(i) Choice of force range applied and force modalities Previous studies which have focused on forces applied to individual TCRs have indicated values ranging between less than 2 pN to 15 pN. The authors may want to state more clearly why they have focused their analysis on forces ranging between 5 and > 100 pN. I consider this an important issue, because the chosen force range appears largely above the observed physiological range, and hence I question the relevance of the reported findings for a better understanding of T-cell recognition. Surely, bulkforce measurements (e.g. traction force microscopy) have invoked such a range, but why would the authors consider them as a guiding range for their stated single molecule force measurements? Another aspect concerns the direction of applied forces. The authors mention and indicate in their graphs pulling forces, more or less in line with the TCR-pMHC length axis. As far as I can tell, the LFC-assay is clearly based on the exertion of shear-forces, which (a) may or may not come to register in synaptic setting and (b) may furthermore exert an influence on the overall TCR-pMHC stability that differs from that resulting from normal forces. In fact, the molecular dynamics simulations assume altogether normal forces. In my view this incongruency needs to be resolved.
(ii) Power and validity of the laminar flow chamber assay There are a number of oddities I perceive when examining the survival-against-time plots in figure 2B. First of all, the plots do not appear to follow (at least in their current form) a single exponential function. I suggest that the authors include such a fit, or alternatively they render the data in form of a semi-exponential plot to give reference to single exponential decay. Another aspect that confuses me is why the survival value does not approach 0 for all plots but takes instead a different value for every force value applied. Moreover, there is no perceivable difference between the 67 pN and 91 pN force plot. Subtracting an off-set, which appears to differ from measurement to measurement (for reasons to be further explained) would change the results plotted in figure 2c with implications for the overall message. Also, why did the authors not display and account for later decay behaviour since there are still a great number of surviving bonds after 5 seconds of recording. Is the decay bi-or multi-phasic by any chance? How often have these experiments been repeated, and would it be possible to indicate a confidence interval? I could not find any answers to my questions. In case they have been provided somewhere in the ms. they should be placed front and centre to facilitate the reading process.
(iii) Validity of the conclusions drawn from the LFC-based experiments I am hesitant to accept the validity of the numerical analysis of koff0 and Χβ as shown for example in figure 2c. First of all, I do not necessarily see single exponential decay in figure 2b (see above), a prerequisite for further analysis. Secondly, there appears to be an anomaly at 14 pN forces, which, if true, indicates catch-bond behaviour. What do the authors then make of this if they formulate for the entire function a slip-bond property? Would the combined data not suggest the existence of a catch-slip bond as put forward by Zhu and colleagues? If true, how could such behavior be described with Χβ being a constant? Would catch-slip bond behaviour not indicate that certain events, which guide the binding behaviour, are not accounted for by the rather simple model? Furthermore, I have doubts regarding the sensitivity of the LFC-based assay which, according to the authors, truthfully reflects single bond behaviour, even for low affinity interactions. What exactly is the noise level in this assay resulting from (a) nonspecific bead-glass interactions and (b) di-or multivalent interactions? I am not aware a single negative control (e.g. TCRmismatched pMHC or pMHC-mismatched TCR) to serve as reference. I find it insufficient to refer to published literature for method validation and would feel more confident about the conclusions drawn if I had been presented with a negative control carried out in parallel with any experiment shown. I also think it would be appropriate to show the survival over time plots for all experiments (including exponential fit) and to feature the camera recordings online. Creating a solid foundation with primary data-related evidence appears critically important, especially if the study reports a finding that goes against intuition (e.g. (a) the catch bond behaviour of A6 + Tax 7Q at 110 pN with a KD of ~ 35 µM and a koff0 of 0.84s-1 or (b) the catch bond behaviour of OT-1 + OVA at 110 pN with a KD of 28 µM and a koff0 of 1.3s-1). How reliable are the NYE 4D measurements? I would assume that given the reduced on-rate, witnessing the occurrence of a single bond may be rather challenging. The scenario appears even more dismal for lowest affinity ligands such as endogenous ligands. Last but not least, more detailed screening of the force range between 0 and 10 pN would be highly informative (see above). But does the LFC-based assay actually support such measurements (maybe through the use of smaller beads)? Such analysis would refocus the study towards physiologically more relevant force ranges.
Referee #3: Pettmann et al. used computational modeling and flow-chamber-based and solution-based binding assays to investigate how mechanical forces affect TCR antigen recognition. They unexpectedly found lower-affinity pMHCs were more resistant to mechanical force than higher-affinity interactions. And this lower-affinities pairs form weak slip or catch bonds, while the higher ones forms strong slip bonds. Based on these characterizations, they suggest force may impair TCR antigen discrimination and ahdesion receptors provide force-shielding role. The computational modeling part is beautiful, but the flow chamber assay has significant technical concerns. So, I feel these conclusions are not strong enough supported by their experiments due to technical concenerns and other comments listed below, their overall conclusion are over interpretated. I do not think this form is suitable for EMBO J.
Major concerns and questions: 1. In the abstract, the statement "...but how force impacts TCR/pMHC off-rates remains unclear" seem not very right, as there are multiple published papers from different groups have clearly demonstrated force can regulate TCR/pMHC dissociation using highly sensitive single-molcule assays (e.g. biomembrane force probe and optical tweezers from Lang, Ellis, Zhu, Garcia, and Chen groups); and in the introduction, authors kept repeating similar statements, which are not appropriate. In the introduction, they also mentioned "how force impact TCR pmhc interaction koff(m) and antigen discrimination is controversial". I highly suggest authors need to carefully revisit those papers and accordingly revise these statements. 2. In the 1st paragraph of introduction, author mentioned ref15 catch bond can be abolished purified form. I went back to look at this reference, I didn't find out related data to support this statement. Can author re-clarify this point? 3. And they further suggest "some catch bond ... may be secondary to TCR signaling rather than intrinsic to TCR/pmHC interaction". Several papers have clearly demonstrated from Ellis and Zhu groups showing that catch bond is clearly essential for triggering TCR signaling. And also the TCR/CD3 complex sitting on the membrane provides an critical biophysical regulation on TCR/pMHC interaction, compared to purified forms. Davis, Zhu, Ellis, Chen, and Garcia's groups have clearly shown these insitu TCR/pMHC binding are more important than the "intrinsic binding" author claimed for TCR triggering and antigen discrimination. I not very clear why authors still kept focusing on "intrinsic ones". I suggest authors should at least study purified TCR/CD3 complex binding with pMHC if possible if they would like to characterize "intrinsic binding kinetics". 4. Technical concerns of flow chamber assay used in this work is a critical issue of this work. Authors used a camera with low temporal resolution (50Hz based on their methods) to capture the TCR/pMHC binding which are of fast kinetics. Based on data obtained by this slow capturing camera, their temporal resolution is much lower compared to those in the optical tweezer and biomembrane force probe assay (Zhu, Ellis, Lang groups).Such that, they missed many fast binding events, leading to significantly bias their final conclusions. I would suggest authors use faster cameras to redo their experiments with at least CD3 complex associated TCRs. 5. Another technical concern, I don't know why authors set a 1s cut off in the analysis of their tethering lifetimes from flowchamber assay. This cut off would also impact the final average lifetimes or koff(m) very much such that the force-dependent TCR/pMHC bond lifetimes would change very much. I do not think this cut off is appropriate. 6. Use flow chamber assay to study single-molecule binding is challenging, although author claimed they did at the singlemolecule level. I am not very clear how they achieved this single-molecule level. I strongly suggest authors to perform singlemolecule binding assay either with optical tweezer or biomembrane force probe to confirm their results (like Zhu and McEver group did before on selectin/ligand binding), otherwise current results are very ambiguous. 7. The lower bound of flow-chamber assay is hard to reach below 10pN range like bioforce assay, I feel author may miss this critical regime, so they obtained very different data from optical tweezer and biomembrane force assay. Actually in their Fig.S6B,Fig.2C,D,G, in the lower bound, their data kind of showing very little catch trends. Again, I suggest they should also repeat their work with single-molecule binding assay. 8. Other concerns of their MD simulation. As they used coarse grained MD simulation instead of all-atom simulations like Garcia and Chen group did before, such that there are some intra-molecular conformational changes are not able to be observed. This might affect the rupture force and Fmax value from the simulation. So, I suggest authors need to repeat these simulations with all-atom MD. How to define and calculate Fmax, as this is very important for calculating Xb and koff. 9. Regarding the data on the OT-1/OVA interaction depicted in Fig.S6, what is the authors' criteria for the definition of low-affinity TCRs? Davis and Zhu groups, two nature papers have clearly shown the in-situ binding affinity is more appropriate than insolution ones for TCR. 10. In the discussion part, many statements are over-interpreted given their data not convinced enough. I strongly suggest authors to revise the discussion accordingly. While referee #2 is satisfied with the introduced changes, referee #3 is not convinced that the analysis "as is" provides enough insight for consideration here.
I have discussed the comments further with referee #2. We both appreciate the raised points and see where the referee is coming from. However, we also find that the analysis provides important insight that will be of value to the field and stimulate further research on this topic. I would therefore like to ask you to submit a revised version by addressing the raised concerns with text changes either in the point-by-point response or in the MS text. Will you also make sure that you have a balanced discussion where you indicate the limitations of the analysis and what the data can tell you and what it can't.
When you submit the revised version will you also take care of the following editorial points: -Please correct the reference format to EMBO journal style -We need a Disclosure and competing interests statement -The movie needs to be renamed to Movie EV1 and called out in the MS text. Please zip the movie with the legend together. This is a very elegant and important study which I consider a much needed contribution to ongoing discussions in the field. I wholeheartedly recommend publication in EMBO Journal.
Referee #3: In general, I do not think authors have really addressed my major concerns, especially the technical concerns of "wet expeirments" as well as "wet data" quality, analysis and interprepation. so their data can not well support their conclusion. More importantly, I do not really think its scientific advances of this study is significant enough for EMBO J.
Major concers are still remains: 1. In Fig.2B, data quality and analysis reamin to be concerned to me. How could the survivall rate be larger than 1.0 when force is 4.1pN? compared to original data in the first submission, the distribution looks normal to me at this force. After authors "socalled" corrected their data after revision, I feel more concerns about the data quality. 2. I feel also very much concerned about the fiting in fig.2C-G panels (force vs off-rate ). Authors use Bell model to fit their data. As Bell model only has one dissociation pathway, it is well known that this model is unable to discribe the catch bond behavior, unless using two state models like Evans (2004 PNAS) or Zhu group did before for selectin's studies. Moreover, for demonstating the goodness of the fitting, at least you should present these data in a semi log to show the data are linearly fit. I don't think many conditions are not well fitted by bell model. Especially, for Fig.2G, A6/Tax7Q panel, I do not think this data can be fitted by bell model? the xb they got is negative , would it be meaningful? If can't be fitted by bell model, authors should really think of other models instead of Bell's. for example, you can look at "Shiwen Guo et al., Communication Biology, 2019; 3. I still hold my significant concerns about their flow chamber assay to characterize single-molecule dissociation kinetics, although they cited several previous flow chamber assay papers in the reponses. The pulling force generated in the assay to the TCR/pMHC bond is really dependent on flow's shear velocity, beads diameter, and molecular length etc. when force increases (c.f. Fig.S1,S2 and standard fluidic mechanics analysis), they needs to increase shear velocity (Fig.S2) given fixed bead size and buffer types (i.e., contant buffer's viscocity). Inevitablly, they would increased the force loading rate at least 10 times more from low force regime to high force regime, given constant molecular stiffness. In contrast, other single-molecule biomechanical assays, they can well control the loading rate in constant. Several papers ( ) clearly shown that force loading rate could affect ligand dissoicaiton kinetics. I'm not sure how would their velocity change affect the their force-dependent kinetics and their extrapulated zero-force off-rate and xb. This uncertainty might also affect their conclusion that force impairs antigen discrimination of TCR. As I said before, the best way is authors to use single-molecule assay to test their TCR-pMHC system and compare with flow-chamber data by themselves. But they directly negelect my suggestion , I don't know why. At least they should test one pair to TCR-pMHC bond. Regarding authors' response to my 1st and 3rd comments, I still don't agree their argument. TCR and CD3 are highly complexed together, especially after recently complexed structure are revealed by Huang and Davis group (Nature,2019; Mol Cell 2022; Cell 2022). TCR/CD3s are clearly tightly associated as a machinery, strongly suggesting that TCR recognition would be regulated by associated CD3. Furthermore, authors also realized that Liu et al., 2015 clearly showed reduced catch bond for a cell-free TCR, further suggesting the importance of cd3 complex for TCR catch bond formation and TCR antigen recognition. So, if we really want to reveal how TCR recognize antigen and how mechanical force regulates this recognition, testing complexed TCR/CD3 binding with pMHC is necessary. So only looking at abTCR (purified form) binding with pMHC to answer TCR recognition problem seems less scientifically meaningful.
Other issues: There are still several issues regarding writing more precisely and correctly. they should be more objectively report what has been published , what is known, what is unknown. For example, in abstract, they wrote" how force impacts the TCR/pMHC offrate remains unclera.". that is not true. Even they selfs have also cited ref.15 : "It is notable that the magnitude of these catchbonds is appreciably reduced (15) or abolished (16) when applying force to purified forms of the same TCRs."

Response to Reviewers
Wednesday November 9, 2022 Dear Dr. Karin Dumstrei, We respectfully resubmit our revised manuscript, 'Mechanical forces impair antigen discrimination by reducing differences in T cell receptor off-rates', for your considering in the EMBO Journal.
We have added additional limitations to the discussion and have provided a point-by-point response below.
We greatly appreciate that you have taken the time to discuss the comments of reviewer 3 with reviewer 2.
We thank you and the three reviewers for taking the time to read our manuscript and for providing constructive comments that have improved it. This is a very elegant and important study which I consider a much needed contribution to ongoing discussions in the field. I wholeheartedly recommend publication in EMBO Journal.
We thank the reviewer for taking the time to read our revised manuscript.

Referee #3:
In general, I do not think authors have really addressed my major concerns, especially the technical concerns of "wet expeirments" as well as "wet data" quality, analysis and interprepation. so their data can not well support their conclusion. More importantly, I do not really think its scientific advances of this study is significant enough for EMBO J.
We thank the reviewer for taking the time to read our revised manuscript.
Major concers are still remains: 1. In Fig.2B, data quality and analysis reamin to be concerned to me. How could the survivall rate be larger than 1.0 when force is 4.1pN? compared to original data in the first submission, the distribution looks normal to me at this force. After authors "so-called" corrected their data after revision, I feel more concerns about the data quality.
The corrected survival distribution can be larger than 1 because we are now subtracting the survival distribution of non-specific interactions (this control was requested by Reviewer 2). Under very low forces (4.1 pN), it is possible for the number of non-specific interactions to be similar to the number of specific interactions and as a result, the survival can appear to be larger than 1. As expected, even at these low forces, the survival is quickly below 1 because there are more specific interactions that have longer survival durations. We note that all of the non-corrected data in the original submission is included in the appendix.
2. I feel also very much concerned about the fiting in fig.2C-G panels (force vs off-rate ). Authors use Bell model to fit their data. As Bell model only has one dissociation pathway, it is well known that this model is unable to discribe the catch bond behavior, unless using two state models like Evans (2004 PNAS) or Zhu group did before for selectin's studies. Moreover, for demonstating the goodness of the fitting, at least you should present these data in a semi log to show the data are linearly fit. I don't think many conditions are not well fitted by bell model. Especially, for Fig.2G, A6/Tax7Q panel, I do not think this data can be fitted by bell model? the xb they got is negative , would it be meaningful? If can't be fitted by bell model, authors should really think of other models instead of Bell's. for example, you can look at "Shiwen Guo et al., Communication Biology, 2019; Bell's model is an empirical model that can describe slip (xb > 0) and catch (xb < 0) bonds. This is demonstrated in Figure 1B of our manuscript. In general, the data points appear above and below the fit of Bell's model (i.e. residuals are randomly distributed), which is a Response to Reviewers hallmark of a good fit. We agree that Bell's model can only capture data where off-rates increase or decrease as force is increased and as a result, we were unable to include 3 interactions in our analysis (Appendix Figure S13). Importantly, the key conclusion that we are making is that forces disproportionately impact the off-rate of higher-affinity interactions compared to lower-affinity interactions, which is clearly seen in the data even before fitting Bell's model. We first explain this point in the text before discussing the result of Bell's model (compare NYE 6V to 4D in Fig 2D or Tax WT to 5H in Fig 2G for example). We have used Bell's model as a way of summarizing this observation using xb (Fig 2F and 2I). We also note that a control of fitting Bell's model is that the extrapolated zero-force off-rate obtained from fitting Bell's model is correlated to the solution affinity as measured by an independent instrument, namely SPR (Fig 2E,F). Therefore, Bell's model is not necessary to observe our key conclusion and the fit of Bell's model does produce reasonable estimates for the off-rate at zero-force.
3. I still hold my significant concerns about their flow chamber assay to characterize singlemolecule dissociation kinetics, although they cited several previous flow chamber assay papers in the reponses. The pulling force generated in the assay to the TCR/pMHC bond is really dependent on flow's shear velocity, beads diameter, and molecular length etc. when force increases (c.f. Fig.S1,S2 and standard fluidic mechanics analysis), they needs to increase shear velocity (Fig.S2) given fixed bead size and buffer types (i.e., contant buffer's viscocity). Inevitablly, they would increased the force loading rate at least 10 times more from low force regime to high force regime, given constant molecular stiffness. In contrast, other single-molecule biomechanical assays, they can well control the loading rate in This uncertainty might also affect their conclusion that force impairs antigen discrimination of TCR. As I said before, the best way is authors to use single-molecule assay to test their TCR-pMHC system and compare with flow-chamber data by themselves. But they directly negelect my suggestion , I don't know why. At least they should test one pair to TCR-pMHC bond.
In AFM and BFP, loading rates are an inevitable consequence of the use of springs as means of force measurement. In LFC (present study), there is an order of magnitude delay between bond kinetics and any force application because of the stretching (straightening) of the ligand and receptor and their associated linkers (consider that there is a ~32 nm long assembly (see Fig EV2 for a schematic) and velocities ranging from 10 to 100 µm/sec, the stretching/straightening duration ranges from 0.3 to 3 ms). Thus, the force is applied only when the assembly is fully stretched/straightened: the loading rate itself is dependent only on the spring constant of the assembly. It follows that force is applied instantly on a bond of which spring constant is an intrinsic parameter (if the antibody linker, invariant in our study and structurally very close to TCR and MHC molecules, is put aside). Taken together, our experimental system does not have the complication of a progressively applied force when we assess bond behaviour at a given force.
We note that the BFP and AFM force ramps durations are in the order of magnitude of one or several hundreds of ms; contact times between both surfaces before pulling are in the Response to Reviewers same range in these methods. It is therefore difficult to assert with certainty what effect on bond lifetime is the consequence of loading rate itself or of other possibilities occurring during force ramp: first, bonds may mature during these hundred of millisecond long durations; second, bond lifetimes are in similar order of magnitude and early breaking of a fraction of them in a time-dependent manner may occur. In LFC, the intrinsic loading is restricted to very short durations with fewer potential artefacts. Lastly, we have used the OT-I TCR, which has heavily been used by the BFP community and have produced the observed catch bond.
4. Regarding authors' response to my 1st and 3rd comments, I still don't agree their argument. TCR and CD3 are highly complexed together, especially after recently complexed structure are revealed by Huang and Davis group (Nature,2019; Mol Cell 2022; Cell 2022). TCR/CD3s are clearly tightly associated as a machinery, strongly suggesting that TCR recognition would be regulated by associated CD3. Furthermore, authors also realized that Liu et al., 2015 clearly showed reduced catch bond for a cell-free TCR, further suggesting the importance of cd3 complex for TCR catch bond formation and TCR antigen recognition. So, if we really want to reveal how TCR recognize antigen and how mechanical force regulates this recognition, testing complexed TCR/CD3 binding with pMHC is necessary. So only looking at abTCR (purified form) binding with pMHC to answer TCR recognition problem seems less scientifically meaningful.
We note that the full TCR-CD3 complex is not necessary to observe a catch bond. First, in our study we have found a catch bond for A6 binding 7Q and OT-I binding OVA using purified TCRa/b without CD3. In the case of A6, we note that the TCR itself was unchanged but rather the ligand was varied, which suggests that catch bonds can be observed as a result of differences in the contact interface between TCR and pMHC (without any impact of CD3). This has also recently been suggested by Chris Garcia (Zhao et al (2022) Science) who used the BFP assay on live T cells to show that changes in the CDR2 loops can change a TCR from exhibiting a slip bond to a catch bond.
The differences between LFC using purified proteins (present study) and the BFP using live T cells are: 1) Use of the full TCR-CD3 complex in BFP (as pointed out by the reviewer), and 2) The repeated use of live T cells in BFP, which means that TCR clustering, cytoskeleton rearrangements, signalling feedbacks, etc can also explain differences between our assays.
We have revised the discussion to include a paragraph on limitations of our assay and included the reviewers point on CD3 and have suggested that future BFP experiments are conducted using the full TCR-CD3 complex on membranes rather than intact live T cells that allow signalling.
Other issues: There are still several issues regarding writing more precisely and correctly. they should be more objectively report what has been published , what is known, what is unknown. For example, in abstract, they wrote" how force impacts the TCR/pMHC off-rate remains unclera.". that is not true. Even they selfs have also cited ref.15 : "It is notable that the Response to Reviewers magnitude of these catch-bonds is appreciably reduced (15) or abolished (16) when applying force to purified forms of the same TCRs." Although the reviewer notes 'several issues', only a single one is explained and we have addressed it by changing "unclear" to "debated" in the abstract. Please note that it is EMBO Journal policy for the transcript of the editorial process (containing referee reports and your response letter) to be published as an online supplement to each paper. If you do NOT want this, you will need to inform the Editorial Office via email immediately. More information is available here: https://www.embopress.org/page/journal/14602075/authorguide#transparentprocess Your manuscript will be processed for publication in the journal by EMBO Press. Manuscripts in the PDF and electronic editions of The EMBO Journal will be copy edited, and you will be provided with page proofs prior to publication. Please note that supplementary information is not included in the proofs.
You will be contacted by Wiley Author Services to complete licensing and payment information. The required 'Page Charges Authorization Form' is available here: https://www.embopress.org/pb-assets/embo-site/tej_apc.pdf -please download and complete the form and return to embopressproduction@wiley.com EMBO Press participates in many Publish and Read agreements that allow authors to publish Open Access with reduced/no publication charges. Check your eligibility: https://authorservices.wiley.com/author-resources/Journal-Authors/openaccess/affiliation-policies-payments/index.html Should you be planning a Press Release on your article, please get in contact with embojournal@wiley.com as early as possible, in order to coordinate publication and release dates.
If you have any questions, please do not hesitate to call or email the Editorial Office. Thank you for your contribution to The EMBO Journal. ** Click here to be directed to your login page: https://emboj.msubmit.net