Modeling and predicting the overlap of B- and T-cell receptor repertoires in healthy and SARS-CoV-2 infected individuals

  1. Maria Ruiz Ortega
  2. Natanael Spisak
  3. Thierry Mora
  4. Aleksandra M. Walczak

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This work is potentially very important to dissect the immune repertoire of T/B cells, which is one of the most critical/difficult parts for the adaptive immune system to achieve antigen specificity. These conclusions and proposed methods will require additional experimental support and further validation in different disease conditions.

    (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. Reviewer #2 agreed to share their name with the authors.)

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Adaptive immunity’s success relies on the extraordinary diversity of protein receptors on B and T cell membranes. Despite this diversity, the existence of public receptors shared by many individuals gives hope for developing population wide vaccines and therapeutics. Yet many of these public receptors are shared by chance. We present a statistical approach, defined in terms of a probabilistic V(D)J recombination model enhanced by a selection factor, that describes repertoire diversity and predicts with high accuracy the spectrum of repertoire overlap in healthy individuals. The model underestimates sharing between repertoires of individuals infected with SARS-CoV-2, suggesting strong antigen-driven convergent selection. We exploit this discrepancy to identify COVID-associated receptors, which we validate against datasets of receptors with known viral specificity. We study their properties in terms of sequence features and network organization, and use them to design an accurate diagnosis tool for predicting SARS-CoV-2 status from repertoire data.

Article activity feed

  1. eLife

    Author Response

    Reviewer #1 (Public Review):

    This interesting work tries to predict and analyze the overlap of BCR and TCR repertoires (mainly in COVID-19 conditions) which is one of the most important aspects of adaptive immunity that is directly related to antigen specificity. However the primary claims were not fully supported by the current data and analysis the authors presented.

    1. Since the authors showed that the TCR/BCR changed with age, whether they corrected their CMV- and CMV+ analysis with age differences?

    Aging and infections both have a similar outcome on the immune repertoire in terms of diversity reduction. We took the three initial age groups (0-25, 26-50, 51-75) and shuffled them, resulting in three new groups with no age structure but the same CMV positive / negative ratio. For each one of these new control groups, we fitted the q parameter. This process gave statistically indistinguishable values of q: q=0.472± 0.006 with no variation among the groups, as was the case for true age separation, meaning that age and not CMV status is the main driver to convergent selection.

    In addition, we added a new figure from an independent cohort (with no CMV information) showing the same effects of age (Figure 2 - figure supplement 3) and a comment about the control for CMV effects in section Results - ageing in TCR repertoires, 2nd paragraph.

    1. TCR repertoire (probably BCR also) changed along with the time during SARS-CoV-2 infection (especially the first several weeks after cleaning the virus). The authors should consider the time points they used in all the COVID-19 studies to validate the method.

    Inferring the dynamics of public repertoires is a very interesting task. However, the availability of longitudinal data (TCR and BCR repertoires belonging to the same people after/before infection) for large numbers of people needed for the sharing analysis remains very limited or inexistent, preventing us from a time-dependent analysis.

    We have added a discussion about this point (see Results - Convergent BCR sharing in COVID-19 donors, first paragraph).

    1. What's the difference between different infections (e.g. CMV vs SARS-CoV-2)? Or does infection lead to the same TCR/BCR changes in the study? A detailed discussion with an analysis of TCR/BCR repertoire regarding different infections CMV vs SARS-CoV-2 needs to be provided.

    For the BCR case, we looked at SARS-CoV-2 infection. Due to the limited amount of repertoires sequenced from cohorts of people suffering from a given disease, the extension of the analysis, while interesting and needed, remains very limited. However, for TCR repertoires we compared data coming from individuals affected by two very different diseases, an acute respiratory infection (SARSCoV-2) and a chronic infection (CMV).
    While it is hard to draw conclusions about the behavior of these two diseases just from the sharing analysis, we can still make some observations (we included them in the main text, see Discussion, 6th paragraph). The comparison of the q factors fitted on each cohort (qcmv = 0.453 ± 0.006, 𝑞sars-cov2 = 0.452 ± 0.002) seems to show compatible values, suggesting for both cases a much less dramatic change than for B cell.

    1. Are there any features along with different infections compared with tumor/autoimmune conditions (I think there are many publications about TCR/BCR dynamics in various diseases)? Analysis of these data is not only important to control for validating their method but also can generate the most interesting data/conclusions on dissecting the specificity of TCR/BCR repertoire.

    TCR response against autoimmune diseases (like ankylosing spondylitis) has previously been discussed in other papers (e.g. Pogorelyy M. et al. PloS Biology, 2019). However, the type of analysis here exposed requires deeper repertoires with several individuals, a characteristic that is not accomplished right now by most of published datasets for autoimmune disease conditions. Moreover, the sharing analysis imposes an overall generalized response to the given antigen. But it is now widely known that many types of cancer likely have an unique genomic molecular signature in each patient, making sharing not the most suitable approach to understand the immune response in this case. We discuss these current limitations in Discussion, 6th paragraph.

  2. eLife

    Evaluation Summary:

    This work is potentially very important to dissect the immune repertoire of T/B cells, which is one of the most critical/difficult parts for the adaptive immune system to achieve antigen specificity. These conclusions and proposed methods will require additional experimental support and further validation in different disease conditions.

    (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. Reviewer #2 agreed to share their name with the authors.)

  3. eLife

    Reviewer #1 (Public Review):

    This interesting work tries to predict and analyze the overlap of BCR and TCR repertoires (mainly in COVID-19 conditions) which is one of the most important aspects of adaptive immunity that is directly related to antigen specificity. However the primary claims were not fully supported by the current data and analysis the authors presented.

    1. Since the authors showed that the TCR/BCR changed with age, whether they corrected their CMV- and CMV+ analysis with age differences?

    2. TCR repertoire (probably BCR also) changed along with the time during SARS-CoV-2 infection (especially the first several weeks after cleaning the virus). The authors should consider the time points they used in all the COVID-19 studies to validate the method.

    3. What's the difference between different infections (e.g. CMV vs SARS-CoV-2)? Or does infection lead to the same TCR/BCR changes in the study? A detailed discussion with an analysis of TCR/BCR repertoire regarding different infections CMV vs SARS-CoV-2 needs to be provided.

    4. Are there any features along with different infections compared with tumor/autoimmune conditions (I think there are many publications about TCR/BCR dynamics in various diseases)? Analysis of these data is not only important to control for validating their method but also can generate the most interesting data/conclusions on dissecting the specificity of TCR/BCR repertoire.

  4. eLife

    Reviewer #2 (Public Review):

    What distinguishes most vertebrate species from invertebrates is the adaptive immune response, the ability of T and B lymphocytes to generate billions of antibody or T cell receptor sequences on the chance that some of them will be specific for a pathogen, even those that have never been seen before in evolution. While modern sequencing technologies have given us the ability to read millions of these sequencing from even a single individual, there has not been a good way to identify those attributable to a given infection en masse. In this paper, Ortega and colleagues develop statistical tools that allow one to extract likely SARS-CoV-2 infection-specific antibody and T cell receptor sequences away from the great bulk of irrelevant sequences and show that these are enriched in some that have been previously identified. These methods can be applied to any infection or vaccine response and thus will be very valuable to the field.

  5. eLife

    Reviewer #3 (Public Review):

    This work provides a new statistical framework to model the sharing of antigen receptors among different individuals in healthy or diseased states. The model incorporates probabilities in convergent recombination and convergent selection to fit the complicated process in V(D)J -recombination and clonal selection. In this work, the statistical model has been used in predicting TCR/BCR sharing in healthy individuals and COVID-19 donors and also linked TCR sharing with CMV disease status and ageing.

    This paper will be of interest to the large class of immunologists who are interested in antigen receptor sharing and their functions in infectious diseases and autoimmune diseases.

  6. ScreenIT

    SciScore for 10.1101/2021.12.17.473105: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    NIH rigor criteria are not applicable to paper type.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    86] database using IgBlast [87].
    IgBlast
    suggested: (IgBLAST, RRID:SCR_002873)
    Using a subset of non-productive, naive IgM sequences of high-quality (with at least 3 reads per consensus sequence), we used IGoR [19] to infer the statistics of the generative process, and to build a model Pgen(σ) that can be used to generate synthetic sequences with no mutations, and free of selection effects that affect real productive sequences.
    IGoR
    suggested: (IGOR Pro, RRID:SCR_000325)

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: An explicit section about the limitations of the techniques employed in this study was not found. We encourage authors to address study limitations.

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    Results from scite Reference Check: We found no unreliable references.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  7. Version published to 10.1101/2021.12.17.473105v1 on bioRxiv