Predicted impact of the viral mutational landscape on the cytotoxic response against SARS-CoV-2

  1. Anna Foix
  2. Daniel López
  3. Francisco Díez-Fuertes
  4. Michael J. McConnell
  5. Antonio J. Martín-Galiano

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

The massive assessment of immune evasion due to viral mutations that increase COVID-19 susceptibility can be computationally facilitated. The adaptive cytotoxic T response is critical during primary infection and the generation of long-term protection. Here, potential HLA class I epitopes in the SARS-CoV-2 proteome were predicted for 2,915 human alleles of 71 families using the netMHCIpan EL algorithm. Allele families showed extreme epitopic differences, underscoring genetic variability of protective capacity between humans. Up to 1,222 epitopes were associated with any of the twelve supertypes, that is, allele clusters covering 90% population. Next, from all mutations identified in ~118,000 viral NCBI isolates, those causing significant epitope score reduction were considered epitope escape mutations. These mutations mainly involved non-conservative substitutions at the second and C-terminal position of the ligand core, or total ligand removal by large recurrent deletions. Escape mutations affected 47% of supertype epitopes, which in 21% of cases concerned isolates from two or more sub-continental areas. Some of these changes were coupled, but never surpassed 15% of evaded epitopes for the same supertype in the same isolate, except for B27. In contrast to most supertypes, eight allele families mostly contained alleles with few SARS-CoV-2 ligands. Isolates harboring cytotoxic escape mutations for these families co-existed geographically within sub-Saharan and Asian populations enriched in these alleles according to the Allele Frequency Net Database. Collectively, our findings indicate that escape mutation events have already occurred for half of HLA class I supertype epitopes. However, it is presently unlikely that, overall, it poses a threat to the global population. In contrast, single and double mutations for susceptible alleles may be associated with viral selective pressure and alarming local outbreaks. The integration of genomic, geographical and immunoinformatic information eases the surveillance of variants potentially affecting the global population, as well as minority subpopulations.

Article activity feed

  1. Version published to 10.1371/journal.pcbi.1009726
  2. ScreenIT

    SciScore for 10.1101/2021.07.04.451040: (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
    A matrix with all inter-family Jaccard coefficients was used for agglomerative hierarchical clustering by clustermap function of seaborn data visualization Python library with default options.
    Python
    suggested: (IPython, RRID:SCR_001658)
    Mutation analyses: Mutations respect to the proteins of the reference Wuhan-1 strain (RefSeq: NC_045512.2) were identified by aligning with Clustal Omega 1.2.1 (40) with an in-house perl script.
    Clustal Omega
    suggested: (Clustal Omega, RRID:SCR_001591)

    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: We detected the following sentences addressing limitations in the study:
    Bioinformatic approaches suffer from intrinsic limitations. These include the possible application of biologically inappropriate thresholds and potentially low predictive performance. Furthermore, alleles considered in algorithms as much as the priceless genome sampling by the worldwide sequencing effort still represent an underestimation of biological variability. Such obstacles were addressed in this study by: (i) utilizing an state-of-the-art algorithm that permits nearly universal fine-grained predictions (~3000 alleles); (ii) the application of stringent cutoffs that reflect the natural strictness of the ligand-HLA binding; (iii) the re-calculation of peptide binding affinity for each mutation; and (iv) the utilization of a large dataset of ~118,000 viral genomes and their corresponding metadata. Mutations were stratified by occurrence, reduction of HLA-binding affinity and geographical dissemination. Thus, the integration of omic data and immunoinformatics in this study very likely capture, despite drawbacks, the principal trends that respond to the posed questions. Large epitope numbers were computationally predicted to be presented by most supertypes. Although all these supermotifs appeared mutated in at least one isolate, most of these mutations did not overcome the supermotif degeneracy. In most cases, the HLA binding affinity was reasonably maintained except from (i) residue substitutions in the second and C-terminal positions of the ligand core, amino acids that u...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We found bar graphs of continuous data. We recommend replacing bar graphs with more informative graphics, as many different datasets can lead to the same bar graph. The actual data may suggest different conclusions from the summary statistics. For more information, please see Weissgerber et al (2015).

    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.

  3. Version published to 10.1101/2021.07.04.451040v1 on bioRxiv