Discriminating mild from critical COVID-19 by innate and adaptive immune single-cell profiling of bronchoalveolar lavages

  1. Els Wauters
  2. Pierre Van Mol
  3. Abhishek Dinkarnath Garg
  4. Sander Jansen
  5. Yannick Van Herck
  6. Lore Vanderbeke
  7. Ayse Bassez
  8. Bram Boeckx
  9. Bert Malengier-Devlies
  10. Anna Timmerman
  11. Thomas Van Brussel
  12. Tina Van Buyten
  13. Rogier Schepers
  14. Elisabeth Heylen
  15. Dieter Dauwe
  16. Christophe Dooms
  17. Jan Gunst
  18. Greet Hermans
  19. Philippe Meersseman
  20. Dries Testelmans
  21. Jonas Yserbyt
  22. Sabine Tejpar
  23. Walter De Wever
  24. Patrick Matthys
  25. CONTAGIOUS collaborators
  26. Maria Bosisio
  27. Michael Casaer
  28. Frederik De Smet
  29. Paul De Munter
  30. Stephanie Humblet-Baron
  31. Adrian Liston
  32. Natalie Lorent
  33. Kim Martinod
  34. Paul Proost
  35. Jeroen Raes
  36. Karin Thevissen
  37. Robin Vos
  38. Birgit Weynand
  39. Carine Wouters
  40. Johan Neyts
  41. Joost Wauters
  42. Junbin Qian
  43. Diether Lambrechts

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

How the innate and adaptive host immune system miscommunicate to worsen COVID-19 immunopathology has not been fully elucidated. Here, we perform single-cell deep-immune profiling of bronchoalveolar lavage (BAL) samples from 5 patients with mild and 26 with critical COVID-19 in comparison to BALs from non-COVID-19 pneumonia and normal lung. We use pseudotime inference to build T-cell and monocyte-to-macrophage trajectories and model gene expression changes along them. In mild COVID-19, CD8 + resident-memory (T RM ) and CD4 + T-helper-17 (T H17 ) cells undergo active (presumably antigen-driven) expansion towards the end of the trajectory, and are characterized by good effector functions, while in critical COVID-19 they remain more naïve. Vice versa, CD4 + T-cells with T-helper-1 characteristics (T H1 -like) and CD8 + T-cells expressing exhaustion markers (T EX -like) are enriched halfway their trajectories in mild COVID-19, where they also exhibit good effector functions, while in critical COVID-19 they show evidence of inflammation-associated stress at the end of their trajectories. Monocyte-to-macrophage trajectories show that chronic hyperinflammatory monocytes are enriched in critical COVID-19, while alveolar macrophages, otherwise characterized by anti-inflammatory and antigen-presenting characteristics, are depleted. In critical COVID-19, monocytes contribute to an ATP-purinergic signaling-inflammasome footprint that could enable COVID-19 associated fibrosis and worsen disease-severity. Finally, viral RNA-tracking reveals infected lung epithelial cells, and a significant proportion of neutrophils and macrophages that are involved in viral clearance.

Article activity feed

  1. Version published to 10.1038/s41422-020-00455-9
  2. ScreenIT

    SciScore for 10.1101/2020.07.09.196519: (What is this?)

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

    Table 1: Rigor

    Institutional Review Board StatementIRB: Ethical approval was obtained from the Research Ethics Committee of KU / UZ Leuven (S63881).
    Consent: All participants provided written informed consent for sample collection and subsequent analyses.
    Randomizationnot detected.
    Blindingnot detected.
    Power Analysisnot detected.
    Sex as a biological variablenot detected.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Single-cell gene expression analysis: Raw gene expression matrices generated per sample were merged and analysed with the Seurat package (v3.1.4)51
    Seurat
    suggested: (SEURAT, RRID:SCR_007322)
    Finally, doublet clusters were identified based on: 1) expression of marker genes from other cell (sub)clusters, 2) higher average UMIs as compared to other (subclusters), and 3) a higher than expected doublets rate (> 20%), as predicted by both DoubletFinder (v2)54 and Scrublet55 and the clustering was re-performed in the absence of the doublet clusters.
    DoubletFinder
    suggested: (DoubletFinder, RRID:SCR_018771)
    Next, this UMAP matrix was fed into SlingShot, with naïve T-cells as a root state for calculation of lineages and pseudotime.
    SlingShot
    suggested: (Slingshot, RRID:SCR_017012)
    Inflammatory pathways and gene set enrichment analysis and tradeSeq: The REACTOME pathway activity of individual cells was calculated by AUCell package (v1.2.4)59.
    REACTOME
    suggested: (Reactome, RRID:SCR_003485)
    The initial application was aimed to identify SARS-CoV-2 reads against thousands of other viruses, and thus the STAR indexes for read alignment were built by combining the human (GRCh38) genome reference with thousands of virus refence genomes from viruSITE.
    STAR
    suggested: (STAR, RRID:SCR_015899)
    Since the likelihood of co-infection with multiple viruses (>2) is low in COVID-19 patients8, we adapted the Viral-Track pipeline to reduce computation time and increase sensitivity.
    Viral-Track
    suggested: None
    Briefly, instead of directly processing raw fastq reads, we took advantage of BAM reads generated for scRNA-seq data, which mapped to human genome by the CellRanger pipeline as described above.
    CellRanger
    suggested: None
    Then the corresponding unmapped BAM reads were extracted using samtools and converted to fastq files using bamtofastq tool to be further processed by UMI-tools for cell barcode assignment before feeding into Viral-Track pipeline.
    samtools
    suggested: (SAMTOOLS, RRID:SCR_002105)
    UMI-tools
    suggested: (UMI-tools, RRID:SCR_017048)
    Cell-to-cell communication of scRNA-seq data: The CellPhoneDB algorithm was used to infer cell-to-cell interactions63.
    CellPhoneDB
    suggested: (CellPhoneDB, RRID:SCR_017054)

    Results from OddPub: Thank you for sharing your data.

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Nevertheless, there are also limitations to our study. For instance, we observed evidence of counter-productive (possibly low-quality) antibody response-related signatures in COVID-19, but failed to perform an in-depth study in this area. Additional studies performing scRNA- and scBCR-seq on serially-collected samples during disease are needed to reinforce this observation. Also, several COVID-19 patients were treated with the antiviral drugs remdesivir, which targets the viral RNA-dependent RNA polymerase, or hydroxychloroquine, which has immunomodulatory traits and is still controversial with respect to its therapeutic effects on disease outcome48–50. Of note, we did not detect major patient-specific cell clusters nor other type of outliers during our analyses. In conclusion, we used single-cell transcriptomics to characterize the innate and adaptive lung immune response to SARS-CoV-2. We observed marked changes in the immune cell compositions, phenotypes as well as immune cross-talks during SARS-CoV-2 infection and identified several distinguishing immunological features of mild versus critical COVID-19. We also documented genetic footprints of several crucial immunological pathways that have been extensively hypothesized, but not always systematically confirmed, to be associated with COVID-19 pathology and SARS-CoV-2 infection biology. We believe that this work represents a major resource for understanding lung-localised immunity during COVID-19 and holds great promise fo...

    Results from TrialIdentifier: We found the following clinical trial numbers in your paper:

    IdentifierStatusTitle
    NCT04327570RecruitingIn-depth Immunological Investigation of COVID-19.

    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.

    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 10.1101/2020.07.09.196519v1 on bioRxiv