PREreview of Single-cell transcriptome and T cell receptor profiling of the tuberculin skin test

Brief summary of the study

This study provides a first-time characterization of human immune response after the tuberculin skin test (TST). To this aim, the authors performed scRNA-seq, ADT-seq and TCR sequencing of skin suction blisters from 31 patients carrying immunological memory to Mtg antigens during the TST. By sequential clustering and Louvian clustering, the authors resolved 101 clusters and derived gene signatures to distinguish different cell types, generating a context-specific and cell-type-specific peudobullk transcriptome dataset that can be used as a future reference to deconvolute TST biopsy. The analysis reveals a T cell-dominated response, highlighting the roles of various T cells and myeloid cells subsets, and the absence of B cells during TST. These data help to improve our understanding of the immune response to Mtb and enable further exploration of bulk transcriptomic data through context-specific cellular deconvolution. Despite the significant biological and clinical interest this work entailed, there are some questions and concerns particularly related to the single-cell RNA seq analysis of T cell subsets and how the analysis compares to the previous characterization of TST immune response. These concerns as well as our suggestions on how to address them are presented below. Major and minor issues are presented separately. 

Major comments 

• The use of single-cell technologies in this study allows for a detailed understanding of the immune landscape of TST. This understanding can be further strengthened by using independent datasets or alternative methods (e.g., flow cytometry) for cell type validation and function annotations. The authors mentioned references 16-18, which provide valuable insights from previous flow cytometry analyses of skin suction blisters from the site of the TST. An explanation of these previous insights, along with a comparison and contrast between previous studies and current single-cell RNA and TCR sequencing data, would further validate the results. For example, in Figure 1A-B, comparing the percentages of different CD4 T cell populations with references 16-18 could guide the reader into a comparison between current and previous studies.

• In Figures 1-2, most of the subclusters are transcriptomically similar. For example, in Figure 2B, the authors identified only five cell types among 37 cytotoxic T cell sub-clusters. The gradual change in cytotoxic activity among clusters may indicate their developmental trajectory towards becoming cytotoxic cells. Could there be any trajectory relationships between different T cells? For instance, are naive T cells becoming cytotoxic T cells or memory T cells? The authors could perform a trajectory analysis to elucidate the relationships between different clusters. This would develop a stronger understanding of the roles of naive T cells and their relationships with other T cell types.

• In Figure 2, the authors can consider visualizing the distribution of TCR clones in response to the Mtb. Do they observe clonal selection of certain TCR clones? What are the transcriptome profiles (and inferred phenotypes) for the dominant TCR clones?

• In Figure S3, the authors described that expressions of type 1 IFN genes were not detected in the dataset. However, probing for evidence of IFN activity using expression of multi-gene signatures specifically induced by either type 1 or type 2 IFNs provided evidence for both type 1 and type 2 IFN activity. What could explain the discrepancy between the lack of IFN gene expression and the multi-gene signature analysis? Could there be a time-dependent response of IFN that is not captured in the dataset at the specific time points?

Minor comments

• Adding schematics of experimental design at the beginning of all figures can help the authors to explain (a) how they collect samples from 31 patients (time points), and (b) how they sub-cluster the data into 101 clusters. More descriptive labels and figure legends can provide further guidance for the readers to understand the complex datasets. 

• Figure 1. The authors could visualize the scRNA-seq data in a UMAP or t-SNE plot with annotation of broadly classified cell types. It helps visualize the general distribution of T types (e.g. T vs B cells) and cell states (e.g. cytotoxic vs naive T cells)

• Figure S1. The authors could label the cluster number with corresponding cell types in figure S1, which will make the figure more self-explanatory.

• Figure S2. The authors may want to include mitochondrial percentages.

• Figure 2B. The authors may want to visualize the cell type distribution using a bar chart or a pie chart (i.e., naive vs cytotoxic vs others), as the number of clusters might confuse readers regarding the actual number of cells.

• Some details in the methods section should be included, such as which surface proteins were selected for ADT sequencing. Additionally, highlighting the selected genes used to define the TST blister signature in the main text would help readers better understand the TST immune response.

• To provide the clinical context, the introduction can elaborate on the significance of the TST as a model for studying TB immunity and its relevance to clinical practice. 

• The discussion can expand on the potential implications of atypical CD8 T cells in TB pathology and Mtb immune evasion strategies and address the absence of B cells in the TST response considering its broader implications for Mtb immunity. 

• The authors can add additional references to describe "atypical" CD8 T-cells, which provide further context for the readers. The interpretation of the high proportion of cytotoxic T-cells as a dominant feature of the TST response is compelling but can be factored in the potential variations among individuals. 

Comments on reporting - information on the statistical analyses or availability of data.

• The statistical analyses are well reported and the supplementary data are comprehensive. 

Suggestions for future studies

• Future studies can investigate the temporal dynamics of the TST response by including additional time points to provide insights into the evolution of the immune response and roles of the different cell types over time. To gain deeper insights into the immune response and functionality of T-cells, investigating beyond the 2-day duration used in this study may be recommended. Functional validation through in vitro and in vivo assays would also strengthen the findings. Comparative studies between individuals with varying Mtb exposure or different TB states (latent versus active) could provide valuable insights into immune correlates of protection and disease progression.

Conflicts of interest of reviewers

• There are no conflicts of interest to declare.

Competing interests

The authors declare that they have no competing interests.