CD8⁺Treg scRNA-seq analysis

The repository contains code of scRNA-seq data analysis of the CD8+HLA-DR+ regulatory T cells as part of a research project "Transcriptional signatures and age-related changes in CD8+HLA-DR+ regulatory T cells" at the Bioinformatics Institute 2023/2024

Introduction

First described in 2014, the CD8⁺HLA-DR⁺ regulatory T lymphocytes (CD8⁺Treg) subset is known for its role in suppressing effector T cells through checkpoint inhibitor molecules, sharing certain features with conventional CD4⁺Treg cells. Despite this, the detailed nature and function of these cells are still not well understood. Understanding this subset is particularly significant in light of age-related alterations in the immune system and the heightened vulnerability of CD8⁺ T lymphocytes to such changes. Preliminary findings indicated that the CD8⁺HLA-DR⁺ population differentiates into two subpopulations based on CD127 (IL7R) surface expression. This led to defining the CD8⁺Treg phenotype as CD3⁺CD8⁺HLA-DR⁺CD127^low. This research aimed to identify transcriptional signatures and examine age-related changes in gene expression within the CD8⁺Treg population, utilizing publicly accessible single-cell RNA-seq (scRNA-seq) data.

Workflow

Data: Single-cell RNA sequencing (scRNA-seq) data from peripheral blood mononuclear cells (PBMCs) of 982 donors, aged 19 to 97, from the OneK1K cohort and 99 healthy donors, aged 22 to 75, from the SLE study were analyzed.

The scanpy package was used for data analysis:

1. Data preprocessing:
   • Normalization;
   • Logarithmization;
   • Scaling;
   • Dimensionality reduction (via PCA);
   • Batch correction (harmony);
   • Leiden clustering, UMAP representation.
2. Automatic cell types annotation with celltypist was employed to annotate the cell clusters and identify CD8⁺ effector T cells.
3. CD8⁺ T cell subsets selection.
4. Return raw counts to the CD8⁺ T cells to better represent them in reduced dimensionality embedding.
5. Repeat preprocessing steps for the CD8⁺ T cells data.
6. Manual annotation of CD8⁺Treg cells based on increased expression of HLA-DRA, HLA-DRB1, and HLA-DRB5 genes and decreased expression of IL7R gene.
7. Differential gene expression analysis (DEA) of CD8⁺Treg at the single-cell level.
8. DEA and gene set enrichment analysis (GSEA) of CD8⁺Treg at the pseudobulk level with edgeR, clusterProfiler and gprofiler2 R packages.

The workflow scheme is represented below:

Results

1. The CD8⁺Treg transcriptional signatures

To determine CD8⁺Treg signatures, scRNA-seq data from donors aged 19 to 60 years (n = 299) were analyzed. According to previous experimental results, T cells in this age range typically do not exhibit distinct age-related changes.

It should be noted that cells potentially corresponding to CD8⁺Treg were not separated into a distinct cluster using the Leiden algorithm but were distributed among CD8⁺ effector T cells. Therefore, the CD8⁺Treg population was defined by their phenotype as CD8⁺ effector T cells with increased expression of HLA-DRA, HLA-DRB1, and HLA-DRB5 genes and decreased expression of IL7R gene (CD127):

DEA at the single-cell level revealed that the CD8⁺Treg population, compared to the main population of CD8⁺ effector T cells, exhibited increased expression of genes associated with MHC-II-mediated antigen presentation, cytotoxicity, and cytoskeletal organization. At the pseudobulk level, GSEA based on the GO terms also characterized the CD8⁺Treg population by increased expression of genes involved in antigen presentation, particularly those mediated by MHC class II molecules.

The single-cell-level DEA and pseudobulk-level GSEA results are represent below respectively:

GO BP terms associated with antigen processing and presentation marked with green frame at GSEA dotplot.

2. Age-related changes in the CD8⁺Treg subset

To assess age-associated changes at the transcriptomic level, CD8⁺Treg populations from young (20-35 years, n = 152) and old (70-97 years, n = 424) donors were analyzed. DEA at the pseudobulk level among older donors showed a shift in the CD8⁺Treg transcriptional profile towards a terminally differentiated phenotype. Specifically, a decrease in the expression of cytotoxic molecules such as LYZ, GZMA, and GZMK was observed, while the expression of GZMH, characteristic of terminally differentiated CD8⁺ T lymphocytes, increased:

Additionally, CD8⁺Treg were characterized by increased expression levels of MHC molecules, as confirmed by GSEA:

Furthermore, GSEA results indicated a decrease in the expression of genes involved in RNA biosynthesis and metabolism, as well as T cell differentiation, activation, and immune response with age. However, no increase in the expression of genes associated with exhaustion, such as PD-1, Tim3, Lag3, TIGIT, CD160, and CD244, was observed. It can be inferred that with age, the CD8⁺Treg population transitions to a terminally differentiated phenotype and exhibits a decreased functional response capacity, but this population does not undergo cellular exhaustion.

Conclusion

Using the scRNA-seq data, we observed that the CD8⁺Treg subpopulation is a heterogeneous group of CD8⁺ effector T lymphocytes. This subpopulation shows increased expression of genes associated with cytotoxicity, cytoskeletal rearrangement, and MHC-II-mediated antigen presentation. This would suggest that CD8⁺Treg-mediated suppression likely involves cell-contact dependent cytolysis of target cells. In older adults, CD8⁺Treg cells exhibit changes in marker gene expression indicative of a terminally differentiated cell phenotype. Despite this shift, there is no evidence of increased expression of exhaustion markers. However, there is a decrease in the expression of genes regulating key processes of T cell activation and function, suggesting that the suppressor function of CD8⁺Treg decreases with age.

Contributors

K. Matveeva^1,2, S. Kolmykov², D. Shevyrev²

• ¹ Bioinformatics Institute, Kantemirovskaya st. 2A, 197342, St. Petersburg, Russia
• ² Sirius University of Science and Technology, Olympic Ave., 1, 354340, Sirius, Russia

Contacts

For any questions or suggestions, please feel free to reach out to Kseniia Matveeva at email or Telegram✨