Mechanisms Of Age-related Changes in Transcriptional Regulation in lymphocytes

National Institute On Aging

Investigators

Linked publications & trials

Abstract

To define transcriptome-based subpopulations of human CD8⁺ T cells, we collected single-cell RNA sequencing (scRNA-seq) data from 53 different tissues and organs, generated using the 10x Genomics platform and publicacal deposited datasets (GSE:126030, 132950, 134578, 148837, 154535, 162180, 181254, 191089, 193449, 206507, 218848, 212966, 213216, 214966, 215405, 216860, 223672, 224143, 232237, 236738, 242477, 250077, 255282, 255690, and 270953, STN49637038, EGAD00001006000, MTAB_11536). Given the variability in quality among publicly available scRNA-seq datasets and the use of different analysis software versions, we first reprocessed all data using CellRanger (v8.01), aligning them to the latest human reference genome (GRCh38.p14) and gene annotation (GENCODE Release 47). This standardization ensured consistent cell identification and updated gene nomenclature across datasets. This high-quality dataset serves as a valuable reference for developing computational pipelines aimed at the standardized classification of human CD8⁺ T cells from any scRNA-seq dataset. To establish a high-quality scRNA-seq dataset, we applied a three-step filtering process: 1) Removal of cells expressing >20% mitochondrial genes, 2) Exclusion of cells with fewer than 1,000 detected genes, and 3) Elimination of potential doublets using the DoubletFinder algorithm. After filtering, we integrated the dataset into a single Seurat object using pc_50/nfeatures_2000. UMAP clustering was performed at resolution 0.3. To annotate the clusters, we utilized scRNA-seq data from sorted human CD8⁺ T cell subsets, including naïve, memory stem cells (SCM), central memory (CM), effector memory (EM), and effector memory CD45RA⁺ (EMRA) T cells. Additionally, we incorporated reported transcriptomic features of known human CD8⁺ T cell subsets. Through this approach, we identified twelve distinct subpopulations of human CD8⁺ T cells, including: Naïve, SCM, CM, Effector memory subsets: EM-GZMK, EM-BATF, EMRA-GZMH, EMRA-CSTW, Resident memory (RM) subsets: RM-KLRB1, RM-HSPA1A, Exhausted subsets: EX-CST3, and Effector subsets: Eff-LDHA and Eff-MKI67. We are in the process to further characterize their function by isolating them using flow cytometry cell sort and in vitro functional assessment. In summary, we have assembled a human CD8⁺ T cell scRNA-seq dataset comprising over one million cells from 53 different tissues and organs from published data and our own unpublished data, identifying twelve distinct subpopulations with unique transcriptomic features. This dataset serves as a foundation for further functional characterization of human CD8⁺ T cells and also provides a valuable reference for developing computational pipelines to enable the standardized classification of human CD8⁺ T cells from any scRNA-seq datasets.