Nature Immunology Paper

Bioinformatics analysis

Back to Main Page

Paper info

  • The transcriptional cofactor Tle3 reciprocally controls effector and central memory CD8+ T cell fates. (Nature Immunology, 2024. X.Zhao, W.Hu, S.R.Park, S.Zhu, S.S.Hu, C.Zang, W.Peng, Q.Shan ,H.H. Xue. co-first (Zhao, W.Hu, Park and Zhu as co-first authors with equal contribution) (https://doi.org/10.1038/s41590-023-01720-w)


Introduction

  1. Antigen-Experienced CD8+ T Cells Differentiation:
    • TEM (Effector Memory T cells)
    • TCM (Central Memory T cells)
  2. Role of Tle3 in Cell Fate Regulation:
    • Tle3 is a transcriptional cofactor that critically regulates the fates and lineage stability of TEM and TCM cells through dynamic redistribution in the genome of antigen-responding CD8+ T cells.
  3. Effects of Tle3 Ablation:
    • Genetic ablation of Tle3 promotes the formation of CD8+ TCM cells at the expense of CD8+ TEM cells.
    • Tle3-deficient CD8+ TEM cells show accelerated conversion into CD8+ TCM cells while maintaining a robust recall capacity.
  4. Functional Interactions of Tle3:
    • Previously Tle3 was known as a corepressor.
    • Acts as a coactivator for Tbet, enhancing chromatin opening at sites characteristic of CD8+ TEM cells and activating transcription of signature genes for CD8+ TEM cells.
    • Interacts with Runx3 and Tcf1 to restrain features characteristic of CD8+ TCM cells.
  5. Overall Role of Tle3:
    • Integrates functions of multiple transcription factors to maintain the lineage fidelity of CD8+ TEM cells.
    • Manipulation of Tle3 activity could be leveraged to favor the production of CD8+ TCM cells.

Fig2

Loss of Tle3 enhances CD8+ TCM cell formation



Fig3

Tle3 promotes TEM and suppresses TCM signature genes



Fig4

De novo Tle3 binding promotes TEM cell chromatin opening



Fig5

Tle3 engages Tbet to promote TEM features



Extended Data Figures

Extended Data Fig2



Extended Data Fig3



Extended Data Fig4



Method

Single Cell RNA-seq Data Processing and Cell Clustering Steps

  1. Data Processing:
    • Raw reads processing: Using Drop-seq Laboratory Protocol v3 and Drop-seq Tools (v2.4.1).
    • Alignment: Aligned to mm10 mouse genome using STAR (v.2.7.9a).
    • Gene expression files: Generated using Drop-Seq Tools (v2.4.1).
    • UMI Calculation: Performed using Seurat (v4.1.2).
  2. Quality Control:
    • Cells with UMI < 1,000 and mitochondrial content > 2.5% were excluded.
  3. Normalization and Transformation:
    • Normalized UMI counts to count-per-million total counts and log-transformed.
  4. Clustering and Dimension Reduction:
    • Used 10 principal components for UMAP manifold learning and clustering analyses.
    • Clustering with Seurat’s FindClusters function (resolution parameter = 0.5).
  5. Marker Gene Identification:
    • Performed using Seurat’s FindAllMarkers function.
  6. Data Visualization of gene expression:
    • Normalized gene expression shown in feature plots or violin plots.
    • Scaled expression data of cluster marker genes in heat maps.
  7. Trajectory Analysis:
    • Performed using Monocle (v3) to learn a principal graph and order cells along it using pseudotime function.
  8. Signature Gene Definition:
    • TEM1 and TEM2 clusters combined as TEM cells, compared with TCM cluster.
    • Criteria: ≥1.5 expression changes and FDR < 0.05.
    • Identified 96 TCM and 53 TEM signature genes.
  9. Impact Characterization of Tle3 Deficiency:
    • Pooled WT and Tle3−/− memory CD8+ T cells for UMAP clustering.
    • TCM and TEM scores calculated using Seurat’s AddModuleScore function and mapped on UMAP or as scatterplots with ggplot2.

RNA-seq and Data Analysis Process

  1. Sample Collection and Preparation:
    • WT and Tle3−/− P14 TCM and TEM Cells: Sorted from recipient mice at 30–35 days post-infection (d.p.i.) in three biological replicates.
    • Ex Vivo Cultured Cells: WT and Tle3ΔCreET TEM cells re-sorted for viable cells, collected in three replicates.
    • RNA Extraction and Library Construction: Using TRIzol RNA isolation reagents (Invitrogen/Thermo Fisher Scientific).Performed by Azenta Life Sciences using SMART-Seq HT Ultra Low Input Kit (Clontech/Takara Bio) for cDNA synthesis, and Nextera XT DNA Library Kit (Illumina) for library preparation.
  2. Sequencing and Data Conversion:
    • Sequencing and Data Conversion: Performed on Illumina HiSeq (4000 or equivalent) in paired-end mode with 150 nucleotide read length.Raw .bcl files converted to fastq and de-multiplexed using bcl2fastq 2.17 (Illumina).
    • Data Deposited: Bulk RNA-seq data deposited at GEO under SuperSeries GSE213041.
  3. Quality Control and Read Mapping:
    • Quality Assessment: Using FastQC (v0.11.9).
    • Read Trimming: Trimmomatic (v.0.39) used to trim 50 bp from the 3′ end to remove low-quality bases.
    • Read Mapping: Mapped to the mm10 mouse genome using STAR-2.7.9a, retaining pairs with MAPQ ≥ 30 and ends mapped to the same chromosome.
  4. Gene Expression Analysis:
    • Expression Matrix: Constructed using featureCounts (v.2.0.1).
    • Expression Estimation: DESeq2 (v1.32.0) used to estimate gene-level FPKM values.
  5. Differential Expression and Clustering Analysis:
    • DEGs Identification: Criteria of ≥1.5-fold changes, FDR < 0.05, and FPKM ≥ 0.5 in the higher-expression condition.
    • K-means Clustering: Applied to define dynamic gene expression patterns from DEGs in three key comparisons.
  6. Signature Analysis and GSEA:
    • Signature Definition: Based on comparative analysis between WT TCM and TEM cell transcriptomes.
    • GSEA: Used to measure relative enrichment of custom gene sets in WT and Tle3-deficient TEM cells.

CUT&RUN Experiment and Data Analysis

  1. Sample Collection and CUT&RUN Protocol:
    • Samples: Naive WT CD8+ T cells, WT Teff (8 days post-infection), WT TEM and TCM cells (≥30 days post-infection), and various deficient conditions for Tbet and Runx3.
    • Replicates: Three biological replicates for naive cells, two for most other conditions, and specific counts for deficient condition experiments.
    • Protocol: Improved protocol with minor modifications for genome-wide mapping of Tle3, Tbet, and Runx3 binding sites.
    • Library Preparation and Sequencing:
      • Quantified using KAPA Library Quantification kit (Roche).
      • Sequenced on Illumina HiSeq 4000 in paired-end mode with 150 nucleotide read length.
    • Data Depositing:
      • CUT&RUN data deposited at GEO under SuperSeries GSE213041.
  2. Data Processing:
    • Quality Assessment: FastQC v0.11.9 used to evaluate the sequencing quality.
    • Read Trimming: Trimmomatic v0.39 used to retain 36 bp from the 5′ end of sequences.
    • Read Alignment: Bowtie2 v2.4.4 aligned sequencing reads to the mm10 mouse genome, retaining only uniquely mapped reads (MAPQ ≥ 30).
    • File Conversion and Sorting:
      • SAMtools v1.13 used to convert SAM files to BAM and sort them.
    • Duplicate Removal:
      • Picard MarkDuplicates 2.26.0 used to remove duplicate reads in BAM files.
    • Peak Calling:
      • MACS v2.2.7.1 used for Tle3 peak calling in paired-end mode (FDR < 0.05).
      • Peaks called using MACS2 applied to pooled reads from biological replicates.

Reproducibility Analysis and Dynamic Tle3 Binding Cluster Identification (C&R peak analysis)

  1. Union Peaks Creation: Merged 34,292 union peaks from nine biological replicates across four cell types.
  2. Normalization: Peaks normalized by length per kilobase, libraries by column sum per million; conducted PCA analysis.
  3. Dynamic Peak Identification: Generated a 34,292 × 9 matrix for Tle3 binding peaks; used DESeq2 (v.1.32.0) with stringent criteria (≥3-fold changes, FDR < 0.1, peak signal score ≥ 0.7).
  4. Clustering: Clustered 3,627 dynamic Tle3 binding peaks using k-means based on expression profiles.
  5. Peak Calling: Criteria included ≥2-fold enrichment, FDR < 0.05, excluding peaks found in IgG CUT&RUN samples.
  6. Comparative Analysis: Analyzed Tle3 binding strength between WT, TRKO early Teff cells, and WT, TbetKO early Teff cells.
  7. Annotation: Used ChIPseeker’s ‘annotatePeak’ to annotate peaks with promoter regions defined as ±1-kb around the TSS.
  8. Motif Analysis: Conducted using HOMER (v4.10.0), extracted top motifs including statistics (P value and percentage of targets).
  9. Functional Annotation: Employed GREAT (v.4.0.4) to annotate genomic regions with biological functions.

ATAC-seq Experiment and Data Analysis

  1. Sample Collection and Library Preparation:
    • Cells: WT and Tle3−/− TEM and TCM cells sorted in 2–3 biological replicates; WT or Tle3ΔCreET TEM cells re-sorted for viability.
    • ATAC-seq Library Sequencing: Libraries quantified, sequenced on Illumina HiSeq 4000 in paired-end read mode (150 nucleotides).
  2. Data Processing:
    • Quality Assessment: FastQC v0.11.9.
    • Read Trimming: Trimmomatic (v0.39) retained 36 bp from the 5′ end.
    • Alignment: Bowtie2 (v2.4.4) aligned reads to mm10 mouse genome; only uniquely mapped reads (MAPQ ≥ 30) retained.
    • Duplicate Removal and File Processing: SAMtools used to convert and sort BAM files; Picard MarkDuplicates removed duplicates.
    • Peak Calling: MACS (v2.2.7.1) for ATAC-seq peak calling, peaks pooled from biological replicates.
  3. Reproducibility Analysis:
    • Union Peaks Creation: 55,158 union peaks created from data across cell types/states, defined as union ChrAcc sites.
    • Normalization: Each peak normalized by length per kilobase, libraries normalized by column sum per million; PCA analysis conducted.
  4. Identification of Differential ChrAcc Clusters:
    • Analysis: DESeq2 (v1.32.0) used to identify differential ChrAcc sites between cell types/states.
    • Criteria for Signature ChrAcc Sites: ≥2-fold changes, FDR < 0.05, ChrAcc site signal score ≥0.15.
    • Stringent Criteria for Tle3 Deficiency Impact: ≥3-fold changes, FDR < 0.05, ChrAcc site signal score ≥0.15.
    • Clustering: 5,553 differential ChrAcc sites clustered into eight distinct clusters using k-means clustering.

Association of Tle3 Binding Sites and Differential ChrAcc Sites with DEGs (C&R + ATAC-seq + RNA-seq Integration)

  1. Assignment Criteria:
    • Proximity to TSS: Sites within ±100-kb of the gene transcription start site (TSS) considered associated with that gene.
    • Promoter-associated Sites: Sites within ±1 kb of the TSS designated as promoter-associated.
    • Distal Sites: Sites beyond ±1 kb from the TSS designated as distal.
  2. Gene Association:
    • A single Tle3 binding or ChrAcc site can be associated with multiple unique Refseq genes, reflecting the complex regulatory landscapes of these genomic regions.

Data Availability

  • GEO Accession Number: The Tle3 CUT&RUN, bulk RNA-seq, single-cell RNA-seq, and ATAC-seq data from CD8+ T cells are available at the Gene Expression Omnibus (GEO) under the accession number GSE213041.


Thank you for reading!