Global analysis of tissue Treg transcriptome across different tissue sources

For the bulk RNA-seq analysis, Tregs and Tconv were isolated from blood, and perfused spleen, lymph nodes, kidney, liver, lung, pancreas, and gut (Peyer’s patches, LPLs, IELs) samples. Samples were prepared from Foxp3Thy1.1 reporter mice, allowing purification based on Foxp3 expression. mRNA was sequencing using the QuantSeq 3’mRNA-Seq Library Prep Kit for Illumina and the QuantSeg data analysis workflow. The relationship between samples at the global level is visualised through PCA (using all samples or tissue means), tSNE and UMAP projections.

Plots may be downloaded by right clicking the plot and selecting 'Save image as...'

Bulk RNA-seq data

Expression heat maps

For the bulk RNA-seq analysis, Tregs and Tconv were isolated from blood, and perfused spleen, lymph nodes, kidney, liver, lung, pancreas, and gut (Peyer’s patches, LPLs, IELs) samples. Samples were prepared from Foxp3Thy1.1 reporter mice, allowing purification based on Foxp3 expression. mRNA was sequencing using the QuantSeq 3’mRNA-Seq Library Prep Kit for Illumina and the QuantSeg data analysis workflow. Heatmaps showing the expression of genes across all samples are based on either the 50 most expressed genes (per tissue) or the (maximum 50) most differentially-expressed genes (per tissue, in comparison to the blood Treg population). Select either the most expressed gene or most differentially-expressed gene option, and then select the tissue of choice, to show the heatmap for all samples based on the selected tissue.
Individual plots may be downloaded by right clicking the plot and selecting 'Save image as...'

Visualise enriched KEGG pathways

Tregs and Tconv were isolated from blood, and perfused spleen, lymph nodes, kidney, liver, lung, pancreas, and gut (Peyer’s patches, LPLs, IELs) samples. Samples were prepared from Foxp3Thy1.1 reporter mice, allowing purification based on Foxp3 expression. Bulk mRNA was sequencing using the QuantSeq 3’mRNA-Seq Library Prep Kit for Illumina and the QuantSeg data analysis workflow. Differential analysis was performed between Tregs in the blood, lymphoid tissues (spleen, lymph nodes), non-lymphoid organs (kidney, liver, lung, pancreas) and gut-associated tissues (Peyer’s patches, LPLs, IELs). KEGG pathways enriched for differential expression are visualised using GAGE and Pathview, with each pairwise differential expression set visualised as log2 fold-change for any gene within the pathway with differential expression.

Effect of genetic perturbation on tissue Treg phenotype

High-dimensional flow cytometry phenotyping was performed on tissue Tregs with genetic deficiency in CD69, ST2, KLRG1, Itgae (CD103), CD11a, amphiregulin (Areg), S1PR2, Hif1α, Blimp1 and BATF. The data for CD69, ST2, KLRG1, CD11a, S1PR2, Hif1α, Blimp1 and BATF are presented from 50%:50% mixed bone-marrow chimeras, where wildtype-derived Tregs serve as the internal control for KO-derived Tregs. The data for Areg and CD103 are presented as analysis of KO mice compared to wildtype littermates. Data can be displayed as either cellular phenotypes or marker expression. Cellular phenotype will allow the selection of the genetic perturbation and the tissue, displaying a tSNE-based quantification of the Treg subsets present in wildtype and KO cells. Marker expression will allow the selection of the marker of choice, with expression displayed as a heatmap and as a tSNE expression overlay, across the different tissues, in wildtype and KO Tregs.

TCR repertoire analysis

Tregs were flow sorted from Foxp3Thy1.1 reporter mice, pre-injected with anti-CD45 via intravenous delivery. Cells were purified from blood, kidney, liver, pancreas and LPLs from the gut. Purified cells were CD4+Foxp3Thy1.1+ as well as negative for intravenous CD45 labelling and for the exclusion markers CD19, CD11b, CD8 and F4/80. Cells were labelled with Hastag TotalSeq and loaded onto the 10x Chromium Controller with sequencing performed on Illumina HiSeq. Data is visualised for each cell where TCRbeta CDR3 expression was observed using Chord diagrams. Chord diagrams represent amino acid-level unique clonotypes on the circular axis, with shared clonotypes represented on the radial axis as shared across different tissues in the same mouse (green), shared across different tissues in different mice (black) and shared between the same tissue in different mice (pink).

t-SNE plots across 6 time-points

Leukocytes were isolated from the perfused tissues of mice aged 8 to 100 weeks, stained for flow cytometry and acquired on a BD FACSymphony. The panel included the markers CD103, CD4, CD45, CD62L, CD8a, CD152 (CTLA-4), CD25, CD44, ICOS, CD3, PD-1, CD19, KLRG1, TCR-beta, CD304 (Neuropilin), T-bet, Helios, CD69, NK1.1, ST2, Foxp3, Ki67 and viability. Tregs were gated as viable CD45⁺ CD3⁺ CD4⁺ TCR^- beta⁺ Foxp3⁺ CD8^- CD19^- lymphocytes, and the tSNE analysis was performed on CD103, CTLA-4, CD25, CD44, ICOS, PD-1, KLRG1, Neuropilin, T-bet, Helios, CD69, ST2 and Ki67 using the Cross-Entropy test script in R. The script is available from GitHub.

t-SNE plots across 3 conditions

The impact of the microbiome on Treg phenotype was assessed by high parameter flow cytometry. SPF mice were compared to gnotobiotic (germ-free) and wilded (cohoused) microbiome mice. For the microbiome enrichment, pet store female mice were wild-exposed prior to cohousing with SPF C57BL/6J mice. Leukocytes were isolated from the perfused tissues of mice, stained for flow cytometry and acquired on a BD FACSymphony. The panel included the markers CD103, CD4, CD45, CD62L, CD8a, CD152 (CTLA-4), CD25, CD44, ICOS, CD3, PD-1, CD19, KLRG1, TCR-beta, CD304 (Neuropilin), T-bet, Helios, CD69, NK1.1, ST2, Foxp3, Ki67 and viability. Tregs were gated as viable CD45⁺ CD3⁺ CD4⁺ TCR^- beta⁺ Foxp3⁺ CD8^- CD19^- lymphocytes, and the tSNE analysis was performed on CD103, CTLA-4, CD25, CD44, ICOS, PD-1, KLRG1, Neuropilin, T-bet, Helios, CD69, ST2 and Ki67 using the Cross-Entropy test script in R. The script is available from GitHub.