Three distinct developmental pathways for adaptive and two IFN-γ-producing γδ T subsets in adult thymus

Abstract

Murine γδ T cells include subsets that are programmed for distinct effector functions during their development in the thymus. Under pathological conditions, different γδ T cell subsets can be protective or can exacerbate a disease. Here we show that CD117, CD200 and CD371, together with other markers, identify seven developmental stages of γδ T cells. These seven stages can be divided into three distinct developmental pathways that are enriched for different TCRδ repertoires and exhibit characteristic expression patterns associated with adaptive (γδTn), IFN-γ-producing (γδT1) and IFN-γ/IL-4-co-producing γδ T cells (γδNKT). Developmental progression towards both IFN-γ-producing subsets can be induced by TCR signalling, and each pathway results in thymic emigration at a different stage. Finally, we show that γδT1 cells are the predominating IFN-γ-producing subset developing in the adult thymus. Thus, this study maps out three distinct development pathways that result in the programming of γδTn, γδT1 and γδNKT cells.

Fig. 1

CD117, CD200 and CD371 are differentially expressed during γδ T cell development. a Gating strategy for analysis of surface makers within the total TCRδ⁺, the Vγ1.1⁺ and the Vγ2⁺ subsets. b Heat map of genes encoding surface-expressed proteins (GO:0009986 or GO:0005886) that were significantly differentially expressed between CD24⁺CD73⁻ and CD24⁺CD73⁺ cells. Data from public data set GSE75920 (three to four individual experiments from sorted cells, each pooled from 30 mice). Expression of genes marked in bold was verified by flow cytometry. Expression of red and blue genes is shown in c–e, and expression of remaining bold genes is shown in Supplementary Fig. 1. c–e Representative flow cytometric plots showing expression of c CD117, d CD200 and e CD371 on TCRδ⁺-gated thymocytes as histograms (left), distribution of marker negative and positive cells within development stages defined by CD73 and CD24 (centre) and expression of markers within development stages of the Vγ1.1⁺ and Vγ2⁺ subsets (right)

Fig. 2

Surface expression of CD117, CD200 and CD371 allows the isolation of seven distinct γδ thymocyte populations a, b t-SNE map of TCRδ⁺ thymocytes coloured by a normalised expression intensities of CD24, CD25, CD73, CD117, CD200 and CD371 or by b FlowSOM automated clustering. Based on flow cytometric analysis of 2 × 5000 TCRδ⁺ cells pooled from two mice in silico. c Heat map of mean expression intensity of markers within the FlowSOM clusters named A to G. d Distribution of cells within populations A to G shown as a percent of the total TCRδ⁺ thymocytes from 1- to 28-week-old mice. Bars depict the mean ± SEM (n = 4, 4, 4, 2, 3 mice). e, f Distribution of cells within populations A to G within the Vγ1.1⁺ and Vγ2⁺ subsets of TCRδ⁺ thymocytes visualised as e t-SNE map coloured by FlowSOM clusters as in b, f stacked bar plots. Bars depict the mean ± SEM (n = 2 mice). g Standard bi-axial TCRδ⁺ gating of population A to G. Numbers denote the mean ± SEM of total TCRδ⁺ thymocytes (n = 2 mice). h, i, j Expression of established surface markers of γδ T cells within populations A to G visualised as h representative histograms or i, j representative t-SNE maps (n = 6 mice). h Dashed lines show background signal by FMO. i, j Colours depict i the location of events assigned to populations A to G within the t-SNE plot or j the normalised expression intensity of each marker

Fig. 3

Developing γδ T cells branch into populations with distinct effector characteristics a, b Pairwise Euclidian distances between each of the A to G populations calculated from whole transcriptome RNA-Seq expressions within a the Vγ1.1⁺ (V1A–V1G) and b the Vγ2⁺ (V2A–V2G) subset. Numbers and colours denote the mean Euclidian distances. c, d Diagrams of the shortest distances between the A to G populations, associating the most transcriptionally similar populations by lines within c the Vγ1.1⁺ and d the Vγ2⁺ subset. Red numbers denote the Euclidian distance between the two connected populations. e Principal component analysis (PCA) of RNA-Seq expression within the A to G populations of the Vγ1.1⁺ and Vγ2⁺ subsets. Lines correspond to the lowest pairwise distances shown in c, d. f Cellular progression predicted by flow cytometric expression of CD24, CD25, CD73, CD117, CD200 and CD371 by Isomap. Each A to G populations was reduced to 100 cells before analysis. Each dot represents a single cell. g Heat map showing expression of key transcription factors and effector molecules related to different γδ T cell effector subsets within the A to G populations of the Vγ1.1⁺ (V1A–V1G) and the Vγ2⁺ (V2A–V2G) subset. h Abundance of known TCRγ and TCRδ CDR3 motifs associated with different γδ effector subsets within the A to G populations of the Vγ1.1⁺ and the Vγ2⁺ subset. Numbers and colour depict the mean percentage of RNA-Seq reads including the indicated CDR3 motifs. Gene expression data are analysed as the mean expression from two independent RNA-Seq libraries, each constructed from RNA isolated from 24 mice across three independent cell sorting runs

Fig. 4

Progression through the D, E and F populations is induced by TCR signalling. a–c Changes in surface marker expression of TCRδ⁺ cells from sorted population A to G cells after 2 days of culture on OP9-DL1 monolayers in the presence or absence of immobilised anti-CD3. The data are visualised as a histograms and b bar plots of the percentage of cells positive for each individual marker as well as c gated into the A to G populations. Bars depict the mean ± SEM from three independent experiments with cells sorted from four to eight mice. Statistical analyses were performed using the paired two-sided Student’s t test, with significance defined as *p < 0.05; **p < 0.01; ***p < 0.001. d Heat map showing expression of genes reported to be induced by γδTCR signalling during development within the A to G populations of the Vγ1.1⁺ (V1A–V1G) and the Vγ2⁺ (V2A–V2G) subset. e TCRδ repertoire of cells within the A to G populations of the Vγ1.1⁺ and the Vγ2⁺ subset. Numbers and colour depict the mean percentage of RNA-Seq reads aligned to the TCRδ locus from two independent RNA-Seq libraries, each constructed from RNA isolated from 24 mice across three independent cell sorting runs

Fig. 5

γδ T cells emigrate from the thymus at the C, E and G stages a, b TCRδ⁺ thymocyte distributions within population A to G after treatment with FTY720 for 5 (5d) or 10 days (10d) visualised using a t-SNE maps and b stacked bar plots of cell numbers within TCRδ⁺, Vγ1.1⁺ and Vγ2⁺ subsets. c Representative CD24 expression within A to G populations after treatment with FTY720 for 5 or 10 days. d, e Expression of CD24 within CD73⁻ and CD73⁺ TCRδ⁺ cells from inguinal lymph nodes after FTY720 treatment for 5 and 10 days as d representative flow cytometry plots and e quantitative bar plots showing the percent of TCRδ⁺ cells within each population. Data from two independent experiments for both 5- and 10-day treatments (n = 4 mice per group per time point). Numbers within gates and bar plots depict the mean ± SEM. Statistical analyses were performed using the unpaired two-sided Student’s t test

Fig. 6

γδT1 cells, but not IL-17A⁺ γδ T cells, accumulate when thymic emigration is inhibited. a–c Expression of IFN-γ and IL-17A within TCRδ⁺ cells after treatment with FTY720 for 5 (5d) or 10 days (10d). a Representative flow cytometry contour plots and normalised quantification of the fold change in b IFN-γ⁺ and c IL-17A⁺ TCRδ⁺ cells (n = 4). d, e Co-expression of IFN-γ with PLZF (n = 2), NK1.1 (n = 4) or CD45RB (n = 4) within TCRδ⁺ cells after treatment with FTY720 for 5 or 10 days. d Flow cytometry plots and e normalised quantification of the fold change within TCRδ⁺IFN-γ⁺ cells. f–i Percentage of cells within the A to G populations expressing f, g Vδ6.3/2 within the Vγ1.1⁺ subset and h, i Vδ4 within the Vγ2⁺ subset after treatment with FTY720 for 5 (n = 2) or 10 (n = 4) days. Bars depict the mean ± SEM from two independent experiments for both 5 and 10 days (n = 2–4 mice per group per time point). Statistical analyses were performed using the unpaired two-sided Student’s t test

Fig. 7

γδTn, γδT1 and γδNKT cells develop through three distinct pathways in the adult thymus. Suggested model of γδ T cell development divided into the A to G stages defined by the expression of CD117, CD200 and CD371 together with CD24, CD25 and CD73. The A stage defines the earliest γδTCR-expressing progenitors, which develop towards the B and C populations and eventually result in the export of TCR-naive adaptive γδ T cells (γδTn). Encounter with cognate TCR ligands induces TCR selection, shifting the cells to progress through the D and F stages. TCR-selected γδ thymocytes will then progress to the E or G stage, resulting in the export of IFN-γ-producing (γδT1) or IFN-γ/IL-4-co-producing (γδNKT) cells, respectively. The C, E and G stages are characterised by the distinct expression of key factors involved in effector programming, including characteristic transcription factor networks and cytokines, and display highly focused TCR Vδ repertoires. IL-17-producing (γδT17) cells do not develop in the adult thymus, but long-lived resident γδT17 cells are included in the G population at homoeostasis and can be distinguished by their expression of CCR6