Partial block in early γδ T cell development and decrease in Vγ4 cells in HEB-deficient mice.

A. Stages of fetal mouse γδ T cell development. Thymocytes at the DN2 (CD4- CD8-CD44+CD25+) and DN3 (CD4-CD8-CD44-CD25+) stages of T cell development rearrange and express TCRγ, TCR8, and TCRβ genes. DN3 cells with productive TCRβ chains expressing a pre-TCR are directed into the αβ-T cell lineage, characterized by upregulation of CD4 and CD8. Cells that receive a γδTCR signal are directed into the γδ T cell lineage (γδTe, early γδ T cells). Surface expression of γδTCR precedes γδTCR signaling; thus, not all γδTCR+ cells have committed to the γδ T cell lineage, and failure to be triggered by a γδTCR ligand can divert them into the αβ T cell lineage (dotted line). γδTCR+ cells receiving a strong signal become CD73+ γδT1 cell progenitors (γδT1p) that mature into IFNγ producing γδT1 cells, whereas cells that receive an intermediate signal become γδT17 cell progenitors (γδT17p) that mature into IL-17 producing γδT17 cells. Downregulation of CD24 and CD27, and upregulation of CD44, marks maturation of γδT17 cells, whereas γδT1 cell maturation is characterized by downregulation of CD24 and maintenance of CD73 and CD27 expression. B. γδ T cell nomenclature. Vγ and V8 chain genes and proteins can be identified by several different naming systems. Here we use the Tonegawa nomenclature to refer to the proteins, and the International Immunogenetics Information System (IMGT) for the genes and transcripts. R-Seurat generated plots use the Mouse Genome Informatics (MGI) nomenclature. The numbering for genes and proteins in these three systems are identical except for the TRDV4 gene, which encodes the Vγ1 protein (highlighted in red). C. Absolute numbers of cells per thymus in WT (blue) and HEB cKO (orange) E18 fetal mice. D, E. Percentages of γδ T cells in WT and HEB cKO fetal thymus. F. Absolute number of γδ T cells per thymus in WT and HEB cKO fetal thymus. G. Flow cytometry plots of Vγ4+ and Vγ1+ cells within the γδ T cell population in WT and HEB cKO fetal thymus. H. Flow cytometry plot of Vγ5 and Vγ6 (Vγ1-Vγ5-) expression on cells within the Vγ1-Vγ4- population. I. Percentages of Vγ1+, Vγ4+, Vγ5+, and Vγ6+ cells out of all γδ T cells in the WT and HEB cKO fetal thymus. J. Absolute numbers of Vγ1+, Vγ4+, Vγ5+, and Vγ6+ cells per WT and HEB cKO fetal thymus. K. Percentages of immature (CD24+) and mature (CD24-) γδ T cells out of all γδ T cells in each Vγ subset in WT and HEB cKO fetal thymus. L. Expression of IL-17A protein in γδ T cells from E18 thymus stimulated with PMA/ionomycin as assessed by intracellular staining. Blue = WT, orange = HEB cKO. * p < 0.05, ** p < 0.01, *** p < 0.001, *** p <0.0001.

Identification of γδ T cell subsets in WT and HEB cKO fetal thymus by scRNA-seq.

γδ T cells were sorted from E18 fetal thymuses from WT and HEB cKO mice, pooled according to genotype, and subjected to scRNA-sequencing and analysis. A. UMAP plots depicting merged WT and HEB cKO cells in 8 clusters (0-7). B. Grouped UMAP showing the distribution of WT (blue) and HEB cKO (orange) cells across all clusters. C. Split UMAP plots showing the distribution of cells in WT (left) and HEB cKO (right) clusters; note that cluster 4 is restricted to WT cells, and cluster 1 is heavily biased towards HEB cKO cells. D. Genes previously identified as signatures for developmental and functional γδ T cell subsets were compiled from previously published reports. The top ten most differentially expressed genes from this list were visualized as a clustered dot plot which was used to assign cluster identities. Two clusters corresponding to early γδ T cells (γδTe) were randomly named γδTe1 and γδTe2. E. Numbers of WT (blue) and HEB cKO (orange) cells per cluster. F. Unbiased clustered dot plot of the top ten most differentially expressed genes across all clusters. In the clustered dot plots, the percentage of cells expressing the gene in each cluster is depicted by the size of the dot, and the color indicates the relative magnitude of expression across clusters.

The αβ T gene program is inflated and the γδT17 precursor gene program is lost in HEB cKO cells.

γδ T cells were sorted from E18 fetal thymuses from WT and HEB cKO mice were pooled according to genotype and subjected to scRNA-sequencing and analysis. A, B. Gene modules were generated from subset-biased genes (Fig. 2), and cells were scored for each module. Module scores are depicted as split feature plots, with WT plots on the left and HEB cKO plots on the right. Module scores that characterize γδ T cell subsets are shown in (A), and a module score for the αβ-T lineage is shown in (B). C-G. Split violin plots of genes that typify different γδ T cell subsets as follows: C) γδTe/γδT17p cells, (D) αβ T cells, (E) γδT17 cells, (F) γδT1p/γδT1 cells, and (G) γδT1 cells. Blue = WT, orange = HEB cKO.

Decreases in T cell effector differentiation and TCR signaling genes in γδT17p cells from HEB cKO mice.

A. Volcano plots showing differential gene expression in γδT17p cells from WT versus HEB cKO fetal thymus. Genes expressed at higher levels in HEB cKO cells are on the left, and genes expressed at lower levels in HEB cKO cells are on the right. Significance (pink) was set at Log2Fc > 0.5 and Log10P <1025. B. Gene ontology analysis of genes significantly reduced in HEB cKO γδT17p cells relative to WT, with significance set at avg Log2Fc > 0.25 and adjusted P value < 0.001. Bar plots show pathway enrichment (Fold enrichment) and significance by false discovery rate (FDR) for each functional category defined in the KEGG pathway list. Minimum genes for pathway inclusion was set at 5 and FDR cutoff was set at 0.05. C. Relative expression of genes associated with strong TCR signaling in WT and HEB cKO cells in each cluster. D. Relative expression of Id3 in immature γδ T cell subsets from WT and HEB cKO mice. E. Split feature plots showing expression of Id3 across all clusters in WT versus HEB cKO cells. F. Relative expression of Maf and Rorc in WT versus HEB cKO γδT cell subsets. WT = blue, HEB cKO = orange.

Fetal γδ T cells from Id3-KO mice are defective in CD73 upregulation and IL-17 production.

A. Absolute numbers of cells per E18 fetal thymus from WT and Id3-KO littermate mice. B, C. Quantification (B) and flow cytometry plots (C) of the percentages of γδ T cells out of all thymocytes. D, E. Flow cytometry plots (D) and quantification (E) of Vγ1+ and Vγ4+ out of all γδ T cells. F. Percentages of Vγ5+ and Vγ6+ out of all γδ T cells. G. Flow cytometry plots of CD24 and CD73 expression in γδTCR+ cells. H. Quantification of mature (CD24-) CD73+ and CD73- γδ T cells out of all γδ T cells. I. Flow cytometry plots of expression of CD27 and CD73 expression in unstimulated (top) and stimulated (bottom) γδ T cells. J. Percentages of CD27+CD73+ cells out of all γδ T cells under unstimulated or stimulated conditions. K, L. Flow cytometry (K) and quantification (L) of the percentages of CD27-CD73-CD24- (primarily mature Vγ6) cells expressing IL-17 in response to stimulation. Blue = WT, pink = Id3-KO. P/I = phorbol 12-myristate 13-acetate (PMA) + ionomycin. * p < 0.05, ** p < 0.01, *** p < 0.001, *** p <0.0001.

Intact γδ T commitment gene program and impaired γδT17 maturation program in Id3-KO mice.

A. Flow cytometry plots showing the percentages of cells expressing PLZF and/or MAF in immature (CD24+) Vγ4 and Vγ6 cells, and in mature (CD24-) Vγ6 cells (note that mature Vγ4 cells are not present in the E18 fetal thymus). B. Quantification of the percentages of cells expressing PLZF and MAF (top), PLZF only (middle) or neither (bottom) within the immature Vγ4 and Vγ6 subsets, and the mature Vγ6 subset. C. Mean fluorescent intensities of PLZF in the PLZF+MAF+ and PLZF+MAF- populations within the immature and mature Vγ subsets. D, E. sRNA-seq UMAP plots of γδ T cells as merged (D) or split (E) into WT versus Id3-KO populations. F. Number of WT and Id3-KO cells per cluster. G. Clustered dot plot of curated gene sets used to assign γδT17p, γδTe, γδT17, and γδT1 identities. H. Expression of γδ T cell commitment genes in WT versus Id3-KO cells by cluster. I. Expression of γδT17 maturation genes in WT versus Id3-KO cells by cluster. Blue = WT, pink = Id3-KO. ** p < 0.01, *** p < 0.001, *** p <0.0001

Synergistic upregulation of Id3 by HEB and CD3 signaling.

A. ChIP-seq data analysis of the binding of HEB, E2A, RNA polymerase, and Egr2, and the extent of H3K27me3 chromatin modification, in DN3 and/or DN4 cells at the Id3 gene locus, obtained from publicly available datasets (see Materials and Methods for accession numbers). The cell type and antibody used in each experiment are indicated to the right of the tracks. Peaks bound by HEB, E2A, and/or Egr2 are indicated in boxes. Inset shows the Id3 exons and the adjacent Gm42329 long non-coding RNA. B. Diagram of experimental design. SCID.adh cells transduced with HEBAlt or control retroviral vectors were cultured for 16 hours in the presence or absence of the anti-TAC antibody, which induces signaling through the CD3 complex. C, E. Flow cytometry plots (C) and quantification (E) of CD25 upregulation with and without stimulation and/or HEB expression. D. Id3 mRNA expression relative to β-actin as determined by quantitative RT-PCR. Rag = Rag2-/- mouse thymocytes, which are arrested at the DN3 stage of development. ** p < 0.01, *** p < 0.001, *** p <0.0001

Defects in αβ T cell development in E18 fetal thymus of HEB cKO mice.

Thymocytes were dissected from E18 WT and HEB cKO littermates and subjected to flow cytometry. At E18 very few cells had become CD4 or CD8 single positives with most cells at the DP stage in WT mice. The DN to ISP and ISP to DP transitions were severely compromised in the HEB cKO fetal thymus, in agreement with the previously reports of HEB-deficient adult mice.

TRGV and TRDV expression profiling reveals depletion of TRDV4 and TRDV5 transcripts and overexpression TRDV4 in the HEB cKO fetal γδ T cells.

A. Violin plots of expression of canonical genes that mark γδ T cell subsets. B. Violin plots showing expression of TRGV and TRDV genes in WT versus HEB cKO by cluster. WT = blue, HEB cKO = orange. C. Blended split feature plots showing cells expressing TRGV chains (blue) or TRDV chains (red), and cells co-expressing TRGV and TRDV chains (pink). Co-expression in WT cells are shown on the top panel of each comparison and HEB cKO cells are shown on the bottom.

Patterns of E protein and Id protein gene expression during γδ T cell development in WT and HEB cKO mice.

A. Relative expression of Tcf12 (HEB), Tcf3 (E2A), Id3, and Id2 in WT versus HEB cKO cells by cluster. WT = blue, HEB cKO = orange. C-F. Co-expression of E protein and Id protein transcripts assessed by blended split feature plots, for (B) Tcf12 and Tcf3, (C) Tcf12 and Id3, (D) Tcf3 and Id3, (E) Tcf12 and Id2, and (F) Tcf3 and Id2. Blue = E protein gene expression, red = Id gene expression, pink = co-expression.

CD73 is upregulated during development of Vγ5 and Vγ1 γδ T cells in WT but not Id3-KO fetal thymus.

Identification of γδ T cell subsets from E18 WT and Id3-KO DN cells using scRNA-seq.

WT and Id3-KO E18 thymocytes were pooled and subjected to magnetic sorting to obtain CD4-CD8- (DN) cells for scRNA-seq. A. UMAP of merged dataset depicting 11 clusters (0-10). B. Expression of lineage-defining genes to assign identities to clusters in merged dataset: Cd3e for T-lineage, Spi1 (encodes PU.1) for myeloid lineage, Sox13 for γδT17 lineage, Maf for myeloid and γδT17 lineages, and Il2rb and Xcl1 for γδT1 lineage. C. Expression of genes defining DN subsets in merged datasets: Cpa3 for DN2 and γδ T cells, Il2ra (encodes CD25) for DN2/3 cells, Ptcra (encodes pre-Tα) for DN3 cells, and Id3 for γδ T cells and DN4 cells. Cd8b1 is upregulated transcriptionally before surface expression and marks αβ-T lineage commitment within DN4 cells. Cd4 was undetectable, validating our MACS enrichment strategy. D. Expression of Rorc in WT versus Id3-KO cells in γδ T cell subsets. E. Rorc expression in all WT and Id3-KO clusters.

γδTCR+ cells from E18 Id3-KO mice include a population of CD4+CD8+ cells, indicating diversion to the αβ-T lineage program.

E18 fetal thymocytes were subjected to flow cytometry. Cells were gated on the TCRγδ+CD3+ population and analyzed for expression of CD4 and CD8 which was quantified in a bar graph depicting the percentage of CD4+CD8+ (DP) cells within the γδTCR+ population.

Model for HEB and Id3 requirements in the development and maturation of γδT17 cells.

A. Strong γδTCR signaling induces high levels of Egr2, which are sufficient to drive Id3 upregulation without HEB, whereas HEB is also required under lower γδTCR signaling conditions. Distinct cytokine signals also participate in Id3 modulation and γδ T cell lineage choice. B. γδT17 development occurs in two stages, the first of which is HEB-dependent, and the second of which is Id3-dependent. HEB induces Id3 during the first stage, which acts in a negative feedback loop to inhibit HEB activity during the second stage. The absence of Id3 allows higher E protein activity, which inhibits second stage regulators but also results in Id2 upregulation, providing partial compensation for the loss of Id3.