Multimodal profiling reveals site-specific adaptation and tissue residency hallmarks of γδ T cells across organs in mice

du Halgouet, Anastasia; Bruder, Kerstin; Peltokangas, Nina; Darbois, Aurélie; Obwegs, David; Salou, Marion; Thimme, Robert; Hofmann, Maike; Lantz, Olivier; Sagar

doi:10.1038/s41590-023-01710-y

Download PDF

Resource
Open access
Published: 04 January 2024

Multimodal profiling reveals site-specific adaptation and tissue residency hallmarks of γδ T cells across organs in mice

Nature Immunology volume 25, pages 343–356 (2024)Cite this article

10k Accesses
18 Altmetric
Metrics details

Subjects

Abstract

γδ T cells perform heterogeneous functions in homeostasis and disease across tissues. However, it is unclear whether these roles correspond to distinct γδ subsets or to a homogeneous population of cells exerting context-dependent functions. Here, by cross-organ multimodal single-cell profiling, we reveal that various mouse tissues harbor unique site-adapted γδ subsets. Epidermal and intestinal intraepithelial γδ T cells are transcriptionally homogeneous and exhibit epigenetic hallmarks of functional diversity. Through parabiosis experiments, we uncovered cellular states associated with cytotoxicity, innate-like rapid interferon-γ production and tissue repair functions displaying tissue residency hallmarks. Notably, our observations add nuance to the link between interleukin-17-producing γδ T cells and tissue residency. Moreover, transcriptional programs associated with tissue-resident γδ T cells are analogous to those of CD8⁺ tissue-resident memory T cells. Altogether, this study provides a multimodal landscape of tissue-adapted γδ T cells, revealing heterogeneity, lineage relationships and their tissue residency program.

FOXO1 is a master regulator of memory programming in CAR T cells

Article Open access 10 April 2024

FOXO1 enhances CAR T cell stemness, metabolic fitness and efficacy

Article Open access 10 April 2024

A single-cell atlas enables mapping of homeostatic cellular shifts in the adult human breast

Article Open access 28 March 2024

Main

Receptor-based classification of T cells segregates them into αβ and γδ T cell lineages. Deciphering the nonredundant functions of γδ T cells compared to their αβ counterparts is a matter of intense research^1,2,3. While αβ T cells constitute a major fraction of T cells in murine organs, γδ T cells dominate in numbers in some tissues, for example, epidermis and the small intestine⁴. γδ T cells are important in various physiological processes and maintain normal tissue functions at steady state^5,6,7,8. During immune threats, γδ T cells carry out diverse functions in different tissues ranging from direct killing of the infected cells to neutrophil recruitment and enhanced antigen presentation^9,10,11. They can also take on protective roles by downmodulating the inflammatory response and promoting tissue repair^{5,12,13,14,15,16,17,18}. However, it is unclear whether these functions correspond to distinct γδ subsets or to a few subsets exhibiting context-dependent activities depending upon their microenvironment. Deciphering this conundrum requires systematic profiling of γδ T cells across multiple organs at single-cell resolution.

Additionally, understanding the tissue adaptation features of γδ T cells can provide essential insights to promote protective immunity. It has been shown that CD8⁺ tissue-resident memory T (T_RM) cells, mucosal-associated invariant T (MAIT) cells and natural killer T (NKT) cells share a common transcriptional program of tissue residency^19,20. By contrast, the existence of a tissue residency program for γδ T cells remains unknown. Using parabiotic mice, it is well established that the skin, intestine, liver and adipose tissue host mainly tissue-resident γδ T cell populations^8,21,22,23. However, parabiosis experiments in these studies were performed using flow cytometry, providing limited insights into their tissue residency features. Hence, deciphering the tissue adaptation features of γδ T cells in multiple organs combining parabiosis with single-cell transcriptomics holds the key to identifying specific tissue-resident γδ subsets and their underlying regulatory programs.

Here, utilizing multimodal single-cell sequencing, we profiled γδ T cells across seven organs and grouped them into eight subsets delineated by the expression of Sell, Ly6c2, Cd160, Gzmb, Rorc, Areg and Klrg1 and cell cycling genes. Single-cell profiling of cell surface proteins further revealed the functional capacities of these identified subsets. Further, we revealed that epidermal and intestinal γδ T cells exhibit a homogenous transcriptional profile and showcase open chromatin at gene loci associated with diverse functions. Single-cell T cell antigen receptor (TCR) profiling revealed the lineage relationships of the identified subsets. Furthermore, using parabiotic mice, we uncovered the cellular and molecular hallmarks of γδ T cell tissue residency across organs. Moreover, comparing the tissue residency features of γδ T cells to those of T_RM and NKT cells revealed analogous transcriptional programs of tissue residency. Altogether, our data represent a cross-organ single-cell multimodal landscape of γδ T cells and provide a highly resolved map of their tissue residency features.

Results

Heterogeneity and tissue-adapted features of γδ T cells

To investigate the tissue-specific heterogeneity and site-adapted features of γδ T cells in mice, we used a single-cell multimodal approach and sorted γδ T cells from various organs using flow cytometry (Fig. 1a,b, Extended Data Fig. 1a and Supplementary Note). Sorted cells were subjected to single-cell RNA sequencing (scRNA-seq), single-cell assay for transposase accessible chromatin during sequencing (scATAC-seq), single-cell TCR sequencing and cell surface profiling using TotalSeq. Integration, clustering and visualization of gene expression data were performed using Harmony and Seurat^24,25 (Extended Data Fig. 1b–e and Supplementary Note). We identified 22 clusters in the dataset, revealing substantial heterogeneity in the γδ T cell compartment across tissues (Fig. 1c). Cells from different tissues contributed differentially to these clusters, indicating site-specific adaptation (Fig. 1d–f). Principal component analysis (PCA) based on average gene expression profiles revealed that γδ T cells from the skin, small intestine and large intestine are remarkably distinct compared to cells from other organs (Fig. 1g).

**Fig. 1: Multi-organ γδ T cell profiling reveals heterogeneity and site-specific adaptation.**

Next, we characterized the identified γδ T cell clusters using differentially expressed genes (Supplementary Table 1). We grouped 22 clusters into 8 major subsets with unique gene expression profiles (Fig. 2a–c). Sell⁺Ly6c2⁻ (clusters 0, 7 and 11) and Sell⁺Ly6c2⁺ (clusters 2 and 8) subtypes exhibited the highest expression of genes associated with lymphocyte migration (Sell and S1pr1) and maturation (Tcf7 and Lef1)^{26,27,28,29,30} (Fig. 2d). Clusters 4, 6, 10, 12 and 14 were classified as Cd160⁺ γδ T cells (Fig. 2a,b,d). CD160 has been shown to control interferon-γ (IFN-γ) secretion by natural killer (NK) cells³¹. We reasoned that Cd160⁺ γδ T cells may represent a distinct site-adapted subset of IFN-γ-producing cells across different organs (Fig. 2e). Clusters 1 and 9 mainly consisted of small and large intestinal γδ intraepithelial lymphocytes (IELs) expressing Cd8a and Itgae (encoding CD103)³², categorized as Gzmb⁺ IELs due to their cytotoxic phenotype (Gzma and Gzmb; Fig. 2d). Clusters 3, 5 and 19 consist of interleukin-17-producing γδ T (γδT17) cells, characterized by the expression of Rorc (encoding RORγt), Il17a and Zbtb16, which controls the development of Vγ6⁺ γδT17 cells^33,34 (Heilig and Tonegawa nomenclature³⁵; Fig. 2d). We termed these cells Rorc⁺ γδ T cells. Interestingly, skin Rorc⁺ γδ T cells (mainly cluster 5) clustered separately from their counterparts in other organs (cluster 3) and exhibited a distinct transcriptional signature (Fig. 1c,d, Extended Data Fig. 1f and Supplementary Table 2). Moreover, skin harbored a transcriptionally distinct cluster 13 characterized by the expression of Areg, Ctla2a and Gem (Fig. 2c–e). We denoted this subtype as the Areg⁺ γδ T cell subset. This cluster represents dendritic epidermal T cells (DETCs) previously described as expressing Areg and playing a tissue repair function following injury^36,37. We further identified two minor groups of γδ T cells—an effector-like subset (cluster 17, Klrg1⁺) and proliferating cells (cluster 15)^38,39 (Fig. 2b,d and Extended Data Fig. 1g). Taken together, these results demonstrate that γδ T cells constitute a heterogeneous group of site-specific T cells that are highly adapted to their local microenvironment.

**Fig. 2: Eight major site-adapted γδ subsets across tissues.**

Maturation states and functional capacities of γδ subsets

We next refined the phenotypes of identified γδ subsets using a panel of TotalSeq antibodies (Fig. 3a,b and Extended Data Fig. 2). Sell⁺Ly6c2⁻ and Sell⁺Ly6c2⁺ cells also expressed CD62L (encoded by Sell), indicating concordance between gene expression and cell surface phenotype (Fig. 3c). Sell⁺Ly6c2⁻ γδ T cells expressed the highest levels of CD62L and CD24, an immature T cell marker (Fig. 3c). Meanwhile, CD122, a marker often associated with IFN-γ-producing cells⁴⁰, was highly expressed by Sell⁺Ly6c2⁺ cells (Fig. 3c)⁴¹. Both subsets expressed CD27, which is also associated with IFN-γ-producing cells⁴² (Fig. 3c). Notably, Cd160⁺ γδ T cells expressed NK-1.1, validating our gene expression-based classification, as this marker has been associated with IFN-γ production^43,44 (Fig. 3c). Gzmb⁺ IELs exhibited the highest levels of CD8a (Fig. 3c). The cell surface expression of KLRG1 was specific to Klrg1⁺ cells (Fig. 3c). Rorc⁺ cells expressed cell surface markers associated with the γδT17 cell lineage, such as CD44, ICOS and CCR6 (Fig. 3c). Interestingly, a fraction of Rorc⁺ cells also expressed Nkp46, classically expressed by NK cells⁴⁵ (Fig. 3c). Furthermore, we used cell surface proteins for dimensionality reduction, identifying many γδ subsets solely through these markers, highlighting the functional relevance of our gene expression-based classification (Fig. 3d). We sought to further validate the markers used to classify the γδ subtypes utilizing antibodies against LY6C and CD160 as well as the RORγt-GFP reporter line⁴⁶. We first assessed the suitability of identified genes to define the eight γδ subsets. We observed that the expression of these genes was exclusive to the defined subtypes, indicating their appropriateness to classify γδ T cell subsets (Fig. 3e and Extended Data Fig. 3). We further quantified the fraction of these subsets in each organ using scRNA-seq as well as flow cytometry revealing a remarkable similarity in the calculated fraction of γδ T cells between both methods (Fig. 3f–i and Extended Data Fig. 4). Overall, flow cytometry analysis validated the gene expression-based quantification of Sell⁺Ly6c2⁻, Sell⁺Ly6c2⁺, Cd160⁺ and Rorc⁺ γδ subsets across organs.

**Fig. 3: Single-cell epitope profiling of γδ T cell subsets.**

Epigenetic features and functional diversity of γδ subsets

To understand the underlying regulatory networks of γδ cell states, we simultaneously profiled gene expression and chromatin accessibility in γδ T cells across different organs. Clustering based on DNA accessibility using Signac⁴⁷ classified them into 17 clusters (Fig. 4a and Extended Data Fig. 5a–d). Importantly, the gene expression data from the same cells were already integrated into the transcriptome-based clustering shown in Fig. 1c, allowing us to map the transcriptome-based cell-type annotation on the chromatin accessibility-based clustering depicted in Fig. 4a. Assigning each scATAC-seq cluster to one of the eight γδ subsets indicated high consistency between the gene expression and chromatin accessibility (Fig. 4b and Extended Data Fig. 5e). We further looked at the differentially accessible peaks in each cluster (Fig. 4c and Supplementary Tables 3 and 4). Clusters 0, 5 and 10, which were classified as Sell⁺Ly6c2⁻ cells, also exhibited open chromatin regions associated with Tcf7, Lef1, Sell and Cd24a (Fig. 4a–c and Extended Data Fig. 5f). Cluster 2, consisting of Sell⁺Ly6c2⁺ cells, exhibited higher accessibility at the Ifng and Tbx21 loci necessary for IFN-γ production⁴⁸ (Fig. 4c and Extended Data Fig. 5g). Similarly to the gene expression-based classification, Cd160⁺ γδ T cells exhibited heterogeneity in their chromatin accessibility landscape and were subdivided into three different subsets—clusters 3, 7 and 4—based on their tissue of origin, indicating site-specific adaptation of this subset (Fig. 4a,b). Clusters 1, 9 and 12, classified as Gzmb⁺ IELs, exhibited open regions across Cd8a, Kit, Gzma, Gzmb and Gzmk loci (Fig. 4c). Uniquely, γδ IELs displayed differentially open chromatin regions across genes encoding various interleukins (Fig. 4c and Extended Data Fig. 6a,b). Notably, we did not detect their transcripts (Extended Data Fig. 6c,d). This suggests that although IELs do not express these cytokines at steady state, they might have the propensity to express them during infection or inflammation which could explain their previously reported protective role^49,50. Based on the chromatin accessibility features, Rorc⁺ γδ T cells were substantially heterogeneous and grouped into five different clusters—clusters 8, 11, 15, 16 and 14. Cluster 14 specifically comprised skin γδT17 cells and exhibited several similar chromatin accessibility features as the Areg⁺ γδ subset, indicating skin-specific adaptation (Fig. 4c). Cluster 15 showed open chromatin at gene loci belonging to both IFN-γ-associated and IL-17-associated programs (Rorc, Il17a and Tbx21; Extended Data Fig. 6e–g). Finally, the skin-derived Areg⁺ γδ T cell subset revealed a unique chromatin accessibility landscape with the highest number of differentially regulated open chromatin regions (Fig. 4c and Extended Data Fig. 6h). These included several genes encoding interleukins and growth factors without evident transcription (Fig. 4c and Extended Data Fig. 6i–n). This suggests that the Areg⁺ γδ subset has the propensity to release factors that may have an important role in promoting tissue repair and wound healing³⁷.

**Fig. 4: Cross-tissue single-cell chromatin landscape of γδ T cell subsets.**

To identify transcription factor (TF) motifs that characterize γδ subsets, we used chromVAR⁵¹. TF motifs belonging to the TCF/LEF family were enriched in clusters 0, 5 and 10 composed of Sell⁺ Ly6c2⁻ cells (Fig. 4d and Extended Data Fig. 7a). Chromatin regions associated with the T-box family of TFs were enriched in clusters 2 and 3 containing Sell⁺Ly6c2⁺ and liver Cd160⁺ cells, respectively (Fig. 4d and Extended Data Fig. 7b). Cluster 7, representing splenic Cd160⁺ cells, displayed specific motif enrichment of the ETS family of TFs (Fig. 4d). The IRF/STAT family of TFs was enriched in two clusters of intestinal γδ IELs (clusters 1 and 9; Fig. 4d and Extended Data Fig. 7c). Another intestinal cluster, cluster 12, had specific enriched motifs of NFAT family TFs (Fig. 4d). Interestingly, Rorc⁺ γδ T cells displayed substantial heterogeneity in enriched TF motifs of RAR-related orphan receptors, GATA3, GATA4, Krüppel-like and nuclear factor-kB family members (Fig. 4d and Extended Data Fig. 7d). Finally, Areg⁺ cells exhibited enriched TF motifs from the RUNX and AP1 families (Fig. 4d and Extended Data Fig. 7e). Altogether, single-cell chromatin accessibility profiling reveals distinct epigenetic features of tissue-adapted γδ T cell subsets (Extended Data Fig. 7f). Moreover, we uncovered that epidermal and intestinal γδ T cells are rather transcriptionally homogeneous, showing simultaneous epigenetic features associated with cytotoxicity, cytokine production, tissue repair and wound healing, indicating functional diversity.

TCR diversity and lineage relationships of γδ subsets

Next, we performed gene expression and TCR repertoire profiling of γδ T cells across organs (Fig. 5a–c, Extended Data Fig. 8a,b and Supplementary Table 5). Using the Shannon diversity score, we observed that the spleen and lymph nodes possessed a more diverse T cell receptor Gamma (TRG) repertoire compared to other organs (Fig. 5d and Extended Data Fig. 8c). Sell⁺Ly6c2⁻ cells (clusters 0, 7 and 13) had small clones with the highest diversity score, highlighting its naïve features (Fig. 5e and Extended Data Fig. 8d). Sell⁺Ly6c2⁺ (clusters 4 and 6) and Cd160⁺ (clusters 3, 8, 11 and 12) subsets comprised medium-sized clones (Fig. 5e). Gzmb⁺ γδ IELs (clusters 2 and 5) were mainly Vγ7⁺, with a few cells exhibiting Vγ1 chain rearrangement, consistent with previous reports (Fig. 5f)^52,53. Of particular interest was the Rorc⁺ subset, which exhibited tissue-specific heterogeneity and formed four clusters (clusters 1, 9, 10 and 18; Fig. 5a,c). Clusters 1 and 10 contained Rorc⁺ cells with Vγ2 and Vγ4 usage (Fig. 5f). Cluster 9 consisted of Vγ6⁺ cells primarily from the liver and lung, exhibiting highly expanded TCR clonotypes and the lowest diversity score (Fig. 5c,e,f and Extended Data Fig. 8d). These Vγ6⁺ cells presented a distinct gene expression signature with upregulation of Cxcr6, genes associated with chronic TCR stimulation (Pdcd1 and Tox) and similar genes to those expressed by DETCs (Areg and Gem; Fig. 5g).

**Fig. 5: TCR features and lineage relationships of γδ subsets across tissues.**

Next, we analyzed the six top expanded clones by plotting them on the uniform manifold approximation and projection (UMAP) representation (Fig. 5h). The first two belonged to Rorc⁺ cells exclusively, showing early lineage segregation (Fig. 5h). The third mainly came from Gzmb⁺ IELs and overlapped exclusively with intestinal Cd160⁺ cells suggesting a shared developmental origin (Fig. 5h). Further analysis of all detected clonotypes in Gzmb⁺ IELs revealed their maximum overlap exclusively with intestinal Cd160⁺ cells, emphasizing the uniqueness of this lineage (Fig. 5i). The next three highly expanded clones were predominantly scattered among Sell⁺Ly6c2⁻, Sell⁺Ly6c2⁺ and Cd160⁺ cells (Fig. 5h). This pattern hints toward a common developmental origin of these three subsets. Interestingly, the fifth highly expanded clonotype was shared among all cell subsets except Rorc⁺ cells, including Gzmb⁺ IELs (Fig. 5h). This finding suggests that Gzmb⁺ IELs with Vγ1 rearrangement and a transcriptional signature similar to the Vγ7 chain might share ontogeny with other γδ subsets. Moreover, the presence of these six highly expanded clones across all organs hints toward the existence of expanded precursors from the thymus seeding different tissues (Fig. 5j). Lastly, quantifying the clonal overlap among the profiled organs revealed that liver and lymph node clones exhibit the highest and least overlap with other tissues, respectively (Extended Data Fig. 8e,f). In summary, TCR clonotype analysis provided valuable insights into the distinct origins of Gzmb⁺ and Rorc⁺ subsets, while also hinting toward a common origin for Sell⁺Ly6c2⁻, Sell⁺Ly6c2⁺ and Cd160⁺ cells.

Tissue residency programs of γδ T cells across organs

To understand the tissue residency features of γδ T cells across different organs, we conducted parabiosis experiments, profiling partner-derived and host-derived γδ T cells across six tissues using scRNA-seq (Fig. 6a). Importantly, the profiled cells were already integrated into the UMAP representation depicted in Fig. 1c. Here, our focus was on single cells from parabiotic mouse pairs (Fig. 6b,c). To distinguish tissue-resident cells from exchanging γδ T cell populations at higher resolution, we utilized the Milo framework, which models cellular states as overlapping neighborhoods on a k-nearest neighbor (KNN) graph⁵⁴. Using this approach, we identified 1,366 neighborhoods that were differentially abundant in partner- and host-derived γδ T cells (Fig. 6d). The skin exhibited mostly tissue-resident neighborhoods, followed by the liver (Fig. 6e). The large intestine, lymph nodes and spleen also contained several tissue-resident neighborhoods (Fig. 6e). The lung, on the other hand, had very few tissue-resident neighborhoods (Fig. 6e). Independent flow cytometry data confirmed these observations (Fig. 6f and Extended Data Fig. 9a,b). To explore the transcriptional programs associated with tissue residency in each organ, we conducted differential gene expression analysis between partner-derived and host-derived neighborhoods. In lymph nodes, genes regulating proliferation (Mki67), the expression of integrin subunits (Itga1 and Itgal) and the γδT17 program (Maf and Rorc) were associated with tissue residency (Fig. 6g). In the spleen and liver, molecular programs associated with cytotoxicity (Gzmb and Gzma) and innate-like IFN-γ production (Ifng, Klrb1c and Cd160) were mainly tissue resident (Fig. 6h,i). In contrast to lymph nodes, many neighborhoods associated with the γδT17 signature (Rorc and Cd163l1) comprised exchanging cellular states in the liver (Fig. 6i). In the large intestine, both cytotoxicity (Gzmb) and γδT17 (Rorc) signatures were associated with tissue residency (Fig. 6j). Importantly, akin to circulating memory CD8⁺ T cells, Tcf7, Lef1, Klf2, Sell and S1pr1 were enriched in exchanging γδ states across organs (Fig. 6g–j). In summary, we reveal that cytotoxicity and rapid IFN-γ production-related molecular programs are general hallmarks of γδ tissue residency, while Tcf7, Lef1 and Klf2 define the core transcriptional program of circulating γδ T cells.

**Fig. 6: Parabiosis reveals molecular features of γδ tissue residency across organs.**

Subset-specific residency features of γδ T cells

To explore the subset-specific tissue residency characteristics of γδ T cells, we performed automatic grouping of γδ neighborhoods and identified 15 groups (Fig. 7a). Groups 5, 6, 7, 9 and 10 were enriched in host-derived neighborhoods, while the rest comprised exchanging cells (Fig. 7b). Groups 5 and 6 consisted of Areg⁺ cells from the skin and Cd160⁺ cells from the liver, respectively, and were highly resident (Fig. 7b–d). Liver and spleen CD160⁺ cells were confirmed as more tissue-resident than CD160⁻ cells via flow cytometry (Fig. 7e,f). Group 7, containing skin Rorc⁺ cells, was also more abundant in resident neighborhoods, unlike Rorc⁺ cells from other organs (group 3; Fig. 7b–d), confirmed by flow cytometry (Fig. 7g). RORγt-GFP⁺ cells from epithelial tissues like lung and liver were not completely exchanging (Fig. 7h), suggesting distinct exchanging and resident γδ subsets in these tissues. We separately analyzed Rorc⁺ cells from the lung and liver, identifying common genes characterizing γδT17 tissue residency (Fig. 7i–k). Resident γδT17 cells exclusively expressed Sdc4 and Pdcd1 (encoding PD-1; Fig. 7l). PD-1 has been shown previously to be expressed by a subset of γδT17 cells that are Vγ6⁺ and display a T_RM phenotype⁵⁵. Next, we investigated γδ subsets in the skin. In the skin, two major γδ subsets were identified: Rorc⁺ (group 7) and Areg⁺ (group 5) cells (Fig. 7c,d). Rorc expression was absent in Areg⁺ cells, indicating the RORγt-GFP⁻ compartment should represent the highly resident Areg⁺ skin cells (Fig. 7m). Indeed, we detected very few partner-derived RORγt-GFP⁻ cells in the skin of parabiotic pairs of mice (Fig. 7m). Group 9, comprising Gzmb⁺ IELs, was also enriched in tissue-resident neighborhoods (Fig. 7b–d). On the other hand, groups 2, 11 and 12, consisting of Sell⁺Ly6c2⁻ and Sell⁺ Ly6c2⁺ cells, were mainly exchanging, supported by flow cytometry analysis (Fig. 7b–d and Extended Data Fig. 9c,d). Overall, we identified γδ subset tissue residency features and their associated molecular programs across tissues.

**Fig. 7: Cell-type-specific residency features of γδ subsets across tissues.**

Analogous programs of γδ and T_RM tissue residency

Next, we sought to systematically compare the tissue residency features of γδ T cells to those of CD8⁺ T_RM cells. Unlike T_RM cells, we did not identify a universal transcriptional program for γδ T cell tissue residency. To bridge this gap, we conducted a supervised analysis using key factors responsible for T_RM formation and maintenance to identify the correspondence between the tissue residency features of γδ T cells and T_RM cells. T_RM cells across many tissues are CD69⁺CD103⁺⁵⁶. Furthermore, human T_RM cells express CD49a encoded by Itga1⁵⁷. We found various γδ subsets expressing Cd69, Itgae and Itga1 (Fig. 8a–d). Gut-resident Gzmb⁺ IELs expressed all three markers, while the liver-resident Cd160⁺ γδ subset expressed Cd69 and Itga1 (Fig. 8a–d). The skin-resident Areg⁺ subset expressed mainly Itgae (Fig. 8a–d). Circulating Rorc⁺, Sell⁺Ly6c2⁻ and Sell⁺Ly6c2⁺ subsets did not express Itgae and Itga1 and exhibited open chromatin and transcriptional programs associated with circulating memory T cells (Fig. 8d and Extended Data Fig. 10). These results indicate that features of T_RM tissue residency are also present in tissue-resident γδ T cells, albeit with some tissue-specific variability. Runx3 has been shown to be a key regulator of T_RM differentiation⁵⁸. Importantly, all tissue-resident γδ subsets expressed Runx3, indicating that Runx3 may play a similar role in establishing γδ tissue residency (Fig. 8d). Moreover, Hobit (Zfp638) and Blimp1 (Prdm1) are also central regulators of various tissue-resident lymphocyte lineages²⁰. In γδ subsets, Hobit expression was restricted to the liver-resident Cd160⁺ γδ T cells, while Blimp1 expression was mainly identified in intestinal Gzmb⁺ γδ IELs (Fig. 8e,f). Few cells in the liver-resident Cd160⁺ compartment also expressed Blimp1 (Fig. 8f). Chromatin accessibility assessment of the regions encoding Hobit and Blimp1 further supported these findings (Fig. 8g,h). These results indicate that, unlike in T_RM and NKT cells, Hobit and Blimp1 may play tissue-specific roles in establishing γδ tissue residency.

**Fig. 8: Analogous tissue residency programs of γδ and T_RM cells.**

Finally, apart from investigating the role of a few marker genes and TFs in establishing γδ tissue residency, we examined the core gene expression signatures defining T_RM residency^20,58 in resident and exchanging γδ neighborhoods (Supplementary Table 6). Our analysis revealed that neighborhoods abundant in tissue-resident γδ T cells across organs exhibited significantly higher expression of T_RM core residency genes, while circulatory memory signatures were more abundant in circulating γδ neighborhoods (Fig. 8i,j). These findings indicate that although there is heterogeneity in the factors required for the establishment of tissue residency among γδ subsets, the overall hallmarks are analogous between γδ and T_RM cells. Overall, these analyses provide a highly resolved view of transcriptional programs governing γδ T cell tissue residency in comparison to their αβ counterparts.

Discussion

While several studies have explored the heterogeneity of γδ T cells at single-cell resolution^{23,36,55,59,60,61}, a comprehensive cross-organ multimodal study detailing their site-specific adaptation and tissue residency features was still lacking. In this study, we provide a multimodal landscape of γδ T cells across several epithelial tissues and lymphoid organs. We demonstrate that γδ T cells in epithelial tissues are epigenetically and transcriptionally unique subsets that are highly adapted in barrier organs. γδ subsets in secondary lymphoid organs substantially differ from their counterparts residing in barrier tissues.

Previous studies have suggested that DETCs and intestinal IELs share several common features in maintaining epithelial barrier integrity and promoting wound healing and regeneration upon damage and inflammation⁴. We clearly demonstrate that both γδ subsets have open chromatin loci associated with a distinct set of interleukins and growth factors, although most of them are not transcribed at steady state. Therefore, we argue that DETCs and intestinal IELs may exert their regenerative roles through the production of interleukins following immune challenge. Although DETCs express Areg and Il13 as described previously³⁶, we did not detect the expression of KGF (encoded by Fgf7) or other growth factors in these cells. However, chromatin regions encoding several growth factors including Fgf7 were significantly open in DETCs, indicating that they may be poised to synthesize these growth factors upon activation during injury. Paradoxically, although implicated in tissue repair functions, intestinal IELs exhibit cytotoxic features and display an open Ifng locus. Therefore, the mechanisms through which DETCs and intestinal IELs strike a balance between limiting pathogens through cytotoxicity while still fostering pro-repair properties remain to be elucidated³². Our data negate the existence of multiple γδ subsets performing these distinct functions and suggest a context-dependent and interleukin-based mechanism, as shown in skin carcinogenesis⁶².

Using the parabiosis mouse model, we lay out a highly resolved single-cell map of tissue residency features of γδ T cells across organs. We did not find a common universal transcriptional program associated with γδ tissue residency. While Itgae was restricted to skin-resident and gut-resident γδ T cells, Itga1 and Hobit were predominantly expressed in liver-resident γδ T cells. Liver T_RM subsets have also been shown to be CD103⁻ and CD49a⁺^20,63. Notably, Runx3 was uniformly expressed in all tissue-resident γδ T cells. Although Runx3 is required for the development of DETCs and regulates CD103 expression⁶⁴, its role in establishing γδ tissue residency in other organs has not been explored. While single genes and TFs linked to T_RM formation displayed tissue-specific regulation in distinct γδ subsets, the core signatures associated with T_RM, NKT and MAIT cell tissue residency closely resemble those of all tissue-resident γδ T cells, suggesting the existence of a core genome-wide transcriptional program associated with tissue residency across all lymphocyte lineages. Furthermore, the circulatory programs associated with effector and central memory αβ T cells (for example, Klf2 and S1pr1) were strikingly similar to circulating γδ T cells.

Although our study primarily focuses on γδ T cells in mice, it highlights several parallels between γδ T cells in mice and humans. For instance, in both mice and humans, tissue-resident γδ T cells in the liver are characterized by the expression of CD49a and CD69⁶⁵. Furthermore, in both species they exhibit the expression of Gzmb and Blimp1, and demonstrate restricted TCR diversity^65,66. In the intestine, γδ IELs in both mice and humans primarily express Itgae, Gzma and Gzmb⁶⁷. Moreover, in humans, the peripheral blood contains a subset of naïve γδ T cells that express TCF7 and LEF1, and these cells exhibit a diverse TCR repertoire similar to Sell⁺Ly6c2⁻ naïve cells observed in mice⁶⁸. A thorough single-cell multimodal profiling of human γδ T cells across diverse organs, combined with the findings in this study, can yield valuable insights into γδ T cell biology across species and clues in utilizing them to enhance protective immunity in diseases.

Methods

Mice

Experiments were performed using mice from two different animal facilities—Max Planck Institute of Immunobiology and Epigenetics in Freiburg (Germany) and Curie Institute in Paris (France). Parabiosis experiments were performed in Paris. C57BL/6J mice in Freiburg were obtained from in-house breeding and were kept in the animal facility of the Max Planck Institute of Immunobiology and Epigenetics in specific-pathogen-free conditions with a 12-h light/12-h dark cycle, a temperature range of 20–23 °C and 60% humidity. All animal experiments were performed in accordance with the relevant guidelines and regulations and approved by the review committee of the Max Planck Institute of Immunobiology and Epigenetics and the Regierungspräsidium Freiburg, Germany. For the experiments performed in Paris, CD45.1/1 and CD45.1/2 animals were generated in-house by crossing CD45.1/1 B6 animals with CD45.2/2 RORγt-GFP B6-MAIT^CAST mice. All experiments were conducted in an accredited animal facility by the French Veterinarian Department following ethical guidelines approved by the ethics committee of the Institut Curie CEEA-IC (Authorization APAF1S no. 24245–2020021921558370-v1 given by National Authority) in compliance with the international guidelines. Mice were housed in a specific-pathogen-free facility at the Curie Institute with a 12-h light/12-h dark cycle, a temperature range of 22–24 °C and 70% humidity.

Cross-tissue single-cell preparation

All animals were euthanized using carbon dioxide or cervical dislocation. All organs were collected fresh (that is, right after euthanasia) in CO₂-independent medium (Gibco) and maintained on ice until processing.

Spleen and lymph nodes

To isolate cells from the spleen and lymph nodes, tissues were dissected and placed on a 40-μm cell strainer (Falcon, Corning) kept on a 50-ml tube (Falcon, Corning) and were mashed on the cell strainer using the back of the 1-ml syringe plunger. Ten milliliters of PBS containing 0.5% BSA and 2 mM EDTA was continuously added while mashing to collect the single-cell suspension in a 50-ml tube. Collected cells were centrifuged at 400g for 5 min at 4 °C. The pellet was resuspended in 10 ml PBS and passed through a 30-μm nylon filter (CellTrics, Sysmex) kept in a 15-ml tube (Falcon, Corning). Cells were again centrifuged at 400g for 5 min at 4 °C. Afterwards, the pellet was resuspended in 100 μl of PBS containing 0.5% BSA and 2 mM EDTA for subsequent FACS staining. Red blood cell lysis was performed for splenic samples using red blood cell lysis buffer (10×, BioLegend) according to the manufacturer’s protocol.

Skin

Skin single-cell suspensions were obtained as previously described⁶⁹. Briefly, dorsal skin tissue was dissected (flattened, epidermis side up) and incubated at 37 °C for 45 min in 1 ml of 500 CU Dispase (Corning). The tissue was then chopped in RPMI 1640 GlutaMAX media supplemented with 1 mM sodium pyruvate, 1 mM nonessential amino acids, 50 μM β-mercaptoethanol, 20 mM HEPES, 100 U ml⁻¹ penicillin, 100 mg ml⁻¹ streptomycin, 0.5 mg ml⁻¹ DNase I (all products from Sigma-Aldrich) and 0.25 mg ml⁻¹ Liberase TL (Roche) and incubated for 1 h 45 min at 37 °C in a 5% CO₂ incubator. After filtering on a 40-μm filter kept in a 50-ml tube, the cells were washed twice in PBS containing 0.5% BSA and 2 mM EDTA, and the cell suspension was removed of skin debris using cell debris removal solution (Miltenyi) following the manufacturer’s instructions.

Liver and lung

To ensure complete lung and liver perfusion (evidenced by organ color change caused by the loss of red blood cells), a 20-ml syringe with a 22-gauge needle was used to inject 1× PBS starting with the right ventricle of the heart (10 ml) followed by the hepatic portal vein (10 ml). After perfusion and dissection of the liver and lung, the tissues were finely minced and digested using collagenase D (0.7 mg ml in PBS) for 30 min at 37 °C on a shaker in Freiburg, while in Paris, the Gentlemacs operating system (Miltenyi) with the m_impTumor_01 program was used. After washing the cell pellet twice at 400g for 5 min, the pellet was resuspended in 8 ml of 44% Percoll density gradient solution and underlaid with 5 ml of 67% Percoll density gradient solution. Centrifugation (without breaks) was performed at 1,600g for 20 min at room temperature. The cell layer containing mononuclear cells at the interface of the 44% and 67% density gradient centrifugation media was removed, transferred and washed. The resulting pellet was resuspended in 100 µl staining buffer (PBS containing 0.5% BSA and 2 mM EDTA).

Intestinal IELs

To isolate IELs from the large and small intestines, tissues were dissected, cleaned to remove feces, cut open and chopped into 2-cm pieces. The pieces were treated with 1 mM 1,4-Dithioerythritol to release IELs (2×, 20 min each at 37 °C, constant shaking). The supernatant was filtered through 70-μm cell strainers (Falcon, Corning) kept in a 50-ml tube (Falcon, Corning) on ice. Cells were washed with PBS containing 0.5% BSA and 2 mM EDTA, and 44% and 67% density gradient centrifugation was performed as described above. After washing, the resulting pellet was resuspended in 100 µl staining buffer (PBS containing 0.5% BSA and 2 mM EDTA).

Antibody staining, flow cytometry and single‐cell sorting

One hundred microliters of antibody staining solution was prepared in PBS containing 0.5% BSA and 2 mM EDTA and added to the isolated cells resuspended in 100 µl staining buffer as described above. Cells were incubated for 20 min on ice, washed three times with 2 ml of 0.5% BSA in PBS and resuspended in 3 ml after the last wash for cell sorting. The following antibodies were used: TCRγδ‐APC (BioLegend, 1:100 dilution), TCRβ‐BV421 (BioLegend, 1:100 dilution), CD45.1-PE (BD Biosciences, 1:100 dilution), CD45.2-AF700 (BioLegend, 1:100 dilution), CD45.2-PerCP5.5 (BD Biosciences, 1:100 dilution), CD160-PECy7 (BioLegend, 1:100 dilution), Ly6C-BV510 (BioLegend, 1:100 dilution) and CD62L-BV421 (BioLegend, 1:100 dilution). Zombie Aqua and Zombie Green fixable viability kits (BioLegend) were used to distinguish dead and living cells. Living TCRγδ⁺ single γδ T cells were sorted in BSA-coated tubes containing 50 µl of PBS using a FACSAria cell sorter (BD Biosciences) equipped with BD FACSDiva software (v8.0.2). Using pulse geometry gates (FSC‐W × FSC‐H and SSC‐W × SSC‐H), doublets/multiplets were excluded. After the completion of sorting, the cells were processed through the different 10x Genomics workflows.

Single-cell RNA sequencing

Single-cell RNA sequencing was performed using 10x Genomics with feature barcoding technology to multiplex cell samples from different organs so that they could be loaded on one well to reduce costs and minimize technical variability. Hashtag oligonucleotides were obtained as purified and already oligo-conjugated in TotalSeq-B (3′ chemistry) and TotalSeq-C (5′ chemistry) formats from BioLegend. Cells were stained with barcoded antibodies together with the staining solution before FACS sorting as described above. The antibody concentrations used were 1 µg per million cells, as recommended by the manufacturer (BioLegend) for flow cytometry applications. After staining, cells were washed three times in PBS containing 2% BSA and 0.01% Tween 20, followed by centrifugation (300g for 5 min at 4 °C) and supernatant exchange. After the final wash, the cells were resuspended in PBS, filtered through 40-µm cell strainers and processed for sorting. Sorted γδ T cells were processed through the 10x Genomics single-cell 3′ or V(D)J workflow according to the manufacturer’s instructions. Libraries were pooled to desired quantities to obtain appropriate sequencing depths as recommended by 10x Genomics and sequenced on a NovaSeq 6000 flow cell.

Single-cell surface protein profiling

To profile γδ T cells using antibodies for the quantification of cell surface proteins on a single-cell level, the following antibodies were obtained as purified, oligo-conjugated TotalSeq-B reagents from BioLegend: CCR6, NKp46, CD117, KLRG1, CCR7, CD8a, CD5, CD122 (IL-2Rβ), CD127 (IL-7Rα), CD278 (ICOS), Ly-6A/E (Sca-1), CD69, CD44, CD27, CD24, CD62L, CD25, NK-1.1, CD4, CCR2, TCR Vγ2, TCRβ, CD11c, CD19, GR-1 and CSF1R. Cells were stained with barcoded antibodies together with the staining solution before FACS sorting as described above. The antibody concentrations used were 1 µg per million cells, as recommended by the manufacturer (BioLegend) for flow cytometry applications. Sorted γδ T cells were processed through the 10x Genomics 3′ workflow according to the manufacturer’s instructions. Libraries were pooled in the desired ratio together with the gene expression libraries to obtain appropriate sequencing depths as recommended by 10x Genomics and sequenced on a NovaSeq 6000 flow cell.

Single-cell simultaneous chromatin accessibility and gene expression profiling

To perform simultaneous measurement of chromatin accessibility and gene expression, we could not barcode different organs with hashtag oligonucleotides; hence, after sorting γδ T cells from different tissues, we pooled them to obtain enough cells (80,000–100,000) to perform nuclear extraction according to the 10x Genomics protocol. Thereafter, single nuclei were processed through the Chromium Single Cell Multiome ATAC + Gene Expression workflow according to the manufacturer’s instructions. Gene expression and chromatin accessibility libraries were sequenced to obtain appropriate sequencing depths as recommended by 10x Genomics using the Illumina NovaSeq 6000 system.

Single-cell simultaneous gene expression and TCR profiling

Simultaneous single-cell RNA and TCR sequencing was performed using 10x Genomics V(D)J workflow with feature barcoding technology to multiplex cell samples from different organs. Hashtag oligonucleotides were obtained as purified and already oligo-conjugated in TotalSeq-C (5′ chemistry) format from BioLegend. Single-cell suspensions from three female mice (aged 8 weeks) for each organ (spleen, liver, lung, lymph node and small intestine) were pooled together before sorting. Cells were stained with barcoded antibodies together with the staining solution before FACS sorting as described above. The antibody concentrations used were 1 µg per million cells, as recommended by the manufacturer (BioLegend) for flow cytometry applications. After staining, cells were washed three times in PBS containing 2% BSA and 0.01% Tween 20, followed by centrifugation (300g for 5 min at 4 °C) and supernatant exchange. After the final wash, the cells were resuspended in PBS, filtered through 40-µm cell strainers and processed for sorting. Single-cell TCR libraries were generated using the following primers⁷⁰. First PCR: 2 μM of forward primer (5′- GATCTACACTCTTTCCCTACACGACGC-3′) and 0.5 μM of each reverse primer (5′-TCGAATCTCCATACTGACCAAGCTTGAC-3′, 5′-GTCTTCAGCGTATCCCCTTCCTGG-3′, 5′-CTTTCAGGCACAGTAAGCCAGC-3′ and 5′-TCTTCAGTCACCGTCAGCCAACTAA-3′). Second PCR: 1 μM of forward primer (5′-GATCTACACTCTTTCCCTACACGACGC-3′) and 1 μM of each reverse primer (5′- CCACAATCTTCTTGGATGATCTGAGACT-3′ and 5′-GTCCCAGTCTTATGGAGATTTGTTTCAGC-3′). Libraries were pooled to desired quantities to obtain appropriate sequencing depths as recommended by 10x Genomics and sequenced on a NovaSeq 6000 flow cell.

Parabiosis experiments

To evaluate the recirculatory and residential properties of γδ T cells, in accordance with published methods⁷¹, congenically distinct (CD45.1/1 or CD45.1/2 and CD45.2/2) aged-matched mice were surgically joined at their olecranon’s and knee joints using a non-absorbable 3-0 suture followed by suturing the skin of both animals together using 5-0 absorbable Vicryl sutures. Five weeks after surgery, animals were euthanized, and organs were collected for subsequent FACS or flow cytometry analysis. Staining was performed with the relevant antibodies in staining buffer containing PBS supplemented with 0.5% BSA, 2 mM EDTA and anti-FcR 2.4G2 (Institut Curie, produced in-house, 0.25 µg per million cells) for 20 min at 4 °C. Flow cytometry acquisition was performed using a Cytoflex (Beckman) cytometer with CytExpert software v2.4. Data were analyzed using FlowJo software (v10.8.0) and GraphPad Prism v8. For cell sorting, organs from different mice were processed separately as described above. For each organ, two pools were obtained by regrouping single-cell suspensions of parabionts with identical congenic markers. These organ pools were then stained with different hashtag oligonucleotides following the manufacturer’s instructions (BioLegend) and regrouped, which resulted in two tubes containing all organs for each congenic marker. FACS was then performed using a FACSAria cell sorter (BD) equipped with BD FACSDiva software v6. Resident or circulating γδ T cells were sorted in two distinct BSA-coated tubes. Sorted cells were processed through the 10x Genomics 3′ workflow according to the manufacturer’s instructions. Libraries were pooled in the desired ratio together to obtain appropriate sequencing depths as recommended by 10x Genomics and sequenced on a NovaSeq 6000 flow cell.

Quantification of gene expression, protein abundance, chromatin accessibility and TCR counts

Quantification of gene expression and/or cell surface protein abundance counts was performed using either cellranger-4.0.0 or cellranger-6.0.0 using the count command, which performs alignment, filtering, barcode counting and UMI counting as well as process feature barcoding data. Simultaneous counting of transcripts and open chromatin regions was performed through cellranger-arc-2.0.1 using the count command, which performs alignment, filtering, barcode counting, peak calling and counting of both ATAC and RNA molecules. Alignments were performed using prebuilt Cell Ranger and Cell Ranger ARC mouse mm10 references. Simultaneous quantification of gene expression, hashtag abundance and TCR counts/repertoire were performed using the multi command of cellranger-7.1.0.

Computational analysis of single-cell gene expression data

We analyzed scRNA-seq data using the R package (v4.1.3 and v4.2.2) Seurat (v4.3.0). We combined five batches of single-cell/single-nucleus gene expression datasets (including cells obtained from parabiotic pairs) and made two observations: there was technical variability that needed to be removed, and we identified various small contaminating clusters of B cells and myeloid cells. To remove batch effects, we integrated the data using Harmony²⁵, an algorithm that uses joint embedding to group cells by cell type rather than dataset-specific conditions. Harmony was executed using the RunHarmony function in Seurat with group.by.vars set to each batch. Furthermore, the clusters containing B cells and myeloid cells were removed. Low-quality cells were removed using the parameters described in Extended Data Fig. 1c. Importantly, ribosomal genes (small and large subunits) as well as predicted genes with Gm-identifier were excluded from the analysis. The normalization method was set to ‘LogNormalize’. Dimensionality reduction was performed using the RunUMAP function, where reduction was set to ‘harmony’ and dims to 1:30. Default resolution was used for clustering. To characterize the clusters, differential gene expression analysis was performed using the FindMarkers function in Seurat.

TotalSeq analysis

Of the five batches, the TotalSeq experiment was performed on one batch. We used the same UMAP coordinates obtained using dimensionality reduction based on gene expression data to visualize the expression of TotalSeq antibodies. Normalization was performed using Seurat with the normalization method set to ‘CLR’, which performs a centered log ratio transformation for normalization.

Single-cell chromatin accessibility analysis

scATAC-seq data analysis was performed using Signac (v1.8.0). Experiments to simultaneously measure chromatin accessibility and gene experiments were performed in two batches from mice belonging to two different animal facilities in Germany and France, which led to technical variability in the datasets. Two-step filtering was applied to scATAC-seq data. Only those nuclei that were also included in scRNA-seq analysis were included in the chromatin accessibility analysis. In the second step, scATAC-seq-based filtering was applied using the parameters described in Extended Data Fig. 5b. Furthermore, common peaks identified in all batches were included in the analysis. Integration was again performed using Harmony with the RunHarmony function. The ‘group.by.vars’ option was set to each batch, and reduction was set to ‘lsi’. Dimensionality reduction was performed using the RunUMAP function, in which reduction was set to ‘harmony’. For clustering of cells, the resolution in the FindClusters function was set to 0.6. To characterize the clusters, differentially accessible peaks were obtained using the FindMarkers function with test.use = ‘LR’. TF motif analysis was performed using the RunChromVAR function in Signac.

Computational analysis of simultaneous single-cell gene expression and TCR data

We analyzed scRNA-seq data using Seurat and single-cell TCR data using scRepertoire (v1.7.0). Clusters containing B cells and myeloid cells were removed. Low-quality cells were removed using the parameters described in Extended Data Fig. 1c. As previously described, ribosomal genes (small and large subunits) as well as predicted genes with Gm-identifier were excluded from the analysis. The normalization method was set to ‘LogNormalize’. Dimensionality reduction was performed using the RunUMAP function, where reduction was set to ‘pca’ and dims to 1:30. Default resolution was used for clustering. To characterize the clusters, differential gene expression analysis was performed using the FindMarkers function in Seurat. In total, 13,753 cells were included in the scRNA-seq analysis. Of 13,753 cells, clonotypes were detected in 4,220 cells. While performing the TCR repertoire analysis using scRepertoire, cloneCall parameter was set to ‘strict’, which uses the V(D)JC genes comprising the TCR plus the nucleotide sequence of the CDR3 region to call the clonotypes. Both γ and δ chains were used for the clonotype analysis wherever detected. Shannon diversity score was calculated based on nucleotide sequences of the TRG repertoire. Apart from the new clonotypes, we identified all the clonotypes of γδ T cells previously summarized⁷², except for those exhibiting the TRAV15-1-DV6-1 usage found in NKT-like IFN-γ/IL-4 double producers (Supplementary Data 1d,e and Supplementary Note).

Neighborhood analysis

To assess the differential contribution of exchanging and resident γδ T cells across organs and cell states, we applied the Milo algorithm⁵⁴ (using the R package miloR 1.2.0), which models cell states as overlapping neighborhoods based on a KNN graph as the basis for abundance testing. The KNN graph was built using the buildGraph function with k = 30 and d = 30, and neighborhoods were defined using the makeNhoods function with prop = 0.1, k = 30, d = 30 and refined = TRUE. Neighborhoods were grouped using groupNhoods with max.lfc.delta = 10. Neighborhoods with a log₂ fold change > 1 were considered resident, while the rest were denoted as circulating.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The primary read files and the raw counts for all single-cell sequencing datasets reported in this paper are available to download from the Gene Expression Omnibus under accession number https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE222454. Processed data can be downloaded from https://github.com/sagar161286/multimodal_gdTcells/.

Code availability

Codes to reproduce the data analysis and figures are available at https://github.com/sagar161286/multimodal_gdTcells/.

References

Sandrock, I. et al. Genetic models reveal origin, persistence and non-redundant functions of IL-17-producing γδ T cells. J. Exp. Med. 215, 3006–3018 (2018).
Article CAS PubMed PubMed Central Google Scholar
Harly, C., Robert, J., Legoux, F. & Lantz, O. γδ T, NKT, and MAIT cells during evolution: redundancy or specialized functions? J. Immunol. 209, 217–225 (2022).
Article CAS PubMed Google Scholar
Darrigues, J., Almeida, V., Conti, E. & Ribot, J. C. The multisensory regulation of unconventional T cell homeostasis. Semin. Immunol. 61–64, 101657 (2022).
Nielsen, M. M., Witherden, D. A. & Havran, W. L. γδ T cells in homeostasis and host defence of epithelial barrier tissues. Nat. Rev. Immunol. 17, 733–745 (2017).
Article CAS PubMed PubMed Central Google Scholar
Sharp, L. L., Jameson, J. M., Cauvi, G. & Havran, W. L. Dendritic epidermal T cells regulate skin homeostasis through local production of insulin-like growth factor 1. Nat. Immunol. 6, 73–79 (2005).
Article CAS PubMed Google Scholar
Komano, H. et al. Homeostatic regulation of intestinal epithelia by intraepithelial gamma delta T cells. Proc. Natl Acad. Sci. USA 92, 6147–6151 (1995).
Article CAS PubMed PubMed Central Google Scholar
Hu, B. et al. γδ T cells and adipocyte IL-17RC control fat innervation and thermogenesis. Nature 578, 610–614 (2020).
Article CAS PubMed PubMed Central Google Scholar
Kohlgruber, A. C. et al. γδ T cells producing interleukin-17A regulate adipose regulatory T cell homeostasis and thermogenesis. Nat. Immunol. 19, 464–474 (2018).
Article CAS PubMed PubMed Central Google Scholar
Ribot, J. C., Lopes, N. & Silva-Santos, B. γδ T cells in tissue physiology and surveillance. Nat. Rev. Immunol. 21, 221–232 (2021).
Article CAS PubMed Google Scholar
Hayday, A. C. γδ T cells and the lymphoid stress-surveillance response. Immunity 31, 184–196 (2009).
Article CAS PubMed Google Scholar
Bonneville, M., O’Brien, R. L. & Born, W. K. γδ T cell effector functions: a blend of innate programming and acquired plasticity. Nat. Rev. Immunol. 10, 467–478 (2010).
Article CAS PubMed Google Scholar
Hsieh, B. et al. In vivo cytokine production in murine listeriosis. Evidence for immunoregulation by gamma delta⁺ T cells. J. Immunol. 156, 232–237 (1996).
Article CAS PubMed Google Scholar
Toulon, A. et al. A role for human skin-resident T cells in wound healing. J. Exp. Med. 206, 743–750 (2009).
Article CAS PubMed PubMed Central Google Scholar
Havran, W. L., Chien, Y. H. & Allison, J. P. Recognition of self antigens by skin-derived T cells with invariant gamma delta antigen receptors. Science 252, 1430–1432 (1991).
Article CAS PubMed Google Scholar
Boismenu, R. & Havran, W. L. Modulation of epithelial cell growth by intraepithelial gamma delta T cells. Science 266, 1253–1255 (1994).
Article CAS PubMed Google Scholar
Boismenu, R., Feng, L., Xia, Y. Y., Chang, J. C. & Havran, W. L. Chemokine expression by intraepithelial gamma delta T cells. Implications for the recruitment of inflammatory cells to damaged epithelia. J. Immunol. 157, 985–992 (1996).
Article CAS PubMed Google Scholar
Gay, D. et al. Fgf9 from dermal gammadelta T cells induces hair follicle neogenesis after wounding. Nat. Med. 19, 916–923 (2013).
Article CAS PubMed PubMed Central Google Scholar
Yang, H., Antony, P. A., Wildhaber, B. E. & Teitelbaum, D. H. Intestinal intraepithelial lymphocyte gamma delta-T cell-derived keratinocyte growth factor modulates epithelial growth in the mouse. J. Immunol. 172, 4151–4158 (2004).
Article CAS PubMed Google Scholar
Salou, M. et al. A common transcriptomic program acquired in the thymus defines tissue residency of MAIT and NKT subsets. J. Exp. Med. 216, 133–151 (2019).
Article CAS PubMed PubMed Central Google Scholar
Mackay, L. K. et al. Hobit and Blimp1 instruct a universal transcriptional program of tissue residency in lymphocytes. Science 352, 459–463 (2016).
Article CAS PubMed Google Scholar
Sugahara, S. et al. Extrathymic derivation of gut lymphocytes in parabiotic mice. Immunology 96, 57–65 (1999).
Article CAS PubMed PubMed Central Google Scholar
Jiang, X. et al. Dermal gammadelta T cells do not freely re-circulate out of skin and produce IL-17 to promote neutrophil infiltration during primary contact hypersensitivity. PLoS ONE 12, e0169397 (2017).
Article PubMed PubMed Central Google Scholar
Tan, L. et al. Single-cell transcriptomics identifies the adaptation of Scart1⁺ Vγ6⁺ T cells to skin residency as activated effector cells. Cell Rep. 27, 3657–3671 (2019).
Article CAS PubMed Google Scholar
Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell 184, 3573–3587 (2021).
Article CAS PubMed PubMed Central Google Scholar
Korsunsky, I. et al. Fast, sensitive and accurate integration of single-cell data with Harmony. Nat. Methods 16, 1289–1296 (2019).
Article CAS PubMed PubMed Central Google Scholar
Arbones, M. L. et al. Lymphocyte homing and leukocyte rolling and migration are impaired in L-selectin-deficient mice. Immunity 1, 247–260 (1994).
Article CAS PubMed Google Scholar
Matloubian, M. et al. Lymphocyte egress from thymus and peripheral lymphoid organs is dependent on S1P receptor 1. Nature 427, 355–360 (2004).
Article CAS PubMed Google Scholar
Germar, K. et al. T-cell factor 1 is a gatekeeper for T-cell specification in response to Notch signaling. Proc. Natl Acad. Sci. USA 108, 20060–20065 (2011).
Article CAS PubMed PubMed Central Google Scholar
Weber, B. N. et al. A critical role for TCF-1 in T-lineage specification and differentiation. Nature 476, 63–68 (2011).
Article CAS PubMed PubMed Central Google Scholar
Okamura, R. M. et al. Redundant regulation of T cell differentiation and TCRα gene expression by the transcription factors LEF-1 and TCF-1. Immunity 8, 11–20 (1998).
Article CAS PubMed Google Scholar
Tu, T. C. et al. CD160 is essential for NK-mediated IFN-gamma production. J. Exp. Med. 212, 415–429 (2015).
Article CAS PubMed PubMed Central Google Scholar
Cheroutre, H., Lambolez, F. & Mucida, D. The light and dark sides of intestinal intraepithelial lymphocytes. Nat. Rev. Immunol. 11, 445–456 (2011).
Article CAS PubMed PubMed Central Google Scholar
Ivanov, I. I. et al. The orphan nuclear receptor RORγt directs the differentiation program of proinflammatory IL-17⁺ T helper cells. Cell 126, 1121–1133 (2006).
Article CAS PubMed Google Scholar
Lu, Y., Cao, X., Zhang, X. & Kovalovsky, D. PLZF controls the development of fetal-derived IL-17⁺Vγ6⁺ γδ T cells. J. Immunol. 195, 4273–4281 (2015).
Article CAS PubMed Google Scholar
Hayday, A. C. et al. Unusual organization and diversity of T-cell receptor alpha-chain genes. Nature 316, 828–832 (1985).
Article CAS PubMed Google Scholar
McKenzie, D. R. et al. Normality sensing licenses local T cells for innate-like tissue surveillance. Nat. Immunol. 23, 411–422 (2022).
Article CAS PubMed PubMed Central Google Scholar
Jameson, J. et al. A role for skin γδ T cells in wound repair. Science 296, 747–749 (2002).
Article CAS PubMed Google Scholar
Omilusik, K. D. et al. Transcriptional repressor ZEB2 promotes terminal differentiation of CD8⁺ effector and memory T cell populations during infection. J. Exp. Med. 212, 2027–2039 (2015).
Article PubMed PubMed Central Google Scholar
Joshi, N. S. et al. Inflammation directs memory precursor and short-lived effector CD8⁺ T cell fates via the graded expression of T-bet transcription factor. Immunity 27, 281–295 (2007).
Article CAS PubMed PubMed Central Google Scholar
Jensen, K. D. et al. Thymic selection determines γδ T cell effector fate: antigen-naive cells make interleukin-17 and antigen-experienced cells make interferon gamma. Immunity 29, 90–100 (2008).
Article CAS PubMed PubMed Central Google Scholar
Wiesheu, R. et al. Ly6C defines a subset of memory-like CD27⁺ γδ T cells with inducible cancer-killing function. Preprint at bioRxiv https://doi.org/10.1101/2020.09.08.287854 (2020).
Ribot, J. C. et al. CD27 is a thymic determinant of the balance between interferon-gamma- and interleukin 17-producing gammadelta T cell subsets. Nat. Immunol. 10, 427–436 (2009).
Article CAS PubMed PubMed Central Google Scholar
Haas, J. D. et al. CCR6 and NK1.1 distinguish between IL-17A and IFN-gamma-producing gammadelta effector T cells. Eur. J. Immunol. 39, 3488–3497 (2009).
Article CAS PubMed Google Scholar
Hatano, S., Murakami, T., Noguchi, N., Yamada, H. & Yoshikai, Y. CD5^-NK1.1⁺ γδ T cells that develop in a Bcl11b-independent manner participate in early protection against infection. Cell Rep. 21, 1191–1202 (2017).
Article CAS PubMed Google Scholar
Sivori, S. et al. p46, a novel natural killer cell-specific surface molecule that mediates cell activation. J. Exp. Med. 186, 1129–1136 (1997).
Article CAS PubMed PubMed Central Google Scholar
Lochner, M. et al. In vivo equilibrium of proinflammatory IL-17⁺ and regulatory IL-10⁺ Foxp3⁺ RORγt⁺ T cells. J. Exp. Med. 205, 1381–1393 (2008).
Article CAS PubMed PubMed Central Google Scholar
Stuart, T., Srivastava, A., Madad, S., Lareau, C. A. & Satija, R. Single-cell chromatin state analysis with Signac. Nat. Methods 18, 1333–1341 (2021).
Article CAS PubMed PubMed Central Google Scholar
Barros-Martins, J. et al. Effector gammadelta T cell differentiation relies on master but not auxiliary T_h cell transcription factors. J. Immunol. 196, 3642–3652 (2016).
Article CAS PubMed Google Scholar
Chen, Y., Chou, K., Fuchs, E., Havran, W. L. & Boismenu, R. Protection of the intestinal mucosa by intraepithelial gamma delta T cells. Proc. Natl Acad. Sci. USA 99, 14338–14343 (2002).
Article CAS PubMed PubMed Central Google Scholar
Tsuchiya, T. et al. Role of gamma delta T cells in the inflammatory response of experimental colitis mice. J. Immunol. 171, 5507–5513 (2003).
Article CAS PubMed Google Scholar
Schep, A. N., Wu, B., Buenrostro, J. D. & Greenleaf, W. J. chromVAR: inferring transcription-factor-associated accessibility from single-cell epigenomic data. Nat. Methods 14, 975–978 (2017).
Article CAS PubMed PubMed Central Google Scholar
Itohara, S. et al. Homing of a gamma delta thymocyte subset with homogeneous T-cell receptors to mucosal epithelia. Nature 343, 754–757 (1990).
Article CAS PubMed Google Scholar
Di Marco Barros, R. et al. Epithelia use butyrophilin-like molecules to shape organ-specific γδ T cell compartments. Cell 167, 203–218 (2016).
Article PubMed PubMed Central Google Scholar
Dann, E., Henderson, N. C., Teichmann, S. A., Morgan, M. D. & Marioni, J. C. Differential abundance testing on single-cell data using k-nearest neighbor graphs. Nat. Biotechnol. 40, 245–253 (2022).
Article CAS PubMed Google Scholar
Edwards, S. C. et al. PD-1 and TIM-3 differentially regulate subsets of mouse IL-17A-producing γδ T cells. J. Exp. Med. 220, e20211431 (2023).
Schenkel, J. M. & Masopust, D. Tissue-resident memory T cells. Immunity 41, 886–897 (2014).
Article CAS PubMed PubMed Central Google Scholar
Cheuk, S. et al. CD49a expression defines tissue-resident CD8⁺ T cells poised for cytotoxic function in human skin. Immunity 46, 287–300 (2017).
Article CAS PubMed PubMed Central Google Scholar
Milner, J. J. et al. Runx3 programs CD8⁺ T cell residency in non-lymphoid tissues and tumours. Nature 552, 253–257 (2017).
Article CAS PubMed PubMed Central Google Scholar
Hahn, A. M. et al. A monoclonal Trd chain supports the development of the complete set of functional γδ T cell lineages. Cell Rep. 42, 112253 (2023).
Article CAS PubMed Google Scholar
Li, Z. et al. Single-cell RNA-seq and chromatin accessibility profiling decipher the heterogeneity of mouse γδ T cells. Sci. Bull. 67, 408–426 (2022).
Article CAS Google Scholar
Hu, Y. et al. Single-cell analysis reveals the origins and intrahepatic development of liver-resident IFN-gamma-producing γδ T cells. Cell Mol. Immunol. 18, 954–968 (2021).
Article CAS PubMed PubMed Central Google Scholar
Dalessandri, T., Crawford, G., Hayes, M., Castro Seoane, R. & Strid, J. IL-13 from intraepithelial lymphocytes regulates tissue homeostasis and protects against carcinogenesis in the skin. Nat. Commun. 7, 12080 (2016).
Article CAS PubMed PubMed Central Google Scholar
Steinert, E. M. et al. Quantifying memory CD8 T cells reveals regionalization of immunosurveillance. Cell 161, 737–749 (2015).
Article CAS PubMed PubMed Central Google Scholar
Woolf, E., Brenner, O., Goldenberg, D., Levanon, D. & Groner, Y. Runx3 regulates dendritic epidermal T cell development. Dev. Biol. 303, 703–714 (2007).
Article CAS PubMed Google Scholar
Zakeri, N. et al. Characterisation and induction of tissue-resident gamma delta T cells to target hepatocellular carcinoma. Nat. Commun. 13, 1372 (2022).
Article CAS PubMed PubMed Central Google Scholar
Hunter, S. et al. Human liver infiltrating γδ T cells are composed of clonally expanded circulating and tissue-resident populations. J. Hepatol. 69, 654–665 (2018).
Article CAS PubMed PubMed Central Google Scholar
Jaeger, N. et al. Single-cell analyses of Crohn’s disease tissues reveal intestinal intraepithelial T cells heterogeneity and altered subset distributions. Nat. Commun. 12, 1921 (2021).
Article PubMed PubMed Central Google Scholar
McMurray, J. L. et al. Transcriptional profiling of human Vδ1 T cells reveals a pathogen-driven adaptive differentiation program. Cell Rep. 39, 110858 (2022).
Article CAS PubMed PubMed Central Google Scholar
du Halgouet, A. et al. Role of MR1-driven signals and amphiregulin on the recruitment and repair function of MAIT cells during skin wound healing. Immunity 56, 78–92 (2023).
Article PubMed PubMed Central Google Scholar
Lee, M. et al. Single-cell RNA sequencing identifies shared differentiation paths of mouse thymic innate T cells. Nat. Commun. 11, 4367 (2020).
Article CAS PubMed PubMed Central Google Scholar
Kamran, P. et al. Parabiosis in mice: a detailed protocol. J. Vis. Exp. 50556 (2013).
Papadopoulou, M., Sanchez Sanchez, G. & Vermijlen, D. Innate and adaptive γδ T cells: how, when, and why. Immunol. Rev. 298, 99–116 (2020).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

We thank D. Grün, T. Boehm and A. Nusser for critical reading of the paper. We thank J. Bodinek-Wersing from the Lighthouse Core Facility at the Medical Faculty, University of Freiburg for the support with cell sorting. Sagar is supported by the Department of Medicine II, Freiburg University Medical Center, Faculty of Medicine, University of Freiburg. This study was supported by the Research Commission from the Faculty of Medicine of the University of Freiburg (Project ID 3095050051 to Sagar), Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) through SFB 1160 (Project ID 256073931 to Sagar), TRR 359 (Project ID 491676693 to Sagar) and 322977937/GRK2344 (to Sagar), Agence Nationale de la Recherche Grant MAITrepair ANR-20-CE15-0028-01 (to O.L.), European Research Council (ERC−2019-AdG-885435; to O.L.) and Fondation pour la Recherche Médicale, grant number FDT202106013036 (to A.H.)

Author information

Anastasia du Halgouet
Present address: National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, MD, USA

Authors and Affiliations

Institut National de la Santé et de la Recherche Médicale U932, PSL University, Institut Curie, Paris, France
Anastasia du Halgouet, Aurélie Darbois, Marion Salou & Olivier Lantz
Department of Medicine II (Gastroenterology, Hepatology, Endocrinology, and Infectious Diseases), Freiburg University Medical Center, Faculty of Medicine, University of Freiburg, Freiburg, Germany
Kerstin Bruder, David Obwegs, Robert Thimme, Maike Hofmann & Sagar
Max Planck Institute of Immunobiology and Epigenetics, Freiburg, Germany
Nina Peltokangas
Faculty of Biology, University of Freiburg, Freiburg, Germany
Nina Peltokangas
Würzburg Institute of Systems Immunology, Max Planck Research Group at the Julius-Maximilians-Universität Würzburg, Würzburg, Germany
Nina Peltokangas
Laboratoire d’Immunologie Clinique, Institut Curie, Paris, France
Olivier Lantz
Centre d’Investigation Clinique en Biothérapie Gustave-Roussy Institut Curie (CIC-BT1428) Institut Curie, Paris, France
Olivier Lantz

Authors

Anastasia du Halgouet
View author publications
You can also search for this author in PubMed Google Scholar
Kerstin Bruder
View author publications
You can also search for this author in PubMed Google Scholar
Nina Peltokangas
View author publications
You can also search for this author in PubMed Google Scholar
Aurélie Darbois
View author publications
You can also search for this author in PubMed Google Scholar
David Obwegs
View author publications
You can also search for this author in PubMed Google Scholar
Marion Salou
View author publications
You can also search for this author in PubMed Google Scholar
Robert Thimme
View author publications
You can also search for this author in PubMed Google Scholar
Maike Hofmann
View author publications
You can also search for this author in PubMed Google Scholar
Olivier Lantz
View author publications
You can also search for this author in PubMed Google Scholar
Sagar
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

A.H. performed parabiosis experiments, tissue preparation for cell sorting and flow cytometry analysis and contributed to single-cell experiments as well as data interpretation and paper writing; K.B. prepared gene expression, chromatin accessibility, cell surface protein and TCR libraries; N.P. contributed to mouse handling and single-cell preparation from tissues; A.D. contributed to parabiosis experiments and single-cell preparation from tissues; D.O. and M.S. contributed to data analysis, R.T. and M.H. provided key scientific inputs and contributed to data interpretation and paper writing; O.L. supervised A.H. and A.D., provided key scientific inputs and contributed to data interpretation and paper writing; S. conceptualized and coordinated the study, acquired funding, conducted experiments, analyzed the data and wrote the paper with the help of A.H. and O.L. All authors read and edited the paper.

Corresponding author

Correspondence to Sagar.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Immunology thanks David Vermijlen and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available. Primary Handling Editor: S. Houston, in collaboration with the Nature Immunology team.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Quality control analysis of the scRNA-seq data.

a, Representative flow cytometry plots showing our strategy to sort γδ T cells for single-cell multiomics experiments using 10x Genomics. b, UMAP representation showing the origin of cells across different batches in different colors after batch correction was performed using Harmony. Note that the experiments were performed in five batches using mice from two different animal facilities, leading to batch-associated variability. c, Table listing the cut-offs used to remove low-quality cells for scRNA-seq analysis. d, UMAP representation showing the origin of cells across different organs in different colors. e, Violin plots showing the number of genes and transcripts quantified per cell from each tissue in the scRNA-seq data. f, Dot plot showing the expression of selected genes differentially expressed between Rorc⁺ clusters 3 and 5. Color represents the mean expression of the gene in the respective cluster, and dot size represents the fraction of cells in that cluster expressing the gene. g, UMAP representation depicting the cell cycle phase of each cell. Cells in S, G₂ and M phases are highlighted using a dark red rectangle.

Extended Data Fig. 2 Quantification of cell surface protein markers.

a-u, UMAP representation showing the normalized expression of 21 cell surface markers profiled using TotalSeq^TM.

Extended Data Fig. 3 Genes identifying various γδ subsets.

a-h, UMAP representation showing the normalized transcript counts of Sell, Ly6c2, Cd160, Gzmb, Klrg1, Mki67, Rorc and Areg used to classify γδ T cell subsets based on scRNA-Seq.

Extended Data Fig. 4 Characterizing γδ T cell subsets using flow cytometry.

a-c, Representative flow cytometry plots showing the fraction of γδ T cells stratified using CD62L and LY6C (a), CD160 (b) and RORγt-GFP (c) across different organs. Quantification of the data is shown in Fig. 3g–i. Ax LN, axillary lymph nodes; Mes LN, mesenteric lymph nodes.

Extended Data Fig. 5 Quality control analysis of the scATAC⁻seq data.

a, UMAP representation showing the origin of cells across two batches in different colors after batch correction was performed using Harmony. b, Table listing the cut-offs applied to remove low-quality cells for scATAC-seq analysis. c, Plot showing the distribution of the TSS enrichment score and Tn5 insertion frequency at TSS sites in the scATAC-Seq data. d, Plots showing the total number of chromatin accessibility counts, TSS distribution and nucleosome signal of the cells included in the analysis after filtering. e, UMAP representation based on chromatin accessibility showing the 22 clusters identified based on gene expression profiles shown in Fig. 1c in different colors. f, g,Chromatin accessibility tracks showing the frequency of Tn5 integration across regions of the genome encoding Cd24a (f) and Ifng (g) for cells grouped by scATAC-seq clusters.

Extended Data Fig. 6 Simultaneous gene expression and chromatin accessibility profiling reveals novel features of γδ subsets.

a, b, Chromatin accessibility tracks showing the frequency of Tn5 integration across regions of the genome encoding Il10 (a) and Il34 (b) for shortlisted γδ subsets. c, d, UMAP representation of scATAC-seq data showing the normalized transcript counts of Il10 (c) and Il34 (d). e-g, Chromatin accessibility tracks showing the frequency of Tn5 integration across regions of the genome encoding Rorc (e), Il17a (f) and Tbx21 (g) for shortlisted γδ scATAC-seq clusters. h, PCA based on the average highly variable chromatin accessibility profiles of eight γδ T subsets revealed that Areg⁺ and Rorc⁺ subsets exhibit distinct transcriptional signatures compared to other subsets. i, Chromatin accessibility tracks showing the frequency of Tn5 integration across regions of the genome encoding Fgf1 for shortlisted γδ subsets. j, UMAP representation of scATAC-seq data showing the normalized transcript counts of Fgf1. k, Chromatin accessibility tracks showing the frequency of Tn5 integration across regions of the genome encoding Fgf7 for shortlisted γδ subsets. l, UMAP representation of scATAC-seq data showing the normalized transcript counts of Fgf7. m, Chromatin accessibility tracks showing the frequency of Tn5 integration across regions of the genome encoding Wnt9a for shortlisted γδ subsets. n, UMAP representation of scATAC-seq data showing the normalized transcript counts of Wnt9a.

Extended Data Fig. 7 Transcription factor families specific to γδ T cell subsets.

a-e, DNA sequence motifs and UMAP representation showing the motif activity score per cell for key overrepresented TF motifs in the scATAC-seq data – Tcf7 (a), Tbx21 (b), Irf9 (c), Rorc (d) and Fos (e). f, UMAP representation showing γδ subsets in different colors overlaid with TF family motifs enriched in each cell subset as identified by chromVAR analysis.

Extended Data Fig. 8 TCR repertoire analysis of γδ T cells across tissues.

a, Violin plots showing the number of genes and transcripts quantified per cell from each tissue in the data where cells were simultaneously profiled for gene expression and TCR rearrangement configurations. b, UMAP representation depicting the cell cycle phase of each cell. Cells in S, G₂ and M phases are highlighted using a dark red rectangle. c, Bar plot quantifying the variable γ chain usage of γδ T cells within each tissue. Seven variable γ chains are depicted in different colors. d, Dot plot showing the Shannon diversity score calculated based on the diversity of TCR γδ clonotypes in different cell types. Dots in different colors represent different cell types. e, Chord diagram depicting the clonal overlap between profiled organs. f, Bar plot quantifying the shared variable γ chain usage of liver γδ T cells with other organs. Seven variable γ chains are depicted in different colors. SI, small intestine; LN, lymph nodes.

Extended Data Fig. 9 Quantification of exchanging and resident γδ T cells across tissues.

a, Bar plot showing the quantification of partner- and host-derived γδ T cells across different tissues using scRNA-seq. b, Representative FACS plots showing the percentage of exchanging and resident γδ T cells across different tissues. Quantification of the data is shown in Fig. 6f. c, Representative FACS plots showing the percentages of exchanging and resident CD62L⁺LY6C^-, CD62L⁺LY6C⁺, CD62L⁻LY6C⁺ and CD62L⁻LY6C⁻γδ T cells across different tissues. d, Plots showing the quantification of partner-derived total, CD62L⁺LY6C⁻, CD62L⁺LY6C⁺, CD62L⁻LY6C⁺ and CD62L⁻LY6C⁻γδ T cells across different tissues (n = 8 mice; four parabiotic pairs). Statistical analysis using Dunnett’s multiple comparisons test (One-way ANOVA), with P values: *P < 0.05; **P < 0.01; ***P < 0.001 and ****P < 0.0001 and ns>0.05 (95% CI). With exact P values for panel d. as follows: I + B LN (*red= 0.0195), MedLN (**red=0.0019; **grey=0.0024), MesLN (**grey=0.0063), Spleen (**red=0.0055; ***orange=0.0006; **grey=0.0027), Liver (*pink=0.0308; *red=0.0110; *orange=0.0103; ***grey= 0.0004), Lung (**red=0.0054), Large intestine (*pink=0.0138; *red=0.0162; **orange=0.0081), Small intestine (**pink=0.0061; *red=0.0226; **orange=0.0091; **grey= 0.0070).

Extended Data Fig. 10 Chromatin landscape of genes associated with circulating memory T cells in γδ subsets.

a-f, Chromatin accessibility tracks showing the frequency of Tn5 integration across regions of the genome encoding Tcf7 (a), Lef1 (b), Klf2 (c), S1pr1 (d), Ccr7 (e) and Sell (f) for six γδ subsets.

Supplementary information

Supplementary Information

Supplementary Note and Supplementary Data 1

Reporting Summary

Peer Review File

Supplementary Tables 1–6

Supplementary Table 1 Table listing differentially expressed genes in γδ scRNA-seq clusters depicted in Fig. 1c. Differentially expressed genes were calculated using a two-tailed Wilcoxon rank-sum test. Adjusted P values are based on Bonferroni correction. Supplementary Table 2 Table listing differentially expressed genes between scRNA-seq clusters 3 and 5 depicted in Fig. 1c. Differentially expressed genes were calculated using a two-tailed Wilcoxon rank-sum test. Adjusted P values are based on Bonferroni correction. Supplementary Table 3 Table listing differentially open chromatin regions in γδ scATAC-seq clusters depicted in Fig. 4c. Differentially accessible regions were calculated using a logistic regression framework. Adjusted P values are based on Bonferroni correction. Supplementary Table 4 Table listing closest genes associated with differentially open chromatin regions in γδ scATAC-seq clusters depicted in Fig. 4c. Supplementary Table 5 Table listing TCR clonotypes detected in single cells depicted in Fig. 5a. Supplementary Table 6 Table listing gene sets of T_RM core residency signatures and exchanging signatures

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

du Halgouet, A., Bruder, K., Peltokangas, N. et al. Multimodal profiling reveals site-specific adaptation and tissue residency hallmarks of γδ T cells across organs in mice. Nat Immunol 25, 343–356 (2024). https://doi.org/10.1038/s41590-023-01710-y

Download citation

Received: 02 April 2023
Accepted: 13 November 2023
Published: 04 January 2024
Issue Date: February 2024
DOI: https://doi.org/10.1038/s41590-023-01710-y