γδ T cell receptor dependencies define a unique immunosurveillance modality

Mice

Adult mice of both sexes were used at 6–24 weeks of age. Mice were bred at the Francis Crick Institute (London, UK) and Animalario del Centro de Investigaciones Biológicas Margarita Salas (Madrid, Spain). All mouse strains were of C57BL/6J genetic background: Trgc1^−/−, Tcrd-creERT2 Trgc1^fl/fl, Rosa26-creERT2 Trgc1^fl/fl, Tcrd-creERT2 Trgc1^fl/fl;Rosa26-mTmG^fl/fl (this mouse strain is the intercross of Tcrd-creERT2 and Rosa26-mTmG^fl/fl, which enables the γδ T cell-specific expression of EGFP after tamoxifen administration), and Nur77-GFP reporter mice (Nr4a1^GFP), which was used to assess antigen receptor signalling activity in vivo through GFP expression downstream of TCR activation. Mice were housed under specific pathogen-free conditions in a 12 h light:12 h dark cycle at an ambient temperature of 20–24 °C and relative humidity of 45–65%. Mice were kept in filter-topped cages with sterilized food and water ad libitum and autoclaved corncob bedding that was changed at least once weekly. All experiments were performed according to UK and Spain animal protection laws.

Conditional Cγ1 deletion

To induce conditional deletion of Cγ1, Trgc1^fl mice were crossed with two different lines expressing tamoxifen-inducible Cre recombinase. Tcrd-creERT2 mice were used for Trgc1 deletion specifically in dermal and epidermal γδ T cells, while Rosa26-creERT2 mice allowed for ubiquitous deletion upon tamoxifen administration including the small intestine. Primer sequences for genotyping are indicated in Supplementary Table 4. Experimental mice received tamoxifen for 7 days, by intraperitoneal injection. Following treatment, epidermis and dermis samples were collected from Tcrd-creERT2 mice, whereas small intestinal IELs were isolated from Rosa26-creERT2 mice for further analysis.

Generation of human γδ T cell lines

Peripheral blood mononuclear cells (PBMCs) from a healthy male donor were isolated by centrifugation on a Ficoll-Paque PLUS (GE Healthcare) gradient. Cells were stained with an anti-TCR V delta 1 Monoclonal Antibody (TS8.2 clone from Invitrogen) and Vδ1⁺ cells isolated by fluorescence-activated cell sorting (FACS)-sorting with a FACSAria Fusion flow cytometer (BD Biosciences) at Flow Cytometry Facilities at Centro de Investigaciones Biológicas Margarita Salas (Madrid, Spain). First, polyclonal Vδ1⁺ cells were stimulated at day 0 with 1 mg ml⁻¹ phytohemagglutinin-L (PHA-L) from Phaseolus vulgaris (Sigma-Aldrich), and co-cultured with irradiated allogeneic feeder cells weekly (PBMCs and Epstein Barr virus-transformed B cells, 40 and 65 Gy, respectively) at 1:2 ratio in Iscove’s Modified Dulbecco’s Medium (IMDM) (GE Healthcare) supplemented with 40 IU ml⁻¹ recombinant human IL-2, 10% AB+ human serum, and 1% l-glutamine and Antibiotic–Antimycotic (Life Technologies). After 3 weeks of allogeneic stimuli, 0,7 million expanded Vδ1⁺ cells were stimulated with 10 μg ml⁻¹ PHA-L (only at day 0) for 24 h and co-cultured with irradiated (150 Gy) HTLV1-producing cell line (MT2) at 1:1 ratio in RPMI 1640 medium from Lonza, supplemented with 100 IU ml⁻¹ recombinant interleukin-2 (rIL-2), 10% fetal bovine serum and 1% l-glutamine and antibiotics-antimycotic. After 4 weeks, cell lines phenotype was checked to be exclusively Vδ1⁺ cells by flow cytometry with a FACSCelesta flow cytometer (BD Biosciences) at the Center for Cytometry and Fluorescence Microscopy of Complutense University of Madrid (Spain).

CRISPR RNP nucleofection and human TCR deletion

Human blood HTLV1-immortalized Vδ1⁺ T cells were cultured in complete RPMI medium (RPMI 1640 with 10% heat-inactivated fetal bovine serum (FBS), 10 mM HEPES, 1× non-essential amino acids (NEAA), 2 mM l-glutamine, 1 mM sodium pyruvate, 50 μM β-mercaptoethanol, 100 U ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin) supplemented 100 U ml⁻¹ of rhIL-2 (PeproTech, 200-02) and subjected to CRISPR to delete TRDV1 gene using a modified protocol published previously⁶⁰. The TRDV1 gene was independently targeted by two different custom designed gRNAs (5′-CATCAGAACCCTGGCGAATA-3′, 5′-ACCACCAACTTGTTTCATAC-3′) synthesized by IDT. In brief, Cas9 RNPs were assembled by complexing two-component 40 μM gRNA (CRISPR RNA (crRNA):trans-activating crRNA (tracrRNA)) with 13 μM Cas9–GFP nuclease (IDT, 10008100) to form an RNP at 20 μM immediately before experiments. Nucleofection was carried out on 1 × 10⁶ cells with 3 μl of Cas9 RNP and P3 buffer (Lonza, V4XP-3032) in a total of 20 μl per reaction using the Lonza Nucleofection system with pulse code EH100. Immediately after nucleofection, 80 μl of pre-warmed complete RPMI medium without penicillin-streptomycin was added directly to the cells in the nucleofection cuvettes and were rested for 10 min at 37 °C. After 10 min, cells were transferred to 48-well cell culture plates and incubated overnight at 37 °C, 5% CO₂ in complete RPMI medium containing 100 U ml⁻¹ of rhIL-2 but without penicillin-streptomycin. After 24 h, cells were sorted based on GFP signal using BD FACS Aria III Cell Sorter (BD) to enrich for transfected GFP⁺ cells and returned to culture. Following 10 days, TRDV1-knockout efficiency was determined by flow cytometry with TCR Vδ1 antibody (Miltenyi Biotec, 130-118-968) using the BD LSR Fortessa instrument (BD).

Isolation of mouse epidermal cells

Whole trunk epidermis was separated from the dermis by incubating skin samples in 2 mg ml⁻¹ Dispase II at 37 °C for 2 h (ref. ⁶¹). Following incubation, the epidermal sheets were carefully peeled off and digested with 0.1 mg ml⁻¹ of DNase I at 37 °C for 15 min shaking at 230 rpm to obtain single-cell suspensions for subsequent analysis.

Isolation of mouse dermal cells

After epidermis separation, dermis tissue was finely minced with scissors and incubated at 37 °C for 1 h with shaking at 700 rpm on a Thermoblock in digestion solution containing 2.7 mg ml⁻¹ collagenase XI, 0.25 mg hyaluronidase, and 0.1 mg ml⁻¹ DNase I in RPMI with 1× HEPES. Digested tissue was then filtered through a 30-μm filter cone (CellTrics 30 μm, 04-0042-2316, Sysmex) into a new tube for subsequent analysis.

Isolation of mouse intestinal IELs

Mouse IELs were isolated from small intestine²¹. In brief, small intestine was opened, washed in PBS, cut into 0.5-cm-long pieces and incubated at room temperature in complete RPMI supplemented with 1 mM DTT. Tissues were then subjected to two rounds of washes with complete RPMI, vortex and filtered through 70-μm cell strainers. The IEL fraction was then purified by Percoll density gradient (80:40:20) centrifugation and stained by flow cytometry.

Spleen, lymph nodes and thymus cells isolation

After spleen, thymus and lymph node dissection, tissue was mechanically disaggregated and filtered through 40-μm cell strainers to obtain single-cell suspensions. Red blood cells were lysed with red blood lysis buffer for 5 min at room temperature. Lymphocytes were then washed with complete RPMI and stained for flow cytometry.

Isolation of primary DETC lines

Primary DETCs were isolated and grown as previously described⁶². After approximately 1 month, cells were assessed for purity by flow cytometry. Cell lines with ≥85% DETCs (short-term DETCs) were used for experiments.

Flow cytometry

To analyse γδ T cell populations, separate tissue-specific cell suspensions were prepared as described above from ears or back skin or gut of individual mice. γδ T cells were blocked with normal hamster IgG plus anti-FcR (2.4G2) and stained with specific antibodies (Supplementary Table 5). For mouse and human experiments, all commercial antibodies were used at a dilution of 1:100. In-house generated antibody was used at a dilution of 1:200. Isotype-matched control antibodies were used at the same concentrations as test antibodies. Analysis was performed with a LSR Fortessa (BD Biosciences) with electronic gates set on live cells by a combination of forward and side light scatter and viability dye exclusion. A minimum of 1,000 live events were collected per sample and data were analysed with FlowJo v10 software (BD). The anti-TCRγδ antibody clone 17D1 was used as previously described⁶¹ to assess Vγ5 epidermal γδ T cells. For spectral analysis, flow cytometry was performed using a Spectral Aurora flow cytometer (Cytek Biosciences, 5-laser system), capable of full-spectrum detection across UV (355 nm), violet (405 nm), blue (488 nm), yellow-green (561 nm) and red (640 nm) lasers. Fluorescence signals were recorded using the Cytek SpectroFlo software. Single-stained controls were included for spectral unmixing, using either compensation beads (UltraComp eBeads, Thermo Fisher) or cells. Negative and fluorescence-minus-one (FMO) controls were included to define gating strategies. Flow cytometry data were acquired using FACSDiva (9.2) with gating performed on live, singlet populations based on FSC-A versus FSC-H and viability dye exclusion. Dimensionality reduction techniques, including t-SNE were applied to visualize high-dimensional datasets using FlowJo v10 (BD Biosciences).

FACS for bulk RNA-seq and ATAC-seq

To isolate mouse γδ T cell populations, separated tissue-specific cell suspensions were prepared as described above, incubated with antibodies for 30 min at 4 °C in FACS buffer (0.5% FBS, 2 mM EDTA in PBS, pH 7.4), then washed twice in FACS buffer. Viable γδ T cell cells were sorted into PBS buffer + 0.04% BSA and retained on ice. Human HTLV1-immortalized Vδ1⁺ were stained with TCR Vδ1 antibody (Miltenyi Biotec, 130-118-968) and sorted into TCR⁺ and TCR⁻ cells. Sorted cells were confirmed to be >85% to 95% pure prior to RNA extraction.

In vitro stimulation assays

For analysis of in vitro production of IFNγ and IL-17 by intracellular flow cytometry, tissue-specific γδ T cells were purified after tissue preparations by FACS as describe above and were then cultured in complete RPMI medium and stimulated for 5 h with 25 ng ml⁻¹ phorbol 12-myristate 13-acetate (PMA) and 1 µg ml⁻¹ Ionomycin in the presence of 10 µg ml⁻¹ brefeldin A (BFA) at 37 °C, 5% CO₂. PMA, Ionomycin and BFA were all purchased from Sigma. For cytokine stimulation, sorted cells were stimulated in complete RPMI with IL-1β (10 ng ml⁻¹) plus IL-23 (50 ng ml⁻¹) or with IL-12 (100 ng ml⁻¹) plus IL-18 (100 ng ml⁻¹) for 21 h. BFA (10 µg ml⁻¹) was added for the final 5 h, and production of IFNγ or IL-17 was immediately assessed by intracellular cytometry. IL-1β, IL-23, IL-12 and IL-18 were purchased from R&D and resuspended following manufacturer’s instructions.

RT–qPCR

Epidermis from back skin was separated from dermis as described above. To isolate specific skin γδ T cell populations, cell suspensions were prepared as described in ‘Flow cytometry’. Samples were directly frozen in RLT buffer prior to RNA purification with DNase digest (RNeasy kit; QIAGEN). cDNA was generated using Superscript-II (Invitrogen, Thermo Fisher Scientific) and analysed using SYBR-green assay (Invitrogen) using a QuantStudio3 Real-time PCR machine (Applied Biosystems, Thermo Fisher Scientific). cDNA was analysed for TCR related genes by combining forward primers specific for each of the Vγ gene segment (Trgv4, Trgv5 or Trgv6) with a reverse primer specific for Trgc1 (or Trgc3). Expression levels of each gene were normalized to cyclophilin A (Ppia) by the dCt method and showed as average dCt of technical replicates per sample. Primer sequences are indicated in Supplementary Table 4.

Apoptosis and necrosis detection assay

Apoptosis and necrosis were evaluated using the Apoptosis/Necrosis Detection Kit following the manufacturer’s instructions (Abcam, ab176749). In brief, cells were collected, washed with PBS, and resuspended in the assay buffer provided with the kit. Cells were then incubated with Apopxin Green Indicator, which labels phosphatidylserine exposed on apoptotic cells, together with the CytoCalcein Violet 450 viability dye and 7-AAD necrosis indicator for 30 min at room temperature protected from light. Following staining, samples were analysed by flow cytometry. Live cells were identified as CytoCalcein Violet 450⁺ Apopxin Green⁻ 7-AAD⁻, apoptotic cells as Apopxin Green⁺ 7-AAD⁻, and necrotic/late-apoptotic cells as 7-AAD⁺. Debris and doublets were excluded based on forward and side scatter characteristics.

Confocal microscopy

Whole depilated split ear sheets were flattened in a histology cassette and fixed with 10% neutral-buffered formalin for 2 h at room temperature and permeabilized in 0.5% Triton X-100 for at least 1 h at room temperature. The epidermal sheets were blocked for 30 min with PBS containing 5% normal donkey serum (blocking buffer) and then incubated with primary antibodies overnight at room temperature in blocking buffer. z-sections were acquired on a Leica SP8 confocal microscope and processed and analysed using Fiji (2.3.0/1.53q)⁶³. Morphological parameter sphericity was calculated as the ratio of cell object border length to its volume, ranging from 0 to 1 with higher values corresponding to more spherical objects. For analysis of mTmG, thicker intestine sections were fixed with 4% paraformaldehyde overnight at at 4 °C overnight and were then included in 4% low-melting temperature agarose (Invitrogen). Sections of 80 μm were obtained with a Leica VT1200/VT1200 S vibratome and embedded in FluoromountG (ThermoFisher Scientific). GFP was acquired on a Leica SP8 confocal microscope (see above). For thinner gut mTmG sections, after overnight fixation guts were included and frozen in OCT prior cryoprotection with 30% sucrose. Using a Cryostat Leica CM3050S guts were sectioned at 20 um and embedded in FluoromountG. Tomato and GFP was acquired using the same confocal microscope.

TEWL measurement

Under inhalation anaesthesia, TEWL of the ear pinnae was measured using a Tewameter TM300 probe (Ck Electronic) under constant temperature and humidity conditions. Values were recorded as the average of ten consecutive measurements once readings were stabilized using MPA software (Ck Electronic).

Ultraviolet radiation model

Under inhalation anaesthesia, mice were positioned under a cardboard box, and one or both ears were exposed and fixed flat using Blu-Tac. UVB radiation (302 nm) was administered to dorsal ear skin for a dose of 45 mJ cm^–2 (C57BL/6) using an XX-15 Series UV bench lamp (UVP) mounted on a Perspex and plastic housing.

Ear swelling measurement

Total ear thickness was calculated as the average of three measurements of the ear pinna with a micrometer (Mitutoyo) after discarding the first measurement.

IMQ-induced psoriasis

To induce psoriasis, ears were treated daily with 5 mg Aldara cream (5% IMQ) per ear for 7 consecutive days. Ear thickness was measured using a micrometer starting 1 day before the first treatment (day 0) and set to 100%. Reddening and scaling were evaluated daily and scored. Reddening and scaling were assessed separately to calculate an average cumulative disease score. For treatment, mice were anaesthetized with isoflurane.

Skin chemical carcinogenesis

DMBA and TPA were purchased from Sigma and dissolved in acetone or 100% ethanol respectively. For cutaneous carcinogenesis, female mice were shaved on the back using hair clippers and allowed to rest for 5 days prior to DMBA initiation. DMBA (300 nmol) was carefully and slowly applied by pipette, in a 75 μl volume to the entire shaved skin area. Mice were rested for 1 week and 15 nmol TPA then applied 3 times a week to the back skin. Hair regrowth during the experiment was gently removed by clipping with trimmers. Mice were monitored daily, and cutaneous tumours were counted and measured with a calliper once weekly. Back skin and tumours were evaluated by visual inspection by an observer blinded to the experimental groups.

Epicutaneous inoculation of B16 melanoma cells

For B16 melanoma inoculation, mice were placed individually in an induction chamber, and anaesthetized with 5% isoflurane (Isoflo, Zoetis) in 100% oxygen with a delivery rate of 5 l min⁻¹ until loss of righting reflex. Lubricating eye gel (Lubrithal Eye Gel 10G Tube) was applied to the eyes to prevent drying. Mice were shaved (with a Wella razor) and depilated using wecprep blades (Pilling). Mouse B16-F10 melanoma cell line was obtained, authenticated and routinely tested for mycoplasma by the Cell Services STP in the Francis Crick Institute. B16-F10 melanoma were collected by washing with PBS, incubating cells at 37 °C for 5 min with 1× trypsin/EDTA solution (0.25% Trypsin-EDTA-Sigma) and washing with Hanks’ balanced saline solution (HBSS). For epicutaneous inoculation, the stratum corneum of back skin was removed by application and removal of cellophane tape (ScotchTM) five to eight times, and the scarified site was wiped with a cotton-tipped applicator soaked in PBS. B16-F10 cells (2 × 10⁵) were suspended in 20 μl of Matrigel basement membrane matrix (Corning) and applied to the scarified region. Mice were rested for about 10 min to allow solidification of Matrigel before application of Tegaderm film (3M) over the gel. It was removed 4 days later. Developing tumours were measured using a digital calliper. Tumour growth and health status were monitored according to institutional ethical guidelines.

Subcutaneous inoculation of B16 melanoma cells

For B16-F10 melanoma cell line inoculation, mice were placed individually in an induction chamber and anaesthetized with 5% isoflurane (Isoflo, Zoetis) in 100% oxygen at a flow rate of 5 l min⁻¹ until the loss of the righting reflex. Lubricating eye gel (Lubrithal Eye Gel 10G Tube) was applied to prevent corneal drying. The injection site, typically the right flank, was shaved using a Wella razor and cleaned with 70% ethanol. B16-F10 melanoma cells were collected by washing with PBS, incubating at 37 °C for 5 min with 0.25% trypsin/EDTA solution (Sigma), and washed with Hanks’ balanced saline solution (HBSS). For subcutaneous inoculation, B16-F10 cells (1 × 10⁶) were suspended in 100 μl of PBS and injected subcutaneously using a 27G needle into the flank region. Mice were observed for 5–10 min post-injection to ensure proper distribution of the suspension and were then returned to their cages. Developing tumours were measured using a digital calliper at regular intervals. Tumour growth and health status were monitored according to institutional ethical guidelines.

AOM plus DSS-induced colorectal tumour model

Colorectal tumorigenesis was induced by the administration of a single intraperitoneal injection of AOM (10 mg kg⁻¹,13.4 M, ≥98% Sigma-Aldrich A5486). One week later, mice received 1.5% dextran sulfate sodium (DSS, molecular mass approximately 40,000, Thermo Scientific Chemicals) in drinking water for 7 consecutive days, followed by 14 days of regular water. This cycle was repeated a total of three times. Body weight, stool consistency and rectal bleeding were monitored daily to assess disease. Mice exhibiting any harmful signs of suffering such as hunched posture, piloerection, rectal bleeding and >20% body weight loss were euthanized immediately. Following the final DSS cycle, mice were maintained on regular drinking water for an additional five weeks, after which colons were collected and analysed.

Colorectal tumour analysis

Colons were dissected, flush with cold PBS, opened longitudinally and gently flattened on a piece of filter paper with the mucosal side facing up. Tissues were then placed in 10% neutral-buffered formalin, and fixed 36 h at room temperature. First, tumours were assessed macroscopically. Each colon was then rolled tightly from the distal end towards the proximal end to form a ‘Swiss roll’. Fixed rolls were then processed for paraffin embedding, sectioned at 5 µm, and stained with haematoxylin and eosin for histological analysis.

Histopathological examination

Four-micrometre-thick, formalin-fixed, paraffin-embedded sections from colon were stained with haematoxylin and eosin and examined by a board-certified Veterinary Pathologists (ASB). Histopathological assessment was performed blind to experimental grouping using a light microscope (Olympus BX43). Four parameters were evaluated and included epithelial hyperplasia and/or goblet cell depletion, leucocyte infiltration into the lamina propria, area affected and markers of severe inflammation, which included crypt abscess formation, submucosal leucocyte infiltration, crypt branching, ulceration and fibrosis. Each parameter was scored on a scale of 0–3 producing a total score ranging from 0 to 12. The presence of neoplasia was semi quantitatively evaluated on a scale 0 to 5 based on the type (adenoma and adenocarcinoma), severity and extension of these lesions.

Bulk RNA-seq

cDNA was prepared from 10 ng input RNA using the NuGEN Ovation RNA-Seq System (V2), and libraries constructed using the NuGEN Ultralow Library System V2. Both these steps followed the manufacturer’s instructions. The resulting libraries were pooled for sequencing on an Illumina HiSeq 2500 platform with single-ended 75-bp reads.

Bulk ATAC-seq

ATAC-seq samples were prepared using the ATAC-seq Kit from Active Motif (53150) following the manufacturer’s instructions. For each sample, 50 000 cells were lysed in ice-cold ATAC Lysis Buffer and immediately spin down at 500g for 10 min at 4 °C. After supernatant removal, nuclei were resuspended in tagmentation master mix and incubated at 37 °C for 30 min while shaking at 800 rpm. Tagmentation reaction was stopped by addition of DNA Purification Binding buffer (Active Motif). Tagmented DNA fragments were purified and eluted in 35 μl of DNA Elution buffer (Active Motif), which were then amplified in a 50 μl PCR reaction using standard ATAC PCR conditions: 72 °C for 5 min; 98 °C for 30 s and thermocycling at 98 °C for 10 s, 63 °C for 30 s and 72 °C for 1 min for 10 cycles. Each 50 μl PCR reaction contained of 33.5 μl Tagmented DNA, 1 μl dNTPs (10 mM), 10 μl Q5 Reaction Buffer (5×), 0.5 μl of Q5 High-Fidelity Polymerase (2 U/μl), 2.5 μl Nextera i5 Indexed primer (25 μM) and 2.5 μl Nextera i7 Indexed primer (25 μM) (Illumina). Following amplification, ATAC-seq libraries were cleaned up using 60 μl of SPRI bead solution (1.2× the sample volume) and two 80% ethanol washes while being placed on a magnetic rack stand before being eluted in 20 μl DNA Purification Elution Buffer (Active Motif). All ATAC-seq libraries were examined on the TapeStation (Agilent) before sequenced on the NovaSeq 6000 system (Illumina), with paired-end read lengths of 50 bp and at least 50 million uniquely mapped reads per sample.

Bulk RNA-seq bioinformatic analysis

Reads alignment and quantification: Raw reads were processed using the publicly available nf-core⁶⁴ RNA-seq pipeline v3.3 with STAR⁶⁴ or RSEM⁶⁵ against mouse genome assembly GRCm38 and Ensembl transcript annotation release 95. Gene-level RSEM abundance estimates were imported into R (v4.4.3) using the DESeq2 1.46.0 package⁶⁶. Transcript counts were normalized for downstream analyses through VST using the default function. Experimental batch effect was removed using limma 3.62.2 package. PCA was performed with VST-normalized counts using plotPCA function from DESeq2.

Differential gene expression analysis based on the negative binomial distribution with Wald test between different genotypes was performed for each dataset using the DESeq function from DESeq2 1.46.0 package. Significance was determined by adjusted P value < 0.05 and a base mean >10. A combined list of significant DEGs from all pairwise comparisons was created and used to identify the number of genes upregulated (log₂ fold change >0) and downregulated (log₂ fold change <0), and to generate heat maps comparing VST-normalized counts across samples with Euclidean distance clustering using the pheatmap 1.0.12 package.

GSEA was performed using the fgsea 1.32.2 package⁶⁷. The analysis was performed on a ranked gene list generated using their level of differential expression (P value) and their log₂ fold change to score magnitude and direction of change. Gene sets were retrieved from the Molecular Signatures Database (MSigDB) using their mouse and human Hallmarks collections with the msigdbr 10.0.1package (https://igordot.github.io/msigdbr/). Additional R packages used for data analysis and plotting include tidyverse 2.0.0, org.Mm.eg.db 3.20.0, ggrepel 0.9.6, magrittr 2.0.3, ggplot2 3.5.1, and RcolorBrewer 1.1-3. All the code to analyse RNA-seq datasets are available at https://github.com/FrancisCrickInstitute/Cg1-KO.

Bulk ATAC-seq bioinformatic analysis

For read alignment and peak calling, sequencing reads were processed using the nf-core ATAC-seq pipeline (v2.1.2)⁶⁴, aligned to the mouse reference genome (mm10), and filtered using the ENCODE blacklist to exclude artefactual regions. Peak calling was performed with MACS2 in broad peak mode. Genotype-specific peak intersections were computed using bedtools intersect (v2.30.0), requiring a minimum of 50% reciprocal overlap between peaks. A peak was only considered to be genuine if present in all but one biological replicates for each genotype per organ group. Read counting was performed using featureCounts (v2.0.3) with–fracOverlap 0.2.

For differential accessibility analysis, read counts were imported into R 4.4.1 and filtered to include only standard chromosomes. GC content bias was addressed using the EDASeq package⁶⁸, utilizing full quantile normalization within and between lanes. Offsets incorporating both library size and GC content were used in a negative binomial generalized linear model framework implemented via edgeR. Differential accessibility was assessed with a likelihood ratio test. Peaks with a false discovery rate (FDR) <0.05 were considered significantly differentially accessible. Genome browser data visualization: Genome browser tracks were generated with bigwig coverage files from RNA-seq and ATAC-seq data and bed files for ATAC-seq peaks initially visualized with Integrative Genomics Viewer (IGV) and finally plotted using the ggbio 1.54.0 package with EnsDb.Mmusculus.v79 2.99.0 as gene database. All the code to analyse ATAC-seq datasets is available at https://github.com/FrancisCrickInstitute/Cg1-KO.

Metabolic analysis

Real-time metabolic profiling was carried out using the Seahorse Extracellular Flux Analyser system (Agilent) as previously described⁶⁹. In brief, human Vδ1 cells were seeded at a density of 100,000 cells per well in poly-d-lysine pre-coated Seahorse XFe96 plates in XF assay buffer (Agilent) and centrifuged at 400g for 5 min to settle cells. Plates were then allowed to equilibrate at 37 °C in a non-CO₂ incubator for 30 min. Mitochondrial function was assessed by measurement of OCR and ECAR under basal conditions and after sequential injection of 1 μM oligomycin, 1.5 μM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP), 100 nM rotenone and 1 μM antimycin A. To assess contribution of FAO to basal mitochondrial oxygen consumption, 300 μM etomoxir was injected in port A prior to injections of 1 μM oligomycin, 1.5 μM FCCP, 100 nM rotenone and 1 μM antimycin A. Data was acquired using the Seahorse XFe96 Extracellular Flux Analyzer and Wave software (Agilent) following manufacturer recommended settings and plotted using Prism v11 software (GraphPad Software).

Cancer cell killing assay

Cancer cell killing assays were performed using the IncuCyte Live-Cell Analysis System (Sartorius) to enable real-time monitoring of target cell viability. Cancer cell lines (HCT116, HT29 and A375) were maintained in culture with complete medium (DMEM for HCT116 and A375, or RPMI for HT29 supplemented with 10% fetal bovine serum, 100 U ml⁻¹ of penicillin and 100 mg ml⁻¹ of streptomycin). Cancer cell lines were authenticated and routinely tested for mycoplasma by the Cell Services STP in the Francis Crick Institute. Cancer cell lines were red-labelled to facilitate cell identification and quantification by stable transduction with a plasmid encoding mCherry and selected using puromycin (1 μg ml⁻¹). Red-labelled cancer cells were purified by FACS and culture with puromycin until assayed. Target cancer cells were seeded in 96-well flat-bottom plates at a density of 2 × 10³ cells per well in complete growth medium and allowed to adhere overnight at 37 °C in a humidified incubator with 5% CO₂. The following day, human γδ T cells were added to target cells at different effector-to-target (E:T) ratios (2:1, 5:1 or 10:1). Cell death was monitored using IncuCyte Cytotox Green fluorescent viability dye added at the manufacturer’s recommended concentration. Plates were placed into the IncuCyte system, and images were acquired every 3 h using a 10× objective. Phase contrast and fluorescence images were collected from 5 fields per well to ensure representative sampling and accurate quantification. Data were analysed using IncuCyte analysis software (Sartorius), and target cell area was quantified over time. All conditions were performed in technical triplicates and experiments were repeated independently at least twice. Data are presented as mean ± s.d.

Data collection software

Conventional flow cytometry data were collected on a BD Fortessa with FACSDiva Software v9.2 (BD). Spectral flow cytometry data were collected on a Aurora flow cytometer with SpectroFlow Software v3.3.0 (Cytek). Gene expression data by quantitative PCR with reverse transcription (RT–qPCR) were collected with QuantStudio Design and Analysis Software vl.5.2 (Applied Biosystems/Thermo Fisher Scientific). Seahorse data were collected with Wave Software v2.4.3.7 (Agilent Technologies). Live imaging lncucyte data were collected with lncucyte Controller Software v2022B Rev3 (Sartorius).

Normalization and statistical analysis

Groups were compared with Prism v11 software (GraphPad Software) using the 2-tailed unpaired Student t-test for comparison of 2 groups or 1-way ANOVA for comparison of 3 groups. For comparisons of three groups across time in multiple experiments, we used a linear mixed-effects model to capture experimental hierarchy on raw values, where time and sample were considered as fixed, and experiment as random effect. Dunnet’s test was used for multiple comparisons correction, where TRDV1-knockout cells were compared against control cells. Data are presented as each data point and mean or mean ± s.d. P < 0.05 was considered significant. The experiments were not randomized, and the investigators were not blinded to allocation during experiments and outcome assessment. No statistical methods were used to predetermine sample size. All experiments were performed at least twice, either with similar results obtained and representative data shown, or with pooled data shown. Graph bars with pooled data from multiple experiments include all experiments performed. Where representative histograms are shown, data reflect ≥2 independent experiments with ≥3 mice per experiment, in which similar results were obtained.

Ethics statement

All animal experiments in the UK were approved by the Home Office. All animal experiments in Spain were approved by the Comunidad de Madrid Animal Experimentation Ethics Committee and authorized by the Dirección General de Agricultura, Ganadería y Alimentación of the Comunidad de Madrid, and were performed in accordance with the relevant national and European legislation. The collection and use of human samples were approved by the Clinical Research Ethics Committee (CEIm) of Hospital Clínico San Carlos (Madrid, Spain) and were conducted in accordance with the principles of the Declaration of Helsinki and applicable national regulations. Written informed consent was obtained from all participants prior to sample collection.

Materials availability

Mouse lines generated in this study will be maintained in the animal house of the current institutes of A.C.H. and M.M.R. and/or stored locally as frozen embryos and can be made available on request.

Reporting summary

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