Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

A unique CD4⁺ T cell subset expressing granzyme K is regulated by transcription factor EOMES and important for T cell-mediated intestinal inflammation

Nature Immunology (2026)Cite this article

Subjects

Abstract

CD4⁺ helper T (TH) cells consist of multiple functional subsets defined by specific effector cytokines and transcription factors. Recently, single-cell transcriptomic analyses have revealed possible existence of additional populations. Here we identify a unique CD4⁺ T cell subset in mouse and human colitis characterized by high levels of granzyme K (Gzmk) expression, designated as THK cells. These cells exhibit unique transcriptional signatures, with minimal expression of classical TH-defining factors but rather prominent Eomesodermin (Eomes) expression. Notably, THK cell differentiation is independent of TH1, TH2 and TH17 lineages in colitis. EOMES is both necessary and sufficient for THK cell induction, by directly driving the expression of Gzmk and associated effector molecules. Genetic ablation of Eomes ameliorates intestinal immunopathology in a T cell-induced colitis model. The THK transcriptional program seems to be conserved across species and in diverse disease contexts. Our findings establish THK cells as a distinct TH cell subtype, and the EOMES–THK axis may serve as a potential therapeutic target in inflammatory diseases.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on SpringerLink
  • Instant access to the full article PDF.

USD 39.95

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Identification of a GZMK-expressing CD4⁺ T cell subset in human intestinal inflammation.
Fig. 2: A distinct Gzmk-expressing CD4⁺ T cell population in murine colitis.
Fig. 3: THK cell differentiation is independent of TH1, TH2 and TH17 programs.
Fig. 4: Eomes is transcriptionally and epigenetically associated with THK cell identity.
Fig. 5: EOMES drives the THK program and constrains alternative TH cell fates.
Fig. 6: Eomes is required for THK cell differentiation and intestinal immunopathology.

Data availability

The raw sequence data reported in this paper have been deposited in National Genomics Data Center under accession code PRJCA048888. The processed RNA-seq, ATAC-seq and CUT&Tag data of this study are available on figshare at https://doi.org/10.6084/m9.figshare.30477716 (ref. 67). Source data are provided with this paper.

Code availability

This article does not report any original code. All functionality used to analyze the sequencing data is publicly available via the software packages described in Methods.

References

  1. Dong, C. Cytokine regulation and function in T cells. Annu. Rev. Immunol. 39, 51–76 (2021).

    Article  CAS  PubMed  Google Scholar 

  2. Voskoboinik, I., Whisstock, J. C. & Trapani, J. A. Perforin and granzymes: function, dysfunction and human pathology. Nat. Rev. Immunol. 15, 388–400 (2015).

    Article  CAS  PubMed  Google Scholar 

  3. Aubert, A., Jung, K., Hiroyasu, S., Pardo, J. & Granville, D. J. Granzyme serine proteases in inflammation and rheumatic diseases. Nat. Rev. Rheumatol. 20, 361–376 (2024).

    Article  CAS  PubMed  Google Scholar 

  4. Donado, C. A. et al. Granzyme K activates the entire complement cascade. Nature 641, 211–221 (2025).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Lan, F. et al. GZMK-expressing CD8(+) T cells promote recurrent airway inflammatory diseases. Nature 638, 490–498 (2025).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Guo, C. L. et al. Granzyme K(+)CD8(+) T cells interact with fibroblasts to promote neutrophilic inflammation in nasal polyps. Nat. Commun. 15, 10413 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Guo, X. et al. Contrasting cytotoxic and regulatory T cell responses underlying distinct clinical outcomes to anti-PD-1 plus lenvatinib therapy in cancer. Cancer Cell 43, 248–268.e9 (2025).

    Article  CAS  PubMed  Google Scholar 

  8. Guo, C. L., Wang, C. S., Wang, X. H., Yu, D. & Liu, Z. GZMK(+)CD8(+) T cells: multifaceted roles beyond cytotoxicity. Trends Immunol. 46, 562–572 (2025).

    Article  CAS  PubMed  Google Scholar 

  9. Kwek, S. S. et al. Identification and regulation of circulating tumor-TCR-matched cytotoxic CD4+ lymphocytes by KLRG1 in bladder cancer. JCI Insight 10, e177373 (2025).

    Article  PubMed  PubMed Central  Google Scholar 

  10. Koga, R. et al. Granzyme K- and amphiregulin-expressing cytotoxic T cells and activated extrafollicular B cells are potential drivers of IgG4-related disease. J. Allergy Clin. Immunol. 153, 1095–1112 (2024).

    Article  CAS  PubMed  Google Scholar 

  11. Aoyagi, R. et al. Single-cell transcriptomics reveals granzyme K-expressing cytotoxic Tfh cells in tertiary lymphoid structures in IgG4-RD. J. Allergy Clin. Immunol. 153, 513–520.e10 (2024).

    Article  CAS  PubMed  Google Scholar 

  12. Elyahu, Y. et al. Aging promotes reorganization of the CD4 T cell landscape toward extreme regulatory and effector phenotypes. Sci. Adv. 5, eaaw8330 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Magen, A. et al. Intratumoral dendritic cell-CD4(+) T helper cell niches enable CD8(+) T cell differentiation following PD-1 blockade in hepatocellular carcinoma. Nat. Med. 29, 1389–1399 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Nie, H. et al. Single-cell meta-analysis of inflammatory bowel disease with scIBD. Nat. Comput. Sci. 3, 522–531 (2023).

    Article  CAS  PubMed  Google Scholar 

  15. Fenton, C. G., Taman, H., Florholmen, J., Sorbye, S. W. & Paulssen, R. H. Transcriptional signatures that define ulcerative colitis in remission. Inflamm. Bowel Dis. 27, 94–105 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  16. Zheng, L. et al. Pan-cancer single-cell landscape of tumor-infiltrating T cells. Science 374, abe6474 (2021).

    Article  PubMed  Google Scholar 

  17. Bai, X. et al. Reciprocal regulation of T follicular helper cells and dendritic cells drives colitis development. Nat. Immunol. 25, 1383–1394 (2024).

    Article  CAS  PubMed  Google Scholar 

  18. Li, Y. et al. Stem-like T cells are associated with the pathogenesis of ulcerative colitis in humans. Nat. Immunol. 25, 1231–1244 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Reis, B. S., Rogoz, A., Costa-Pinto, F. A., Taniuchi, I. & Mucida, D. Mutual expression of the transcription factors Runx3 and ThPOK regulates intestinal CD4⁺ T cell immunity. Nat. Immunol. 14, 271–280 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Mucida, D. et al. Transcriptional reprogramming of mature CD4(+) helper T cells generates distinct MHC class II-restricted cytotoxic T lymphocytes. Nat. Immunol. 14, 281–289 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Li, J., He, Y., Hao, J., Ni, L. & Dong, C. High levels of eomes promote exhaustion of anti-tumor CD8(+) T cells. Front. Immunol. 9, 2981 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Endo, Y. et al. Eomesodermin controls interleukin-5 production in memory T helper 2 cells through inhibition of activity of the transcription factor GATA3. Immunity 35, 733–745 (2011).

    Article  CAS  PubMed  Google Scholar 

  23. Ichiyama, K. et al. Transcription factor Smad-independent T helper 17 cell induction by transforming-growth factor-β is mediated by suppression of Eomesodermin. Immunity 34, 741–754 (2011).

    Article  CAS  PubMed  Google Scholar 

  24. Lupar, E. et al. Eomesodermin expression in CD4+ T cells restricts peripheral Foxp3 induction. J. Immunol. 195, 4742–4752 (2015).

    Article  CAS  PubMed  Google Scholar 

  25. Johnston, R. J. et al. Bcl6 and Blimp-1 are reciprocal and antagonistic regulators of T follicular helper cell differentiation. Science 325, 1006–1010 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Sun, Q. et al. BCL6 promotes a stem-like CD8(+) T cell program in cancer via antagonizing BLIMP1. Sci. Immunol. 8, eadh1306 (2023).

    Article  CAS  PubMed  Google Scholar 

  27. Tan, S. N. et al. Regulatory T cells converted from Th1 cells in tumors suppress cancer immunity via CD39. J. Exp. Med. 222, e20240445 (2025).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Zander, R., Khatun, A., Kasmani, M. Y., Chen, Y. & Cui, W. Delineating the transcriptional landscape and clonal diversity of virus-specific CD4(+) T cells during chronic viral infection. eLife 11, e80079 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Zhang, P. et al. Eomesodermin promotes the development of type 1 regulatory T (T(R)1) cells. Sci. Immunol. 2, eaah7152 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  30. Gruarin, P. et al. Eomesodermin controls a unique differentiation program in human IL-10 and IFN-γ coproducing regulatory T cells. Eur. J. Immunol. 49, 96–111 (2019).

    Article  CAS  PubMed  Google Scholar 

  31. Bonnal, R. J. P. et al. Clonally expanded EOMES(+) Tr1-like cells in primary and metastatic tumors are associated with disease progression. Nat. Immunol. 22, 735–745 (2021).

    Article  CAS  PubMed  Google Scholar 

  32. Zhang, P. et al. Eomesodermin+ CD4+ T cells are critical for curative immunotherapy outcomes. Immunity 58, 3024–3039.e7 (2025).

    Article  CAS  PubMed  Google Scholar 

  33. Thelen, B. et al. Eomes is sufficient to regulate IL-10 expression and cytotoxic effector molecules in murine CD4(+) T cells. Front. Immunol. 14, 1058267 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Ansaldo, E. et al. T-bet-expressing Tr1 cells driven by dietary signals dominate the small intestinal immune landscape. Proc. Natl Acad. Sci. USA 123, e2520747122 (2026).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Devarajan, P. et al. Cytotoxic CD4 development requires CD4 effectors to concurrently recognize local antigen and encounter type I IFN-induced IL-15. Cell Rep. 42, 113182 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Marshall, N. B. et al. NKG2C/E marks the unique cytotoxic CD4 T cell subset, ThCTL, generated by influenza infection. J. Immunol. 198, 1142–1155 (2017).

    Article  CAS  PubMed  Google Scholar 

  37. Takeuchi, A. et al. CRTAM determines the CD4+ cytotoxic T lymphocyte lineage. J. Exp. Med. 213, 123–138 (2016).

    Article  CAS  PubMed  Google Scholar 

  38. Mazzoni, A. et al. Eomes controls the development of Th17-derived (non-classic) Th1 cells during chronic inflammation. Eur. J. Immunol. 49, 79–95 (2019).

    Article  CAS  PubMed  Google Scholar 

  39. Curran, M. A. et al. Systemic 4-1BB activation induces a novel T cell phenotype driven by high expression of Eomesodermin. J. Exp. Med. 210, 743–755 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Raveney, B. J. et al. Eomesodermin-expressing T-helper cells are essential for chronic neuroinflammation. Nat. Commun. 6, 8437 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Stienne, C. et al. Foxo3 transcription factor drives pathogenic T helper 1 differentiation by inducing the expression of Eomes. Immunity 45, 774–787 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Zhang, C. et al. Extrapituitary prolactin promotes generation of Eomes-positive helper T cells mediating neuroinflammation. Proc. Natl Acad. Sci. USA 116, 21131–21139 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Raveney, B. J. E. et al. Involvement of cytotoxic Eomes-expressing CD4(+) T cells in secondary progressive multiple sclerosis. Proc. Natl Acad. Sci. USA 118, e2021818118 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Chen, S. et al. Eomesodermin expression in CD4(+) T-cells associated with disease progression in amyotrophic lateral sclerosis. CNS Neurosci. Ther. 30, e14503 (2024).

    Article  CAS  PubMed  Google Scholar 

  45. Joulia, E. et al. Eomes-dependent mitochondrial regulation promotes survival of pathogenic CD4+ T cells during inflammation. J. Exp. Med. 221, e20230449 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Kanazawa, T. et al. Pathogenic potential of Eomesodermin-expressing T-helper cells in neurodegenerative diseases. Ann. Neurol. 95, 1093–1098 (2024).

    Article  CAS  PubMed  Google Scholar 

  47. Elyahu, Y. et al. CD4 T cells acquire Eomesodermin to modulate cellular senescence and aging. Nat. Aging 5, 1970–1982 (2025).

    Article  CAS  PubMed  Google Scholar 

  48. Roessner, P. M. et al. EOMES and IL-10 regulate antitumor activity of T regulatory type 1 CD4(+) T cells in chronic lymphocytic leukemia. Leukemia 35, 2311–2324 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Schnell, A., Littman, D. R. & Kuchroo, V. K. T(H)17 cell heterogeneity and its role in tissue inflammation. Nat. Immunol. 24, 19–29 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Liu, Z. et al. Epigenetic reprogramming of Runx3 reinforces CD8+ T-cell function and improves the clinical response to immunotherapy. Mol. Cancer 22, 84 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Chi, X. et al. RORγt expression in mature T(H)17 cells safeguards their lineage specification by inhibiting conversion to T(H)2 cells. Sci. Adv. 8, eabn7774 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Liu, X. et al. Genome-wide analysis identifies Bcl6-controlled regulatory networks during T follicular helper cell differentiation. Cell Rep. 14, 1735–1747 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Alquicira-Hernandez, J. & Powell, J. E. Nebulosa recovers single-cell gene expression signals by kernel density estimation. Bioinformatics 37, 2485–2487 (2021).

    Article  CAS  PubMed  Google Scholar 

  54. Hao, Y. et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat. Biotechnol. 42, 293–304 (2024).

    Article  CAS  PubMed  Google Scholar 

  55. 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 

  56. Chen, S., Zhou, Y., Chen, Y. & Gu, J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884–i890 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  57. Kim, D., Paggi, J. M., Park, C., Bennett, C. & Salzberg, S. L. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 37, 907–915 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Danecek, P. et al. Twelve years of SAMtools and BCFtools. Gigascience 10, giab008 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  59. Liao, Y., Smyth, G. K. & Shi, W. The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote. Nucleic Acids Res. 41, e108 (2013).

    Article  PubMed  PubMed Central  Google Scholar 

  60. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  61. Xu, S. et al. Using clusterProfiler to characterize multiomics data. Nat. Protoc. 19, 3292–3320 (2024).

    Article  CAS  PubMed  Google Scholar 

  62. Ramirez, F., Dundar, F., Diehl, S., Gruning, B. A. & Manke, T. deepTools: a flexible platform for exploring deep-sequencing data. Nucleic Acids Res. 42, W187–W191 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  64. Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  66. Wang, Q. et al. Exploring epigenomic datasets by ChIPseeker. Curr. Protoc. 2, e585 (2022).

    Article  CAS  PubMed  Google Scholar 

  67. Xie, T. Processed RNA-seq, ATAC-seq and CUT&Tag data. figshare https://doi.org/10.6084/m9.figshare.30477716 (2026).

Download references

Acknowledgements

We thank Y. Qiu (Peking University, Beijing) for providing Stat6−/− mice, F. Shao (National Institute of Biological Sciences, Beijing) for Gzmk−/− mice and Y. Sun (Beijing Institute of Genomics, Chinese Academy of Sciences, Beijing) for Runx3fl/fl mice. We thank all Dong laboratory members for their help. This work was supported by grants from the National Natural Science Foundation of China (32550002 to C.D., 82402058 to Q.L. and 92574302 to Q.L.), the Zhejiang Provincial Natural Science Foundation of China (LD25C120002 to C.D.), the National Key R&D Program of China (2023YFC2508200 to H. Yuan) and the Construction Fund of Key Medical Disciplines of Hangzhou on Immune Diseases and Precision Therapy (2025HZGF01 to C.D.). C.D. was a New Cornerstone Investigator.

Author information

Author notes

  1. Xue Bai

    Present address: Department of Gastroenterology, Shengjing Hospital of China Medical University, Shenyang, China

Authors and Affiliations

  1. Institute for Immunology and School of Basic Medical Sciences, Tsinghua University, Beijing, China

    Tian Xie, Hao Zhang, Kun Wei, Xinxin Chi, Xue Bai & Yujie Fu

  2. Westlake University School of Medicine and New Cornerstone Laboratory, Hangzhou, China

    Tian Xie, Yizhou Du, Qihan Wang, Hao Zhang, Yujie Fu, Zhilin Peng, Yicheng Zhu, Qiuyan Lan & Chen Dong

  3. Peking University-Tsinghua University-National Institute of Biological Sciences Joint Graduate Program, Tsinghua University, Beijing, China

    Tian Xie

  4. Shanghai Immune Therapy Institute, Shanghai Jiao Tong University School of Medicine-affiliated Renji Hospital, Shanghai, China

    Yicheng Zhu

Authors
  1. Tian Xie
    View author publications

    Search author on:PubMed Google Scholar

  2. Yizhou Du
    View author publications

    Search author on:PubMed Google Scholar

  3. Qihan Wang
    View author publications

    Search author on:PubMed Google Scholar

  4. Hao Zhang
    View author publications

    Search author on:PubMed Google Scholar

  5. Kun Wei
    View author publications

    Search author on:PubMed Google Scholar

  6. Xinxin Chi
    View author publications

    Search author on:PubMed Google Scholar

  7. Xue Bai
    View author publications

    Search author on:PubMed Google Scholar

  8. Yujie Fu
    View author publications

    Search author on:PubMed Google Scholar

  9. Zhilin Peng
    View author publications

    Search author on:PubMed Google Scholar

  10. Yicheng Zhu
    View author publications

    Search author on:PubMed Google Scholar

  11. Qiuyan Lan
    View author publications

    Search author on:PubMed Google Scholar

  12. Chen Dong
    View author publications

    Search author on:PubMed Google Scholar

Contributions

T.X. and C.D. conceived of the study. T.X. performed the experiments, conducted bioinformatic analyses and drafted the paper. Y.D., Q.W., H.Z., K.W., X.C., X.B., Y.F., Z.P., Y.Z. and Q.L. provided experimental support. C.D. supervised the research and revised the paper.

Corresponding author

Correspondence to Chen Dong.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Immunology thanks the anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available. Primary Handling Editor: L. A. Dempsey, 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 A GZMK-high CD4⁺ T cell population in integrated pan-cancer scRNA-seq data.

UMAP visualization of CD4⁺ T cells from a published integrated pan-cancer scRNA-seq dataset. Density plots show expression of SELL, CCR7, TBX21, IFNG, GATA3, RORC, BCL6, CXCR5, FOXP3, IL2RA, GZMB, and GZMK.

Extended Data Fig. 2 Analysis of Gzmk-tdTomato reporter expression across intestinal and lymphoid tissues in steady-state mice.

Flow cytometric analysis of lymphocytes from small intestinal intraepithelial lymphocytes (SI-IEL), small intestinal lamina propria lymphocytes (SI-LPL), large intestinal IEL (LI-IEL), large intestinal LPL (LI-LPL), Peyer’s patches (PP), and mesenteric lymph nodes (mLN) of Gzmk-tdTomato reporter mice at steady-state. Left panels: representative plots showing tdTomato signal in CD4⁺, CD8αα⁺, CD8αβ⁺, and γδ T cell subsets. Right panels: frequency of tdTomato⁺ cells within each subset (n = 5). Data are shown as mean ± s.e.m. and are representative of three independent experiments.

Source data

Extended Data Fig. 3 Phenotypic and functional characterization of Gzmk-expressing CD4⁺ T cells.

a, Flow cytometry analysis of Gzmk-tdTomato expression in mesenteric lymph nodes (mLN) CD4⁺ T cells of Rag1−/− colitis mice adoptively transferred with naive CD4⁺ T cells from Gzmk-tdTomato reporter mice. Upper: representative plots (y-axis: Gzmk-tdTomato; x-axis: T-bet, GATA3, RORγt, BCL6, or FOXP3). Lower: quantification of MFI of T-bet (n = 8), GATA3 (n = 3), and BCL6 (n = 8), and frequency of RORγt⁺ (n = 8) and FOXP3⁺ (n = 8) cells in Gzmk-tdTomato⁺ versus Gzmk-tdTomato⁻ cells. b–d, Flow cytometry analysis of perforin (b), GZMB (c), and IFNγ (d) in mLN CD4⁺ T cells. Left: representative plots (y-axis: Gzmk-tdTomato; x-axis: target protein). In d, right: mean fluorescence intensity (MFI) of IFNγ in IFNγ⁺Gzmk-tdTomato⁻ and IFNγ⁺Gzmk-tdTomato⁺ subsets (n = 8). e, Violin plots showing expression of Gzmk, Prf1, Gzmb, and Ifng across CD4⁺ T cell clusters (cluster 0, n = 5,932; cluster 1, n = 3,502; cluster 2, n = 2,703; cluster 3, n = 1,703; cluster 4, n = 1,610; cluster 5, n = 133). Violin plots with embedded box plots showing median (center line) and 25th–75th percentiles (box bounds). Numbers above: adjusted P values (Wilcoxon rank-sum test, two-tailed, Cluster 1 versus all others); below: average expression levels. f, Proportion of Gzmk-tdTomato⁺ cells among sorted populations after 19-h co-culture with P815 target cells. g, In vitro cytotoxicity assay measuring proportion of dead P815 cells at varying effector-to-target ratios, with or without anti-CD3 antibody (n = 3). Data are shown as mean ± s.e.m. and are representative of at least two independent experiments. Exact P values are shown in each graph. Statistical significance was assessed by two-tailed paired t-test (a,d).

Source data

Extended Data Fig. 4 THK cell differentiation is independent of TH1, TH2, and TH17 programs.

a,b, Flow cytometry analysis of GZMB (a) and perforin (b) in mesenteric lymph nodes (mLN) CD4⁺ T cells from mice co-transferred with wild-type and Tbx21−/− cells (n = 6). Representative plots and quantification as in Fig. 3a, b. c,d, Flow cytometry analysis of perforin in mLN CD4⁺ T cells from recipients of wild-type and Stat6−/− (n = 7) (c) or Cd4cre Rorcfl/fl (n = 6) (d) cells; quantification as in Fig. 3e. e, Gzmk and Prf1 mRNA expression in sorted wild-type and Cd4cre Bcl6fl/fl CD4⁺ T cells from the colonic LP, measured by RT–qPCR (n = 3). Data are shown as mean ± s.e.m. and are representative of two independent experiments. Exact P values are shown in each graph. Statistical significance was assessed by two-tailed paired t-test (a–d) or two-tailed unpaired t-test (e).

Source data

Extended Data Fig. 5 THK cells are distinct from CD4⁺CD8α⁺ cytotoxic T cells.

a–c, Violin plots showing expression of Runx3 (a), Zbtb7b (b) and Cd8a (c) across CD4⁺ T cell clusters (cluster 0, n = 5,932; cluster 1, n = 3,502; cluster 2, n = 2,703; cluster 3, n = 1,703; cluster 4, n = 1,610; cluster 5, n = 133). Violin plots with embedded box plots showing median (center line) and 25th–75th percentiles (box bounds). Numbers above: adjusted P values (Wilcoxon rank-sum test, two-tailed, Cluster 1 versus all others); below: average expression levels. d, Flow cytometric analysis showing CD8α expression in Gzmk-tdTomato⁺ CD4⁺ T cells compared to other CD4⁺ T cell populations in lamina propria (LP). e,f, Flow cytometry analysis of CD8α (n = 4) and GZMB (n = 5) in LP CD4⁺ T cells from recipients of wild-type and Vav1cre Runx3fl/fl CD4⁺ T cells g, Gzmk mRNA expression in sorted wild-type and Vav1cre Runx3fl/fl CD4⁺ T cells from the colonic LP, measured by RT–qPCR (n = 3). Data are shown as mean ± s.e.m. and are representative of two independent experiments. Exact P values are shown in each graph. Statistical significance was assessed by two-tailed paired t-test (d–f) or two-tailed unpaired t-test (g).

Source data

Extended Data Fig. 6 Eomes is transcriptionally and epigenetically associated with THK cell identity.

a, UMAP embedding of scRNA-seq data from colitic CD4⁺ T cells, colored by ‘THK cell signature’ enrichment score. b, Eomes expression across CD4⁺ T cell clusters (cluster 0, n = 5,932; cluster 1, n = 3,502; cluster 2, n = 2,703; cluster 3, n = 1,703; cluster 4, n = 1,610; cluster 5, n = 133). Violin plots with embedded box plots showing median (center line) and 25th–75th percentiles (box bounds). Numbers above: adjusted P values (Wilcoxon rank-sum test, two-tailed, Cluster 1 versus all others); below: average expression levels. c, Density plots showing GZMK and EOMES expression in human pan-cancer dataset. d, Correlation analysis between GZMK and all other genes in bulk RNA-seq data from ulcerative colitis patients and healthy controls (GSE128682). Shown is R2 from Pearson correlation; P values are unadjusted. e, Time-course analysis of EOMES, Gzmk-tdTomato, T-bet, and RORγt expression in CD4⁺ T cells from the lamina propria (LP) and mesenteric lymph nodes (mLN) at days 7, 14, 21, 28, and 35 after naive CD4⁺ T cell transfer. Left: Representative flow cytometry plots showing coexpression of EOMES and Gzmk-tdTomato (top), T-bet (middle), and RORγt (bottom) in LP. Right: Quantification of EOMES⁺Gzmk-tdTomato⁺, T-bet⁺, and RORγt⁺ CD4⁺ T cells in LP and mLN (n = 3 for each time point). f–i, Flow cytometry analysis of EOMES expression in mLN CD4⁺ T cells from Rag1−/− recipients co-transferred with wild-type (CD45.1⁺) and knockout (CD45.2⁺) naive CD4⁺ T cells. Tbx21−/− (n = 6) (f), Stat6−/− (n = 7) (g), Cd4cre Rorcfl/fl (n = 6) (h), and Cd4cre Bcl6fl/fl (n = 5) (i). Left: representative plots. Right: frequency of EOMES⁺ cells within donor-derived populations. Data are shown as mean ± s.e.m. and are representative of two independent experiments. Exact P values are shown in each graph. Statistical significance was assessed by two-tailed paired t-test (f–i).

Source data

Extended Data Fig. 7 EOMES drives the THK program and constrains alternative TH cell fates.

a, RT–qPCR analysis of Gzmk, Prf1, and Gzmb expression in control or Eomes-overexpressing CD4⁺ T cells activated under non-polarizing conditions. Data normalized to Actb and presented relative to control (n = 3 per group; one undetected replicate in the Gzmk control group). b, Reanalysis of published RNA-seq data (GSE122889) from control or Eomes-overexpressing CD8⁺ T cells. Expression levels of Gzmk, Prf1, and Gzmb shown as FPKM. c, EOMES and Gzmk-tdTomato expression in CD4⁺ T cells under in vitro TH1, TH2, TH17, Treg polarizing conditions, and in vivo in adoptively transferred OT-II×Gzmk-tdTomato CD4⁺ T cells from OVA-immunized recipient mice. d,e, Flow cytometry analysis of T-bet (d) and GATA3 (e) in control or Eomes-overexpressing CD4⁺ T cells cultured under TH1- and TH2-polarizing conditions, respectively (n = 3). Left: representative plots. Right: quantification of frequency. f,g, Flow cytometry analysis of RORγt (f) and IL-17A (g) in control or Eomes-overexpressing CD4⁺ T cells under TH17-polarizing conditions (n = 3). Left: representative plots. Right: quantification of frequency. h, Flow cytometry analysis of FOXP3 in control or Eomes-overexpressing CD4⁺ T cells under Treg-polarizing conditions (n = 3). Left: representative plots. Right: quantification of frequency. i,j, Flow cytometry analysis of PD-1 and CXCR5 (i) and BCL6 and CXCR5 (j) in donor-derived OT-II CD4⁺ T cells 7 days after transfer into wild-type recipients and immunization with OVA protein (n = 8 per group). Left: representative plots. Right: quantification of frequency. k, Expression of EOMES and BLIMP1-EYFP in CD4⁺ T cells isolated from the lamina propria (LP) and mesenteric lymph nodes (mLN) of colitis mice. Data are shown as mean ± s.e.m. and are representative of at least two independent experiments. Exact P values are shown in each graph. Statistical significance was assessed by two-tailed unpaired t-test.

Source data

Extended Data Fig. 8 EOMES directly binds to regulatory elements of THK-associated genes.

a, Genomic annotation of EOMES-bound regions identified by CUT&Tag in Eomes-overexpressing CD4⁺ T cells. Pie chart shows distribution across promoters, introns, distal intergenic regions, and other features. b, Transcription factor motif enrichment analysis in EOMES-bound peaks. Shown are fold enrichment (target/background) and−Log(p value) (natural log) from HOMER (binomial test). Dashed gray lines: P = 0.05 and fold enrichment = 1. c, Reanalysis of published ChIP-seq data (GSE122888) showing EOMES occupancy at the Gzmk and Prf1 loci in control or Eomes-overexpressing CD8⁺ T cells. Tracks display EOMES ChIP-seq signal and control signal (mm10).

Extended Data Fig. 9 Eomes ablation alters CD4⁺ T cell differentiation in vivo.

a–m, Flow cytometric analysis of transcription factors and effector molecules in donor-derived CD4⁺ T cells from colonic lamina propria (LP) and mesenteric lymph nodes (mLN) (n = 5 or 6). Representative plots showing EOMES and Gzmk-tdTomato expression, with quantification of the frequency of Gzmk-tdTomato⁺ cells in mLN (a). Representative plots (left) and frequencies (right) of perforin⁺ (b) and GZMB⁺ (c) cells in mLN. d–m, Frequencies of T-bet⁺ (d,e), IFNγ⁺ (f,g), RORγt⁺ (h,i), IL-17A⁺ (j,k), and FOXP3⁺ (l,m) cells in LP (upper), mLN (lower). n, Frequencies of CD8α⁺ cells in CD4⁺ T cells in LP. o,p, Disease severity in Rag1−/− recipients of naive CD4⁺ T cells from Gzmk−/− (n = 5) or control (n = 5) donors. Body weight change (o); colon length at endpoint (p). Data are shown as mean ± s.e.m. and are representative of three independent experiments. Each data point represents an individual mouse. Exact P values are shown in each graph. Statistical significance was assessed by two-tailed unpaired t-test (a–n,p) and two-way ANOVA with Bonferroni’s post hoc test (o).

Source data

Extended Data Fig. 10 Analysis of the THK transcriptional program across species and diseases.

a,b, Differential gene expression analysis comparing human GZMKhigh CD4⁺ T cells to other CD4⁺ subsets in published scRNA-seq datasets from inflammatory bowel disease (IBD) patients (a; data from Fig. 1a) and pan-cancer samples (b; data from Extended Data Fig. 1). Differential expression was assessed using the Wilcoxon rank-sum test (two-tailed) implemented in the Seurat package. P values were adjusted using the Bonferroni correction. Red dots: upregulated in GZMKhigh CD4⁺ T cells; blue dots: upregulated in other CD4⁺ T cells. Horizontal dashed line: adjusted P = 0.05; vertical dashed lines: log2(fold change) = ±1. c, Overlap between human orthologs of upregulated genes from bulk RNA-seq of Gzmk-tdTomato⁺ versus Gzmk-tdTomato⁻ CD4⁺ T cells in murine colitis (Fig. 4a) and upregulated genes in human GZMKhigh subsets (a,b). d–f, UMAP visualization of CD4⁺ T cells from murine disease models: Hepa1-6 tumor-infiltrating lymphocytes (GSE285225; d), CNS-infiltrating CD4⁺ T cells in EAE (GSE156196; e), and splenic CD4⁺ T cells during chronic LCMV cl13 infection (GSE201730; f). Gzmkhigh subsets are highlighted; Gzmk and Eomes expression are shown. g–i, Gene set enrichment scores of the ‘THK cell signature’ in Gzmkhigh versus other CD4⁺ T cell subsets in tumor (g; Gzmkhigh, n = 2,943; others, n = 42,419), EAE (h; Gzmkhigh, n = 1,882; others, n = 18,003), and LCMV cl13 (i; Gzmkhigh, n = 341; others, n = 7,031) models. Violin plots with embedded box plots show median (center line) and 25th–75th percentiles (box bounds). Numbers above: adjusted P values (Wilcoxon rank-sum test, two-tailed, Gzmkhigh versus others); below: mean scores. j–l, Volcano plots of differentially expressed genes in Gzmkhigh CD4⁺ T cells versus other CD4⁺ T cell subsets in tumor (j), EAE (k), and LCMV cl13 (l) models. Differential expression was assessed using the Wilcoxon rank-sum test (two-tailed) implemented in the Seurat package. P values were adjusted using the Bonferroni correction. Red dots: upregulated in Gzmkhigh CD4⁺ T cells; blue dots: upregulated in other CD4⁺ T cells. Horizontal dashed line: adjusted P = 0.05; vertical dashed lines: log2(fold change) = ±1. m, Venn diagram showing intersection of significantly upregulated genes across all four murine disease models. All upregulated genes in this figure were defined using the thresholds: adjusted P < 0.05 and log2(fold change) > 1.

Supplementary information

Source data

Source Data Fig. 1 (download XLSX )

Statistical source data.

Source Data Fig. 2 (download XLSX )

Statistical source data.

Source Data Fig. 3 (download XLSX )

Statistical source data.

Source Data Fig. 4 (download XLSX )

Statistical source data.

Source Data Fig. 5 (download XLSX )

Statistical source data.

Source Data Fig. 6 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 2 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 3 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 4 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 5 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 6 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 7 (download XLSX )

Statistical source data.

Source Data Extended Data Fig. 9 (download XLSX )

Statistical source data.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Xie, T., Du, Y., Wang, Q. et al. A unique CD4⁺ T cell subset expressing granzyme K is regulated by transcription factor EOMES and important for T cell-mediated intestinal inflammation. Nat Immunol (2026). https://doi.org/10.1038/s41590-026-02479-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • DOI: https://doi.org/10.1038/s41590-026-02479-6

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing