henhao
Sign in
ImmunologybioRxiv
Preprint

Cysteine reactivity profiling links metabolic rewiring to the dysfunctional human T cell proteome.

Hiroyuki KojimaiD, Charlotte R. WayneiD, Luis F. Somarribas PattersoniD, Henry SanfordiD, Tzu-Jou CheniD, Ya-Hui LiniD, Joshua David SchoenfeldiD, Lisa H.F. McGaryiD, Yanting CheniD, Korbinian Nepomuk KroppiD, Beatrice D. ZhangiD, Jahan A. RahmaniD, Tiffany L. ZhangiD, Nathalie RopekiD, Cameron G. RobertsiD, Yuxi AiiD, Kartikeya M. MenoniD, A. Ari HakimiiD, Jiankun LyuiD, Christopher Austin KlebanoffiD, Omar Ibrahim Abdel-WahabiD, Santosha Adipudi VardhanaiD, Ekaterina V. VinogradovaiD

Abstract

T cells exposed to persistent antigen in the context of chronic viral infections or tumors lose cytotoxic capacity. Post-transcriptional regulatory mechanisms contributing to human T cell dysfunction remain poorly understood. Complementary proteomic and cysteine reactivity profiling of human T cells undergoing chronic antigen-driven dysfunction identified post-translational changes associated with activation of multiple stress pathways, including the stress kinase MAP2K4, which were validated in primary TILs. Functional analysis revealed that many reactivity changes occurred within nucleotide-binding proteins, reflecting ATP loss due to chronic antigen-driven suppression of oxidative TCA cycle metabolism. Targeting MAP2K4 restored T cell cytotoxic capacity during prolonged tumor cell co-culture. These findings reveal a mechanism by which metabolic rewiring alters protein function and lay the groundwork for cysteine-directed therapeutics to enhance cancer immunotherapy.

PDFDOI
Full-text digest ↓

02 /Specification & Classification

Click any tag to browse matching papers
Citations
1
OpenAlex Topics

In-depth summary

✦ Multi-agent verified full-text synthesis

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (2 Oct 2026) · CC BY 4.0

In chronically stimulated human T cells, suppressed oxidative metabolism and nucleotide loss accompany MAP2K4 activation. Inhibiting MAP2K4 (with the MKK4/7 inhibitor BSJ-04-122) restored tumor-cell killing in vitro.

The paper says post-transcriptional mechanisms in human T cell dysfunction are poorly understood, and that neither the functional state of the exhausted T cell proteome nor its dependence on metabolic state had been established. It pairs cysteine reactivity profiling with metabolomics to connect nucleotide depletion to specific protein-level changes and to stress-pathway activation.

Key results

  1. ShownAn in vitro chronic-stimulation system for human T cells reproduced transcriptional and functional hallmarks of exhaustion, including impaired serial killing of NY-ESO-1-expressing melanoma cells.

    Progressive rise in a 43-gene chronic activation score by bulk RNA-seq; nearly all D15 chronic CD8+ cells mapped to tumor exhausted CD8+ states by scRNA-seq label transfer; D8C killing impaired vs D8A (p < 0.01, n = 3-5 donors). · Transcriptome-independent functional proteomic…; Figure legends

  2. ShownCysteine reactivity profiling found changes that expression data did not show, concentrated in nucleotide-binding proteins, including higher reactivity of MAP2K4_C246 and MAP2K3_C227 in chronic cells.

    More than 150 reactivity changes (cutoff >2-fold vs D2, n = 5 donors), the most in D8C; MAP2K3/4 activation confirmed by Western blot in D8C but not D8A. · Transcriptome-independent functional proteomic…; Figure legends

  3. SuggestedChronic stimulation suppressed oxidative TCA metabolism and mitochondrial ATP production, consistent with a collapse of the nucleotide pool; ATP add-back to lysates (2 donors) decreased cysteine reactivity at MAP2K3/4 and other nucleotide-binding proteins, more so in chronic cells. Mitochondrial ROS did not rise.

    Apparent ATP EC50 for MAP2K4 0.077 mM (D2) vs 0.300 mM (D8C); MAP2K3 0.134 vs 0.328 mM. Lower m+2/m+4 13C labeling of TCA intermediates; mitochondrial ATP production reduced (n = 4 donors); GSSG:GSH unchanged. · Impaired nucleotide synthesis alters the…; Metabolic rewiring to limit ROS accumulation…; Figure legends

  4. SuggestedPhosphorylation of MAP2K4 and JNK was also raised in exhausted T cells from mouse LCMV Clone 13 infection, KPC tumors and human renal cell carcinoma TILs, as seen in vitro.

    Western blots of sorted mouse CD8+ T cells and TILs from two RCC patients; proteomic PCA used five RCC patients and showed that in vivo cells differ from in vitro cells. · Stress kinase activation in in vivo exhausted T…; Figure legends

  5. ShownThe covalent MKK4/7 inhibitor BSJ-04-122 shifted the chronic-cell proteome toward the acute state, lowered exhaustion markers and restored tumor killing; thapsigargin induced exhaustion-like changes, while ISRIB did not.

    Proteomics in n = 4 donors (4,198 proteins); lower LAG3, PDCD1, GZMB vs DMSO; BSJ-04-122 significantly reduced tumor-cell abundance over 8 days of co-culture (one-way repeated measures ANOVA; exact p-value not given in the text). · Stress pathway activation limits the cytotoxic…; Figure legends

Methods, data and reproducibility

Primary human T cells were activated and then expanded with or without continued anti-CD3 stimulation, then profiled by TMT-multiplexed proteomics, iodoacetamide-probe cysteine reactivity (TMT-ABPP, with ATP add-back), RNA-seq, metabolomics, isotope tracing and flux assays. Findings were checked in mouse and human tumor T cells and with small-molecule perturbations.

  • In vitro activated, acute and chronically stimulated T cells: unenriched proteomics (n = 6 donors), TMT-ABPP (n = 5), RNA-seq (n = 3), polar metabolomics (n = 4) · human · n = 3 to 6 donors by assay
  • ATP add-back TMT-ABPP and detergent-soluble/insoluble proteomics · human · n = 2 donors each
  • Sorted CD8+ T cells from LCMV-Armstrong, LCMV-Clone 13 and KPC tumor-bearing mice vs in vitro mouse T cells · mouse
  • PD-1+ CD3+ TILs from renal cell carcinoma tumors vs in vitro T cells (low-input proteomics) · human · n = 5 patients
  • Serial cytotoxicity assays with NY-ESO-1 TCR T cells against A375 and SK-MEL-37 melanoma cells · human · n = 3-5 donors
Methods & tools
TMT-ABPP, IA-DTB probe, IP2 / ProLuCID, DESeq2, STAR, GSEAPy, scikit-learn, Seurat, MassHunter Profinder, BSJ-04-122, ISRIB, thapsigargin
Reagents & environment
Versions given include STAR v2.7.0a, DESeq2 v1.42, scikit-learn 1.4.2, GSEAPy v1.1.4, imputeLCMD 2.1, Seurat v5.1.0, UniProt release 2024_03.
Reproducibility gaps
Raw proteomic and RNA-seq data are said to be deposited in PRIDE and GEO, but no accessions or URLs are printed. Code and the web portal 'will be deposited' on GitHub; other information is available on request.
Limitations (4)
  • The in vitro system does not capture tumor-microenvironment inputs; proteomics of mouse and human tumor T cells shows additional environmentally driven changes not reproduced in culture. · stated by the authors
  • ATP add-back was done in lysates, which lose subcellular organization and metabolite gradients, so it may not reflect ATP effects in intact cells. · stated by the authors
  • Mitochondrial ROS is not excluded as a contributor in vivo or over longer chronic exposure; thapsigargin effects may involve calcium signaling. · stated by the authors
  • Several key datasets are small: ATP add-back and fractionation proteomics use 2 donors, and in vivo MAP2K4-JNK validation relies on Western blots (two RCC patients). · noted in this summary

Who should read:Read this if you want a cysteine chemoproteomics case study connecting T cell metabolic state to nucleotide-sensitive protein regulation and to MAP2K4 as a candidate target.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISYosef '25Mitra '26Mandal '26Bresser '26Hiemer '19Meston '20
64% semantic matchImmunology
Omri Yosef, Leonor Cohen‐Daniel, Oded Shamriz+11 more·Nature Communications·May 8, 2025· 7 citations