Search
Search articles by title, abstract,author or doi
Advanced search

Clonotype-resolved T-cell response monitoring via repertoire-wide multidimensional TCR decoding

Hongyu Liu , Shian Ouyang , Rong Yin , Wenhao Yu , Xinlu Zhao , Zihan Zhao , Mingxin Fan , Jing Li , Wenqi Zhu , Jie Wang , Xingwang Xie , Rong Yang , Zijian Guo , Peng R. Chen , Jie P. Li

Vita ››

PDF (1697KB)
Vita > Correspondence > DOI: 10.15302/vita.2026.07.0054
Vita Open AccessPublished:

Clonotype-resolved T-cell response monitoring via repertoire-wide multidimensional TCR decoding

Author information +
History +
Vita () Cite this article
PDF (1697KB)

登录浏览全文

4963

注册一个新账户 忘记密码

Dear Editor,
Antigen-specific T-cell responses across the T-cell receptor (TCR) repertoire are central to immunotherapies including immune checkpoint blockade, tumor vaccines, and adoptive T-cell therapy1–3. However, current methods cannot simultaneously achieve broad antigen coverage, high TCR resolution, and high throughput, and therefore provide only a partial view of T-cell responses. Peptide–major histocompatibility complex (pMHC)-multimers are restricted to predefined epitopes and cannot assess the response capacity of isolated TCRs4. Conversely, while antigen-presenting cell (APC)/T-cell co-culture assays detect responses to unknown antigens through activation signaling, they lack sufficient specificity and remain challenging to integrate into high-throughput sequencing pipelines5,6. Here, we introduce T-SCOPE (TCR Signaling-COupled response Potency Evaluation), a multidimensional platform for the repertoire-wide monitoring of T-cell response potency across the entire TCR repertoire (Fig. 1a).
To develop an activity-based T-cell response recording platform, we considered a chemical dye (dibromofluorescein, DBF) as the labeling catalyst in T-SCOPE. DBF can be non-genetically displayed on various cell types, including patient-derived primary APCs, via mgSrtA-mediated conjugation in a biocompatible manner7. In DBF-anchored APC/T-cell co-cultures, antigen-specific pMHC–TCR recognition triggers immune synapse formation. DBF-mediated proximity labeling then selectively tags T cells forming these contacts, leaving non-specific cells unlabeled. Thus, this approach provides a functional, synapse-based readout of dynamic cell–cell interactions.
Guided by this rationale, we first used 1G4-3 Jurkat cells (Supplementary Fig. S3a) recognizing the NY-ESO-1 epitope SLLMWITQC (SLL-1G4) to compare T-cell responses elicited by T-SCOPE versus direct SLL-pentamer stimulation. Notably, antigen-specific T-SCOPE labeling was consistently accompanied by CD69 upregulation versus non-specific labeling (E629–38), whereas SLL-pentamer staining failed to induce this activation marker (Fig. 1b; Supplementary Fig. S1a–d). Transcriptomic profiling further revealed that only the T-SCOPE-labeled (Bio+) population exhibited a broad activation program, characterized by the upregulation of TCR signaling-responsive genes such as NR4A1, DUSP2, NFATC1, and PDCD1 (Supplementary Fig. S1e–j). These findings demonstrate that, unlike conventional pentamer staining, T-SCOPE labeling is directly correlated with functional T-cell activation.
We further investigated whether T-cell response was essential for T-SCOPE capture. Inhibiting TCR-proximal signaling via Lck or Src family kinase inhibitors (dasatinib, PP2 and RK24466) significantly reduced T-SCOPE labeling in both mouse and human T cells (Fig. 1c; Supplementary Fig. S2). In contrast, perturbing downstream pathways (JAK or calcium signaling) had minimal impact. Additionally, disrupting actin polymerization and immune synapse formation with cytochalasin D (CytoD) completely abolished T-SCOPE capture (Fig. 1c; Supplementary Fig. S2b–h). Crucially, none of these treatments affected pMHC-pentamer staining (Fig. 1d; Supplementary Fig. S2i–k), demonstrating that T-SCOPE requires TCR-proximal signaling and synapse formation, whereas pentamer staining does not.
Next, we examined whether T-SCOPE labeling reflects T-cell response potency. Using 1G4-TCR variants specific to SLL-1G4, we observed a TCR avidity-dependent labeling pattern where T-SCOPE-labeled ratios positively correlated with their known avidities (Fig. 1e). This correlation was consistent with SLL-pentamer results and remained robust across various APC types and antigens (Supplementary Fig. S3a–e). To further validate these findings, we employed NFAT-GFP Jurkat cells expressing GP100-TCR variants recognizing the YLEPGPVTV epitope (YLE-GP100) (Supplementary Fig. S3f). The results further confirm that the Bio+ ratio closely matches measured TCR avidities regardless of antigen quantity.
Furthermore, we found that the T-SCOPE Bio+ ratio also correlated with downstream TCR responses, such as CD69 expression and NFAT-regulated GFP signal (Supplementary Fig. S3g–j). Bulk RNA-seq revealed that Bio+ cells with higher-avidity TCRs exhibited a more extensive activation program, where the expression of key markers like NR4A1 positively aligned with the Bio+ ratio (Supplementary Fig. S4 and Table S1). Moreover, the Bio+ ratio accurately predicted TCR response scores based on transcription factor gene sets. In contrast, pMHC-pentamer staining elicited only minimal gene expression changes and significantly weaker signaling responses. These findings demonstrate that the Bio+ ratio effectively reflects both TCR avidity and functional antigen-response potency. T-cell states also fundamentally influence antigen response capacities (Supplementary Fig. S5). Memory-like OT-1 cells exhibited more robust responses and higher T-SCOPE labeling efficiency, whereas exhausted T (Tex) cells showed diminished responses and lower capture. Crucially, perturbing TCR signaling or immune synapse formation abolished labeling in both phenotypes.
Then, we investigated simultaneous functional monitoring of multiclonal TCRs within a mixed pool. Jurkat subsets expressing distinct TCR variants were barcoded with fluorescent dyes, pooled, and subsequently demultiplexed for analysis. The Bio+ ratio for each variant positively correlated with its functional avidity across both SLL-1G4 and YLE-GP100 panels, even in skewed mixtures (Fig. 1f; Supplementary Fig. S6). T-SCOPE achieved reliable resolution at peptide concentrations as low as 1 nM and maintained sensitivity when antigen-specific cells were diluted to 1/1,000, demonstrating high performance in complex populations (Supplementary Fig. S6d). Thus, T-SCOPE leverages intercellular labeling to quantitatively monitor diverse antigen response potencies within mixed TCR pools by measuring the Bio+ ratio of specific TCR subpopulations.
To resolve antigen-induced T-cell responses by high-throughput sequencing, we integrated T-SCOPE with single-cell RNA sequencing (scRNA-seq), paired scTCR-seq, and cell-surface DNA barcoding to simultaneously profile gene expression, labeling intensity, sample identity, and TCR sequences at single-cell resolution (Fig. 1g; Supplementary Fig. S7). We applied this platform to evaluate the SIINFEKL (N4)-specific repertoire in ovalbumin-vaccinated mice, revealing a 23% response rate in the N4 group (Supplementary Fig. S7b). Next, we coupled the T-SCOPE-labeled pool with single-cell sequencing by indexing N4-primed and unprimed groups with distinct sample tags8,9.
To quantify the signal, covalent biotin tags were converted into sequenceable DNA information using DNA-barcoded streptavidin. To systematically analyze the biotin-labeling status of each TCR clonotype at the single-cell level, we defined a metric named TScore, calculated via three steps. First, to calibrate independent experiments, the biotin-tag cutoff for the N4 group was determined by mapping the wet-lab flow cytometry Bio+ ratio directly onto the sequencing data distribution of each sample. Second, T cells were re-grouped into clonotypes by TCR sequences as barcodes to calculate both the Bio+ ratio and average biotin-tag level (MeanBiotintag), filtering out clones with fewer than three cells to ensure robustness. Third, the TScore for each clonotype was calculated using the formula: TScore = MeanBiotintag × (Bio+% + 1). Conceptually and computationally, unlike traditional abundance-based enrichment scores, TScore mirrors the real-world antigen recognition process by uniquely integrating wet-lab-calibrated population-level positive ratios with single-cell intensity (Supplementary Fig. S7a and Methods). By ranking all detectable TCRs (clone size > 2) according to their TScores, we characterized the overall antigen response potency of the SIINFEKL-specific repertoire. Notably, our analysis revealed that TScore did not directly correlate with TCR frequency (Fig. 1h; Supplementary Table S2).
To assess whether the TScore reflects the potency of antigen responses in the native T-cell repertoire, we designed three assays. First, we validated the functional activity of TCRs with varying TScores through wet-lab experiments. In Jurkat cell co-cultures, TCRs with TScores exceeding 1.62 exhibited functional responses to SIINFEKL, with T-cell activation positively correlating with TScore rankings (Fig. 1i; Supplementary Fig. S7f). Notably, the T-SCOPE-identified TCR-1 exhibited killing potency comparable to the potent, non-natural OT-1 TCR, confirming the platform's ability to identify high-quality natural TCRs (Fig. 1j).
Second, we compared TScore with pentamer-seq by indexing N4-pentamer signals via DNA-barcoded antibodies (Supplementary Fig. S8 and Table S3). Among shared TCRs with clone size exceeding 2, we identified 4 pentamer+ TCRs (three had TScores of > 1.62), and 16 pentamer− TCRs (12 had TScores of < 1.62), yielding approximately 75% pentamer referenced accuracy. Unlike pentamer signals that failed to reliably reflect functional response, TScore accurately captured the functional hierarchy (Fig. 1k), establishing itself as a complementary method to pMHC-multimers for response potency ranking within sequencing datasets.
Third, we established Re-stimu-seq by stimulating the SIINFEKL-TCR pool with N4-pulsed dendritic cells (DCs) for scRNA-seq profiling (Supplementary Fig. S9). This was integrated with T-SCOPE-seq by merging two datasets and regrouping cells by TCR sequences to assign TScores to matching clonotypes. Higher TScores correlated with increased expression of activation and cytotoxicity genes alongside reduced expression of the memory marker Sell (Supplementary Fig. S9d). High-TScore TCRs were enriched in the CD8Teff_Ifng subset and upregulated antigen-response programs upon re-exposure (Supplementary Fig. S9e, f). Functional scores positively correlated with TScores, confirming that high-ranking TCRs exhibit superior responsiveness (Supplementary Fig. S9g–o and Table S4). Notably, while TCR-2 showed the largest clone size (n = 25) and effector (96% in the TCRs-III group) dominance in Re-stimu-seq (data not shown), both functional avidity and TScore ranking indicated that its antigen-response capacity was not the strongest (Fig. 1i, k). In summary, through side-by-side comparisons with gold-standard assays, T-SCOPE enables single-TCR-resolved evaluation and ranking of response potency.
A comprehensive understanding of global tumor-responsiveness is crucial, as pre-existing tumor-reactive CD8+ T cells are closely associated with the efficacy of cancer immunotherapies1,2. We applied T-SCOPE to evaluate responses to pan-tumor-associated antigens using tumor lysate-primed monocyte-derived DCs (Fig. 1l; Supplementary Fig. S10). By conjugating DBF to these autologous APCs, we identified T-cell tumor-responsive populations in draining lymph nodes (dLNs) and tumor-infiltrating lymphocytes (TILs) from human clinical samples. Compared to normal tissue lysate-primed controls, tumor lysate-primed DCs captured significant Bio+ populations in lung cancer CD8+ (17% to 36% across individual variations) and CD4+ (15% to 30%) TILs (Fig. 1m; Supplementary Fig. S10c, d). Pre-existing tumor-responsive T cells were also identified in dLNs across lung and bladder cancer samples (Fig. 1m; Supplementary Fig. S10c–f). Collectively, T-SCOPE effectively monitors T-cell responses to unknown tumor antigens in human intratumoral TILs and dLNs.
Next, we applied T-SCOPE to monitor pan-tumor antigen responsiveness across the lung cancer TIL repertoire at clonotype resolution. An optimized 5' single-cell sequencing pipeline categorized 8,021 T cells, and major CD8+ subsets included effector memory (Tem), tissue-resident memory (Trm), exhausted, and proliferating cells (Supplementary Fig. S11). Highly expanded clonotypes predominantly displayed Tex and Tem phenotypes.
We further validated the link between T-SCOPE labeling and tumor response capacity in CD8+ TILs by assigning TScores (Fig. 1n; Supplementary Fig. S11e and Table S5). First, we compared T-SCOPE labeling with potential marker genes and established tumor-specific gene sets6,10,11, finding that TScore correlated positively with both marker gene expression and gene set scores (Supplementary Fig. S12a, b and Table S7). Then, to validate TScore, we compared gene expression between control group (without T-SCOPE labeling) and the stimulated TL group using highly expanded TCRs (clone size > 20 in the TL group). Antigen stimulation upregulated genes associated with T-cell activation, proliferation, adhesion, and TCR signaling. Functional gene set scores from these pathways correlated positively with TScores (Fig. 1o; Supplementary Fig. S13 and Table S6), confirming that TScore-based ranking effectively quantified TCR responses to pan-tumor antigens within the entire repertoire.
Integrating Bio+ readouts with transcriptomes reveals links between global tumor responsiveness and phenotypes. In CD8+ TILs, high TScores were enriched in Trm and Tex clusters, whereas GZMK+ Tem clusters exhibited the lowest TScores (Fig. 1n; Supplementary Fig. S11d). At clonotype resolution, high TScore TCRs predominantly displayed Trm phenotypes, moderate TScores showed Tex, and lower TScores showed GZMK+ Tem (Fig. 1p; Supplementary Fig. S13k and Table S6). These results indicate that tumor-responsive TCRs preferentially adopt a Trm state, while high-GZMK Tem cells exhibit weaker reactivity, which is consistent with recent studies showing that T cells adopt Trm and Tex phenotypes potentially driven by continuous and chronic exposure to tumor antigens1,12. These findings were further corroborated by Trm and progenitor signature scores (Supplementary Fig. S12c, d and Table S7). Additionally, submaximal response in Tex cells suggests reduced capacity due to functional exhaustion despite capable antigen recognition.
Although pMHC-tetramer-13, trogocytosis-14, large-scale synthesis-15 and proximity-labeling-based16,17 methods exist for deciphering TCR-antigen recognition, they are limited by predefined antigens and struggle to quantify response potency or directly analyze patient-derived primary tumor samples (Supplementary Table S8). To address this, the T-SCOPE strategy offers an integrated, clonotype-resolved, and repertoire-wide T-cell response monitoring platform with the following advantages. First, as an activity-based method, it records the genuine response of T cells to antigens within a specific biological context (i.e., the tumor), rather than reflecting merely pMHC–TCR recognition. Second, by converting activation readouts into biotin labeling, T-SCOPE provides a unified workflow that integrates T-cell functional responses into a single high-throughput sequencing run, thereby obviating the need for time-consuming and labor-intensive TCR synthesis and parallel functional validation. Third, by using TCRs as barcodes, T-SCOPE links T-cell activity and transcriptional profiles with TCR sequences to establish a “three-dimensional” TCR code, thereby providing a quantitative TScore metric rather than a binary classification. Fourth, beyond predefined antigens, T-SCOPE anchors labeling catalysts directly onto patient-derived DCs, enabling the utilization of tumor lysates as a comprehensive tumor antigen source to process complex antigens without prior knowledge, thereby mapping the landscape of TIL responsiveness. These findings suggest that T-SCOPE holds potential for diverse cancer vaccine studies involving both predefined and unknown antigens, a capability highly valuable in clinical settings where immune-driven antigen spreading expands the T-cell repertoire toward novel, non-targeted epitopes. In the future, T-SCOPE requires further optimization for low-abundance samples and faces practical barriers due to high requirements for fresh tissue and blood volumes. The TScore threshold is system-specific and determined by the given biological context. Accordingly, T-SCOPE serves as a response monitoring platform rather than a TCR-antigen mapping tool. Moving forward, further research will combine existing experimental platforms for antigen identification and computational approaches to bridge this gap. Looking ahead, by expanding the breadth and depth of T-cell response monitoring, T-SCOPE could become a foundational platform adopted across diverse immunotherapies and T-cell-centric clinical settings.

DATA AVAILABILITY

The paired single-cell RNA-seq and TCR-seq data that support the findings of this study have been deposited into the China National GeneBank Sequence Archive (CNSA) with accession numbers CNP0009798 for the mouse datasets and CNP0009804 for the human datasets.

[1]

Caushi, J.X. et al. Nature 596, 126–132 (2021).

[2]

Chen, Y.Q. et al. Cancer Cell 42, 1268–1285.e7 (2024).

[3]

Ott, P.A. Nat. Biotechnol. 44, 740–751 (2026).

[4]

Joglekar, A.V. & Li, G.D. Nat. Methods 18, 873–880 (2021).

[5]

Pauken, K.E. et al. Trends Immunol. 43, 180–194 (2022).

[6]

van der Leun, A.M., Thommen, D.S. & Schumacher, T.N. Nat. Rev. Cancer 20, 218–232 (2020).

[7]

Liu, H.Y. et al. J. Am. Chem. Soc. 144, 5517–5526 (2022).

[8]

Zhao, X.L. et al. RSC Chem. Biol. 3, 1052–1060 (2022).

[9]

Deng, T. et al. Immunity 58, 2336–2352.e12 (2025).

[10]

Lowery, F.J. et al. Science 375, 877–884 (2022).

[11]

Hanada, K.I. et al. Cancer Cell 40, 479–493.e6 (2022).

[12]

Pai, J.A. et al. Cancer Cell 41, 776–790.e7 (2023).

[13]

Ma, K.Y. et al. Nat. Immunol. 22, 1590–1598 (2021).

[14]

Li, G.D. et al. Nat. Methods 16, 183–190 (2019).

[15]

Gaglione, S.A. et al. Immunity 59, 477–493.e9 (2026).

[16]

Liu, Z.L. et al. Cell 183, 1117–1133.e19 (2020).

[17]

Liu, S.B. et al. J. Exp. Med. 223, e20251779 (2026).

RIGHTS & PERMISSIONS

The Author(s) 2026. Published by Higher Education Press. This is an Open Access article distributed under the terms of the CC BY license (https://creativecommons.org/licenses/by/4.0/).

Cite this article

Download citation ▾
Liu, H. et al.  Clonotype-resolved T-cell response monitoring via repertoire-wide multidimensional TCR decoding  Vita https://doi.org/10.15302/vita.2026.07.0054 ()
AI Summary AI Mindmap
PDF (1697KB)

Supplementary Files

Supplementary Materials

Supplementary Table_S1

Supplementary Table_S2

Supplementary Table_S3

Supplementary Table_S4

Supplementary Table_S5

Supplementary Table_S6

Supplementary Table_S7

Sections
Figures
References

457

Accesses

0

Citation

Detail

Recommended

AI思维导图

/

〈 〉