Precision modeling of tumor antigen-specific T-cell responses in humanized mice for preclinical assessment of cancer immunotherapies.

Precision modeling of tumor antigen-specific T-cell responses in humanized mice for preclinical assessment of cancer immunotherapies - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice J Immunother Cancer . 2026 Apr 15;14(4):e013989. doi: 10.1136/jitc-2025-013989 Search in PMC Search in PubMed View in NLM Catalog Add to search Precision modeling of tumor antigen-specific T-cell responses in humanized mice for preclinical assessment of cancer immunotherapies Samuel Gebhardt Samuel Gebhardt 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland 2 Institute of Experimental Immunology, University of Zurich, Zürich, Switzerland Find articles by Samuel Gebhardt 1, 2, ✉ , Marine Le Clech Marine Le Clech 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Marine Le Clech 1 , Barbara Höllbacher Barbara Höllbacher 3 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Basel, Basel, Switzerland Find articles by Barbara Höllbacher 3 , Sarah Brunner Sarah Brunner 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Sarah Brunner 1 , Ahmet Varol Ahmet Varol 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Ahmet Varol 1 , Simone Lang Simone Lang 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Simone Lang 1 , Silvia Jenni Silvia Jenni 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Silvia Jenni 1 , Stefanie Briner Stefanie Briner 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Stefanie Briner 1 , Emilio Yángüez Emilio Yángüez 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Emilio Yángüez 1 , Tamara Hüsser Tamara Hüsser 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Tamara Hüsser 1 , Ioana Domocos Ioana Domocos 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Ioana Domocos 1 , Ria Tauscher Ria Tauscher 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Ria Tauscher 1 , Jan Eckmann Jan Eckmann 4 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Munich, Penzberg, Germany Find articles by Jan Eckmann 4 , Christian Münz Christian Münz 2 Institute of Experimental Immunology, University of Zurich, Zürich, Switzerland Find articles by Christian Münz 2 , Johannes Sam Johannes Sam 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland Find articles by Johannes Sam 1 Author information Article notes Copyright and License information 1 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Zurich, Schlieren, Switzerland 2 Institute of Experimental Immunology, University of Zurich, Zürich, Switzerland 3 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Basel, Basel, Switzerland 4 Roche Pharma Research and Early Development (pRED), Roche Innovation Center Munich, Penzberg, Germany Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise. Additional supplemental material is published online only. To view, please visit the journal online ( https://doi.org/10.1136/jitc-2025-013989 ). Several patent applications with relevance to this work have been filed by Roche. SG, MLC, SaB, BH, SL, AV, SJ, ID, EY, TH, StB, RT, JE and JS are current or former employees of Roche and hold ownership of Roche stocks. CM has no conflict of interest with respect to the presented research. ✉ Dr Samuel Gebhardt; [email protected] Received 2025 Oct 17; Accepted 2026 Mar 27; Collection date 2026. Copyright © Author(s) (or their employer(s)) 2026. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ Group. This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See https://creativecommons.org/licenses/by-nc/4.0/ . PMC Copyright notice PMCID: PMC13084978  PMID: 41986069 Abstract Background Humanized mice are highly valuable models for the preclinical development of cancer immunotherapies. However, conventional models often fail to mount robust and trackable tumor antigen-specific (TA-specific) T-cell responses, which hinders the comprehensive evaluation of therapies that depend on this endogenous antitumor immunity. Methods To address this limitation, we generated a humanized mouse model by engrafting immunodeficient mice with human hematopoietic stem cells that were transduced with predefined T-cell receptor (TCR) specificities. This approach enables the de novo generation of naïve, functional TA-specific T cells at adjustable frequencies. Results We found that transgene-bearing human thymocytes have a developmental advantage in the murine thymus, likely due to the early formation of the transgenic TCR and associated autonomous signaling. In tumor studies, the presence of these TA-specific T cells delayed tumor growth and promoted increased phenotypic diversity of intratumoral T cells, including the formation of PD-1 + /TCF1 + precursor exhausted T cells, which were absent in control mice. We demonstrate the utility of this model in two distinct therapeutic contexts. First, we found that high baseline frequencies of TA-specific T cells blunted the efficacy of a T-cell bispecific antibody (TCB), suggesting that pre-existing exhaustion limits TCB activity. Furthermore, the model enabled the evaluation of costimulatory agonists (FAP-CD40 and FAP-4-1BBL), and demonstrated their dependence on this T-cell compartment for antitumor efficacy. Conclusions This adaptable and physiologically relevant model provides a platform to dissect the complex interplay between human tumors, immune cells, and immunotherapies and has the potential to significantly improve the translation of preclinical findings to the clinic. Keywords: Immunotherapy, Bispecific T cell engager - BiTE, Combination therapy, T cell Receptor - TCR, Tumor infiltrating lymphocyte - TIL WHAT IS ALREADY KNOWN ON THIS TOPIC Conventional humanized mouse models usually lack a pre-existing, functional repertoire of trackable tumor antigen-specific (TA-specific) T cells. This severely limits the preclinical evaluation of immunotherapies designed to modulate endogenous antitumor immunity, such as costimulatory agonists and immunomodulators, whose mechanisms depend on the presence of this T-cell compartment. It also obscures the assessment of the impact of broadly acting agents, like T-cell bispecifics, on the TA-specific T-cell subset. WHAT THIS STUDY ADDS This study establishes a humanized mouse platform that overcomes this limitation by supporting the de novo development of TA-specific T cells at controllable frequencies. We demonstrate that these TA-specific T cells are a prerequisite for the intratumoral development of the clinically relevant programmed cell death protein 1 (PD-1) + /T cell factor 1 (TCF1) + precursor exhausted T-cell population. Furthermore, the model revealed that a high baseline frequency of these T cells leads to increased T-cell exhaustion after T-cell bispecific therapy, and was associated with lower efficacy, a finding not previously observable in conventional humanized mouse models. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY This platform provides a robust in vivo tool to mechanistically dissect and compare immunotherapies that rely on endogenous T-cell responses, enabling the rigorous preclinical testing of immunotherapeutic agents like FAP-CD40 and FAP-4-1BBL that were previously untestable in this context. The findings on T-cell bispecific resistance provide a strong preclinical rationale for investigating baseline T-cell infiltration and exhaustion status as a patient stratification biomarker in solid tumor clinical trials. Background Tumor antigen-specific (TA-specific) T cells play a critical role in antitumor immunity, and their presence and function within the tumor microenvironment (TME) have been correlated with improved outcomes in several cancers. 1 , 3 Despite their importance, current preclinical in vivo models struggle to fully recapitulate the complex interplay between human TA-specific T cells, the TME, and cancer cells. 4 5 These limitations may hinder the translation of preclinical findings into clinical settings, highlighting the need for improved in vivo models that accurately reflect human TA-specific T-cell biology. TA-specific T cells represent a central component of many cancer immunotherapies, as their activation and expansion are essential for generating effective antitumor responses. 6 Various immunotherapeutic strategies rely on these cells, including checkpoint inhibitors (eg, anti-programmed cell death protein 1 (PD-1)), 7 costimulatory agonists (eg, PSMA-CD28, 8 FAP-4-1BBL), 9 targeted myeloid cell agonists (eg, FAP-CD40), 10 and targeted cytokines (eg, PD-1-IL2v), 11 which are particularly effective in immunologically “hot” tumors with high T-cell infiltration. 12 These “hot” tumors are often considered more responsive to immunotherapies due to the baseline presence of tumor-reactive T cells. 7 12 13 While tumor-reactive CD8 + T cells are rare in some indications, they can represent a dominant fraction in others, with frequencies of T-cell clones recgnizing a single tumor epitope ranging from 1% to 30%, and the total frequencies of tumor-specific T cells in cases like melanoma showing ranges up to 50–80%. 14 In contrast, T-cell bispecific antibodies (TCBs) can induce tumor cell lysis even in “cold” tumors through polyclonal T-cell engagement via CD3ε, independent of T-cell receptor (TCR) specificities. 15 However, the impact of such continuous CD3 stimulation on the function, persistence and exhaustion of potentially chronically stimulated TA-specific T cells remains poorly understood, which limits the possibility to optimize TCB dosing, patient stratification, and immunotherapy combination strategies. Humanized mice have emerged as valuable preclinical models for evaluating human immune responses and immunotherapies in vivo . 5 These models are generated by engrafting immunodeficient mice with human hematopoietic stem cells (HSCs), resulting in the development of a diverse repertoire of human immune cells, including T cells, B cells, and myeloid cells. This approach offers a more physiologically relevant system than traditional rodent models, enabling the study of human-specific immune mechanisms using the same therapeutic molecules used in patients. 5 HSC-humanized mice have proven particularly instrumental for the development of TCBs. 16 , 18 Despite these promises, one key limitation of most currently available humanized mouse models is the inability to generate robust, reproducible, and trackable populations of human TA-specific T cells. This is partly due to the chimeric thymic environment, where developing human T cells undergo selection processes influenced by a combination of human and murine factors, including major histocompatibility complex (MHC) molecules, cytokines, and cell–cell interactions. 19 Although humanized mice have successfully modeled antigen-specific T-cell responses against viral infections like Epstein-Barr virus (EBV), 20 the generation of robust and trackable antigen-specific T-cell responses against tumors remains an ongoing challenge in the field. Several strategies have been explored to generate human TA-specific T-cell responses in humanized mice, including vaccination strategies, 21 adoptive transfer of TCR-engineered T cells, 22 targeted dendritic cell activation 23 or the use of TCR-engineered HSCs. 24 , 27 Most of these approaches have generally relied on the use of HLA-transgenic mouse strains, aiming to enhance HLA-restricted T-cell selection, improve T-cell priming, and maintain T-cell homeostasis. However, these approaches were often complex and resulted in low and variable frequencies of TA-specific T cells, limiting their suitability as models for larger-scale drug discovery applications. 2124 , 27 Additionally, there is limited experimental evidence from in vivo tumor studies to evaluate the functional impact of the generated TA-specific T cells in this context. Last but not least, while BLT mice (bone marrow, liver, thymus) offer a more complete human immune environment, their complexity and requirement for human fetal tissue pose significant challenges for their widespread use. 28 To address this critical unmet need, we further adapted a previously introduced humanized mouse model based on the engraftment of HSCs lentivirally transduced with defined TCR sequences targeting clinically relevant TAs, such as NY-ESO-1 and MART-1. 23 This approach generates robust and trackable populations of naïve, functional human TA-specific T cells in vivo . This platform enables investigation of the complex interplay between the TME, human TA-specific T-cell responses and cancer immunotherapies in a highly controlled and physiologically relevant setting. Specifically, this model facilitates detailed studies of TCR dynamics, including affinity, activation, and exhaustion profiles, ultimately supporting the translation of preclinical findings to clinical benefit. Materials and Methods Detailed descriptions of the materials and methods are provided in the supplemental materials and methods ( online supplemental file 1 ). Results Humanization with TCR-transduced human HSCs leads to the development of naïve and functional antigen-specific human T cells in immunodeficient mice To establish a humanized mouse model with human TA-specific T cells, we used the well-characterized, high affinity 1G4 NY-ESO-1 TCR, which recognizes the SLLMWITQC peptide sequence presented on MHC class I (HLA-A2.1). 23 Human HLA-A2.1 + HSCs were transduced with lentiviral particles coding for this TCR sequence ( figure 1A,B ). At the time of injection into immunodeficient mice at day 2 after thawing, the engineered HSCs maintained a high expression of the stem cell markers CD34 and CD133.1, which was not affected by the TCR-transduction ( figure 1C ). Small aliquots of TCR-transduced (TCR-TD) and non-transduced (Non-TD) HSCs were kept in vitro to monitor HSC marker kinetics and assess transduction efficiency. On day 6, CD34 and CD133.1 were partially downregulated at similar frequencies in TCR-TD and Non-TD conditions ( figure 1D ). The TCR-transduction of HSCs achieved a transduction efficiency of ~40% as evaluated by the frequency of green fluorescent protein (GFP) + cells ( figure 1E ). On day 2, TCR-TD and Non-TD HSCs were injected into immunodeficient BRGS-CD47 mice that had been pretreated with busulfan ( figure 1F ). Flow cytometry analysis of peripheral blood at week 14 post-HSC injection revealed engraftment of TCR-TD HSCs with no major difference in engraftment efficiency and immune cell composition between TCR-TD and Non-TD groups ( figure 1G ). CD4 + and CD8 + T-cell frequencies were not affected by the TCR-transduction. In TCR-TD mice, approximately 60% of human CD8 + T cells in the blood exhibited stable GFP + expression ( figure 1H , left, online supplemental figure 1 ), which correlated with NY-ESO-1 dextramer staining ( figure 1H , right), indicating stable TCR expression and efficient binding of the cognate MHC-peptide complex. In contrast, around 40% of CD4 + T cells showed GFP positivity ( figure 1H , left); however, dextramer staining was minimal, which may be explained by the absence of CD8 co-receptor stabilization, suggesting suboptimal MHC-peptide complex binding. Interestingly, only a small fraction of circulating human B cells expressed GFP. Phenotypic assessment revealed no influence of TCR-transduction, as CD8 + T cells exhibited a predominantly naïve (CD45RA + /CCR7 + ) while CD4 + T cells displayed a more effector memory (CD45RA − /CCR7 − ) phenotype in both TCR-TD and Non-TD settings ( figure 1I ). Flow cytometry analysis of mesenteric lymph nodes (LN), spleen, and bone marrow (BM) at 18 weeks post-injection revealed presence of transgenic cell subsets in all organs, with highest frequencies in T cells ( figure 1J ). GFP signal was also detectable in several CD11b + myeloid cell subsets, mast cells and in CD34 + /CD133.1 + HSCs in the BM. Ex vivo functional assessment of isolated T cells from peripheral blood of TCR-TD and Non-TD mice revealed strong granzyme B upregulation and proliferation of transgenic T cells in the presence of T2 cells pulsed with NY-ESO-1 peptide. No proliferation and granzyme B upregulation were observed in the absence of the peptide and in the Non-TD group ( figure 1K ). This finding suggests that the engraftment with TCR-TD HSCs results in the generation of naïve and functional TA-specific T cells able to bind the cognate MHC-peptide complex. While dextramer staining of CD4 + T cells was minimal, we confirmed that the NY-ESO-1 TCR can be activated on both CD4 + and CD8 + T cells in vitro ( online supplemental figure 1 ). Figure 1. TCR transduction of human HSCs and subsequent engraftment in immunodeficient mice leads to development of naïve, functional, TCR-expressing T cells in humanized mice. ( A ) Design of the transgenic insert coding for the 1G4 NY-ESO-1 TCR. The EF1α promoter drives expression of TCR-β and TCR-α chains, separated by T2A self-cleaving peptide. An E2A peptide site further separates the TCR-α chain from the GFP reporter. ( B ) Experimental workflow for HSC transduction. Human CD34 + /HLA-A2.1 + cord blood HSCs were transduced with TCR-encoding lentivirus on day 1 post-thawing. ( C ) Representative flow cytometry dot plot illustrating CD133.1 and CD34 expression (left) and frequency of CD133.1 + /CD34 + (right) in TCR-TD and Non-TD HSCs at day 2 (n=5) and ( D ) at day 6 post thawing. ( E ) Transduction efficiency assessed by GFP expression; representative flow cytometry dot plot (left) and frequency of GFP + HSCs at day 6 (n=5). ( F ) Experimental design for humanization. BRGS-CD47 (4–5 weeks old) were pretreated with busulfan (15 mg/kg) followed by intravenous injection of 1×10 5 TCR-TD or Non-TD HSCs. ( G ) Human cell engraftment analysis by flow cytometry at week 14 post HSC transfer showing human CD45 + cells (frequency and counts/µL), B cells (CD19 + ), T cells (CD3 + ), and CD4 + /CD8 + T-cell distribution. TCR-TD: n=44, Non-TD: n=12. ( H ) TCR expression analysis. Left: representative flow cytometry plots showing GFP expression and NY-ESO-1 dextramer binding in CD4 + and CD8 + T cells of TCR-TD and Non-TD mice. Right: quantification of GFP + cells across lymphocyte populations in TCR-TD mice. ( I ) Representative flow cytometry plots of CD45RA/CCR7 expression profiles of CD8 + and CD4 + T cells in peripheral blood. Right: frequency of naïve (CD45RA + /CCR7 + ) cells among CD4 + and CD8 + T cells. ( J ) Analysis of GFP + cell subsets in LN, spleen (pDCs: CD123 + /CD303 + ; cDC1s: CD141 + ; cDC2s: CD11c + /CD1c + ) and BM (HSCs: CD34 + /CD133.1 + ; Mast cells: CD117 + /FcεRI + ) (TCR-TD n=3–5, Non-TD n=3). ( K ) Functional assessment of TCR-specific responses by granzyme B expression (left) and proliferation dye dilution (right) of peripheral blood cells of TCR-TD and Non-TD mice after incubation with T2 cells pulsed with the short NY-ESO-1 peptide or without peptide (n=3 per group, E:T ratio=1:1). For panels C, D, G, H, I, J: bar graphs show mean+SD. Statistical analysis was performed using a non-paired t-test for two groups, or Šídák’s multiple comparisons test for more groups. For panel H: statistical analysis was performed using Fisher’s LSD test. *p>0.05; **p≤0.05; ***p≤0.01; ****p≤0.001; *****p≤0.0001. BM, bone marrow; E:T, effector-to-target ratio; GFP, green fluorescent protein; HSC, hematopoietic stem cell; LN, lymph nodes; LSD, Least significant difference ; Non-TD, non-transduced; ns, not significant; TCR, T-cell receptor; TD, transduced. Open in a new tab The TCR-transduction model can be customized for diverse TCRs and antigen-specific T-cell frequencies To demonstrate the versatility of the TCR transduction model, we generated additional lentiviral constructs that encode a low-affinity NY-ESO-1 TCR (4A2) and the DMF5 MART-1 TCR, each tagged with red fluorescent protein (RFP) to distinguish them from the GFP-tagged high-affinity NY-ESO-1 TCR ( figure 2A ). 29 30 Using these constructs, we established either single TCR-TD mice for the different TCRs, or combinations of TCRs recognizing different tumor antigens (NY-ESO-1 and MART-1), or the same antigen but with varying affinities (low and high, figure 2B ). Successful HSC transduction, as indicated by GFP or RFP expression, confirmed the absence of double-positive RFP + /GFP + HSCs, ensuring that each HSC contained maximally one TCR transgene ( figure 2C ). Following HSC transduction, TCR-TD and Non-TD HSCs were injected into busulfan-pretreated BRGS-CD47 mice ( figure 2D ). Flow cytometry analysis of peripheral blood at 16 weeks post-injection confirmed successful engraftment of human CD45 + cells, with comparable B and T-cell frequencies across all conditions ( figure 2E ). The presence of GFP + and/or RFP + T cells in addition to the respective dextramer staining demonstrated successful development of transgenic T cells and the surface expression of the TCRs ( figure 2F ). Of note, while reporter expression and dextramer binding frequency correlated strongly across cohorts, reporter fluorescence (GFP/RFP) consistently captured a broader population of functional transgenic T cells than dextramer staining alone, particularly in CD4 + and low-affinity CD8 + (4A2) groups ( online supplemental figure 2 ). Consequently, we used GFP/RFP expression as the primary metric for identifying the total functional transgenic T-cell population. While similar levels of transgenic T cells were observed in mice expressing a single TCR, in conditions with two TCRs, predominantly the high affinity NY-ESO-1 TCR was expressed, which may be explained by differences in transduction efficiency. To assess TCR functionality, isolated splenic T cells of humanized mice were stimulated ex vivo with T2 cells pulsed with titrated concentrations of MART-1 or NY-ESO-1 peptide. Dose-dependent T-cell activation was observed for all three TCRs, as indicated by the upregulation of CD25, GZMB, and Ki-67 compared with Non-TD cells stimulated with the corresponding peptides ( figure 2G ). Notably, the low-affinity NY-ESO-1 TCR group exhibited a less pronounced increase in T-cell activation. Furthermore, the generation of TA-specific T cells was independent of the mouse background strain used, as similar engraftment, T-cell frequencies and functionality were observed in both NSG and BRGS-CD47 mice ( online supplemental figure 2 ). Lastly, to demonstrate the flexibility of the model in terms of TA-specific T-cell frequencies, we mixed high-affinity NY-ESO-1 TCR-TD HSCs with Non-TD HSCs at various ratios, allowing for titration of the fraction of transgene bearing HSCs ( figure 2H ). Analysis of peripheral blood 14 weeks post-HSC injection demonstrated a linear correlation between the initial frequency of transduced HSCs and the frequency of GFP + CD4 + , CD8 + , and CD19 + cells ( figure 2I ). Interestingly, higher transduction rates of HSCs correlated with lower overall human cell engraftment ( online supplemental figure 2 ). Figure 2. The TCR-transduction model can be customized for diverse TCRs and antigen-specific T-cell frequencies. ( A ) Design of the transgenic inserts coding for high-affinity NY-ESO-1 (1G4, hiNY), low-affinity NY-ESO-1 (4A2, loNY) and MART-1 (DMF5) TCR. All constructs are driven by the EF1α promoter; TCR-β and TCR-α chains are separated by a T2A self-cleaving peptide. An E2A peptide site further separates the TCR-α chain from the GFP or RFP reporter. ( B ) Schematic illustration of the experimental workflow for dual TCR humanization. HSCs were transduced with individual TCR constructs and mixed in a 1:1 ratio prior to injection. ( C ) Representative flow cytometry plots showing GFP/RFP expression in single-transduced (left) and dual-transduced (right) HSCs at day 6 post-transduction. ( D ) Experimental design for humanization. BRGS-CD47 (4–5 weeks old) were pretreated with busulfan (15 mg/kg) followed by intravenous injection of different conditions of TCR-TD HSCs or Non-TD HSCs (1×10 5 cells). ( E ) Flow cytometry analysis of human cell engraftment at week 16 showing human CD45 + cells (frequency and counts/µL), B cells (CD19 + ), and T cells (CD3 + , high NY TCR-TD n=5; MART-1 TCR-TD n=9; low NY TCR-TD n=10; high NY+MART-1 TCR-TD n=7; high+low NY TCR-TD n=8; Non-TD n=7). ( F ) TCR expression analysis; representative plots showing GFP/RFP expression and respective dextramer binding in CD8 + T cells of peripheral blood at week 16 for each condition (left) and quantification of transgene-positive CD8 + T cells across groups (right). ( G ) Functional assessment of TCR-specific responses by stimulating splenic T cells with peptide-pulsed T2 cells. Bar graphs show log (fold change) of CD25, granzyme B and Ki-67 expression relative to Non-TD controls across peptide concentrations. ( H ) Experimental design for titration of baseline frequency of antigen-specific T cells. HSCs with varying GFP + frequencies (8.5–41.2%) were injected into busulfan-pretreated mice. ( I ) Correlation between initial HSC GFP + frequency and resulting transgene-positive lymphocyte populations in peripheral blood at week 14. Linear regression analysis is shown for CD8 + T cells ( r ²=0.7157), CD4 + T cells ( r ²=0.7300), and CD19 + B cells ( r ²=0.6920; 8.5% n=8; 15.1% n=10; 22.6% n=10; 29.5% n=9; 41.2% n=10). For panels in E, F and G: bar graphs show mean+SD. For panels in I, box and whisker plots show median and 5–95% percentiles. Statistical analysis was performed using one-way ANOVA combined with Tukey’s multiple comparison testing for panel E, two-way ANOVA combined with Tukey’s multiple comparison testing for panel F and G. For panel H: r ² was calculated by simple linear regression. Ns or no line *p>0.05; **p≤0.05; ***p≤0.01; ****≤0.001; *****≤0.0001. ANOVA, analysis of variance; GFP, green fluorescent protein; HSC, hematopoietic stem cell; Non-TD, non-transduced; ns, not significant; RFP, red fluorescent protein; TCR, T-cell receptor; TD, transduced. Open in a new tab Transgenic thymocytes exhibit developmental advantage and maturation also in the absence of cognate MHC class I To investigate human T-cell development in the murine thymus, we performed single-cell RNA sequencing and protein analysis (feature barcoding) and T-cell receptor sequencing (TCR-seq) on thymic tissue from TCR-TD and Non-TD humanized mice. These analyses revealed a diverse repertoire of human hematopoietic cells, including thymocytes at various developmental stages (double-negative (DN), double-positive (DP), single-positive (SP) CD4 + and CD8 + , regulatory T cells (Tregs), and γδ T cells; figure 3A , online supplemental figure 3 ). The TCR transgene was expressed in all human hematopoietic cell types, although at lower frequencies in B cells ( figure 3B ), consistent with our flow cytometry data ( figure 1H ). Given the potential interaction of thymocytes with both murine and human MHC molecules, we investigated the HLA expression on human immune cells. B cells exhibited markedly higher HLA-ABC and HLA-DR expression compared with other hematopoietic subsets ( figure 3C ). Furthermore, immunohistochemistry (IHC) revealed a distinct medullary localization of B cells, coinciding with areas enriched in SP thymocytes ( figure 3D , online supplemental figure 4 ). TCR-seq analysis of endogenous TCRs revealed a considerably higher proportion of transgene-positive thymocytes expressing an endogenous TCR-α chain but lacking a corresponding TCR-β chain (“orphan VJ”) compared with transgene-negative thymocytes ( figure 3E , online supplemental figure 4 ). In contrast, transgene-negative thymocytes predominantly expressed both endogenous TCR-α and β chains, with a smaller subset expressing only TCR-β chains. While the transgenic TCR-α chain uses TRAV21 and TRAJ6 segments, we observed no enrichment of these segments in endogenous TCR-α chains ( online supplemental figure 4 ), indicating spontaneous rearrangement. The frequency of GFP-positive cells progressively increased across thymocyte developmental stages, with the highest levels observed in SP cells (DN<DP<SP; figure 3F ), suggesting an enrichment of transgenic thymocytes during T-cell development. To further investigate the localization of GFP + cells, we performed three-dimensional immunophenotyping of thymic sections, which revealed an accumulation of GFP + cells in the medullary regions. These regions were further characterized by strong TCR-α/β expression in SP-positive thymocytes ( figure 3G , online supplemental figure 4 ). To determine whether thymocytes expressing the transgenic HLA-A2.1-restricted NY-ESO-1 TCR require interaction with cognate HLA-A2.1-expressing cells for development, we engrafted TCR-TD HSCs from both HLA-A2.1 + and HLA-A2.1 − HSC donors ( figure 3H ). Surprisingly, GFP + T cells expressing the NY-ESO-1 TCR developed in both HLA-A2.1 + and HLA-A2.1 − settings ( figure 3I , online supplemental figure 5 ), indicating that interaction of the transgenic TCR with HLA-A2.1 is not required for the development of these NY-ESO-1-specific thymocytes. Furthermore, we investigated thymocyte developmental trajectories using pseudotime analysis, which revealed distinct trajectories in TCR-TD and Non-TD mice ( figure 3J ). In TCR-TD mice, direct transitions from proliferating DP thymocytes to SP stages could be observed, with a reduced proportion of quiescent DP thymocytes (~18% in TCR-TD vs ~50% in Non-TD mice, figure 3K ) and increased CD8 + SP population suggesting a faster maturation process of transgene-positive thymocytes in TCR-TD mice. Figure 3. Transgenic thymocytes exhibit developmental advantage and maturation also in the absence of cognate MHC class I. ( A ) UMAP visualization of human cell subsets in the thymus of humanized mice colored by cell subset clusters (TCR-TD n=4, Non-TD n=2, total cells n=56,933). ( B ) UMAP projection showing NY-ESO-1 TCR transgene expression across cell populations. ( C ) UMAP projection visualizing the expression of HLA-ABC and HLA-DR as measured by feature barcoding. ( D ) Sequential IHC images of thymic architecture with representative images showing CD20, CD1a and CD4/CD8 expression. Scale bars: 500 µm; scale bars of higher magnification images: 50 µm. ( E ) Analysis of endogenous TCR sequences in transgene-positive and transgene-negative thymocytes. Only cells with detected endogenous TCR sequences are shown. VJ: TCR-α chain; VDJ: TCR-β chain. ( F ) Flow cytometric analysis of GFP + cell frequency during T-cell development: DN (CD4 − CD8 − ), DP (CD4 + CD8 + ), and SP stages. ( G ) Three-dimensional immunofluorescence imaging of thymic organization. Murine ICAM1 (white), human TCR-αβ (cyan), and GFP (green). Scale bar: 200 µm; scale bars of higher magnification images: 40 µm. ( H ) Experimental design for humanization of immunodeficient, busulfan-pretreated BRGS-CD47 mice with TCR-TD HSCs of HLA-A2.1 + and HLA-A2.1 − donors and Non-TD HSCs. ( I ) Flow cytometry analysis at week 14 post-transfer showing human CD45 + cells, T-cell frequency, and GFP + frequencies in CD4 + /CD8 + T cells (left) and representative plots of GFP and NY-ESO-1 dextramer staining in CD8 + T cells from HLA-A2.1-derived donor-derived cells (right). ( J ) Force-directed graph visualization of T-cell development colored by subset identity and pseudotime in Non-TD and TCR-TD mice. ( K ) Quantitative comparison of thymic subsets between TCR-TD and Non-TD mice based on scRNA-seq analysis (TCR-TD n=4, Non-TD n=2). For panels F, I, K: bar graphs show mean+SD. For panel I: statistical analysis was performed using one-way ANOVA and Turkey’s Multiple Comparisons test; for panel K: statistical analysis was performed using two-way ANOVA and Šídák’s multiple comparisons test. No line or ns. *p>0.05; **p≤0.05; ***p≤0.01; ****p≤0.001; *****p≤0.0001. Act, activated; ANOVA, analysis of variance; C, cortex; CD8αα, unconventional CD8αα T cells; DN, double negative (CD4 − /CD8 − ); DP, double positive (CD4+/CD8+); DP(P), proliferating DP; DP(Q), quiescent DP; Entry, entry stage; FA, force atlas; GFP, green fluorescent protein; HSC, hematopoietic stem cell; IHC, immunohistochemistry; ILC3, Type 3 innate lymphoid cells; M, medulla; MHC, major histocompatibility complex; Non-TD, non-transduced; ns, not significant; pDC, plasmacytoid dendritic cell; prog, progenitor cell; scRNA-seq, single-cell RNA sequencing; SP, single-positive; T, T cell; TCR, T-cell receptor; TD, transduced; Th17, T helper 17 cells; Treg, regulatory T cell; UMAP, Uniform Manifold Approximation and Projection; αβ, TCR-αβ; γδT, gamma delta T cells. Open in a new tab Transgenic NY-ESO-1 specific T cells influence tumor growth and promote precursor exhausted T-cell differentiation in vivo To confirm the specificity of the NY-ESO-1 TCR, we assessed its in vitro activity against a panel of human tumor cell lines with varying NY-ESO-1 and HLA-A2.1 expression. These cell lines were co-incubated with CD8 + T cells engineered to express the 1G4 NY-ESO-1 TCR, with endogenous TCRs knocked out. Tumor cell lysis was directly proportional to NY-ESO-1 expression and was observed only in HLA-A2.1-positive cell lines, confirming the specificity of the TCR ( online supplemental figure 6 ). The in vivo antitumor activity of NY-ESO-1-specific T cells was assessed by challenging TCR-TD and Non-TD mice with a panel of tumor cell lines ( figure 4A ). Tumor growth was significantly delayed or prevented in TCR-TD mice challenged with HLA-A2.1-positive, NY-ESO-1-expressing cell lines, but not with the HLA-A2.1-negative NCI-H929 cell line. IHC confirmed consistent NY-ESO-1 expression in tumors ( figure 4B ), with NY-ESO-1 H-scores correlating with transcript levels (nPTM, normalized transcripts per million) from the Human Protein Atlas ( online supplemental figure 6 ). U266B1 showed the highest NY-ESO-1 expression and the highest observed tumor rejection in TCR-TD mice. Notably, sustained MHC class I and NY-ESO-1 expression in all tumor models at study termination indicated that tumor growth was not due to loss of the target antigen or MHC class I downregulation ( online supplemental figure 7 ). Both the A375 and A375-HLA2.1 tumor models were chosen for subsequent preclinical efficacy studies, as they still show high engraftment rates despite the baseline antitumor reactivity of the transgenic T cells, allowing for assessment of therapeutic efficacy. We next investigated the impact of baseline TA-specific T-cell frequency on tumor growth. We observed a clear correlation between peripheral GFP + CD8 + T-cell frequency and A375-HLA2.1 tumor growth kinetics ( online supplemental figure 8 ), demonstrating the ability of the TCR-TD model to control for baseline levels of TA-specific tumor-infiltrating T cells. To characterize the T-cell phenotypes, we performed flow cytometry analysis of tumor-infiltrating lymphocytes (TILs) across four different tumor models in tumor-bearing TCR-TD and Non-TD mice ( figure 4C ). Unbiased clustering analysis revealed distinct CD8 + T-cell populations unique to TCR-TD mice, most notably a prominent cluster of precursor exhausted T cells (Tpex) characterized by co-expression of PD-1, T cell factor 1 (TCF1), and Kiel-67 (Ki-67), but lacking granzyme B (GZMB, Cluster 9, online supplemental figure 8 ). Consistent with this finding, manual gating analysis also demonstrated increased T-cell infiltration in TCR-TD mice ( figure 4D ) and identified the exclusive presence of the Tpex population in the presence of TA-specific T cells across all tested tumor models ( figure 4E , online supplemental figure 8 ), suggesting that sustained TCR signaling is necessary for the development of this phenotype. This finding, along with increased frequencies of activated and cytotoxic CD25 + /GZMB + T cells (Cluster 10) and FoxP3 + T cells subsets (Clusters 4 and 6), suggests a complex interplay of effector and regulatory T cell subsets within the TCR-TD TME. Ex vivo Phorbol 12-myristate 13-acetate (PMA)/ionomycin restimulation of TILs and splenocytes from A375 tumor-bearing mice revealed higher production of tumor necrosis factor (TNF)-α, interferon (IFN)-γ, and interleukin (IL)-2 by intratumoral T cells in TCR-TD mice, indicating a more robust inflammatory cytokine response ( figure 4F ). Interestingly, Non-TD mice exhibited higher frequencies of degranulating GZMB + /CD107a + cytotoxic T cells, likely reflecting transient activation due to alloreactivity rather than a sustained, antigen-directed antitumor response. Figure 4. Transgenic NY-ESO-1 specific T cells influence tumor growth and promote precursor exhausted T-cell differentiation in vivo . ( A ) Experimental design showing set-up of tumor challenges of TCR-TD and Non-TD mice (left). Right: individual tumor volume (mm 3 ) shown as spider plots: U266B1, MDA-MB-231_NY-ESO-1, NCI-H522, SHP-77, A375-HLA2.1, A375 (all NY-ESO-1 + /HLA-A2.1 + ), and NCI-H929 (NY-ESO-1 + /HLA-A2.1 − ). Statistical significance was assessed using linear mixed-effects models (Sidak’s post hoc correction). Reported p values represent the overall group effect (P group ) and the interaction between group and time (P interaction ); models were fit to the longitudinal data up to the point of final animal attrition of the vehicle group. Asterisks on the X-axis denote the first day of significant therapeutic divergence (p<0.05), if applicable. Ratios indicate successful tumor engraftment in TCR-TD mice. ( B ) NY-ESO-1 expression analysis by immunohistochemistry across tumor models. Lower panels show higher magnification of boxed regions. Scale bars: 2 mm, higher magnification: 100 µm. ( C ) UMAP visualization of unbiased flow cytometry analysis of pooled tumor-infiltrating lymphocytes of TCR-TD and Non-TD mice (left) and protein expression heatmap by cluster (right; n=8 per group). ( D ) Quantification of T-cell infiltration (counts/mg) per tumor model in TCR-TD and Non-TD mice by flow cytometry. ( E ) Representative flow cytometry dot plots showing PD-1 + /TCF1 + Tpex in TCR-TD versus Non-TD mice (left) and quantification across tumor models (right). ( F ) Flow cytometry analysis of functionality of tumor-infiltrating T cells by intracellular cytokine production following PMA/ionomycin stimulation of A375 tumor and spleen single-cell suspensions from TCR-TD and Non-TD mice. For panels D, E and F: bar graphs show mean+SD; For panels D and E: statistical analysis was performed using two-way ANOVA (panels D and E) or ordinary one-way ANOVA (panel F) and Fisher’s LSD multiple comparisons test; no line or ns *p>0.05; **p≤0.05; ***p≤0.01; ****p≤0.001; *****p≤0.0001. ANOVA, analysis of variance; CM, central memory; dext + , dextramer positive; EM, effector memory; imfp, intra mammary fat pad; LSD, Least significant difference; Non-TD, non-transduced; ns, not significant; PD-1, programmed cell death protein 1; PMA, Phorbol 12-myristate 13-acetate; s.c., subcutaneous; TCF1, T cell factor 1; TCR, T-cell receptor; TD, transduced; Tpex, precursor exhausted T cells; Tregs, regulatory T cells; UMAP, Uniform Manifold Approximation and Projection. Open in a new tab Baseline TA-specific T-cell frequency modulates TCB therapy efficacy The established TCR-TD humanized mouse model was used to evaluate the efficacy and immunopharmacodynamic effects of the melanoma-associated chondroitin sulfate proteoglycan (MCSP)-TCB in the A375-HLA2.1 xenograft ( figure 5A ). This model was chosen due to the broad and homogenous expression of MCSP ( figure 5B ). Mice were randomized into groups with “GFP low ” and “GFP high ” frequencies (10% cut-off) of TA-specific T cells based on the frequency of GFP + CD8 + T cells in peripheral blood, to reflect high and low baseline frequencies of TA-specific T cells, and injected with tumor cells. Once tumors reached an average size of 150–200 mm³, mice received weekly intravenous injections of 2.5 mg/kg MCSP-TCB or vehicle control (histidine buffer). MCSP-TCB demonstrated efficacy in mice with lower TA-specific T-cell frequencies, significantly delaying tumor growth and extending survival ( figure 5C ). However, no significant effect on tumor growth or survival was observed in mice with higher frequencies of TA-specific T cells ( figure 5D ). Notably, control mice in the GFP high group already exhibited slower tumor growth compared with the GFP low group, suggesting higher pre-existing antitumor reactivity as described above. Flow cytometry analysis at study termination revealed distinct immune profiles between the GFP low and GFP high groups, both at baseline and following MCSP-TCB treatment. At baseline, the GFP high group showed a higher T-cell infiltration and higher frequencies of dextramer-positive T cells and Tregs ( figure 5E ). Additionally, Ki-67 expression was elevated in dextramer-positive T cells, indicating pre-existing T-cell activation and proliferation ( figure 5F ). In contrast, the GFP low group showed lower T-cell infiltration and lower Treg frequencies and Ki-67 expression at baseline. MCSP-TCB treatment led to enhanced cytotoxicity and proliferation in the GFP low group, as evidenced by an increase in granzyme B and Ki-67 expressing cells, particularly within the dextramer-positive population ( figure 5F ). A modest increase in granzyme B expression was also observed in dextramer-negative cells. This enhanced cytotoxicity likely contributed to the observed efficacy and extended survival in this group. While MCSP-TCB did lead to increased T-cell infiltration in the GFP high group, this increase was not associated with a corresponding increase in TA-specific T cells. Instead, the frequency of dextramer-positive T cells decreased, suggesting no preferential expansion of TA-specific T cells, which is further supported by the ratio of dextramer-positive cells in tumors vs spleens, which is close to one in the TCB-treated groups. Furthermore, MCSP-TCB treatment promoted further T-cell exhaustion in the GFP high group, as evidenced by upregulation of PD-1 and T cell immunoglobulin and mucin-domain containing-3 (Tim-3) on TILs, particularly within the dextramer-positive subset. This upregulation of exhaustion markers was not observed to the same extent in the GFP low group. Finally, TCR downregulation, potentially as a consequence of chronic stimulation and exhaustion, was observed in the GFP high group after TCB therapy, as indicated by decreased CD3 expression ( figure 5F ). Figure 5. Baseline TA-specific T-cell frequency modulates TCB therapy efficacy. ( A ) Schematic illustration of the experimental design. TCR-TD humanized mice were stratified by peripheral blood GFP + CD8 + T-cell frequency (GFP low : <10%; GFP high : >10%), implanted with A375-HLA2.1 tumors, and treated with MCSP-TCB (2.5 mg/kg) or vehicle. (Vehicle: GFP low n=8 per group; MCSP-TCB: GFP low n=8; GFP high n=9) ( B ) MCSP expression in A375-HLA2.1 tumors by immunohistochemistry. Right panel shows a higher magnification of the boxed region. Scale bars: 2 mm and 200 µm. ( C ) Tumor growth kinetics and Kaplan-Meier analysis of time to reach 500 mm 3 of the GFP low (HR=0.148; 95% CI 0.041 to 0.535; p=0.0036) and ( D ) GFP high group (HR=0.846; 95% CI 0.256 to 2.79; p=0.784). ( E ) Flow cytometry analysis of tumor-infiltrating lymphocytes showing CD3 + T-cell counts per mg tumor; frequency of dextramer positive CD8 + T cells; spleen/tumor ratio of dextramer + CD8 + T cells; Treg counts per mg. ( F ) Phenotypic flow cytometry analysis of Dext + and Dext − tumor-infiltrating T cells showing cytotoxicity (granzyme B); proliferation (Ki-67) and exhaustion markers (PD-1, Tim-3, Lag-3). For panels C and D: tumor growth curves show mean+SEM; statistical significance was assessed using linear mixed-effects models (Sidak’s post hoc correction), reported p values represent the overall group effect (P group ) and the interaction between treatment and time (P interaction ), models were fit to the longitudinal data up to the point of final animal attrition in the vehicle group. Asterisks on the X-axis denote the first day of significant therapeutic divergence (p<0.05), if applicable. Time to event was monitored using Kaplan-Meier analysis. Statistical significance was determined using the Mantel-Cox (log-rank) test. The HR and 95% CIs were calculated using the log-rank approach to quantify the reduction in the risk of reaching 500 mm 3 . In panels E and F: bar graphs show mean+SD. Statistical analysis was performed using ordinary one-way ANOVA (for panel E) or two-way ANOVA (for panel F) and Fisher’s LSD multiple comparisons test; no line or ns *p>0.05; **p≤0.05; ***p≤ 0.01; ****p≤0.001; *****p≤0.0001. ANOVA, analysis of variance; dext − , dextramer negative; dext + , dextramer positive; GFP, green fluorescent protein; GZMB, Granzyme B; Lag-3, Lymphocyte activation gene 3; LSD, Least significant difference; MCSP, melanoma-associated chondroitin sulfate proteoglycan; MFI, Mean fluorescent intensity; ns, not significant; s.c., subcutaneous; PD-1, programmed cell death protein 1; TA, tumor antigen; TCB, T-cell bispecific antibody; TCF1, T cell factor 1; TCR, T-cell receptor; TD, transduced; Tim-3, T cell immunoglobulin and mucin-domain containing-3; Tregs, regulatory T cells. Open in a new tab While MCSP-TCB demonstrated efficacy in the A375-HLA2.1 model, particularly in settings with low pre-existing immunity, its activity in the parental A375 model (lacking HLA-A2.1 overexpression) was limited. MCSP-TCB monotherapy did not significantly delay tumor growth ( online supplemental figure 9 ). To overcome this resistance, MCSP-TCB was combined with either FAP-CD40 or FAP-4-1BBL, targeting fibroblast activation protein (FAP) expressed in the tumor stroma. Both combination therapies significantly delayed tumor growth ( online supplemental figure 10 ), but with distinct mechanisms. FAP-CD40 enhanced TA-specific T-cell frequency in the spleen and augmented cytotoxicity of tumor-infiltrating T cells (increased granzyme B expression). In contrast, FAP-4-1BBL promoted a polyclonal CD8 + T-cell response, reduced exhaustion markers (Tim-3, PD-1), and increased the frequency of PD-1 + /TCF1 + Tpex cells within the tumor. FAP-CD40 and FAP-4-1BBL enhance the activity of TA-specific T cells in TCR-TD humanized mice To demonstrate the utility of the TCR-TD model for evaluating immunotherapies relying on endogenous antitumor immunity, the activity of FAP-CD40 and FAP-4-1BBL was assessed in TCR-TD ( figure 6A ) and Non-TD ( figure 6B ) mice bearing A375 melanoma tumors. These molecules use an FAP-targeting moiety that cross-reacts with both human and murine FAP, thereby anchoring the CD40 or 4-1 BBL binding arms to the murine stromal fibroblasts within the xenograft microenvironment. Mice were randomized into the different treatment arms and treatment was initiated when tumors reached an average size of 150–200 mm³. In TCR-TD mice, both FAP-CD40 and FAP-4-1BBL, administered as monotherapies or in combination, significantly delayed tumor growth and extended survival ( figure 6C ). In contrast, the combination therapy had no effect on tumor growth in Non-TD mice ( figure 6D ), highlighting the essential role of TA-specific T cells in mediating their therapeutic efficacy. In the spleen, both therapies led to an expansion of CD8 + T cells, and FAP-CD40 monotherapy increased, while FAP-4-1BBL monotherapy decreased the frequency of dextramer-positive T cells ( figure 6E ). The combination of FAP-CD40 and FAP-4-1BBL resulted in an increase in splenic dextramer-positive T cells. Ex vivo flow cytometry analysis of explanted tumors at termination revealed highest total CD8 + T-cell counts in the combination group ( figure 6F ). Intratumoral dextramer-positive T cells were increased by FAP-CD40 monotherapy but were further enhanced by the addition of FAP-4-1BBL, reaching approximately fivefold higher levels in the combination group compared with the vehicle group. The PD-1 + /TCF1 + Tpex population, only present in TCR-TD mice as shown earlier, showed FAP-CD40 driven expansion, which was also further enhanced in the combination group. Further analysis of the therapy effects on TA-specific T cells revealed an elevated granzyme B and Ki-67 expression of intratumoral dextramer-positive cells after FAP-CD40 and FAP-4-1BBL therapies, with the highest granzyme B levels in the combination group ( figure 6G , online supplemental figure 11 ). PD-1 and Tim-3 were not upregulated after any of the therapies; in fact, PD-1 levels were lower in treatment groups compared with the vehicle group, indicating a potential delay or prevention of T-cell exhaustion. Cytokine analysis of the tumor lysates revealed distinct cytokine abundances of the two therapies. FAP-4-1BBL primarily drove the secretion of TNF-α, IFN-γ, Interferon Gamma-induced Protein 10 (IP-10), and IL-2, while a synergistic increase in IL-6 was observed only in the combination group and mainly in the TCR-TD setting ( figure 6H , online supplemental figure 11 ). Ex vivo restimulation of tumor-infiltrating lymphocytes with PMA/ionomycin confirmed increased frequencies of TNF-α and IFN-γ positive cells in the FAP-CD40 and FAP-4-1BBL groups ( online supplemental figure 11 ). Figure 6. FAP-CD40 and FAP-4-1BBL enhance the activity of TA-specific T cells in TCR-TD humanized mice. ( A ) Schematic illustration of the experimental design in TCR-TD and ( B ) Non-TD mice. A375 cells were injected subcutaneously, followed by treatment with vehicle, FAP-CD40 (13.3 mg/kg), FAP-4-1BBL (1 mg/kg) or combination of both (n=12 mice per group), using a murine/human cross-reactive FAP-binder on both molecules. ( C ) Tumor growth kinetics and Kaplan-Meier analysis of time to reach 1000 mm 3 of TCR-TD (vehicle vs combination therapy: HR: 0.292, 95% CI 0.111 to 0.767, p=0.0125; overall p=0.0297) and ( D ) Non-TD mice (HR=0.9291; 95% CI 0.392 to 2.20; p=0.866). ( E ) Flow cytometry analysis of splenic T cells at day 29 showing frequencies of CD8 + and Dext + T cells. ( F ) Flow cytometry analysis of tumor-infiltrating T cells showing counts per mg of CD8 + T cells, dextramer + CD8 + T cells and PD-1 + /TCF1 + T cells (pooled from day 29 and termination). ( G ) Phenotypic flow cytometry analysis of Dext + and Dext − tumor-infiltrating T cells showing cytotoxicity (granzyme B); proliferation (Ki-67) and exhaustion markers (PD-1, Tim-3). ( H ) Heatmap showing relative levels of cytokines and chemokines in tumor lysates across treatment groups. For panels C and D: tumor growth curves show mean+SD; statistical significance was assessed using linear mixed-effects models (Sidak’s post hoc correction), reported p values represent the overall group effect (P group ) and the interaction between treatment and time (P interaction ), models were fit to the longitudinal data up to the point of last animal attrition of the vehicle group. Asterisks on the X-axis denote the first day of significant therapeutic divergence (p<0.05), if applicable. Time to event was monitored using Kaplan-Meier analysis. Statistical significance was determined using the Mantel-Cox (log-rank) test. The HR and 95% CIs were calculated using the log-rank approach to quantify the reduction in the risk of reaching 1000 mm 3 . In panels E, F and G: Bar graphs show mean+SD. Statistical analysis was performed using ordinary one-way ANOVA (for panels E and F) or two-way ANOVA (for panel G) and Fisher’s LSD multiple comparisons test; no line or ns *p>0.05; **p≤0.05; ***p≤0.01; ****p≤0.001; *****p≤0.0001. ANOVA, analysis of variance; dext − , dextramer negative; dext + , dextramer positive; FAP, fibroblast activation protein; GZMB, Granzyme B; IFN, interferon; IL, interleukin; IP-10, Interferon Gamma-induced Protein 10; MFI, Mean fluorescent intensity; Non-TD, non-transduced; ns, not significant; s.c., subcutaneous; PD-1, programmed cell death protein 1; TCF1, T cell factor 1; TCR, T-cell receptor; TD, transduced; Tim-3, T cell immunoglobulin and mucin-domain containing-3; TNF, tumor necrosis factor. Open in a new tab Discussion The successful development of effective cancer immunotherapies critically depends on our ability to predict and optimize therapeutic responses, yet preclinical models struggle to recapitulate the complex interactions between human immune cells and the TME. Current humanized mouse models, while valuable, often fail to generate robust and trackable populations of TA-specific T cells, limiting their utility in large-scale drug development. 5 Here, we demonstrate that lentiviral delivery of predefined TCR specificities to human CD34 + HSCs creates a versatile and reproducible platform for studying human TA-specific T-cell biology in vivo . This approach overcomes key limitations of existing models, including the need for fetal tissue, 24 26 extensive HSC manipulation, 21 or HLA-transgenic mouse strains, 23 31 while enabling precise control over the frequency and specificity of tumor-reactive T cells. The utility of this approach is enhanced by several technical adaptations that improve its accessibility and scalability. By minimizing in vitro manipulation through a streamlined 48-hour protocol, we maintain high HSC quality and engraftment rates, 32 as evidenced by sustained CD34 and CD133.1 expression. The incorporation of transduction enhancers 33 eliminates the need for cell sorting while achieving efficient gene delivery. Importantly, despite widespread transgene expression across hematopoietic lineages driven by the EF1α promoter, functional TCR expression remains restricted to T cells due to CD3-dependent surface transport. 34 In B cells and other non-T lineages, the absence of CD3 subunits (γ, δ, ε, ζ) prevents the assembly and trafficking of the TCR-α and β chains to the cell surface, providing inherent specificity without requiring T cell-specific promoters. This refined methodology enables systematic investigation of human TA-specific T-cell responses across various therapeutic contexts, revealing that Tpex cells develop exclusively in the presence of TA-specific T cells and that high baseline frequencies of TA-specific T cells correlate with reduced TCB efficacy, findings that inform immunotherapy development and patient stratification. One particularly intriguing finding is the efficient development of HLA-A2.1-restricted TCR transgenic T cells in humanized mice, irrespective of cognate HLA expression in the host or donor HSCs. This observation challenges the canonical model where positive selection relies on interactions with self-MHC molecules expressed by cortical thymic epithelial cells. While cross-reactivity with murine MHC is formally possible, it is unlikely given the successful development observed with multiple distinct TCR specificities. Our results align with previous studies that demonstrate murine MHC-independent human T-cell development in humanized mice. 35 36 Instead, human hematopoietic cells likely contribute significantly to shaping the overall human immune environment in these models. We observed prominent thymic B-cell populations localized near the cortico-medullary junction expressing high levels of MHC class I and II, consistent with their capacity for antigen presentation and known roles in negative selection. 37 While human hematopoietic cells presumably influence the development and selection of the endogenous T-cell repertoire through MHC interactions, the efficient maturation of transgenic HLA-A2.1-restricted T cells even from HLA-A2.1-negative donors demonstrates that their positive selection occurs independently of cognate TCR-HLA interactions. Instead, the developmental advantage and HLA-independence of transgenic thymocytes strongly suggest autonomous signaling driven by the early-expressed transgenic TCR. TCR-seq confirmed the efficient suppression of the endogenous TCRβ locus, likely due to allelic exclusion mediated by the transgenic TCR-β chain. 25 38 39 This early expression, combined with engineered cysteine modifications promoting preferential pairing of transgenic α and β chains, 40 likely facilitates premature assembly of a functional TCR complex. We hypothesize that this complex delivers constitutive survival and developmental signals, functionally analogous to the ligand-independent signaling driven by the pre-TCR complex, 41 42 thereby enabling transgenic thymocytes to bypass classical MHC-dependent checkpoints and avoid death by neglect. This hypothesis is supported by pseudotime analysis, which revealed direct transitions from proliferating DP to SP stages in TCR-TD mice, circumventing the quiescent DP phase typical of conventional positive selection. 43 Intriguingly, we observed the development of both CD4 + and CD8 + T cells expressing the MHC class I-restricted transgenic TCR. This raises the question of how lineage commitment is determined in the hypothesized absence of cognate MHC-TCR interactions which usually dictate the T-cell commitment. 44 While the precise mechanism remains to be investigated, our results suggest that the transgenic TCR does partially influence lineage fate, favoring CD8 + T cells. However, the presence of a significant transgenic CD4 + population indicates that this influence is not absolute. Several factors could contribute to this outcome, such as heterogeneity in TCR signaling strength, variations in TCR expression level, availability of co-receptors, cytokines or morphological and spatial localization in the thymic microenvironment. 45 Further studies are needed to better dissect the exact mechanisms behind the lineage commitment of these transgenic thymocytes. The injection of human NY-ESO-1 + HLA-A2.1 + xenograft models into TCR-TD mice resulted in complete tumor growth rejections or delayed growth kinetics compared with non-TD mice, suggesting antitumor responses by the TA-specific T cells. The growth differences correlated with the peripheral frequency of TA-specific T cells; a variable that can be controlled with the TCR-TD model presented here. Therefore, this model enables the induction of varied tumor inflammation levels, reflecting diverse tumor states or individual patient profiles. Additionally, the system may offer insights into novel mechanisms of tumor evasion from TA-specific T-cell pressure and may facilitate new target identification, as several xenografts continued to grow despite a high inflammation status. Additionally, we observed an increased phenotypic diversity of intratumoral T cells in TCR-TD mice, most notably a prominent population of PD-1 + /TCF1 + Tpex cells exclusively within these tumors. This finding suggests that chronic TCR signaling in TA-specific T cells is essential for intratumoral Tpex cell development and/or persistence. Their presence in the TCR-TD model is especially important as they have been shown to play a key role in various therapeutic approaches, 1146 , 48 yet their detailed role, especially in a human context, remains an area of ongoing research that could be potentially further explored in this system. Furthermore, the role of transgenic CD4 + T cells expressing the MHC class-I restricted TCR is not yet fully clear. High-affinity TCRs can achieve activation independently of CD8 co-receptor stabilization, 49 50 a phenomenon supported by our in vitro and ex vivo activation data. While these CD4 + cells likely contribute to the antitumor response through localized cytokine help, their specific in vivo contribution remains difficult to isolate from the one of CD8 + effectors, primarily due to the limited dextramer binding on CD4 + T cells, despite functional competence. This limitation could be addressed in the future by incorporating MHC class II-restricted TCRs. To showcase the application of the model in preclinical research of cancer immunotherapies, we investigated the impact of TA-specific T cells on TCB activity. Our findings indicate that more inflamed or “hot” tumors, characterized by a higher presence of TA-specific T cells, exhibit reduced responsiveness to MCSP-TCB compared with less inflamed tumors. Furthermore, TCB treatment did not preferentially expand NY-ESO-1-specific T cells; in contrary, it led to decreased frequencies of dextramer + CD8 + T cells, suggesting infiltration of polyclonal T cells and thereby a polyclonal activity of the TCB. Whether these findings result from chronic TCR stimulation leading to exhaustion and dysfunction of TA-specific T cells, or from a TME with higher T-cell diversity, including Tregs and other T-cell phenotypes, requires further investigation. However, clinical data for glofitamab, a CD20-TCB, indicate that patients with progressive disease have higher baseline PD-1 signature scores compared to responders, suggesting that putatively exhausted T cells may be less responsive to TCB engagement. 51 Furthermore, in line with our results, other reports in syngeneic mouse models indicate no preferential clonal expansion of OVA-specific T cells on TCB treatment. 52 These findings are highly relevant for future TCB optimizations and have the potential to support patient stratification, treatment regimens, and combination strategies. Beyond TCBs, the TCR-TD platform provides a valuable opportunity to evaluate immunotherapies designed to target pre-existing human TA-specific T-cell immunity in vivo , addressing an unmet need in preclinical development. We exemplified this capability by assessing an FAP-targeted myeloid agonist (FAP-CD40) and costimulator (FAP-4-1BBL). Significant antitumor efficacy for both agents, alone or combined, was observed only in TCR-TD mice harboring TA-specific T cells, demonstrating the necessity of this cellular compartment for their therapeutic activity. The ability to track and phenotype the defined TA-specific population within this model enabled the dissection of their distinct modes of action, which was not possible in conventional humanized mice. FAP-CD40 preferentially expanded TA-specific T cells, including the functionally important Tpex population, whereas FAP-4-1BBL appeared to elicit a broader CD8 + T-cell response potentially mitigating exhaustion. Their combination yielded the most potent expansion of TA-specific effectors. Furthermore, these findings offer a compelling preclinical rationale for combination strategies; integrating molecules like FAP-CD40 or FAP-4-1BBL with TCBs may overcome the limited efficacy of TCB monotherapy observed in solid tumors, representing a promising therapeutic strategy for this challenge. While this study primarily compares TA-specific TCR-TD mice to Non-TD baselines, we used HLA-mismatched tumor models to demonstrate that antitumor efficacy and the development of the Tpex phenotype are strictly dependent on cognate MHC/peptide recognition rather than autonomous TCR signaling. Nevertheless, the future use of tumor-irrelevant, virus-specific TCR-TD controls could further refine our understanding of the interplay between tonic signaling and antigen-driven exhaustion in this system. The TCR-TD humanized mouse model presented here offers a powerful platform for dissecting the complex human biology of TA-specific T cells in vivo . While sharing conceptual roots with murine transgenic systems like the OT-I model, our platform provides a significant translational advantage by enabling the evaluation of actual human clinical lead molecules on human T-cell responses. With precise control over TA-specific T-cell frequency and clonality, this model allows researchers to explore how pre-existing immunity drives tumor progression and escape, shapes the microenvironment, and influences therapeutic responses to both TCBs and non-TCB agents. It is a valuable tool for optimizing therapeutic molecules during development, supporting combination assessments, and informing patient selection strategies. Future improvements could involve optimized mouse strains with enhanced human myeloid compartments to study antigen presenting cells (APC)-TA-specific T-cell interactions. Additionally, the model could be expanded to include MHC class II-restricted TCRs for studying CD4 + T cell-mediated responses, or even transition to other research fields such as autoimmune diseases or virology. By bridging the gap between preclinical insights and clinical applications, this model has the potential to increase the translatability of preclinical findings to clinical settings. Supplementary material online supplemental file 1 jitc-14-4-s001.pdf (329.5KB, pdf) DOI: 10.1136/jitc-2025-013989 online supplemental file 2 jitc-14-4-s002.pdf (3.1MB, pdf) DOI: 10.1136/jitc-2025-013989 Acknowledgements We thank Leo Kunz for his help in setting up a suitable 3DIP panel and the 360 Genomics lab for the sequencing of the thymus samples. Furthermore, we thank Julia Fossati for her help with the CRISPR/Cas9-mediated knock-outs. Footnotes Funding: This study was funded by F. Hoffmann-La Roche Ltd. As employees of the funder, the authors were responsible for the study design, data collection, analysis, interpretation, and the writing of the manuscript. The funder did not influence the results/outcomes of the study despite author affiliations with the funder. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: All animal experiments were performed in compliance with the Swiss Animal Welfare Act and were approved by the Cantonal Veterinary Office of Zurich. Data availability free text: The complete dataset generated for this study and that support the findings of this study are available from the corresponding author upon reasonable request. Data availability statement Data are available upon reasonable request. References 1. Chen DS, Mellman I. Oncology meets immunology: the cancer-immunity cycle. Immunity. 2013;39:1–10. doi: 10.1016/j.immuni.2013.07.012. [ DOI ] [ PubMed ] [ Google Scholar ] 2. Gajewski TF, Schreiber H, Fu Y-X. Innate and adaptive immune cells in the tumor microenvironment. Nat Immunol. 2013;14:1014–22. doi: 10.1038/ni.2703. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Fridman WH, Zitvogel L, Sautès-Fridman C, et al. The immune contexture in cancer prognosis and treatment. Nat Rev Clin Oncol. 2017;14:717–34. doi: 10.1038/nrclinonc.2017.101. [ DOI ] [ PubMed ] [ Google Scholar ] 4. Mhaidly R, Verhoeyen E. Humanized Mice Are Precious Tools for Preclinical Evaluation of CAR T and CAR NK Cell Therapies. Cancers (Basel) 1915;12:1915. doi: 10.3390/cancers12071915. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Chuprin J, Buettner H, Seedhom MO, et al. Humanized mouse models for immuno-oncology research. Nat Rev Clin Oncol. 2023;20:192–206. doi: 10.1038/s41571-022-00721-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Sharma P, Allison JP. The future of immune checkpoint therapy. Science. 2015;348:56–61. doi: 10.1126/science.aaa8172. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Tumeh PC, Harview CL, Yearley JH, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature New Biol. 2014;515:568–71. doi: 10.1038/nature13954. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Siddiqui BA, Carneiro BA, Stein MN, et al. A phase 1/2 study of nezastomig (anti-PSMA×CD28) with or without cemiplimab (anti–PD-1) in patients (pts) with metastatic castration-resistant prostate cancer (mCRPC) and clear cell renal cell carcinoma (ccRCC) JCO. 2025;43 doi: 10.1200/JCO.2025.43.5_suppl.TPS293. [ DOI ] [ Google Scholar ] 9. Claus C, Ferrara C, Xu W, et al. Tumor-targeted 4-1BB agonists for combination with T cell bispecific antibodies as off-the-shelf therapy. Sci Transl Med. 2019;11:eaav5989. doi: 10.1126/scitranslmed.aav5989. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Sum E, Rapp M, Fröbel P, et al. Fibroblast Activation Protein α-Targeted CD40 Agonism Abrogates Systemic Toxicity and Enables Administration of High Doses to Induce Effective Antitumor Immunity. Clin Cancer Res. 2021;27:4036–53. doi: 10.1158/1078-0432.CCR-20-4001. [ DOI ] [ PubMed ] [ Google Scholar ] 11. Codarri Deak L, Nicolini V, Hashimoto M, et al. PD-1-cis IL-2R agonism yields better effectors from stem-like CD8+ T cells. Nature New Biol. 2022;610:161–72. doi: 10.1038/s41586-022-05192-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Wu B, Zhang B, Li B, et al. Cold and hot tumors: from molecular mechanisms to targeted therapy. Signal Transduct Target Ther. 2024;9:274. doi: 10.1038/s41392-024-01979-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Sade-Feldman M, Yizhak K, Bjorgaard SL, et al. Defining T Cell States Associated with Response to Checkpoint Immunotherapy in Melanoma. Cell. 2019;176 doi: 10.1016/j.cell.2018.12.034. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Scheper W, Kelderman S, Fanchi LF, et al. Low and variable tumor reactivity of the intratumoral TCR repertoire in human cancers. Nat Med. 2019;25:89–94. doi: 10.1038/s41591-018-0266-5. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Zhou S, Liu M, Ren F, et al. The landscape of bispecific T cell engager in cancer treatment. Biomark Res. 2021;9:38. doi: 10.1186/s40364-021-00294-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Bacac M, Colombetti S, Herter S, et al. CD20-TCB with Obinutuzumab Pretreatment as Next-Generation Treatment of Hematologic Malignancies. Clin Cancer Res. 2018;24:4785–97. doi: 10.1158/1078-0432.CCR-18-0455. [ DOI ] [ PubMed ] [ Google Scholar ] 17. Leclercq G, Steinhoff N, Haegel H, et al. Novel strategies for the mitigation of cytokine release syndrome induced by T cell engaging therapies with a focus on the use of kinase inhibitors. Oncoimmunology. 2022;11:2083479. doi: 10.1080/2162402X.2022.2083479. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Eckmann J, Fauti T, Biehl M, et al. Forimtamig, a novel GPRC5D-targeting T-cell bispecific antibody with a 2+1 format, for the treatment of multiple myeloma. Blood. 2025;145:202–19. doi: 10.1182/blood.2024025987. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Lee JY, Han A-R, Lee DR. T Lymphocyte Development and Activation in Humanized Mouse Model. Dev Reprod. 2019;23:79–92. doi: 10.12717/DR.2019.23.2.079. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Strowig T, Gurer C, Ploss A, et al. Priming of protective T cell responses against virus-induced tumors in mice with human immune system components. J Exp Med. 2009;206:1423–34. doi: 10.1084/jem.20081720. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Najima Y, Tomizawa-Murasawa M, Saito Y, et al. Induction of WT1-specific human CD8+ T cells from human HSCs in HLA class I Tg NOD/SCID/IL2rgKO mice. Blood. 2016;127:722–34. doi: 10.1182/blood-2014-10-604777. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Provasi E, Genovese P, Lombardo A, et al. Editing T cell specificity towards leukemia by zinc finger nucleases and lentiviral gene transfer. Nat Med. 2012;18:807–15. doi: 10.1038/nm.2700. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Masterman K-A, Haigh OL, Tullett KM, et al. Human CLEC9A antibodies deliver NY-ESO-1 antigen to CD141+ dendritic cells to activate naïve and memory NY-ESO-1-specific CD8+ T cells. J Immunother Cancer. 2020;8:e000691. doi: 10.1136/jitc-2020-000691. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Vatakis DN, Koya RC, Nixon CC, et al. Antitumor activity from antigen-specific CD8 T cells generated in vivo from genetically engineered human hematopoietic stem cells. Proc Natl Acad Sci U S A . 2011;108:E1408–16. doi: 10.1073/pnas.1115050108. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Giannoni F, Hardee CL, Wherley J, et al. Allelic exclusion and peripheral reconstitution by TCR transgenic T cells arising from transduced human hematopoietic stem/progenitor cells. Mol Ther. 2013;21:1044–54. doi: 10.1038/mt.2013.8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Vatakis DN, Bristol GC, Kim SG, et al. Using the BLT Humanized Mouse as a Stem Cell based Gene Therapy Tumor Model. JoVE . 2012;e4181 doi: 10.3791/4181. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Hu Z, Xia J, Fan W, et al. Human melanoma immunotherapy using tumor antigen-specific T cells generated in humanized mice. Oncotarget. 2016;7:6448–59. doi: 10.18632/oncotarget.7044. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Lan P, Tonomura N, Shimizu A, et al. Reconstitution of a functional human immune system in immunodeficient mice through combined human fetal thymus/liver and CD34+ cell transplantation. Blood. 2006;108:487–92. doi: 10.1182/blood-2005-11-4388. [ DOI ] [ PubMed ] [ Google Scholar ] 29. Bethune MT, Li X-H, Yu J, et al. Isolation and characterization of NY-ESO-1-specific T cell receptors restricted on various MHC molecules. Proc Natl Acad Sci U S A . 2018;115:E10702–11. doi: 10.1073/pnas.1810653115. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Johnson LA, Heemskerk B, Powell DJ Jr, et al. Gene transfer of tumor-reactive TCR confers both high avidity and tumor reactivity to nonreactive peripheral blood mononuclear cells and tumor-infiltrating lymphocytes. J Immunol. 2006;177:6548–59. doi: 10.4049/jimmunol.177.9.6548. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Gschweng EH, McCracken MN, Kaufman ML, et al. HSV-sr39TK positron emission tomography and suicide gene elimination of human hematopoietic stem cells and their progeny in humanized mice. Cancer Res. 2014;74:5173–83. doi: 10.1158/0008-5472.CAN-14-0376. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Uchida N, Hsieh MM, Hayakawa J, et al. Optimal conditions for lentiviral transduction of engrafting human CD34+ cells. Gene Ther. 2011;18:1078–86. doi: 10.1038/gt.2011.63. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Schott JW, León-Rico D, Ferreira CB, et al. Enhancing Lentiviral and Alpharetroviral Transduction of Human Hematopoietic Stem Cells for Clinical Application. Mol Ther Methods Clin Dev. 2019;14:134–47. doi: 10.1016/j.omtm.2019.05.015. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Call ME, Wucherpfennig KW. Molecular mechanisms for the assembly of the T cell receptor-CD3 complex. Mol Immunol. 2004;40:1295–305. doi: 10.1016/j.molimm.2003.11.017. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Läderach F, Zdimerova H, Rieble L, et al. MHC class II‐deficient mice allow functional human CD4 + T‐cell development. Eur J Immunol. 2023;53:e2250313. doi: 10.1002/eji.202250313. [ DOI ] [ PubMed ] [ Google Scholar ] 36. Darguzyte M, Antczak P, Bachurski D, et al. Long-Term Human Immune Reconstitution, T-Cell Development, and Immune Reactivity in Mice Lacking the Murine Major Histocompatibility Complex: Validation with Cellular and Gene Expression Profiles. Cells. 2024;13:1686. doi: 10.3390/cells13201686. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Castañeda J, Hidalgo Y, Sauma D, et al. The Multifaceted Roles of B Cells in the Thymus: From Immune Tolerance to Autoimmunity. Front Immunol. 2021;12:766698. doi: 10.3389/fimmu.2021.766698. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Khor B, Sleckman BP. Allelic exclusion at the TCRbeta locus. Curr Opin Immunol. 2002;14:230–4. doi: 10.1016/s0952-7915(02)00326-6. [ DOI ] [ PubMed ] [ Google Scholar ] 39. Brady BL, Steinel NC, Bassing CH. Antigen receptor allelic exclusion: an update and reappraisal. J Immunol. 2010;185:3801–8. doi: 10.4049/jimmunol.1001158. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Cohen CJ, Li YF, El-Gamil M, et al. Enhanced antitumor activity of T cells engineered to express T-cell receptors with a second disulfide bond. Cancer Res. 2007;67:3898–903. doi: 10.1158/0008-5472.CAN-06-3986. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Saint-Ruf C, Ungewiss K, Groettrup M, et al. Analysis and expression of a cloned pre-T cell receptor gene. Science. 1994;266:1208–12. doi: 10.1126/science.7973703. [ DOI ] [ PubMed ] [ Google Scholar ] 42. von Boehmer H. Unique features of the pre-T-cell receptor α-chain: not just a surrogate. Nat Rev Immunol. 2005;5:571–7. doi: 10.1038/nri1636. [ DOI ] [ PubMed ] [ Google Scholar ] 43. Park J-E, Botting RA, Domínguez Conde C, et al. A cell atlas of human thymic development defines T cell repertoire formation. Science. 2020;367:eaay3224. doi: 10.1126/science.aay3224. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Singer A, Adoro S, Park J-H. Lineage fate and intense debate: myths, models and mechanisms of CD4- versus CD8-lineage choice. Nat Rev Immunol. 2008;8:788–801. doi: 10.1038/nri2416. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Singer A, Bosselut R. CD4/CD8 coreceptors in thymocyte development, selection, and lineage commitment: analysis of the CD4/CD8 lineage decision. Adv Immunol. 2004;83:91–131. doi: 10.1016/S0065-2776(04)83003-7. [ DOI ] [ PubMed ] [ Google Scholar ] 46. Siddiqui I, Schaeuble K, Chennupati V, et al. Intratumoral Tcf1+PD-1+CD8+ T Cells with Stem-like Properties Promote Tumor Control in Response to Vaccination and Checkpoint Blockade Immunotherapy. Immunity. 2019;50:195–211. doi: 10.1016/j.immuni.2018.12.021. [ DOI ] [ PubMed ] [ Google Scholar ] 47. Im SJ, Hashimoto M, Gerner MY, et al. Defining CD8+ T cells that provide the proliferative burst after PD-1 therapy. Nature New Biol. 2016;537:417–21. doi: 10.1038/nature19330. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Miller BC, Sen DR, Al Abosy R, et al. Subsets of exhausted CD8+ T cells differentially mediate tumor control and respond to checkpoint blockade. Nat Immunol. 2019;20:326–36. doi: 10.1038/s41590-019-0312-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 49. Chervin AS, Stone JD, Holler PD, et al. The Impact of TCR-Binding Properties and Antigen Presentation Format on T Cell Responsiveness. J Immunol. 2009;183:1166–78. doi: 10.4049/jimmunol.0900054. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Laugel B, van den Berg HA, Gostick E, et al. Different T Cell Receptor Affinity Thresholds and CD8 Coreceptor Dependence Govern Cytotoxic T Lymphocyte Activation and Tetramer Binding Properties. J Biol Chem. 2007;282:23799–810. doi: 10.1074/jbc.M700976200. [ DOI ] [ PubMed ] [ Google Scholar ] 51. Bröske A-ME, Korfi K, Belousov A, et al. Pharmacodynamics and molecular correlates of response to glofitamab in relapsed/refractory non-Hodgkin lymphoma. Blood Adv. 2022;6:1025–37. doi: 10.1182/bloodadvances.2021005954. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 52. Hutter-Karakoc I, Varypataki EM, Neelakandhan A, et al. Chronic antigen stimulation in solid tumors induces t cell exhaustion and limits efficacy of t cell bispecific therapies. bioRxiv. 2025 doi: 10.1101/2025.01.29.635556. Preprint. [ DOI ] [ PMC free article ] [ PubMed ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials online supplemental file 1 jitc-14-4-s001.pdf (329.5KB, pdf) DOI: 10.1136/jitc-2025-013989 online supplemental file 2 jitc-14-4-s002.pdf (3.1MB, pdf) DOI: 10.1136/jitc-2025-013989 Data Availability Statement Data are available upon reasonable request. Articles from Journal for Immunotherapy of Cancer are provided here courtesy of BMJ Publishing Group ACTIONS View on publisher site PDF (10.6 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top