Background
Chimeric antigen receptor (CAR) T cell-based therapy has greatly improved cancer treatment, especially for hematologic cancers.1 By providing T cells with synthetic receptors that target tumor antigens, like CD19 in B cell malignancies, CAR T cells are enabled to identify and destroy cancer cells with high precision and potency.2 Despite this progress, hurdles such as antigen loss, T-cell dysfunction, and limited success in solid tumors underscore the need for deeper mechanistic insights into CAR T cell biology.
Recent research indicates that the spatial architecture of membrane proteins is crucial for regulating T-cell signaling and activation.3-5 In the context of CAR T cells, the nanoscale localization and clustering of chimeric antigen receptors on the plasma membrane can influence basal signaling activity, modulate receptor sensitivity, and influence immune synapse assembly.6-8 The organization of receptors within membrane microdomains, along with the dynamic formation of multiprotein complexes, is thought to fine-tune both the intensity and duration of signaling, thereby influencing T-cell persistence, effector function, and therapeutic efficacy.9-11
To explore this complex molecular landscape, proteomic approaches that additionally provide information about protein organization are needed. By mapping the networks and membrane domains associated with the CAR, we can begin to understand how its function is regulated. This understanding could guide the development of strategies to optimize CAR T-cell activity and enhance the therapeutic potential.
Methods
Here, we harnessed the Proximity Network Assay to investigate the membrane architecture of CD19-targeting CAR T cells, both in their resting state and during tumor engagement (Figure 1). The Proximity Network Assay is a high-multiplex, sequencing-based assay that maps protein organization at single-cell resolution by linking DNA-barcoded antibodies into spatial interaction networks.12 The result is a detailed view of immune receptor organization across thousands of cells simultaneously. Commercially available CAR T cells (CD19 scFv-FLAG-TM-CD28-CD3ζ) were sourced from PROMAB and challenged with CD19-positive tumor cells (Raji). Cells were profiled using the Pixelgen Proxiome Kit, Immuno 155, spiked in with a barcoded anti-FMC63 or anti-FLAG detection antibody to enable CAR detection.
Sample processing and data generation were handled using the publicly available ‘nf-core/pixelator’ pipeline, which calculates both protein abundance and spatial proximity scores for each cell and protein. Downstream data analysis was conducted using the open-source R package pixelatorR (also available as a Python package called ‘pixelgen-pixelator’) and publicly available R packages.
Identifying the CD19 CAR proxiome
To study the local protein environment surrounding the CD19 CAR, we applied the Proximity Network Assay. CAR-positive cells could be efficiently detected in a complex cell mix containing 90% resting PBMCs and 10% CAR T cells (Figure 2A). These CD19 CAR-positive cells typically presented a large number of CAR molecules, well distributed all over the cell surface in accordance with previous reports (Figure 2B).7 To enable identification of various CARs, FMC63-based detection was compared to FLAG-based detection of double-positive CAR cells.
As expected, similar fractions of CAR+ cells were identified using the two strategies (Figure 2C). The spatial organization of CD8+ CAR+ cells was evaluated focusing on 64 proteins expressed on the cells (Figure 2D). The CAR molecule exhibited significant colocalization with canonical signaling and adhesion proteins, including TCRαβ, CD5, CD6, CD44, and ICAM-1/2/3, highlighting its incorporation into functional membrane domains (Fig. 2D, E). Conversely, the CAR was excluded from tetraspanins such as CD53 and CD81, and from lipid raft-associated proteins like CD52, CD55, and CD59. The segregation from lipid rafts mirrors native TCR organization in resting cells, suggesting a conserved spatial mechanism regulating synthetic and endogenous receptors.
Response to tumor cells
To explore how membrane composition changes in response to target engagement, CD19 CAR T cells were cocultured with CD19+ Raji tumor cells for either four or 24 hours, followed by fixation and analysis via the Proximity Network Assay. Pronounced alterations in protein profiles were observed in CD8+ T cells at both time points. Early activation at four hours was marked by upregulation of CD69 and 4-1BB, whereas by 24 hours, elevated levels of PD-1, GITR, and OX40 indicated sustained activation and regulatory signaling. Simultaneously, reductions in CD28, 2B4, and TCRαβ expression suggested altered costimulatory capacity and TCR involvement. Notably, the CAR signal itself diminished over time, possibly due to receptor internalization, antigen masking, or shedding. These observations reveal how CAR T cells dynamically reconfigure their surface proteome during antigen engagement.
Trogocytosis induces membrane remodeling
A striking finding was the marked increase in ICAM-1 (CD54) expression on CD8 T cells post-coculture (Figure 3A). Three-dimensional visualization revealed that ICAM-1 was not uniformly distributed but rather appeared in concentrated patches on the cell surface (Figure 3B). These patches were rich in B-cell markers such as CD40 and CD21, while showing exclusion from T cell-specific CD8, indicating acquisition of membrane fragments from tumor cells via trogocytosis (Figure 3B, C). The spatial resolution of the Proximity Network Assay makes it uniquely suited to distinguish between upregulated expression and membrane acquisition.
To quantify trogocytosis, we developed an approach to count tumor-derived patches on individual T cells (Figure 3D). Activated CD25+ cells demonstrated increased patch acquisition, and PD-1+ (exhausted) cells harbored the highest number of tumor patches (Figure 3E), aligning with previous findings linking trogocytosis to progressive T-cell dysfunction.13 These insights emphasize the importance of membrane transfer in shaping CAR T-cell phenotype and efficacy over time.
Summary
Traditional methods such as flow cytometry and CITE-seq offer valuable information on protein expression but overlook spatial context. Here, we present the Proximity Network Assay as a next-generation platform for CAR T-cell profiling. This method provides both quantitative and spatially resolved insights into surface receptor landscapes, capturing multiprotein domains, receptor clustering, and trogocytosis events. By resolving receptor organization at single-cell resolution, the assay offers a high-definition view of CAR T cell biology and functional states, empowering future efforts to enhance engineered T-cell therapies.
References
1. Cappell KM, Kochenderfer JN. Long-term outcomes following CAR T cell therapy: what we know so far. Nat Rev Clin Oncol. 2023;20(6):359-371. doi:10.1038/s41571-023-00754-1
2. Zhang X, Zhu L, Zhang H, Chen S, Xiao Y. CAR-T Cell Therapy in Hematological Malignancies: Current Opportunities and Challenges. Front Immunol. 2022;13:927153. doi:10.3389/fimmu.2022.927153
3. Li B, Lu Y, Zhong MC, et al. Cis interactions between CD2 and its ligands on T cells are required for T cell activation. Sci Immunol. 2022;7(74):eabn6373. doi:10.1126/sciimmunol.abn6373
4. Lacouture C, Chaves B, Guipouy D, et al. LFA-1 nanoclusters integrate TCR stimulation strength to tune T-cell cytotoxic activity. Nat Commun. 2024;15(1):407. doi:10.1038/s41467-024-44688-3
5. Jung Y, Wen L, Altman A, Ley K. CD45 pre-exclusion from the tips of T cell microvilli prior to antigen recognition. Nat Commun. 2021;12(1):3872. doi:10.1038/s41467-021-23792-8
6. Sarén T, Saronio G, Marti Torrell P, et al. Complementarity-determining region clustering may cause CAR-T cell dysfunction. Nat Commun. 2023;14(1):4732. doi:10.1038/s41467-023-40303-z
7. Chen J, Qiu S, Li W, et al. Tuning charge density of chimeric antigen receptor optimizes tonic signaling and CAR-T cell fitness. Cell Res. 2023;33(5):341-354. doi:10.1038/s41422-023-00789-0
8. Chockley PJ, Ibanez-Vega J, Krenciute G, Talbot LJ, Gottschalk S. Synapse-tuned CARs enhance immune cell anti-tumor activity. Nat Biotechnol. 2023;41(10):1434-1445. doi:10.1038/s41587-022-01650-2
9. Balagopalan L, Raychaudhuri K, Samelson LE. Microclusters as T Cell Signaling Hubs: Structure, Kinetics, and Regulation. Front Cell Dev Biol. 2021;8. doi:10.3389/fcell.2020.608530
10. Manz BN, Jackson BL, Petit RS, Dustin ML, Groves J. T-cell triggering thresholds are modulated by the number of antigen within individual T-cell receptor clusters. Proc Natl Acad Sci. 2011;108(22):9089-9094. doi:10.1073/pnas.1018771108
11. Wu L, Wei Q, Brzostek J, Gascoigne NRJ. Signaling from T cell receptors (TCRs) and chimeric antigen receptors (CARs) on T cells. Cell Mol Immunol. 2020;17(6):600-612. doi:10.1038/s41423-020-0470-3
12. Karlsson F, Simonetti M, Galonska C, et al. Single-Cell Protein Interactomes by the Proximity Network Assay. bioRxiv. Preprint posted online June 24, 2025:2025.06.19.660329. doi:10.1101/2025.06.19.660329
13. Chen Y, Xin Q, Zhu M, et al. Trogocytosis in CAR immune cell therapy: a key mechanism of tumor immune escape. Cell Commun Signal. 2024;22(1):521. doi:10.1186/s12964-024-01894-2
Hanna van Ooijen, PhD, is an immunology application scientist, Vincent van Hoef, PhD, is a bioinformatics scientist, and Divya Thiagarajan, PhD, is a senior scientist at Pixelgen Technologies AB in Stockholm, Sweden.
