Decoding CAR t cell membrane architecture with the proximity network assay

Abstract Membrane protein organization plays a critical role in shaping CAR T cell function, yet current tools lack the resolution to study these interactions at scale and in single cells. Here, we present the Proximity Network Assay, a novel sequencing-based platform for single-cell protein interactomics, capable of resolving the spatial organization and abundance of 155 surface proteins in thousands of individual cells without the use of optics. This method leverages DNA-barcoded antibodies, proximity-dependent ligation, and high-throughput sequencing to generate nanoscale “proxiomes” that describe how membrane proteins cluster and interact within single cells. Applying the assay to CD19 CAR T cells, we profiled both resting cells and those cocultured with CD19⁺ tumor targets. At baseline, the CD19 Chimeric Antigen Receptor co-localized with CD8, CD5, and ICAM-1, and was spatially excluded from lipid raft-associated proteins such as CD59 and CD55—mirroring endogenous TCR organization. Upon tumor engagement, we detected significant remodeling of the surface proteome, including upregulation of activation markers and checkpoint molecules like 4-1BB, PD-1, and GITR. Importantly, the assay's spatial resolution enabled us to distinguish true upregulation from trogocytosis, revealing that exhausted CAR T cells preferentially accumulate tumor-derived membrane fragments. The Proximity Network Assay opens a new dimension in immune cell profiling by integrating spatial context with multiplex protein quantification. This approach offers powerful capabilities for dissecting the functional architecture of immune cells and holds promise for advancing both basic research and translational applications in immunotherapy.