Single-cell profiling of physically interacting immune cells using image-guided sorting

Abstract

Single-cell transcriptomics of cell-cell complexes offers key insights into in vivo immune interactions, yet distinguishing true physical contacts from artifacts remains challenging. Here, we integrated spectral imaging cell sorting with physical cell deconvolution to separate circulating T cell-monocyte doublets from tuberculosis patients into physically interacting (in-contact) and non-interacting (coincidental) populations. Single-cell sequencing of individual cell partners revealed that physical contact is associated with robust signatures distinct from singlets, including unidirectional monocyte-to-T cell RNA transfer, heightened metabolic activity and T cell cytotoxicity, and a trend toward antigen-specific enrichment. These signatures spanned all T cell and monocyte subpopulations, reflecting a global program of T cell-monocyte engagement. Coincidental doublets displayed intermediate transcriptional profiles between singlet and in-contact states, suggesting the presence of recently disrupted or forming physical pairs. This framework provides high-fidelity deconvolution of cellular interactions, uncovering signatures governing T cell-monocyte communication in human disease.