Raman spectral signatures reflect altered metabolic state of exhausted T cells

Abstract

T cell exhaustion within the tumor microenvironment drives CD8+ T cells into a dysfunctional state characterized by progressive loss of proliferative capacity and effector functions, thereby limiting anti-tumor immunity and therapeutic efficacy. To investigate the biochemical alterations associated with exhaustion, an in vitro model of CD8+ T cell exhaustion was established through chronic antigenic stimulation of murine OT-1 CD8+ T cells. Phenotypic, functional, metabolic, and transcriptional characterization confirmed the acquisition of an exhausted state. Single-cell Raman spectroscopy was subsequently employed to generate biochemical signatures of activated, and exhausted CD8+ T cells. Principal component analysis (PCA) of the Raman spectral data revealed distinct separation of these two cell subsets, reflecting underlying biochemical differences associated with their functional states. Differential Raman spectral features corresponding to nucleic acids, carbohydrates, proteins, and lipids contributed significantly to this segregation, reflecting altered metabolic and biosynthetic states during exhaustion progression. Classification of the spectral data using machine-learning algorithms enabled accurate segregation of activated and exhausted T cells. Collectively, this study demonstrates that single-cell Raman spectroscopy can distinguish exhausted CD8+ T cells in a label-free and non-destructive manner, highlighting its potential as a platform for immune profiling and monitoring T cell dysfunction.