Deep-UV microscopy for label-free lymphocyte characterization and subtyping

Characterization of T-cell phenotypes, including their activation state and subtype, is critical for monitoring physiological immune responses to diseases and treatments (e.g., chemotherapy), and assessing manufacturing pipelines for emerging CAR- and adoptive-T cell therapies. The current gold standard for T-cell phenotyping is flow cytometry, which uses expensive commercial analyzers and class-specific fluorophores requiring trained laboratory personnel and biochemical reagents. Due to the use of exogenous fluorophores, conventional flow cytometry assays are also destructive (i.e., end point measurements). Recent advances in optical techniques, including fluorescence lifetime imaging microscopy and quantitative phase imaging, have enabled label-free, non-destructive analysis of lymphocyte populations with varying degrees of specificity. However, these techniques still have inherent drawbacks such as the requirement for expensive instrumentation, limited lateral resolution, and/or low throughput, limiting the analysis that can be performed and capability for translation to the clinic. To address these limitations, we apply deep-UV microscopy, a high-resolution, label-free molecular imaging technique previously demonstrated for complete blood count analysis and bone marrow hematopathology, among other applications. We show that static deep-UV images of T-cells can be used to phenotype cell viability and activation state using both feature-based algorithms and supervised machine learning models. Furthermore, by quantifying dynamic intracellular activity from a single-wavelength deep-UV time series, we demonstrate subtyping of CD4/8+ T-cells. Cells remain viable after the brief UV light exposure. Ultimately, we highlight the potential of deep-UV microscopy as a simple, label-free alternative for fast and accurate T-cell characterization.