Computational flow cytometry for characterizing the immune response in vaccine studies

Flow cytometry is a powerful technique used to quantify the expression of multiple extracellular and intracellular molecules on single cells, allowing the phenotypic and functional characterization of cell populations. When many parameters are investigated simultaneously, it is not feasible to analyze all the possible bi-dimensional combinations of marker expression with classical manual analysis and the use of advanced automated tools to process and analyze high-dimensional data sets becomes necessary. To overcome this problem, novel computational techniques have been developed in the recent years, and computational flow cytometry has become a novel discipline useful for providing a set of tools to analyze, visualize and interpret large amounts of cell data in a more automated and unbiased way. The present thesis is focused on the automated analysis of high-dimensional flow cytometric data to characterize, in an unbiased and data-driven way, the B and T cell immune responses elicited in the mouse model by different vaccine formulations. An in depth investigation of the automated tools currently available for the analysis of multiparametric flow cytometric data, discussing the advantages and the limits of the most commonly used algorithms was also conducted (Chapter 3). The different B cell subsets elicited by immunization with or without the vaccine adjuvant CAF01 were characterized by automated analysis employing the FlowSOM clustering approach (Chapter 4). The computational analysis allowed to identify different B cell populations, including plasmablasts, plasma cells, germinal center B cells and their intermediate subsets. Among these reactivated cells, the frequency of plasma cells was significantly higher in lymph nodes of mice immunized with the adjuvanted formulation compared to antigen alone. Integration of clustering and dimensionality reduction approaches allowed also the identification of a sub-population of germinal center memory B cells, that could not be identified with a single automated tool (Chapter 5). The polyfunctional activity of antigen-specific CD4+ T cells was analyzed in mice immunized with heterologous prime-boost vaccine formulations. The automated analysis allowed to visualize clusters of cells producing different patterns of cytokines, according to the different adjuvants used for priming and boosting, expanding the knowledge on the adjuvant role in modulating the T helper effector function (Chapter 6). In conclusion, the computational approach has allowed to characterize in an unbiased way the B and T cell responses following immunization with different vaccine strategies, detecting clusters of cells that would be hardly identified with traditional analysis. Automated tools address many needs associated with high-dimensional datasets analysis and these results strengthen their use in vaccination studies.