Hierarchical modeling for rare event detection and cell subset alignment across flow cytometry samples.

Flow cytometry is the prototypical assay for multi-parameter single cell analysis, and is essential in vaccine and biomarker research for the enumeration of antigen-specific lymphocytes that are often found in extremely low frequencies (0.1% or less). Standard analysis of flow cytometry data relies on visual identification of cell subsets by experts, a process that is subjective and often difficult to reproduce. An alternative and more objective approach is the use of statistical models to identify cell subsets of interest in an automated fashion. Two specific challenges for automated analysis are to detect extremely low frequency event subsets without biasing the estimate by pre-processing enrichment, and the ability to align cell subsets across multiple data samples for comparative analysis. In this manuscript, we develop hierarchical modeling extensions to the Dirichlet Process Gaussian Mixture Model (DPGMM) approach we have previously described for cell subset identification, and show that the hierarchical DPGMM (HDPGMM) naturally generates an aligned data model that captures both commonalities and variations across multiple samples. HDPGMM also increases the sensitivity to extremely low frequency events by sharing information across multiple samples analyzed simultaneously. We validate the accuracy and reproducibility of HDPGMM estimates of antigen-specific T cells on clinically relevant reference peripheral blood mononuclear cell (PBMC) samples with known frequencies of antigen-specific T cells. These cell samples take advantage of retrovirally TCR-transduced T cells spiked into autologous PBMC samples to give a defined number of antigen-specific T cells detectable by HLA-peptide multimer binding. We provide open source software that can take advantage of both multiple processors and GPU-acceleration to perform the numerically-demanding computations. We show that hierarchical modeling is a useful probabilistic approach that can provide a consistent labeling of cell subsets and increase the sensitivity of rare event detection in the context of quantifying antigen-specific immune responses.

Vascular cell adhesion molecule-1 (VCAM-1) and intercellular adhesion molecule-1 (ICAM-1) provide co-stimulation in positive selection along with survival of selected thymocytes.

T-cell differentiation in the thymus depends on positive selection of CD4+CD8+ double positive (DP) thymocytes by thymic major histocompatibility complex (MHC) molecules. Positive selection allows maturation of only those thymocytes that are capable of self-peptide-MHC recognition. Thymocytes that fail to bind self-peptide-MHC die by apoptosis. An important question in thymocyte differentiation is whether co-stimulation is required for positive selection and on which cells co-stimulatory molecules may be expressed in the thymus. The vascular cell adhesion molecule (VCAM-1) and the intercellular cell adhesion molecule (ICAM-1) are known to be potent co-stimulatory molecules in activation of peripheral T-cells by interacting with the integrins VLA-4 and LFA-1, respectively. We were prompted to investigate whether VCAM-1 and ICAM-1 may also act as co-stimulators during selection of thymocytes. By using recombinant proteins of murine VCAM-1 and ICAM-1 fused to the Fc region of human IgG1 (rVCAM-1, rICAM-1) we examined the capacity of VCAM-1 and ICAM-1 to act as co-stimulatory molecules in positive selection in vitro. Triggering the CD3/TCR complex together with co-stimulation applied by rVCAM-1 or rICAM-1 induced the generation of CD4+ single positive (SP) thymocytes from CD4+CD8+ DP thymocytes whereas either signal alone did not result in generation of CD4+ SP thymocytes. VCAM-1 and ICAM-1 act therefore as co-stimulatory molecules in thymocyte positive selection in vitro. The generation of CD4+ SP cells is accompanied by cell survival both when it was co-stimulated with rVCAM-1 and with rICAM-1. Importantly we show here that VCAM-1 expression in the murine thymus is restricted to cortical F4/80 positive hematopoietic antigen presenting cells (hAPC) present exclusively in the cortex whereas expression of ICAM-1 has been reported on the epithelium both in cortex and medulla. This suggests that not only the cortical epithelium may use the co-stimulatory molecule ICAM-1 to mediate positive selection, but also cortical hAPCs may contribute to positive selection of thymocytes by using the co-stimulator VCAM-1.

Dynamic populations of dendritic cell-specific ICAM-3 grabbing nonintegrin-positive immature dendritic cells and liver/lymph node-specific ICAM-3 grabbing nonintegrin-positive endothelial cells in the outer zones of the paracortex of human lymph nodes.

In the paracortex of lymph nodes, cellular immune responses are generated against antigens captured in peripheral tissues by dendritic cells (DCs). DC-SIGN (dendritic cell-specific ICAM-3 grabbing nonintegrin), a C-type lectin exclusively expressed by DCs, functions as an antigen receptor as well as an adhesion receptor. A functional homologue of DC-SIGN, L-SIGN (liver/lymph node-SIGN, also called DC-SIGN-related), is expressed by liver sinus endothelial cells. In lymph nodes, both DC-SIGN and L-SIGN are expressed. In this study, we analyzed the distribution of these two SIGN molecules in detail in both normal and immunoreactive lymph nodes. DC-SIGN is expressed by mature DCs in paracortical areas and in addition by DCs with an immature phenotype in the outer zones of the paracortex. L-SIGN expression was also detected in the outer zones on sinus endothelial cells characterized by their expression of the lymphatic endothelial markers LYVE-1 and CLEVER-1. During both cellular and humoral immune responses changes in the amount of DC-SIGN+ immature and mature DCs and L-SIGN+ endothelial cells were observed, indicating that the influx or proliferation of these cells is dynamically regulated.