Characterizing Phenotypes and Signaling Networks of Single Human Cells by Mass Cytometry

Abstract

Single cell mass cytometry is revolutionizing our ability to quantitatively characterize cellular biomarkers and signaling networks. Mass cytometry experiments routinely measure 25-35 features of each cell in primary human tissue samples. The relative ease with which a novice user can generate a large amount of high quality data and the novelty of the approach have created a need for example protocols, analysis strategies, and datasets. In this chapter, we present detailed protocols for two mass cytometry experiments designed as training tools. The first protocol describes detection of 26 features on the surface of human peripheral blood mononuclear cells. In the second protocol, a mass cytometry signaling network profile measures 25 node states comprised of five key signaling effectors (AKT, ERK1/2, STAT1, STAT5, and p38) quantified under five conditions (Basal, FLT3L, SCF, IL-3, and IFNγ). This chapter compares manual and unsupervised data analysis approaches, including bivariate plots, heatmaps, histogram overlays, SPADE, and viSNE. Data files in this chapter have been shared online using Cytobank ( http://www.cytobank.org/irishlab/ ).

Keywords: Human; Immunophenotyping; Mass cytometry (CyTOF); Phospho-specific flow cytometry (phospho-flow); Signaling network profile; Single cell biology.

Conflict of interest statement

Conflict of interest disclosure: J.M.I. declares a competing financial interest (cofounder and board member of Cytobank Inc.).

Figures

Fig. 1

Phenotyping human PBMC subsets with traditional bivariate gating & heatmap analysis. (a) Bivariate plots compare features measured on healthy human PBMCs by mass cytometry. Heat corresponds to proportional cell abundance in a plot region for the population indicated above the plot. Intact cells (grey gate) were defined using event length and an iridium based cell marker (Ir-191 intercalator, NA). Among the intact cells, leukocytes were defined as CD45+ events. CD3 and CD19 were then used to identify T cells and B cells, respectively. Subsets of T cells and B cells were identified using additional markers (CD4, CD8, CD45RA, and IgM). The non-T non-B cells (CD45 + CD3 − CD19 −) were gated as monocytes, dendritic cells, and natural killer (NK) cell using CD14, CD11b, CD11c, HLA-DR, CD16, and CD56. (b) A heatmap compares expression of 27 measured features on the same cells populations shown in (a) using a log-like arcsinh15 scale. The heat corresponds to the arcsinh15 fold difference in median expression for a given marker compared to the table minimum

Fig. 2

SPADE clusters PBMCs into populations based on similar marker expression. SPADE plots show clustered populations of healthy human PBMCs (circles) connected using a minimum spanning tree. Each circle represents a population of cells with a similar phenotype for the 21 markers shown. The size of each circle is proportional to the number of cells in that population. The heat color for each circle corresponds to the median expression of the indicated marker on the cells within that circle. Heat corresponds to the arcsinh15 fold difference in median expression for a given marker. Note the scale min and max differ for each marker. Black outlines termed “bubbles” and associated population labels derive from manual interpretation of cellular identity based on marker expression

Fig. 3

viSNE arranges cells in a 2D map representing phenotypic similarity. viSNE maps show healthy human PBMCs arranged according to phenotypic similarity for the 21 displayed markers measured by mass cytometry. The axes are unitless dimensions that reflect phenotypic differences. The distance between any two cells on the map corresponds to how similar or different the cells are from each other in high-dimensional space. Heat corresponds to the arcsinh15 fold difference in median expression for a given marker. This allows for a global single-cell view of every parameter in every cell. Cell populations can then be identified through various techniques, including automated clustering or manual analysis of well-characterized markers

Fig. 4

Mass cytometry phospho-flow analysis of AML cell signaling responses. (a) A heatmap compares phospho-protein abundance in Kasumi-1 AML cells following 15 min of stimulation by FLT3L, SCF, IL-3, or IFNγ. Each row in the heatmap or histogram overlay corresponds to a stimulation condition and each column corresponds to a phospho-protein (p-AKT, p-ERK1/2, p-STAT5, p-STAT1, or p-p38). Heat corresponds to the arcsinh15 fold difference in median expression for a given marker compared to the unstimulated condition (Unstim.). (b) The same data as in (a) are shown in histogram overlay format. Histogram overlays illustrate the distribution of marker expression within a population and highlight heterogeneity (e.g., p-STAT5 response to IL-3)