Mass Cytometry Defines Virus-Specific CD4+ T Cells in Influenza Vaccination

Abstract

The antiviral response to influenza virus is complex and multifaceted, involving many immune cell subsets. There is an urgent need to understand the role of CD4⁺ T cells, which orchestrate an effective antiviral response, to improve vaccine design strategies. In this study, we analyzed PBMCs from human participants immunized with influenza vaccine, using high-dimensional single-cell proteomic immune profiling by mass cytometry. Data were analyzed using a novel clustering algorithm, denoised ragged pruning, to define possible influenza virus-specific clusters of CD4⁺ T cells. Denoised ragged pruning identified six clusters of cells. Among these, one cluster (Cluster 3) was found to increase in abundance following stimulation with influenza virus peptide ex vivo. A separate cluster (Cluster 4) was found to expand in abundance between days 0 and 7 postvaccination, indicating that it is vaccine responsive. We examined the expression profiles of all six clusters to characterize their lineage, functionality, and possible role in the response to influenza vaccine. Clusters 3 and 4 consisted of effector memory cells, with high CD154 expression. Cluster 3 expressed cytokines like IL-2, IFN-γ, and TNF-α, whereas Cluster 4 expressed IL-17. Interestingly, some participants had low abundance of Clusters 3 and 4, whereas others had higher abundance of one of these clusters compared with the other. Taken together, we present an approach for identifying novel influenza virus-reactive CD4⁺ T cell subsets, a method that could help advance understanding of the immune response to influenza, predict responsiveness to vaccines, and aid in better vaccine design.

FIGURE 1.. Schematic representation of DRP algorithm.

(A) The original data matrix has individual cells as rows and cell surface markers as columns. (B) A principal components analysis is performed on this original data matrix, and the data matrix is denoised by hard eigenvalue thresholding (red line). (C) An initial tree is created based on an agglomerative hierarchical clustering of the denoised data. (D) This tree is then pruned back to create a set of clusters (circles of different colors) that equal or exceed the first minimum cell count (e.g., illustrated in this study with minimum of four). (E) This pruning process is repeated for each successively larger minimum cell count. (F) The optimal cluster solution is from that minimum cell count of greatest improvement in cluster quality relative to next smallest minimum cell count. Together, (D) and (F) illustrate the formation of nested clusters (see Materials and Methods). Actual trees will be many times larger than depicted in this study because total cell counts will equal or exceed O [10³]. For this reason, DRP is computer-memory demanding. The figure was generated using R package dendextend (64), R package MASS (65), base R (43), and Microsoft PowerPoint and Word and Windows Paint (Microsoft Corporation, Redmond, WA). See Statistical Methods and also see Section S1.2.2 in Bruggner et al. (24).

FIGURE 2.. Estimated proportions of all six clusters identified by DRP.

The estimated proportions of each cluster identified by DRP at day 0 and day 7 are represented graphically. Left panels show the influenza peptide–stimulated condition, and right panels show the unstimulated condition. Each line represents a single study participant. *p < 0.05. Black star symbols are estimated mean proportions for that cluster, stimulation condition, and day with the numeric values reported to four to five decimal places. The estimates of mean proportions differ from Table III because they are without correction for regression covariates. The figure was prepared in SAS ODS Graphics (SAS Institute).

FIGURE 3.. Star plots of three of the six total clusters identified by DRP.

Star plots of Clusters 3, 4, and 6 show the expression of various markers represented in two-dimensional space. Individual cells from each cluster are represented on the star plot as colored dots (3, blue; 4, green; and 6, magenta). Arrows indicate the expression of individual markers. Clockwise from top left shows day 0 unstimulated, day 0 stimulated, day 7 stimulated, and day 7 unstimulated. Star plots were produced through stratified random downsampling to 82 cells for each combination of cluster, visit, and condition, which allowed equal weighting of all three clusters for both visits and both conditions. For this reason, these star plots can only be used to assess patterns in expression and not in abundance. Software packages for producing star plots were base R plus R packages matrixcalc, plotrix, JPEN, sampling, and VCA, as indicated in the Materials and Methods. Stim day 7 modified from Holmes et al. (34) with permission from Mary Ann Liebert, Inc., New Rochelle, NY. Stim, stimulated with HA1NP peptide mix; Unstim, unstimulated.

FIGURE 4.. Heat map of expression profiles for all six Clusters.

The heat map shows mean expression level [Arsinh (y/5), y = raw expression] across stimulation conditions and visits in color scale that ranges from white (high) to black (low) for all markers on the panel (labeled at bottom). Cluster label numbers are shown on the right. This heat map allows approximate determination of the mean phenotypic and functional characteristics of the cells that constitute the represented clusters. The heat map was produced with R package heatmap3, as indicated in the Materials and Methods.

FIGURE 5.. Expression profiles of Cluster 3 and Cluster 4 represented as dot plots.

The expression (raw intensity) of some important characterizing markers expressed by Cluster 3 (blue) and Cluster 4 (green) are shown in the form of dot plots. Each dot is an individual cell. Markers shown are (A) CCR7 and CD45RA, (B) IL-2 and IFN-γ, (C) IL-17 and TNF-α, and (D) CD154 and ICOS. The figure was prepared in SAS ODS Graphics (SAS Institute).

FIGURE 6.. Distribution of Cluster 3 and Cluster 4 within each participant.

The distribution of cluster abundance (square root of proportion of CD4⁺ T cells) for Clusters 3 and 4 is shown by participant, separately for each combination of visit and stimulation condition. Each line represents a single study participant. Color-coding is by individual to show how the abundance of cells differs between clusters for that individual. Square-root transformation (vertical axes) facilitates visual separation of individuals. The figure was prepared in SAS ODS Graphics (SAS Institute). Stim, stimulated with HA+NP peptide mix; Unstim, unstimulated.