Skip to main content
. 2020 May 22;9:e53226. doi: 10.7554/eLife.53226

Figure 2. The cytokine co-expression is altered in PD-1+ T cells in young and old mice.

(A) Disease status of NZBxW mice was scored according to their age and proteinuria (PU). The disease score was correlated with the frequencies of PD-1+ T cells and serum levels of anti-ds-DNA-IgG antibodies. (B) Disease-associated changes of the protein expression shown by histogram overlays (top) and statistical analyses (bottom). (C) Representative gating scheme of PD-1-, PD-1low and PD-1hi subpopulations in CD4+ CD44+ T cells from mice of disease score 1 and 5, respectively. (D) Statistical analyses of the frequencies of IL-2, TNF-α, IFN-γ and IL-10 producers in PD-1 subpopulations from mice with disease score of 1 (light grey) and 5 (dark grey). (E) Frequencies of TNF-α + cells of the PD-1+ IFN-γ+ subpopulation in disease score 1 and 5. (F) Representative bin plots of disease score 1 and 5 with PD-1 (x-axis), IFN-γ (y-axis) displaying the frequencies of TNF-α, IL-2 and IL-10, respectively, per bin. Cut-off for PD-1hi cells is marked with dashed lines. Data represent two independent experiments with (A), n = 2 mice for each score, (B– F) n = 6–7 mice per group. Samples were compared using the Mann Whitney test (B), a repeated measure two-way ANOVA with Geisser-Greenhouse correction and Dunnett‘s multiple comparison test (D) and a two-sided unpaired t-test (E). Data are presented as the mean ± SEM.

Figure 2—source data 1. Figure 2A: Serum levels of anti-ds-DNA-IgG vs. PD-1 frequency.
Figure 2—source data 2. Figure 2B: Frequencies of protein expression of mice in disease score 1 and 5, respectively.
Data represent two independent experiments with n = 6 or seven mice per group.
Figure 2—source data 3. Figure 2D: Frequencies of IL-2, TNF-α, IFN-γ and IL-10 producers in PD-1 subpopulations of mice in disease score 1 and 5, respectively.
Data represent two independent experiments with n = 7 mice per group.
Figure 2—source data 4. Figure 2E: Frequencies of TNF-α cells of the PD-1+ IFN-γ+ subpopulation in mice of disease score 1 and 5, respectively.
Data represent two independent experiments with n = 5 mice per group.

Figure 2.

Figure 2—figure supplement 1. The combinatorics of the co-expression of IFN-γ, TNF-α and IL-2 in CD44+ T cells is altered with disease progression.

Figure 2—figure supplement 1.

(A) Gating strategy to identify CD44+ CD4+ T cells. (B–D) Quantification of cytokine co-expression in disease stages by pie charts (B), bar plots (C) and bin plots (D). Bin plots visualize density (top row), frequency of IFN-γ producers (middle row) and expression level of IFN-γ (MFI+, bottom row). (D) To compare IFN-γ expression levels, a common scaling was used to depict the MFI+ (IFN-γ) for all disease scores (1–5) together. Cell frequencies per quadrant are calculated on the number of cells per sample (black) and number of Z+ cells per sample (green). Grey bins contain less than 10 Z+ cells. Data represent one experiment with PMA/ionomycin stimulated splenic T cells: (B–D), n = 2 mice per group. Data are presented as mean in (C).
Figure 2—figure supplement 1—source data 1. Figure 2—figure supplement 1B: Frequencies of boolean combinations of the co-expression of cytokines in mice of disease score 1 to 5.
Data represent one experiment with PMA/ionomycin stimulated splenic T cells: n = 2 mice per group.
Figure 2—figure supplement 2. PD-1+ T cells exhibit differential cytokine expression levels compared to PD-1- T cells in young and old diseased mice.

Figure 2—figure supplement 2.

(A) Representative histogram overlays for disease score 1 (top) and 5 (bottom) showing the cytokine expression in the PD-1 subgroups. Controls are colored in grey. (B) Barplots with mean fluorescence intensities (MFI) of IL-2, TNF-α, IFN-γ and IL-10 per PD-1 subpopulation in disease score 1 and 5. (C, D) Bin pots for CD44+ T cells of disease score 1 (top) and 5 (bottom) with PD-1 (x-axis), IFN-γ (y-axis) show cell density and MFI+ of TNF-α, IL-2 (C) and IL-10 producing cells (D). (D) To analyze IL-10 expression level, files of four samples were concatenated for both disease scores. The MFI range in (C, D) was normalized between the two scores, but determined individually for each cytokine. PD-1hi cut-off is marked with dashed lines. Cell frequencies per quadrant are calculated on the number of cells per sample (black) and number of Z+ cells per sample (green). Grey bins contain less than 10 Z+ cells. (E) Pie charts representing the boolean combinations of the co-expression of IL-2, TNF-α, IFN-γ and IL-10 in PD-1 subpopulations in disease score 1 and 5, respectively. Data represent two independent experiments with (A, C), n = 6 mice per group, (B), n = 4 mice per group, and (E), n = 7 mice per group. (B) Samples were compared using a repeated measure two-way ANOVA with Geisser-Greenhouse correction and Dunnett‘s multiple comparison test. Data are presented as the mean ± s.e.m.
Figure 2—figure supplement 2—source data 1. Figure 2—figure supplement 2B: Mean fluorescence intensity in PD-1 subpopulations of mice in disease score 1 and 5, respectively.
Data represent two independent experiments with n = 4 mice per group.
Figure 2—figure supplement 2—source data 2. Figure 2—figure supplement 2E: Frequencies of boolean combinations of the co-expression of IL-2, TNF-α, IFN-γ and IL-10 in PD-1 subpopulations in disease score 1 and 5, respectively.
Data represent two independent experiments with n = 7 mice per group.
Figure 2—figure supplement 3. Bin plot patterns are reproducible and statistically robust.

Figure 2—figure supplement 3.

(A) Representative contour plots for the co-production of IL-10 with other cytokines by PMA/ionomycin stimulated T cells of diseased mice. (B–C) Comparison of the IL-10 cell expression patterns of four individual (B) and three concatenated (C) mice (different experiments) of each score (score 1, top row; score 5, bottom row). Cell frequencies per quadrant are calculated on the number of cells per sample (black) and number of Z+ cells per sample (green). (D) Statistical analysis of the frequencies shown in the upper right quadrant in the bin plots from (B). (E) Different cut-off values for IL-10 were used (histogram, left) to create the respective bin patterns for IL-10 (right). (F) Different settings of minimum of cells per bin were used to visualize the relative standard error of the mean (RSEM) of the IL-10 signal per bin. Samples in (D) were compared using two-sided unpaired t-test with Welch-correction. Data are presented as the mean ± s.e.m.
Figure 2—figure supplement 3—source data 1. Figure 2—figure supplement 3D: Frequencies shown in the upper right quadrant in the bin plots of mice in disease score 1 and 5, respectively.
Data represent two independent experiments with n = 4 mice per group. Samples were compared using an unpaired two-tailed t-test.