Figure 6.
The dominant effect of RA on human alloresponses is to increase GI-tropic RARαhi CD8 effector T cells and is potentiated in IL-23–rich conditions. (A) Proportion of CD8 and CD4 T cells proliferating after allostimulation in the absence or presence of RA (0.1 and 1 µM). Results are shown for 9 independent experiments. (B) Heat map of percentage of alloproliferative CD8 and CD4 T cells expressing GI-tropic molecules, high RARα, the activation marker CD25, the effector T-cell transcription factor T-bet, and the CD4 regulatory T-cell transcription factor FOXP3 after allostimulation with or without exogenous RA (1 µM). Results depict median values for 6 to 25 independent experiments. (C) Fold-change in frequency of alloproliferative GI-tropic TP CD8 effector T cells (coexpressing high RARα, T-bet, IL-23R) after allostimulation with or without exogenous RA (1 µM) and the RARα-specific inhibitor ER-50891 (ER). Results depict 8 independent experiments. (D) Heat map of percentage of alloproliferative CD8 and CD4 T cells the molecules depicted in panel B after allostimulation with exogenous RA (1 µM) with or without IL-23. Results depict median values for 12 to 27 independent experiments except for α4β7 coexpression (3 independent experiments). (E) Percentage of live alloproliferative CD8 T cells expressing both α4 and β7 integrins allostimulated with or without exogenous RA (1 µM) with or without IL-23. Results depict 3 to 9 independent experiments. (F) Ratio of live alloproliferative cells with a CD8 effector T-cell (Teff):CD4 regulatory T-cell (Treg) phenotype after allostimulation with exogenous RA (1 µM) with or without IL-23. Results depict 24 independent experiments. (G) Fold-change in frequency of live alloproliferative GI-tropic TP CD8 effector T cells after allostimulation with or without exogenous RA (1 µM) and/or LPS. Results depict 13 independent experiments. (H) Unsupervised phenotypic clustering of live PBMCs after allostimulation with or without exogenous RA (1 µM) and LPS displayed as phenograph-generated tSNE plots. Clusters are derived from 7 independent experiments. (I) Heat map of relative expression levels of RARα, T-bet, IL-23R, and the GI-tropic molecules β7 and CCR9 of distinct cell trace violet (CTV)dim alloproliferative CD3 T-cell clusters. (J) Volcano plot depicting fold-change and statistical significance of abundance of distinct T-cell clusters within live PBMCs allostimulated with or without RA (1 µM) and LPS. Cluster colors are depicted in panel I. (K) Percentage of live mononuclear cells in cluster 7 in PBMCs allostimulated in the presence or absence of RA and LPS. (L) Phenotypic heat map for in vitro cluster 7 in human PBMCs allostimulated with RA and LPS. The phenotype for cluster 4 seen in peripheral blood of GI-GVHD patients is also shown for comparison. Horizontal lines are medians. nd, not done. *P < .05; **P < .01; ***P < .001; ****P < .0001, Wilcoxon matched pairs signed-rank test (B,D,E,J-K) or mixed effects models with posttest correction (A,C,E,G).

The dominant effect of RA on human alloresponses is to increase GI-tropic RARαhi CD8 effector T cells and is potentiated in IL-23–rich conditions. (A) Proportion of CD8 and CD4 T cells proliferating after allostimulation in the absence or presence of RA (0.1 and 1 µM). Results are shown for 9 independent experiments. (B) Heat map of percentage of alloproliferative CD8 and CD4 T cells expressing GI-tropic molecules, high RARα, the activation marker CD25, the effector T-cell transcription factor T-bet, and the CD4 regulatory T-cell transcription factor FOXP3 after allostimulation with or without exogenous RA (1 µM). Results depict median values for 6 to 25 independent experiments. (C) Fold-change in frequency of alloproliferative GI-tropic TP CD8 effector T cells (coexpressing high RARα, T-bet, IL-23R) after allostimulation with or without exogenous RA (1 µM) and the RARα-specific inhibitor ER-50891 (ER). Results depict 8 independent experiments. (D) Heat map of percentage of alloproliferative CD8 and CD4 T cells the molecules depicted in panel B after allostimulation with exogenous RA (1 µM) with or without IL-23. Results depict median values for 12 to 27 independent experiments except for α4β7 coexpression (3 independent experiments). (E) Percentage of live alloproliferative CD8 T cells expressing both α4 and β7 integrins allostimulated with or without exogenous RA (1 µM) with or without IL-23. Results depict 3 to 9 independent experiments. (F) Ratio of live alloproliferative cells with a CD8 effector T-cell (Teff):CD4 regulatory T-cell (Treg) phenotype after allostimulation with exogenous RA (1 µM) with or without IL-23. Results depict 24 independent experiments. (G) Fold-change in frequency of live alloproliferative GI-tropic TP CD8 effector T cells after allostimulation with or without exogenous RA (1 µM) and/or LPS. Results depict 13 independent experiments. (H) Unsupervised phenotypic clustering of live PBMCs after allostimulation with or without exogenous RA (1 µM) and LPS displayed as phenograph-generated tSNE plots. Clusters are derived from 7 independent experiments. (I) Heat map of relative expression levels of RARα, T-bet, IL-23R, and the GI-tropic molecules β7 and CCR9 of distinct cell trace violet (CTV)dim alloproliferative CD3 T-cell clusters. (J) Volcano plot depicting fold-change and statistical significance of abundance of distinct T-cell clusters within live PBMCs allostimulated with or without RA (1 µM) and LPS. Cluster colors are depicted in panel I. (K) Percentage of live mononuclear cells in cluster 7 in PBMCs allostimulated in the presence or absence of RA and LPS. (L) Phenotypic heat map for in vitro cluster 7 in human PBMCs allostimulated with RA and LPS. The phenotype for cluster 4 seen in peripheral blood of GI-GVHD patients is also shown for comparison. Horizontal lines are medians. nd, not done. *P < .05; **P < .01; ***P < .001; ****P < .0001, Wilcoxon matched pairs signed-rank test (B,D,E,J-K) or mixed effects models with posttest correction (A,C,E,G).

Close Modal

or Create an Account

Close Modal
Close Modal