Pathogenic T Cells in Celiac Disease Change Phenotype on Gluten Challenge: Implications for T-Cell-Directed Therapies

Abstract

Gluten-specific CD4⁺ T cells being drivers of celiac disease (CeD) are obvious targets for immunotherapy. Little is known about how cell markers harnessed for T-cell-directed therapy can change with time and upon activation in CeD and other autoimmune conditions. In-depth characterization of gluten-specific CD4⁺ T cells and CeD-associated (CD38⁺ and CD103⁺ ) CD8⁺ and γδ⁺ T cells in blood of treated CeD patients undergoing a 3 day gluten challenge is reported. The phenotypic profile of gluten-specific cells changes profoundly with gluten exposure and the cells adopt the profile of gluten-specific cells in untreated disease (CD147⁺ , CD70⁺ , programmed cell death protein 1 (PD-1)⁺ , inducible T-cell costimulator (ICOS)⁺ , CD28⁺ , CD95⁺ , CD38⁺ , and CD161⁺ ), yet with some markers being unique for day 6 cells (C-X-C chemokine receptor type 6 (CXCR6), CD132, and CD147) and with integrin α4β7, C-C motif chemokine receptor 9 (CCR9), and CXCR3 being expressed stably at baseline and day 6. Among gluten-specific CD4⁺ T cells, 52% are CXCR5⁺ at baseline, perhaps indicative of germinal-center reactions, while on day 6 all are CXCR5^- . Strikingly, the phenotypic profile of gluten-specific CD4⁺ T cells on day 6 largely overlaps with that of CeD-associated (CD38⁺ and CD103⁺ ) CD8⁺ and γδ⁺ T cells. The antigen-induced shift in phenotype of CD4⁺ T cells being shared with other disease-associated T cells is relevant for development of T-cell-directed therapies.

Keywords: RNA-Seq; T cells; celiac disease; gluten challenge; mass cytometry.

Figure 1

Study design and initial clustering of gluten HLA‐DQ2.5:gluten tetramer‐binding (Tet⁺) CD4⁺ T cells. A) Depiction of study design. In part I, gut homing (integrin β7⁺) Tet⁺ and Tet⁻ effector memory T (T_EM) (CD45RA⁻, CD62L⁻) cells were bulk‐sorted and analyzed with RNA‐seq (n = 6). In Part II, mass cytometry was performed (n = 10). Based on surface markers previously reported to characterize gluten‐specific CD4⁺ T cells in untreated celiac disease (CeD), i.e., T‐cell autoimmune (T_A) markers, and with addition of surface markers identified from RNA‐seq results, a mass cytometry staining panel was designed. Labeled antibodies were used to stain PBMCs obtained before (baseline, BL) and 6 days (d6) after a 3 day gluten challenge. B) Increase in numbers of Tet⁺ integrin β7⁺ cells per million CD4⁺ T cells (paired t‐test, median value indicated) as detected with flow cytometry. Patients chosen for RNA‐seq analysis are indicated. C,D) Graph plots of high‐dimensional analysis (UMAP and t‐sne) with adjunct histograms depicting all CD4⁺ T cells and visualizing Tet⁺ cells on d6 (red), baseline (dark gray) and all CD4⁺ T cells merged from baseline and d6 (light gray). Red and dark gray arrows indicate clusters of Tet⁺ cells on d6 and BL, respectively. Color maps indicate the distribution of cells stained for CD45RA, CD62L and integrin β7⁺ confirming location of Tet⁺ cells on baseline and d6 as these markers were used to sort cells for RNA‐seq.

Figure 2

Phenotype of HLA‐DQ2.5:gluten tetramer‐binding (Tet⁺) cells before and after oral gluten challenge. A) Mass cytometry‐derived UMAP plot of integrin β7⁺ effector memory T (T_EM) cells of challenge patients (n = 10). Cluster 1A and cluster 1B define Tet⁺ at baseline (dark gray) and cluster 2 define Tet⁺ cells on d6 (red). Light gray represents integrin β7⁺ T_EM cells merged from baseline and day 6. B) Percentage of Tet⁺ and integrin β7⁺ T_EM cells on baseline (BL) and day 6 (d6) which locate within clusters 1A, 1B, and 2, respectively. The 28 markers used to create the plots are visualized with their cellular distributions in Figure S3 (Supporting Information). C–F) Heat map depicting log₂ fold change (FC) of homing markers, chemokine and interleukin (IL) receptors as analyzed with RNA‐seq and mass cytometry. Tet⁺ cells at d6, Tet⁺ cells at BL in clusters 1A and 1B are compared with integrin β7⁺ T_EM cells d6. To the right of the heat maps, absolute staining intensities for each individual marker as found with mass cytometry staining of Tet⁺ cells d6 and integrin β7⁺ T_EM cells are shown. In panels (D) and (F), log₂ FC RNA expression of some additional differentially expressed homing markers not analyzed by mass cytometry are shown.

Figure 3

Phenotype of HLA‐DQ2.5:gluten tetramer‐binding (Tet⁺) cells of untreated CeD versus treated CeD before and after gluten challenge. A) Mass cytometry‐derived UMAP plot (generated from plot of Figure 2A) visualizing only integrin β7⁺ effector memory T (T_EM) cells from untreated CeD patients. Location of cluster 3, defining most Tet⁺ cells in untreated celiac disease (UCeD) patients versus location of clusters 1A, 1B, and 2, which defined Tet⁺ cells on baseline and day 6 after gluten challenge. B) Percentage of Tet⁺ from 4 UCeD patients locating within clusters 2 and 3, respectively. The 28 markers used to create the plot are visualized with their staining distribution in Figure S3 (Supporting Information). C,D) Heat map depicting log₂ fold change of markers analyzed by mass cytometry comparing cell types as indicated. All data include integrin β7⁺ CD4⁺ T_EM cells only.

Figure 4

Ranking T‐cell markers defining gluten‐specific CD4⁺ T cells on day 6 after gluten challenge. A) Stepwise logistic regression analysis with receiver operating characteristics (ROC) comparing three set of markers for their ability to distinguish gluten‐specific CD4⁺ T cells from other gut‐homing (integrin β7⁺) effector memory (CD45RA⁻ CD62L⁻) cells. The blue line represents a previously established set of markers (previous),^{[ 6 ]} the green line represents a set of markers identified from RNA‐seq results (i.e., this study) (new) and the red line represents a mix of markers from the two sets (mixed) (comparison including also integrin β7, CD45RA and CD62L in Figure S6 in the Supporting Information). The predictor with the highest p‐value was removed at each step and model performance was evaluated using 10‐fold cross validation. B) Correlation of the 32 markers constituting the mixed panel looking at gluten tetramer binding (Tet⁺) and nontetramer binding (Tet⁻) cells 6 days after gluten challenge (d6). Samples were down‐sampled to be proportional between sample types (Tet^+/−) and donors. C) Top 8 markers to define Tet⁺ on d6 of gluten challenge and D) for Tet⁺ cells on d6 and in untreated CeD (UCeD) combined. P values reflect the ability of each indicated marker to predict the Tet⁺ cells.

Figure 5

Comparing phenotypic characteristics of celiac disease‐associated (CD38⁺, CD103⁺) CD8⁺ and γδ ⁺ T cells with gluten‐specific CD4⁺ T cells at day 6 after gluten challenge. A) Mass cytometry‐derived t‐sne and UMAP plots depicting all CD3⁺ T cells, with CD8⁺, γδ ⁺, and CD4⁺ T‐cell types in addition to the CD38⁺ CD103⁺ subsets of CD8⁺ (turquoise arrows) and γδ ⁺ (red arrows) T cells color coded, on day 6 after gluten challenge. B) Percentage of CD103⁺ CD38⁺ cells among CD8⁺ T cells in a representative CeD patient C) and in n = 7 CeD patients on baseline and day 6. D) UMAP plot highlighting CD103⁺ CD38⁺ cells among CD8⁺ T cells, E) with percentage of cells within cluster 4 at baseline and day 6 (n = 7). F) Percentage of CD103⁺ CD38⁺ cells among γδ ⁺ T cells in a representative CeD patient G) and in n = 7 CeD patients on baseline and day 6. H) UMAP plot highlighting CD103⁺ CD38⁺ cells among γδ ⁺ T cells, I) with percentage of cells within cluster 6 at baseline and day 6 (n = 7). J) Correlation between mass cytometry‐derived log₂ fold‐change expression of indicated markers on cluster 4 cells versus all CD8⁺ T cells (x axis, n = 6, including only those with increase in (C) and cluster 6 cells versus all γδ ⁺ T cells (y axis, n = 4, including only those with detectable increase in (G). K) Correlation between log₂ fold‐change expression of indicated markers on cluster 4 versus all CD8⁺ T cells (x axis) and HLA‐DQ2.5:gluten tetramer‐positive gut‐homing effector memory cells in cluster 2 (see Figure 2) versus all CD4⁺ T cells (y axis, n = 6). The median log₂ fold changes from the CeD patients analyzed on day 6 of oral gluten challenge are depicted J,K).