Elucidating Immune Cell Changes in Celiac Disease: Revealing New Insights from Spectral Flow Cytometry

Abstract

Celiac disease (CD) is an immune-mediated enteropathy of the small intestine triggered by gluten ingestion. Although the small bowel is the main organ affected, peripheral blood cell alterations have also been described in CD. We aimed to investigate immunological cell patterns in the blood of treated CD patients and in response to a 3-day gluten challenge (GC). Blood samples were collected from 10 patients with CD and 8 healthy controls on a gluten-free diet at baseline and 6 days after initiating the GC. All the samples were analyzed by spectral flow cytometry using a 34-marker panel. We found that patients with CD displayed a lower proportion of memory B cells compared to healthy controls, both at baseline and post-GC. Additionally, we observed the previously reported activated gut-homing CD4⁺, CD8⁺, and TCRγδ⁺ T lymphocytes on day 6 post-GC, and found the CD8⁺ subpopulation to be the most readily identifiable by flow cytometry. Importantly, the CCR9 marker proved effective in enhancing the selection of these gluten-responsive T cells, offering the potential for increased diagnostic accuracy. Spectral flow cytometry involves a complex data analysis, but it offers valuable insights into previously unexplored immunological responses and enables in-depth cell characterization.

Keywords: B cells; T cells; gluten challenge; gluten-free diet; spectral cytometry.

Figure A1

Density plots and marker expression gradients across the dimensionality reductions for (A) CD45⁺ and (B) CD3⁺ subsets.

Figure A2

Graphical representation of the four cell subpopulations differentially expressed between patients with celiac disease (CD) and healthy controls (HC) stratified by sex. (A) CD27⁺ CD45RA⁺ HLA-DR⁺ CD19⁺ CD20⁺ in relation to the total CD19⁺ CD20⁺; (B) CD49d+β7^hi CD38⁺ CD4⁺ in relation to the total CD4⁺; (C) CD103+β7^hi CD38⁺ CD8⁺ in relation to the total CD8⁺; and (D) CD103+β7^hi CD38⁺ TCRγδ⁺ in relation to the total TCRγδ⁺. P-values were calculated using an unpaired Student’s t-test. Only significant p-values are shown.

Figure A3

Correlation between age and the frequency of each differentially expressed cell subtype in samples collected on day 6 following the 3-day gluten challenge in celiac disease (CD) and healthy controls (HC). (A) CD27⁺ CD45RA⁺ HLA-DR⁺ CD19⁺ CD20⁺ in relation to the total CD19⁺ CD20⁺; (B) CD49d+β7^hi CD38⁺ CD4⁺ in relation to the total CD4⁺; (C) CD103+β7^hi CD38⁺ CD8⁺ in relation to the total CD8⁺; and (D) CD103+β7^hi CD38⁺ TCRγδ⁺ in relation to the total TCRγδ⁺.

Figure 1

Flow cytometry gating strategy used to distinguish major immune cell populations, including CD4⁺ T cells, CD8⁺ T cells, TCRγδ⁺ T cells, B cells, monocytes, dendritic cells, basophils, and innate lymphoid cells (ILCs).

Figure 2

Unsupervised analysis of CD45⁺ cells. (A) Overlay of the 70 clusters generated by FlowSOM in the dimensionality reduction. (B) Clusters mapped by group and time point, highlighting the five significant clusters identified and their location in the dimensionality reduction. (C) Boxplots displaying significant comparisons (p < 0.05) between groups at each time point (CD: celiac disease; HC: healthy controls) and between time points (baseline vs. day 6) within each group. (D) Dimensionality reduction plot showing the overlap between the significant clusters identified by the unsupervised analyses and the manually gated populations. (E) Visualization of the significant clusters with their relative expression of cell lineage and phenotype markers. Lineage markers that were expressed and contributed to the identification of the representative population of each cluster are presented, along with all markers related to cell trafficking, cell nature, activation, and immune checkpoints.

Figure 3

Unsupervised analysis of CD3⁺ cells. (A) Overlay of the 40 clusters generated by FlowSOM in the dimensionality reduction. (B) Clusters mapped by group and time point, highlighting the two significant clusters identified. (C) Boxplots showing significant comparisons (p < 0.05) between groups on day 6 (CD: celiac disease; HC: healthy controls) and between time points (baseline vs. day 6) in the CD group. (D) Dimensionality reduction showing the overlap between the significant clusters and the manually gated populations. (E) Visualization of the significant clusters with their relative expression of T cell lineage and phenotype markers.

Figure 4

Percentage of the CD27⁺ HLA-DR⁺ CD45RA⁺ CD20⁺ CD19⁺ cells with respect to the total CD20⁺ CD19⁺ cells in patients with celiac disease (CD) and healthy controls (HC) on a gluten-free diet (baseline) and on day 6 following a 3-day gluten challenge.

Figure 5

Percentage of activated gut-homing T cells in relation to the total CD4⁺, CD8⁺, and TCRγδ⁺ corresponding to patients with celiac disease (CD) and healthy controls (HC) on a gluten-free diet (baseline) and on day 6 following a 3-day gluten challenge. Only significant p-values are shown.

Figure 6

Percentage of the activated gut-homing T cells of interest across the three T cell subsets analyzed: CD4⁺, CD8⁺, and TCRγδ⁺ T cells. Only significant p-values are shown.

Figure 7

Mean percentage and standard error of the mean of phenotypic marker expression in manually identified activated gut-homing CD4⁺, CD8⁺, and TCRγδ⁺ T cells. Only significant p-values are shown, with p-values for differences between days displayed in blue and p-values for differences between cell types shown in black.