Volume 132, Issue 7, Pages 2371-2382 (June 2007) Reduced Chemokine Receptor 9 on Intraepithelial Lymphocytes in Celiac Disease Suggests Persistent Epithelial Activation  Richard W. Olaussen, Malin R. Karlsson, Knut E.A. Lundin, Jørgen Jahnsen, Per Brandtzaeg, Inger N. Farstad  Gastroenterology  Volume 132, Issue 7, Pages 2371-2382 (June 2007) DOI: 10.1053/j.gastro.2007.04.023 Copyright © 2007 AGA Institute Terms and Conditions

Figure 1 Flow-cytometric analyses of duodenal cell suspensions from a healthy control subject (no. 6: A, D, G, and J) vs a patient with treated (no. 7: B, E, H, and K) or untreated (no. 19: C, F, I, and L) celiac disease (CD). Displayed is the degree of CCR9 expression among viable (A–F) CD3+ or (G–L) CD8+ cells in the lymphocyte region of the (A–C and G–I) epithelial or (D–F and J–L) lamina propria suspensions, respectively. Percentages of cells in each quadrant are shown in the upper right corner of the plots. Cursors were set according to plots obtained with concentration- and isotype-matched irrelevant antibodies for each subject and compartment. Gastroenterology 2007 132, 2371-2382DOI: (10.1053/j.gastro.2007.04.023) Copyright © 2007 AGA Institute Terms and Conditions

Figure 2 Scatter plots of compiled flow-cytometric data displaying CCR9 expression on viable (A and B) CD3+ or (C and D) CD8+ cells in the lymphocyte region of duodenal (A and C) epithelial or (B and D) lamina propria suspensions from healthy control subjects (Contr.) vs treated (Treat.) or untreated (Untreat.) patients with celiac disease. Closed symbols are used for patient 1 (closed square; only A and B), who had multiple food allergies; patient 8 (closed triangle), who had been on a GFD for 12 years; patient 9 (closed circle), who accidentally ingested gluten 2 weeks before biopsy; and patient 10 (closed rectangle), who was on low-dose prednisone treatment for well-controlled systemic lupus erythematosus diagnosed 17 years before biopsy. All other subjects are marked with open symbols (circles). CD8 staining was not performed for patient 1 (missing in C and D), and the CCR9 level for lamina propria CD8+ cells is missing for patient 14 (in D). The median value for each subject group is marked with a horizontal bar. Kruskal–Wallis test was significant for all plots, and P values were obtained with the 2-tailed Mann–Whitney U test and Bonferroni correction for multiple testing of independent variables (P values multiplied by 3). Gastroenterology 2007 132, 2371-2382DOI: (10.1053/j.gastro.2007.04.023) Copyright © 2007 AGA Institute Terms and Conditions

Figure 3 (A and B) Flow-cytometric dot plots and (C–F) compiled flow-cytometric data displaying the combined distribution of CCR9 and CD45RO on (A, C, and E) CD4+ or (B, D, and F) CD8+ T cells from peripheral blood of healthy control subjects (Contr.) vs treated (Treat.) or untreated (Untreat.) patients with celiac disease. The flow-cytometric dot plots (A and B) are from patient 12, and percentages of cells in each quadrant are shown in the upper right corner of the plots. Cursors were set according to plots obtained with concentration- and isotype-matched irrelevant antibodies. (C–F) Scatter plots show the frequency of CCR9 expression among (C and D) CD45RO+ or (E and F) CD45RO− peripheral blood (C and E) CD4+ or (D and F) CD8+ T cells harvested immediately after duodenal biopsy had been performed. Blood was not drawn from patients 1 and 16, and CD4 data are missing for patient 15. Patient 19 had a CCR9 level of 1.9% in plot F (not shown). The median value for each subject group is marked with a horizontal bar. Kruskal–Wallis test was not significant for any plot, suggesting lack of differences among the groups. Gastroenterology 2007 132, 2371-2382DOI: (10.1053/j.gastro.2007.04.023) Copyright © 2007 AGA Institute Terms and Conditions

Figure 4 Flow-cytometric data showing the effect of stimulation with (A–G) PMA/ionomycin, (H, I, and O) anti-CD3/TCR, or (L, M, and P) anti-NKG2D on CCR9 expression of viable peripheral blood lymphocytes from healthy volunteers. In addition, coexpression of CCR9 and NKG2D on freshly isolated (J and N) CD4+ and (K and N) CD8+ peripheral blood T cells is shown. Percentages of cells in each quadrant and CCR9 MFI are indicated in the upper right corners and to the right (not in J and K), respectively, in the flow-cytometric dot plots. Gates were set for the lymphocyte region, and dead cells were excluded with propidium iodide staining except in J and K. Cursors were set from control plots of respective unstimulated and stimulated cells at each time point obtained with concentration- and isotype-matched irrelevant mAbs (not shown). (A, B, and D) Unstimulated and (C and E) PMA/ionomycin-stimulated cells were phenotyped and examined for CCR9 expression after (A) 0, (B and C) 4, and (D and E) 22 hours. Cells in the lower left corner of A–E were non-T cells that were reduced after stimulation (C and E). (F) Frequencies and (G) MFI values of CCR9+ T cells are shown for 3 similar experiments (triangles, squares, and circles). Values at 0 hours were set as 100%, and the increase or decrease normalized accordingly was calculated for stimulated (closed symbols) and unstimulated (open symbols) cells at each time point. Data collectively show that CCR9 expression on T cells was inversely related to cellular activation. PBMCs from a healthy volunteer were stimulated for 3 days with (H) irrelevant or (I) anti-CD3/TCR mAbs and subsequently phenotyped and examined for CCR9 expression. CD7 was used as a phenotypic marker because CD3 was unavailable after TCR stimulation. The mean ± SD from 3 similar experiments is shown for CCR9 MFI values from unstimulated or TCR-stimulated cells (O). NKG2D and CCR9 on peripheral blood T cells from a healthy volunteer showed inverse expression on CD4+ T cells (J) but striking coexpression on CD8+ T cells (K). The mean with 95% confidence interval from 3 similar experiments is shown for the number of NKG2D+ cells among CD4+CCR9+ and CD8+CCR9+ T cells (N). The CCR9 MFI value of CD7+CD8+ PBMCs from a healthy volunteer is shown after stimulation with irrelevant (Unstimulated; L) or anti-NKG2D (Stimulated; M) mAbs. Similar results were obtained for a total of 6 volunteers tested in 2 triplicate experiments performed at different time points; paired data for unstimulated and stimulated cells from each volunteer are linked with a line (P). Gastroenterology 2007 132, 2371-2382DOI: (10.1053/j.gastro.2007.04.023) Copyright © 2007 AGA Institute Terms and Conditions