Increased immune cell infiltration of the exocrine pancreas: a possible contribution to the pathogenesis of type 1 diabetes

Abstract

Type 1 diabetes (T1D) results from a complex interplay between genetic susceptibility and environmental factors that have been implicated in the pathogenesis of disease both as triggers and potentiators of β-cell destruction. CD8 T cells are the main cell type found in human islets, and they have been shown in vitro to be capable of killing β-cells overexpressing MHC class I. In this study, we report that CD8 T cells infiltrate the exocrine pancreas of diabetic subjects in high numbers and not only endocrine areas. T1D subjects present significantly higher CD8 T cell density in the exocrine tissue without the presence of prominent insulitis. Even T1D donors without remaining insulin-containing islets and long disease duration show elevated levels of CD8 T cells in the exocrine compartment. In addition, higher numbers of CD4(+) and CD11c(+) cells were found in the exocrine tissue. Preliminary data in type 2 diabetic (T2D) subjects indicate that overall, there might be a spontaneous inflammatory infiltration of the exocrine tissue, common to both T1D and T2D subjects. Our study provides the first information on the precise tissue distribution of CD8 T cells in pancreata from T1D, T2D, autoantibody-positive, and healthy control subjects.

Figure 1

Quantification of CD8⁺ T cells in pancreatic tissue sections. Representative quantification of CD8 T cells in pancreata is shown (see Research Design and Methods for details). A–E: Frozen pancreas sections from a T1D donor with <5 years of disease duration were stained for glucagon and CD8. Images were acquired using a BIOREVO BZ-9000 slide scanner system (Keyence, Osaka, Japan). Representative images of exocrine and endocrine pancreatic CD8 T cell quantification in healthy (F), Ab⁺ (G), T1D (H), and T2D donors (I). Numbers represent CD8 T cell quantification in the exocrine pancreas (in red), in the periphery of the islet (in blue), and in the endocrine tissue (in yellow). The contour of the islets is defined by a dashed line. Scale bars, 500 μm in A–C and 100 μm in D–I.

Figure 2

Diabetic donors present high exocrine infiltration. A, B, and D: Mean exocrine density values (cells/mm²) for each donor and group are shown for CD4⁺, CD11c⁺, and CD8⁺. (CD4: n = 10 for control subjects, n = 8 Ab⁺, n = 6 T2D, n = 8 T1D <5 years, and n = 11 T1D >5 years. CD11c: n = 10 control subjects, n = 9 Ab⁺, n = 8 T2D, n = 9 T1D <5 years, and n = 10 T1D >5 years. CD8: n = 15 control subjects, n = 14 Ab⁺, n = 11 T2D, n = 12 T1D <5 years, and n = 19 T1D >5 years.) Open symbols indicate donors in which the nPOD Pathology Core found signs of insulitis. C: Linear regression analysis of CD4 and CD11c density (n = 42). Outliers are shown as open circles. The corresponding R² and P values are indicated in the graph. E: Mean density value (cells/mm²) for CD8⁺, CD4⁺, and CD11c⁺ cells for the exocrine pancreas is shown. For statistical analysis, nonparametric Kruskal-Wallis test followed by Dunn multiple comparisons test was used to determine significance. Linear regression analysis was used in C. Outliers were identified using the Rout method (FDR of 1% [Q = 1]). Two and one outliers from the control group were identified in A and D, respectively; in B, one control, one T2D, and two T1D donors were identified as outliers. *Significant difference between groups (P ≤ 0.05); **significant difference between groups (P ≤ 0.01); ***significant difference between groups (P ≤ 0.001); ****significant difference between groups (P ≤ 0.0001).

Figure 3

Distribution of CD8 infiltration in the pancreas of healthy, Ab⁺, T1D, and T2D donors. Mean CD8 density values (cells/mm²) for each donor and group are shown in the islets in A and D, in the periphery in B, in the exocrine pancreas in E, and as total CD8 T cell density in C. Open symbols indicate donors in which the nPOD Pathology Core found signs of insulitis. A–C: n = 15 control subjects, n = 14 Ab⁺, n = 11 T2D, n = 12 T1D <5 years, and n = 19 T1D >5 years. D and E: T1D Ins⁻, T1D donors with only IDIs; T1D Ins⁺, T1D donors with ICIs (n = 15 control subjects, n = 14 T1D Ins⁺, and n = 17 T1D Ins⁻). F: Mean CD8 density values (cells/mm²) for each donor are shown for donors with pancreatitis (n = 21) and donors without pancreatitis (n = 50). For statistical analysis, one-way ANOVA with Holm-Sidak multiple comparisons test and Kruskal-Wallis followed by Dunn multiple comparisons test were used to determine significance. A Student t test was applied in F. Mean ± SD is shown in A–F. Outliers were identified using the Rout method (FDR of 1% [Q = 1]). In A and B, one outlier T2D and two outliers T1D >5 years were identified; in D, one outlier control was identified. *Significant difference between groups (P ≤ 0.05); **significant difference between groups (P ≤ 0.01); ***significant difference between groups (P ≤ 0.001). w, with; wo, without.

Figure 4

T1D donors present lower islet area and pancreatic weight compared with healthy control subjects. A: Bars represent mean relative islet area (%) for each group based on insulin and glucagon staining. Differences among the groups were not statistically significant, but a strong tendency to lower area in T1D and T2D donors was seen. B: Bars represent mean pancreatic weight for each group based on information provided by nPOD (n = 14 control subjects, n = 14 Ab⁺, n = 11 T2D, n = 11 T1D <5 years, and n = 19 T1D >5 years). For statistical analysis, one-way ANOVA with Holm-Sidak multiple comparisons test and Kruskal-Wallis followed by Dunn multiple comparisons test were used to determine significance in A and B, respectively. Mean ± SEM values are shown. **Significant difference between groups (P ≤ 0.01); ***significant difference between groups (P ≤ 0.001).

Figure 5

CD8 T cells are similarly distributed in the pancreas of diabetic donors and control subjects. A: Bars represent mean CD8 T cell density (cells/mm²) for each region (head, body, and tail) and group (n = 23 control subjects, n = 32 Ab⁺, n = 21 T2D, and n = 33 T1D). Mean ± SEM values are shown. B: Population mean and 95% CIs of interimage variation (CV) in total CD8 T cell density are shown for each donor and group (n = 15 control subjects, n = 14 Ab⁺, n = 11 T2D, and n = 31 T1D). For statistical analysis, one-way ANOVA with Holm-Sidak multiple comparisons test in B and Kruskal-Wallis followed by Dunn multiple comparisons test in A were used to determine significance.

Figure 6

Correlation between CD8 T cell density and clinical parameters. Correlation analysis of CD8 T cell density and clinical parameters, corresponding P and r values are shown for control subjects (black), Ab⁺ (red), T1D (blue), and T2D (green). A: CD8 T cell density correlates with age (r = 0.2850; P = 0.0160; n = 71). B: CD8 T cell density does not correlate with BMI (r = 0.1897; P = 0.1271; n = 66). C: There was no correlation between CD8 T cell density and time spent in ICU (r = 0.2186; P = 0.0778; n = 66). CD8 T cell density does not correlate with relative islet area (r = −0.1978; P = 0.0982; n = 71) (D) or pancreas weight (r = −0.07899; P = 0.5188; n = 69) (E). Linear regression analysis was used to determine significance. *Significant correlation (P ≤ 0.05).