Islet cell dedifferentiation is a pathologic mechanism of long-standing progression of type 2 diabetes

Abstract

Dedifferentiation has been implicated in β cell dysfunction and loss in rodent diabetes. However, the pathophysiological significance in humans remains unclear. To elucidate this, we analyzed surgically resected pancreatic tissues of 26 Japanese subjects with diabetes and 11 nondiabetic subjects, who had been overweight during adulthood but had no family history of diabetes. The diabetic subjects were subclassified into 3 disease stage categories, early, advanced, and intermediate. Despite no numerical changes in endocrine cells immunoreactive for chromogranin A (ChgA), diabetic islets showed profound β cell loss, with an increase in α cells without an increase in insulin and glucagon double-positive cells. The proportion of dedifferentiated cells that retain ChgA immunoreactivity without 4 major islet hormones was strikingly increased in diabetic islets and rose substantially during disease progression. The increased dedifferentiated cell ratio was inversely correlated with declining C-peptide index. Moreover, a subset of islet cells converted into exocrine-like cells during disease progression. These results indicate that islet remodeling with dedifferentiation is the underlying cause of β cell failure during the course of diabetes progression in humans.

Keywords: Beta cells; Diabetes; Endocrinology; Insulin; Metabolism.

Figure 1. Morphologic changes in individual islets and correlation with β cell function.

Representative images of pancreatic islets immunostained with chromogranin A (ChgA) (shown in red) and (A) insulin (shown in green) and (B) glucagon (shown in green) in each group. Scale bar: 20 μm. Quantitative analyses of (C) ChgA-positive cells per islet, (D) β cells per islet, (E) α cells per islet, and (F) the α cell/β cell ratio per islet. Data are means ± SD (n = 11 for non-DM, n = 12 for early-DM, n = 11 for advanced-DM, and n = 3 for intermediate-DM). ***P < 0.001 by 1-way ANOVA followed by Bonferroni’s post hoc test. (G) Quantitative analyses of insulin and glucagon double-positive cells in the pancreatic sections of specimens from the indicated subjects in each group. Data are means ± SD (n = 6 for non-DM, n = 7 for early-DM, and n = 5 for advanced-DM). *P < 0.05, **P < 0.01 by 1-way ANOVA followed by Bonferroni’s post hoc test. We performed single regression analysis (Spearman’s correlation coefficient) to assess relationships between the C-peptide index obtained from 18 subjects (n = 5 for non-DM, n = 4 for early-DM, and n = 9 for advanced-DM) and (H) β cells/islet, (I) α cells/islet, and (J) α cell/β cell ratio per islet. Open circles, non-DM control subjects. Closed triangles, early-DM subjects. Closed squares, advanced-DM subjects. CPI, C-peptide index.

Figure 2. Evaluation of dedifferentiation and correlations with islet morphology and clinical parameters.

(A) Representative images of pancreatic islets immunostained with ChgA (shown in red) and endocrine cocktail (insulin, glucagon, somatostatin, and pancreatic polypeptide [4H], shown in green). Scale bar: 20 μm. (B) Quantitative analysis of dedifferentiated cells (dedifferentiation score), calculated as ChgA⁺4H^–/ChgA⁺ cells per islet. Data are means ± SD. **P < 0.01, ***P < 0.001 by 1-way ANOVA followed by Bonferroni’s post hoc test (n = 11 for non-DM, n = 12 for early-DM, n = 3 for intermediate-DM, and n = 11 for advanced-DM). Single regression analysis (Spearman’s correlation coefficient) of correlations between dedifferentiation score and (C) β cells/islet, (D) α cells/islet, and (E) the C-peptide index obtained from 21 subjects (n = 5 for non-DM, n = 4 for early-DM, n = 3 for intermediate-DM, and n = 9 for advanced-DM), and (F) diabetes duration. (G) Comparison of dedifferentiation score between the subjects receiving insulin treatment (n = 6) and those treated with sulfonylurea but not insulin (n = 7). Open circles, control subjects. Closed triangles, early-DM subjects. Open squares, intermediate-DM subjects. Closed squares, advanced-DM subjects. CPI, C-peptide index. SU, sulfonylurea.

Figure 3. Analysis of correlations of dedifferentiation with islet morphology and age in early-DM subjects.

(A) Single regression analysis (Spearman’s correlation coefficient) of correlations between islet morphology parameters and islet area with dedifferentiation scores in early-DM subjects (n = 12). (B) Single regression analysis (Spearman’s correlation coefficient) of correlations between age and islet plasticity parameters in non-DM control (n = 11) and early-DM subjects (n = 12). Open circles, non-DM control subjects. Closed triangles, early-DM subjects.

Figure 4. Immunohistochemical evidence of conversion from endocrine to exocrine cell phenotype in failing islets.

Islet area–matched male subjects without pancreatic cancer were selected from each group. The 9 subjects were classified into 3 comparison sets according to the fraction of islet area as shown in Supplemental Table 3. Their pancreatic sections were examined. Representative images of pancreatic islets immunostained with ChgA (green) and amylase (red) of 9 subjects in 3 comparison sets representing different islet areas are shown (A–C, D–F, and G–I). Insets demonstrate representative cells showing immunoreactivity for ChgA and amylase. Scale bar: 20 μm. (J) Z-stack of pancreatic islets immunostained with ChgA (green) and amylase (red) of advanced-DM. Multiple Z-plane fluorescent images of pancreatic sections of the subjects in C, F, and I were captured. The representative Z-stack images are shown in the same order as in C, F, and I. The representative islet cells expressing both amylase (red) and ChgA (green) are shown in the lower images with a high magnification. Scale bar: 20 μm.