Islet remodeling in female mice with spontaneous autoimmune and streptozotocin-induced diabetes

Abstract

Islet alpha- and delta-cells are spared autoimmune destruction directed at beta-cells in type 1 diabetes resulting in an apparent increase of non-beta endocrine cells in the islet core. We determined how islet remodeling in autoimmune diabetes compares to streptozotocin (STZ)-induced diabetes. Islet cell mass, proliferation, and immune cell infiltration in pancreas sections from diabetic NOD mice and mice with STZ-induced diabetes was assessed using quantitative image analysis. Serial sections were stained for various beta-cell markers and Ngn3, typically restricted to embryonic tissue, was only upregulated in diabetic NOD mouse islets. Serum levels of insulin, glucagon and GLP-1 were measured to compare hormone levels with respect to disease state. Total pancreatic alpha-cell mass did not change as autoimmune diabetes developed in NOD mice despite the proportion of islet area comprised of alpha- and delta-cells increased. By contrast, alpha- and delta-cell mass was increased in mice with STZ-induced diabetes. Serum levels of glucagon reflected these changes in alpha-cell mass: glucagon levels remained constant in NOD mice over time but increased significantly in STZ-induced diabetes. Increased serum GLP-1 levels were found in both models of diabetes, likely due to alpha-cell expression of prohormone convertase 1/3. Alpha- or delta-cell mass in STZ-diabetic mice did not normalize by replacement of insulin via osmotic mini-pumps or islet transplantation. Hence, the inflammatory milieu in NOD mouse islets may restrict alpha-cell expansion highlighting important differences between these two diabetes models and raising the possibility that increased alpha-cell mass might contribute to the hyperglycemia observed in the STZ model.

Figure 1. Changes in islet endocrine cell populations during progression of autoimmune diabetes.

Alpha, beta and delta-cells were quantified as a percentage of total islet area (A) and total endocrine cell mass (B) in 4-wk old diabetes-prone (black bars), 12-wk old insulitic (hatched bars) and 18–24 wk old diabetic (white bars) female NOD mice. Endocrine cell area (C) and mass (D) was quantified similarly in age-matched female Balb/c mice. Significant changes among groups are indicated as: *p<0.01, **p<0.001 and ***p<0.0001.

Figure 2. Islet endocrine cell changes during progression of autoimmune diabetes as leukocyte infiltration increases and dissipates.

Serial pancreas sections from 4(red) and somatostatin (green), or glucagon (red) and CD45-positive leukocytes (green). Scale bar = 10 µm.

Figure 3. Correlation of insulitis progression with alpha-cell remodeling in individual islets in diabetic NOD mice.

Composition of the insulitic lesion in 4-wk old diabetes-prone (islets from 5 mice), 12-wk old insulitic (islets from 7 mice) and 18–24 wk old diabetic (islets from 7 mice) female NOD mice (A, Insulitis score: white = 0 (pre-inflammation), lightest grey = 1 (less than 1/3 of the islet infiltrated), light grey = 2 (between 1/3 and 2/3 of the islet infiltrated), grey = 3 (more than 2/3 of the islet infiltrated), dark grey = 4 (islets with full insulitis) and black = 0 (post-inflammation)). Correlation between the severity of insulitis and degree of glucagon-positive alpha-cells in 4 wk old (diamond), 12 wk old (square) and 20 wk old diabetic NOD mice (triangle; B). Linear regression analysis revealed a negative correlation between the glucagon and insulitis score in the 12 wk old (dotted line R² = 0.79 and p<0.0005) and 20 wk old diabetic (solid line; R² = 0.85 and p<0.0001) groups. Note: islets with an insulitis score of zero may be either pre- or post-inflammation. Islets that were termed “post-inflammation” were easily identifiable; that is, the insulitis had dissipated following destruction of the beta-cells and largely been replaced by alpha and delta-cells.

Figure 4. Changes in islet endocrine cell populations in mice with STZ-induced diabetes.

Alpha, beta and delta-cells were quantified as a percentage of total islet area (A) and total endocrine cell mass (B) in 12-wk old Balb/c mice (black bars) and 12-wk Balb/c mice with STZ-induced diabetes (white bars) Significant changes among groups are indicated as: *p<0.01, **p<0.001 and ***p<0.0001. Increased proportions of alpha (red) and delta (green) cells coincides with decreased insulin expression (data not shown, C) in 12 wk old female Balb/c mice with STZ-induced diabetes. This was observed irrespective of whether the mice were left untreated (middle panel) or received a syngeneic islet transplantation to restore euglycemia (right panel). Age- and gender-matched Balb/c control islets demonstrated normal endocrine cell distribution with peripherally located alpha and delta-cells and beta-cells throughout the core of the islet (left panel). Scale bar = 10 µm.

Figure 5. Proliferation of endocrine cells and leukocytes in NOD and Balb/c mouse islets.

Co-immunostaining for the nuclear proliferation marker BrdU (green) and islet hormones (red) in 4, 12 and 20 wk old female NOD (A) and age-matched Balb/c mice (B). Sections are counterstained with DAPI to visualize nuclei (and infiltrating immune cells). White arrows indicate double-positive cells, and white arrowheads indicate single BrdU-positive cells within the islet mantel. Scale bar = 10 µm.

Figure 6. Quantification of proliferating endocrine cells in NOD and Balb/c mice with and without STZ-induced diabetes.

Number of proliferating alpha, beta and delta-cells were quantified in relation to total respective endocrine cell number in 4-wk old diabetes-prone (black bars), 12-wk old insulitic (hatched bars) and 18–24 wk old diabetic (white bars) female NOD mice (A). Number of proliferating endocrine cells was quantified similarly in age-matched female Balb/c mice (B), and in 12 wk old Balb/c mice without (black bars) or with (white bars) STZ-induced diabetes (C). Significant changes among groups are indicated as: *p<0.01, **p<0.001 and ***p<0.0001.

Figure 7. Identification of beta-cell markers in islets from diabetic NOD mice compared to STZ-induced diabetic Balb/c mice.

Representative pancreas sections from 4/c mice without or with STZ-induced diabetes were immunostained for glucagon (red) and Glut-2, PDX-1, Ngn3, PC-1/3 and GLP-1 (green, as indicated). Sections are counterstained with DAPI to visualize nuclei (and infiltrating immune cells). The insert on the glucagon and PC-1/3 panel indicates double-positive cells in an area where all glucagon cells are positive for PC-1. Scale bar = 10 µm.

Figure 8. Serum hormone levels reflect changes in endocrine cell mass.

Levels of glucagon, insulin and GLP-1 in serum from 4-wk old diabetes-prone (black bars), 12-wk old insulitic (hatched bars) and 18–24 wk old diabetic (white bars) female NOD mice (A), and in 12 wk old Balb/c mice without (black bars) or with (white bars) STZ-induced diabetes (B). The ratios of the proglucagon gene products in percent is given for both the NOD and Balb/c mice (C). Significant changes among groups are indicated as: *p<0.01, **p<0.001 and ***p<0.0001.