Figure 1.
MLDS-induced islet inflammation causes lymphangiogenesis.
MLDS treated BALB/c mice received FTY720, sunitinib, or anti-VEGFR3 mAb starting from the first STZ injection. (A) Whole mount immunohistochemistry of isolated normal islets of BALB/c mice. Blood vessels: CD31; lymphatic vessels: LYVE-1. Scale bar: 200 pixels. 200× magnification. (B) Immunofluorescent analysis of beta-cells (insulin), T cells (CD3) and lymphatic vessels (LYVE-1) in pancreas 7 days or 13 days after initiation of MLDS treatment. Scale bars: 32 µm. (D) Immunofluorescent analysis of beta-cells (insulin), blood vessels (MECA32) and lymphatic vessels (LYVE-1) 7 days after initiation of MLDS treatment. Scale bars: 32 µm. (C) and (E) Quantitative analysis of insulin, CD3, LYVE-1 and MECA32 staining of pancreas 7 days after initiation of MLDS treatment. 12–15 islets for insulin and CD3, 12–15 areas around islets for LYVE-1, and 14–21 islets or areas around islets for MECA32; 2 slides/mouse; 2–4 mice/group. * P≤0.05, ** P≤0.01, *** P≤0.001; ns, not significantly. (F) and (G) Normal BALB/c mice received indicated treatment for 7 days. (F) Immunofluorescent analysis of beta-cells (insulin), lymphatic vessels (LYVE-1) and blood vessels (MECA32). 200× magnification. (G) Quantitative analysis of LYVE-1 and MECA32 staining of pancreas. 18–21 areas around islets; 2 slides/mouse; 2–4 mice/group. P>0.1 vs untreated mice. Mean ± SD. 200× magnification.
Figure 2.
Inhibition of lymphatic function alters draining LN size and prevents MLDS induced LN lymphangiogenesis.
MLDS treated BALB/c mice received FTY720, sunitinib, or anti-VEGFR3 mAb starting from the first STZ injection. Mice were euthanized 7 days after initial MLDS treatment. (A) Weight (n = 5 mice/group) and cell number (pooled 2 LNs) of draining pancreatic LNs. (B) Immunofluorescent analysis of T cells (CD3), lymphatic vessels (LYVE-1) and HEVs (PNAd) in pancreatic LNs 7 days after initiation of MLDS treatment. 100× magnification. (C) Quantitative analysis of LYVE-1 and PNAd staining in draining pancreatic LNs. 2 slides/mouse, 2–4 mice/group. Mean ± SD. * P≤0.01, **, P≤0.001.
Figure 3.
Inhibition of lymphatic function prevents diabetes by MLDS treatment.
(A) MLDS treated BALB/c mice received FTY720, sunitinib, or anti-VEGFR3 mAb starting from the first STZ injection for 2 weeks. Blood glucose profiles, compared to control of MLDS treated mice receiving the indicated treatments. P-values for all groups compared to control, <0.001. (B) Incidence of diabetes. MLDS treated mice received ALK1-Fc, control human IgG1 or PBS starting from the first STZ injection for 4 weeks. (C and D) MLDS treated BALB/c mice received anti-VEGFR2 mAb or control rat IgG1 starting from the first STZ injection for 2 weeks. (C) Blood glucose profiles, p-values all <0.05, anti-VEGFR2 compared to rat IgG1 or normal. (D) Incidence of diabetes, p = 0.0002.
Figure 4.
Expression of VEGFs and chemokines by pancreas and lymphatic endothelial cells.
(A) mRNA expression of chemokines in pancreas of MLDS treated BALB/c mice determined by qRT-PCR before treatment and on day 7. Values expressed as fold change. 3 mice/group, data representative of 2 separate experiments. (B) Flow sorting gates. LEC gated on FSC-A vs SSC-A, FSC-A vs FSC-W, CD45+, DAPI−, CD31+/− and LYVE-1+. Right upper panel, sorted LEC (LYVE-1+CD31+/−CD45−) cultured on Matrigel for 5 days. 100× magnification. (C) mRNA expression profile of chemokines and VEGFs in sorted pancreatic LEC. CD45−CD31+/−LYVE-1+ cells sorted from normal or MLDS treated CX3CR1GFP/+ (day 3 after initial treatment), 6–7 mice/group. Data representative of 2 separate experiments. PCR performed in triplicate. * P≤0.05, ** P≤0.01, *** P≤0.001. All data are represented as mean ± SD.
Figure 5.
Phenotype of pancreatic macrophages.
(A) Identification of macrophage subsets in pancreas of CX3CR1GFP/+ C56BL/6 mice. Pancreatic single cell suspensions were gated on FSC-A vs FSC-W and CD45+. Histograms show receptor expression profile of CX3CR1hiLYVE-1− (green line) and CX3CR1loLYVE-1+ (red line) macrophages. (B) Sorted CX3CR1loLYVE-1+ and CX3CR1hiLYVE-1− macrophages spun onto glass slides and stained with Wright's stain. 1000× magnification. (C) mRNA expression profile of chemokine receptors, chemokines and VEGFs in pancreatic macrophage subsets. mRNA levels examined by qRT-PCR in duplicate or triplicate. 6–7 mice/group, data representative of 2–4 separate experiments.
Figure 6.
Interaction of LEC and pancreatic macrophages.
(A) CX3CR1hi macrophages, LYVE-1+ macrophages, CD11b−-lymphocytes and LEC were sorted from pancreatic single cell suspensions of CX3CR1GFP/+ mice. Co-cultured CX3CR1hi macrophages-CFSE, LYVE-1+ macrophages-eFlour670, or lymphocytes-CFSE with/without LEC on Matrigel for 5 days. Scale bars: 60 µm. 100× magnification. (B) Immunofluorescent staining of CD68 and LYVE-1 in pancreas from C57BL/6 mice. Scale bars: 160 µm. 50× magnification. (C) Sorted CX3CR1hi macrophages labeled with CFSE and co-cultured with LEC (left upper panel) or without LEC (left lower panel) for 5 days, and stained for CD11b (red) and LYVE-1 (yellow). Scale bars: 10 µm. 630× magnification. Right panel, quantitative analysis, total cells from 5 fields (1344×1024 pixels) were counted. All data representative of 2 to 4 separate experiments.
Figure 7.
Macrophages infiltrate into inflamed islets.
MLDS treated BALB/c mice received sunitinib, or anti-VEGFR3 mAb starting from the first STZ injection 3 days. (A) Immunofluorescent analysis of CD68+LYVE-1+ (yellow arrows) and CD68+LYVE-1− (red arrows) macrophage subsets migrating near islets. 200× magnification. Scale bars: 30 µm. (B) Quantitative analysis of CD68+LYVE-1+ and CD68+LYVE-1− cells surrounding islets. Each symbol represents one islet. 43–52 islets/group, 4–5 slides/mice, 3 mice/group. *** P≤0.001. Mean ± SD.