Fig 1.
Characterization of BM-derived cells in retinas of GFP+ BM chimeras at 6 months after transplantation.
GFP+ cells (green) colocalized with (A) microglial marker Iba-1 (red), (B) endothelial cell marker Collagen IV (red), (C) Müller cell marker Glutamine synthetase (red), (D) pericyte marker PDGFR-β (red) but not with (E) astrocyte marker GFAP (red). Colocalization of GFP+ cells with the respective markers was observed as yellow stain. Scale bars are 10 μm. (F) Percentages of GFP+ cells expressing markers for specific cell types, N = 4–5.
Fig 2.
Characterization of BM-derived cells in retinal sections of GFP+ BM chimeras at 6 months after transplantation.
Nuclei were counterstained with DAPI (blue). Overlays of confocal images of GFP+ cells (green) immunoreactive for (A) Iba1 (red) labeling microglia, (B) Collagen IV (red) labeling endothelial cells, (C) Glutamine synthetase (red) labeling Müller glia, (D) PDGF-Rβ (red) labeling pericytes. Colocalization of GFP+ cells with the respective markers was observed (yellow) in (A-D). (E) GFAP (red) labeling astrocytes, (F) Tyrosine hydroxylase (red) labeling amacrine cells, (G) Rhodopsin (red) labeling rod photoreceptors. No colocalization of GFP+ cells with these markers was observed in (E-G). Scale bars are 20 μm. N = 4–5.
Fig 3.
BM-derived cells in the retina of chimeric mice.
(A) Number of BM-derived cells per mm2 area of control or diabetic retina. Representative flow charts of GFP+ cells in the retina shown below. (B) ~ 93% of GFP+ cells detected in the retina are CD45- cells. Diabetes does not change the number of CD45- cells in the retina. Representative flow charts gated on GFP+ cells of CD45- and CD45+ cells shown below. N = 4.
Fig 4.
Diabetes alters BM-derived microglia in retina.
(A) Percentage of BM-derived microglia per mm2 area of control or diabetic retina. N = 7–8. (B) GFP+ microglia (gated as GFP+ cells that are Thy1-, Ly6G-, Ly6C-, CD45dim CD11b+ cells) are expressed as a percentage of total GFP+ retinal cells in control and diabetic retina of chimeric mice, and indicate increase in BM-derived microglia in diabetic mice. Representative flow charts of GFP+ microglia in the retina are shown. N = 3–5, ** p< 0.01. (C) Confocal images of retina isolated from control or diabetic GFP+ BM-transplanted mice. Microglial marker Iba-1+ staining (red) with GFP+ (green) cells, showing colocalization (yellow) in retina. Increased retraction of processes observed in BM-derived microglia in diabetic GFP+ chimeric mouse retina compared to ramified, resting phenotype of microglia in control retinas (white arrowheads). Scale bars are 50 μm. (D) Quantification of dendrite length of microglia in diabetic and control chimeric mouse retinas is shown. N = 4–5, *** p< 0.001.
Fig 5.
Effect of diabetes on BM-derived monocytes.
GFP+ monocytes (CD11b+ Ly6C+, CD11b+ Ly6Chi and CD11b+ Ly6Clo) are expressed as a percentage of total GFP+ cells in blood. N = 4–5, ** p< 0.01 and * p<0.05.
Fig 6.
Diabetes alters response of BM-derived cells and splenocytes to LPS stimulation.
Increased secretion of cytokines IL-1β and TNF-α in (A) BM-derived dendritic cell-enriched population (B) splenocytes stimulated with LPS. N = 4–5, * p< 0.05.
Fig 7.
Diabetes reduces number of BM-derived endothelial cells in retina.
GFP+ endothelial cells (gated as CD45- Tie-2+ CD31+ cells) are expressed as a percentage of total GFP+ retinal cells, and indicate a significant decrease in BM-derived endothelial cells in diabetic mice. Representative flow charts of GFP+ endothelial cells in the retina are shown. N = 4–5.
Fig 8.
Diabetes alters release of vascular reparative cells into circulation.
(A) CACs as a percentage of total bone marrow cells, *** p<0.001 (B) CACs as a percentage of total splenocytes. (C) CACs as a percentage of total blood cells. CACs are gated as Lin- CD34+ CD309+ cells. Representative flow charts of GFP+ CACs in control and diabetic BM, spleen and blood are shown below. * p<0.05, N = 4–8.
Fig 9.
Diabetes alters homing efficiency of CACs.
(A) Scheme for testing homing efficiency of diabetic CACs to the BM. Male C57BL/6-Tg(CAG-EGFP) mice were made diabetic by STZ injections. After 8 months of diabetes, 10,000 CACs were isolated from BM of diabetic and control GFP+ mice, and injected into the vitreous of healthy mice. Seven days post injection the BM and retinas were collected and analyzed by flow cytometry and confocal microscopy for presence of GFP+ CACs. (B) Control or diabetic GFP+ CACs (green) injected into the vitreous of mice with healthy retinal vasculature stained red using anti-collagen IV antibody. Quantitation of area of green CACs observed in confocal retinal images is shown. Scale bars are 50 μm, N = 10–12. (C) Quantification of GFP+ cells from bone marrow of tibia and femurs of recipient mice shows homing of diabetic GFP+ CACs to the recipient bone marrow was significantly lower compared to control GFP+ CACs. N = 4–6. (D) Diabetes significantly alters expression of integrins β2 and β3 on diabetic CACs in blood. N = 3, * p< 0.05, ** p<0.01.