Type 1 diabetes predisposes to enhanced gingival leukocyte margination and macromolecule extravasation in vivo

Abstract

Background and objective: Diabetes predisposes to periodontal disease. However, the cellular and molecular mechanisms linking the two conditions are not clear. The impact of chronic hyperglycemia on leukocyte margination and macromolecule extravasation was determined in gingival vessels in vivo.

Materials and methods: Gingival intravital microscopy was employed to measure extravasation of fluorescein isothiocyanate (FITC)-dextran in diabetic Akita and healthy wild-type (WT) mice. Rhodamine 6G and FITC-LY6G were injected for nonspecific and polymorphonuclear-specific leukocyte labeling, respectively. Surface expression of leukocyte adhesion molecules was determined with flow cytometry and western blotting.

Results: Vascular permeability was significantly increased in Akita gingival vessels compared with WT [permeability index (PI): WT, 0.75 ± 0.05; Akita, 1.1 ± 0.03: p < 0.05). Wild-type gingival vessels reached comparable permeability 2 h after intragingival injection of tumor necrosis factor α (TNFα), used here as positive control (PI, 1.17 ± 0.16). The number of rolling leukocytes was significantly elevated in diabetic gingiva (WT, 25 ± 3.7 cells/min; Akita, 42 ± 8.5 cells/min; p < 0.03). Similar rolling cell counts were obtained in WT after intragingival injection of TNFα (10 ng TNFα, 47 ± 1.3 cells/min; 100 ng TNFα, 57.5 ± 5.85 cells/min). The number of leukocytes firmly attached to the endothelium was similar in WT and Akita mice. Leukocyte cell-surface expression of P-selectin glycoprotein ligand-1 and CD11a was increased in Akita mice, while L-selectin remained unchanged when compared with WT. Moreover, P-selectin expression in Akita gingival tissues was elevated compared with that of WT.

Conclusion: Chronic hyperglycemia induces a proinflammatory state in the gingival microcirculation characterized by increased vascular permeability, and leukocyte and endothelial cell activation. Leukocyte-induced microvascular damage, in turn, may contribute to periodontal tissue damage in diabetes.

Fig. 1

In vivo assessment of gingival vascular permeability and leukocyte margination. Rhodamine 6G (R6G; leukocyte labeling) and FITC– dextran (vasculature visualization and tracer molecule) or FITC–LY6G (PMN labeling) were injected intravenously, and intravital microscopy was performed at a controlled body temperature as specified in the Material and methods section. The black arrows indicate the outer side of the vessel wall. (A) The FITC–dextran is retained in the gingival vasculature of a WT mouse. (B) Increased FITC–dextran extravasation in Akita gingiva within the same time period. (C) Rhodamine 6G-positive leukocytes inside gingival postcapillary venules. The FITC–dextran (green) and R6G (red) images were merged offline to obtain this picture. (D) Overlapping images of R6G-positive (red) and LY6G-positive (green) cells indicate that the majority of leukocytes 2 h after tumor necrosis factor α (TNFα) stimulation are polymorphonuclear leukocytes (PMNs; yellow). The white arrows indicate intravascular leukocytes as follows: 1,rolling leukocyte (red); 2, rolling PMN (green); and 3, attached PMN (yellow). Scale bar represents 50 μm; original magnification ×400.

Fig. 2

Microvascular permeability in the gingiva. Wild-type (WT) and Akita mice were injected with a 50 μL intravenous bolus of FITC–dextran (molecular weight, 150 kDa; 6 mg/kg), and 30 min later pictures were taken under the FITC filter. Tumor necrosis factor α (TNFα; 10 ng) was administered to some WT mice 2 h before the experiment. The permeability index represents the ratio of extravascular vs. intravascular green fluorescence intensity as specified in the Material and methods section. Permeability index values were normalized to WT control values. Results are means + SEM; *p < 0.05; n = 3 in each group.

Fig. 3

Leukocyte rolling in gingival venules. Leukocytes were labeled with an intravenous injection of R6G. A 100 μm segment of a superficial unbranched venule was selected, and intravital videos were recorded using a rhodamine filter. The number of R6G-labeled cells rolling along the endothelium was counted during the 30 s observation period. Akita gingival vessels displayed a significantly higher number of rolling leukocytes than WT. Stimulation with 10 ng TNFα in WT gingiva 2 h before the experiment resulted in a comparable elevation in leukocyte rolling. Results are means + SEM; *p < 0.03; n = 4 in each group.

Fig. 4

Leukocyte adhesion molecule expression in chronic hyperglycemia. (A) Flow cytometric analysis of peritoneal leukocytes showed increased surface expression of PSGL-1 (n = 6). (B,C) L-Selectin (n = 7) and CD11a (n = 4) expression ws not significantly different between WT and Akita leukocytes. Black line, WT; gray line, Akita: and black line with gray shaded area, IgG control. Representative histograms are shown on the left. (D) Densitometric analysis of P-selectin gingival expression assessed by western blot in Akita mice relative to WT. Data for P-selectin were normalized to total protein and expressed as a percentage of WT values (n = 5; *p < 0.05). FL2, orange fluorescence signal channel.

Fig. 5

Leukocyte and vascular changes in chronic hyperglycemia. In physiological conditions, circulating leukocytes are constantly passing through narrow capillaries with smaller diameters than their own, resulting in direct contact with endothelial cells. This results in temporary rolling along the inner vessel wall via P-selectin–PSGL-1 interactions. In the postcapillary segment, they detach and continue circulating (left side, under ‘normal’). High blood glucose levels and the resulting advanced glycation end-products (AGEs) lead to both priming of leukocytes and endothelial cell changes which favor longer interaction between the two. We showed here that P-selectin on endothelial cells and its ligand PSGL-1 on leukocytes are overexpressed, which may in part explain the observed high rolling rate in vivo. Prolonged leukocyte–endothelial cell interaction may also promote release of reactive oxygen species (ROS) by polymorphonuclear leukocytes (PMNs) directly against the vessel wall by favoring both proximity and cell activation. Furthermore, changes in endothelial cell phenotype and the basement membrane structure may explain the increased permeability for large molecules (right side, under ‘diabetes’).