Intraocular Pressure-Induced Endothelial Dysfunction of Retinal Blood Vessels Is Persistent, but Does Not Trigger Retinal Ganglion Cell Loss

Abstract

Research has been conducted into vascular abnormalities in the pathogenesis of glaucoma, but conclusions remain controversial. Our aim was to test the hypothesis that retinal endothelial dysfunction induced by elevated intraocular pressure (IOP) persists after IOP normalization, further triggering retinal ganglion cell (RGC) loss. High intraocular pressure (HP) was induced in mice by episcleral vein occlusion (EVO). Retinal vascular function was measured via video microscopy in vitro. The IOP, RGC and their axons survival, levels of oxidative stress and inflammation as well as vascular pericytes coverage, were determined. EVO caused HP for two weeks, which returned to baseline afterwards. Mice with HP exhibited endothelial dysfunction in retinal arterioles, reduced density of RGC and their axons, and loss of pericytes in retinal arterioles. Notably, these values were similar to those of mice with recovered IOP (RP). Levels of oxidative stress and inflammation were increased in HP mice but went back to normal in the RP mice. Our data demonstrate that HP induces persistent endothelial dysfunction in retinal arterioles, which persists one month after RP. Oxidative stress, inflammation, and loss of pericytes appear to be involved in triggering vascular functional deficits. Our data also suggest that retinal endothelial dysfunction does not affect RGC and their axon survival.

Keywords: glaucoma; oxidative stress; retinal arterioles; retinal ganglion cells; vascular endothelial dysfunction.

Figure 1

Follow up of IOP throughout the whole experiment. EVO increased IOP for two weeks. Control group refers to the sham group. Values are expressed as mean ± SEM (**** p < 0.0001; n = 8 per group).

Figure 2

Relative changes in the luminal dimeters of retinal arterioles in response to different intravascular pressures and vasoactive agents. The changes in response to stepwise increases in intraluminal pressure (A), to the thromboxane mimetic, U46619 (B), to the endothelium-dependent vasodilator, acetylcholine, (C) and to the endothelium-independent vasodilator, nitroprusside (D). EVO: episcleral vein occlusion; HP: high intraocular pressure; RP: recovered intraocular pressure. Control group refers to the sham group. Values are expressed as mean ± SEM (* p < 0.05, ** p < 0.01, Control vs. EVO (HP); ⁺⁺ p < 0.01, ⁺⁺⁺ p < 0.001, ⁺⁺⁺⁺ p < 0.0001, Control vs. EVO (RP); n = 8 per group).

Figure 3

DHE-stained retinal cryosections and intensity quantification for individual retinal layers. (A,D,G) show the DHE-stained retinal cryosections in the control, EVO (HP), and EVO (RP) groups, respectively. (B,C,E,F,H,I) show the staining intensity quantified from vessels, GCL, IPL, INL, OPL, and ONL, respectively. The white arrows point to retinal blood vessels. GCL: ganglion cell layer; IPL: inner plexiform layer; INL: inner nuclear layer; OPL: outer plexiform layer; ONL: outer nuclear layer. EVO: episcleral vein occlusion; HP: high intraocular pressure; RP: recovered intraocular pressure. Control group refers to sham group. Values are expressed as mean ± SEM (* p < 0.05; n = 5–6 per group, scale bar = 50 μm).

Figure 4

Immunostaing of anti-NOX2 and quantitative analysis of its fluorescence intensity in retinal cryosections. The anti-NOX2 stained retinal cryosections from Control group (A,B), EVO (HP) group (E,F) and EVO (RP) group (I,J). (C,D,G,H,K,L) show the staining intensity quantified from vessels, GCL, IPL, INL, OPL and ONL respectively. The white arrows point to retinal blood vessels. GCL: ganglion cell layer; IPL: inner plexiform layer; INL: inner nuclear layer; OPL: outer plexiform layer; ONL: outer nuclear layer. EVO: episcleral vein occlusion; HP: high intraocular pressure; RP: recovered intraocular pressure. Control group refers to sham group. Values are expressed as mean ± SEM (* p < 0.05; n = 6 per group, scale bar = 50 μm).

Figure 5

Messenger RNA expression of prooxidant gene (NOX2; A) and proinflammatory cytokine (TNF-α, B) in retina. EVO: episcleral vein occlusion; HP: high intraocular pressure; RP: recovered intraocular pressure. Control group refers to the sham group. Values are expressed as mean ± SEM (* p < 0.05, ** p < 0.01; n = 5–6 per group).

Figure 6

Immunostaining of anti-α-SMA and quantitative analysis of its fluorescence intensity in retinal flatmounts. The immunostained retinal flatmounts from the control group (A), the EVO (HP) group (B), and the EVO (RP) group (C). The staining intensity comparison among three groups (D). EVO: episcleral vein occlusion; HP: high intraocular pressure; RP: recovered intraocular pressure. Control group refers to the sham group. Values are expressed as mean ± SEM (* p < 0.05; n = 6 per group, scale bar = 50 μm).

Figure 7

Immunostaining and RGC density in the retinal flatmounts. (A,D,G,J) show the RGC density in the center of the retina. (B,E,H,K) exhibit the RGC density in the middle of the retina. (C,F,I,L) show the RGC density in the periphery of the retina. (M) represents the mean RGC density in the whole retina, averaged from the three different parts. Control group refers to the sham group. Values are expressed as mean ± SEM. (* p < 0.05, ** p < 0.01; n = 5–6 per group, scale bar = 50 μm).

Figure 8

Scoring of PPD-stained optic nerve cross-sections. The normal stained axons in the control mice (A). The collapsed and more darkly stained abnormal axons in glaucoma mice from both the EVO (HP) group (B) and the EVO (RP) group (C). Arrows point to the abnormal axons. Higher scores were assessed both in glaucoma mice with HP and those with RP (D). EVO: episcleral vein occlusion; HP: high intraocular pressure; RP: recovered intraocular pressure. Control group refers to the sham group. Values are expressed as mean ± SEM. (*** p < 0.001; n = 5–6 per group, scale bar = 50 μm).