Fig 1.
Comparison of MRI and OCT regions-of-interest (ROI’s).
Four non-overlapping retinal regions were evaluated from MRI data A): ± 400–1000 μm and ± 1000–2000 μm from the optic nerve head for their transretinal 1/T1 profiles (shown in Figs 3 and 6). Two regions in central retina were studied from OCT data B): ± 350–630 μm from the optic nerve head to determine outer nuclear layer (ONL) thickness, and the external limiting membrane–retinal pigment epithelium (ELM-RPE) thickness.
Fig 2.
Timing diagrams for the three groups studied by OCT.
A) Sildenafil is administered after 4 hr of room Light and studied 1 hr later in the Light (SLL). B) Sildenafil is given in the Dark then Light-adapted and studied 5 hr later (SDL). C) Sildenafil is given in the Dark and maintained as Dark-adapted during OCT examination (SDD). MB, methelyene blue; ALA; α-lipoic acid. Control groups were given saline.
Table 1.
Model details for OCT layer thickness analysis.
Fig 3.
QUEST MRI showing oxidative stress localized to peripheral superior retina.
Modeled 1/T1 profiles approximately 1 hour post-sildenafil IP in light-adapted mice given either saline (n = 4 mice, black line) or anti-oxidants (AO, n = 5 mice, red line) in these four retinal regions: A) 1000 to 2000 inferior, B) 1000 to 2000 superior, C) 400–1000 inferior, and D) 400–1000 superior; representative OCT images (bottom of each graph) provide spatial orientation. Only significant (horizontal black bar) AO reductions in 1/T1 indicative of oxidative stress are shown. Each profile has a solid line indicating the mean and a shaded region indicating 95% confidence intervals.
Fig 4.
The effect of sildenafil on generation of oxidative stress in a representative ex vivo data set of unfixed cryosections stained with DCF.
DCF staining for reactive oxygen species is shown in green; nuclei staining (DAPI) is shown in blue. The same intensity scaling is used in all images; scale bars are 100 μm.
Fig 5.
1 hr post-sildenafil after 4 hr of light-adaption mice (SLL, Fig 2) shows thinner, dark-like ELM-RPE.
Modeled A) ONL inferior retina, B) ONL superior retina, C) ELM-RPE thickness inferior retina, and D) ELM-RPE thickness superior retina in uninjected control dark (CD, n = 11 mice, grey bar), control light (CL, n = 5 mice, white bar), SLL+saline (SLL, n = 5 mice, green bar), and SLL+AO (n = 5 mice, red bar) in the two different retinal regions. ONL is invariant to condition. ELM-RPE is significantly (horizontal black bar) thinner in the dark than in the light (CD vs. CL) as expected [44, 53, 54]. In SLL+saline and SLL+AO groups ELM-RPE thickness is not different from CD (and was thinner than CL); no evidence for oxidative stress was found. The points in each plot represent the estimated mean for each mouse based on the model. Error bars indicate 95% confidence intervals. Note the same control data sets are presented in each graph to facilitate comparisons.
Fig 6.
5 hr post-sildenafil in light-adaption mice (SDL, Fig 2) shows light-like ELM-RPE.
Modeled A) ONL inferior retina, B) ONL superior retina, C) ELM-RPE thickness inferior retina, and D) ELM-RPE thickness superior retina in uninjected control dark (CD, n = 11 mice, grey bar), control light (CL, n = 5 mice, white bar), SDL+saline (SDL, n = 6 mice, green bar), and SDL+AO (n = 6 mice, red bar) in the two different retinal regions. ONL is unresponsive to condition. In SDL+saline and SDL+AO groups ELM-RPE thickness is not different from CL (and was thicker than CD); no evidence for oxidative stress was found. The points in each plot represent the estimated mean for each mouse based on the model. Error bars indicate 95% confidence intervals. Note the same control bars are presented in each graph to facilitate comparisons.
Fig 7.
QUEST MRI in dark-adapted mice do not show outer oxidative stress.
Modelled 1/T1 profiles approximately 1 hour post-sildenafil IP in dark-adapted mice given either saline (n = 6 mice, black line) or anti-oxidants (AO, n = 3 mice, red line) in these four retinal regions: A) 1000 to 2000 inferior, B) 1000 to 2000 superior, C) 400–1000 inferior, and D) 400–1000 superior; representative OCT images (bottom of each graph) provide spatial orientation. Only significant (horizontal black bar) AO reduction in 1/T1 indicative of oxidative stress are shown. Each profile has a solid line indicating the mean and a shaded region indicating 95% confidence intervals.
Fig 8.
1 hr post-sildenafil in dark-adapted mice (SDD, Fig 2) shows thinner ELM-RPE.
Modeled A) ONL inferior retina, B) ONL superior retina, C) ELM-RPE thickness inferior retina, and D) ELM-RPE thickness superior retina in uninjected control dark (CD, n = 11 mice, grey bar), control light (CL, n = 5 mice, white bar), SDD+saline (SDD, n = 3 mice, green bar), and SDD+AO (n = 3 mice, red bar) in the two different retinal regions. ONL is constant regardless of condition. In SDD+saline and SDD+AO groups, ELM-RPE thickness is not different from CD (and was thinner than CL); no evidence for oxidative stress was found. The points in each plot represent the estimated mean for each mouse based on the model. Error bars indicate 95% confidence intervals. Note the same control bars are presented in each graph to facilitate comparisons.
Fig 9.
Contrast sensitivity is transiently reduced at 5 hr post-sildenafil.
Compared to uninjected 5 hr light adapted mice (white bar), sildenafil + saline (green bars) A) decreased contrast sensitivity only under SDL conditions but B) did not change spatial frequency threshold (SFT). AO (red bars) did not correct this reduced contrast sensitivity. Black horizontal bars indicate significant differences (P < 0.05). The points in each plot represent the estimated mean for each mouse based on the model. Error bars are 95% confidence intervals. Number of mice used per group: uninjected, n = 6 mice; SLL+saline, n = 5 mice; SLL+AO, n = 6 mice; SDL+saline, n = 5 mice; SDL+AO, n = 5 mice; 24 hour+saline, n = 6 mice; 24 hour+AO, n = 6 mice.
Fig 10.
Working model of factors that regulate the ELM-RPE thickness.
Oxidative stress is hypothesized to induce an acidosis (dotted red line) that would be expected to convert a thicker “light” ELM-RPE phenotype into a thinner “dark-like” phenotype (see text for details); this expectation is supported experimentally [24, 44, 63].