17β-estradiol ameliorates oxygen-induced retinopathy in the early hyperoxic phase

Abstract

Retinopathy of prematurity (ROP) is a major and leading cause of blindness in premature infants. It has been realized that early treatment for ROP is important. However, all the early treatments of ROP are focusing on peripheral retinal ablation which does not surmount the limit of extinguishing retinal neovascularization and protecting the retinas of children with ROP from the injury of ablation. In this study, we investigated the morphological changes of retina and oxidative stress alterations in the early phase of oxygen-induced retinopathy (OIR) and tested the effects of 17β-estradiol (17β-E2), a nonselective estrogen receptor (ER) agonist, on early phase OIR development. We found that large central capillary-free areas were induced in the retinas of pups exposed to hyperoxia on postnatal day 9 (P9), whereas vascularization was almost complete in the retinas of pups exposed to normoxia at the same age. The concentrations of malondiadehyde (MDA), an end-product of oxidative stress, and nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, a major enzyme producing free radicals, as well as the activity of NADPH oxidase were significantly elevated in the retinas of pups exposed to hyperoxia on P9 and postnatal day 13 (P13) compared to those in age matched pups exposed to normoxia. Treatment with 17β-E2 decreased not only the percentage of the central capillary-free area to total retina area but also the concentrations of MDA and NADPH oxidase as well as the activity of NADPH oxidase in a dose-dependent manner in pups exposed to hyperoxia on p9 and P13. The concentration of VEGF was significantly decreased on P9 but increased on P14 in the retinas of pups exposed to hyperoxia, whereas it was significantly elevated on P9 but decreased on P14 in the retinas of pups treated with 17β-E2. The effect of 17β-E2 could be reversed by the co-treatment with ICI182780, a high affinity estrogen receptor antagonist, which suggested that 17β-E2 might exert its effect on early hyperoxic phase of OIR through estrogen receptor. Our results suggest that treatment with antioxidant drugs at early hyperoxic phase of ROP even before the appearance of retinal neovascularization may be more effective than their application to ROP at late phase, which may abolish the deleterious factors that contribute to retinal neovascularization and promote retinal blood vessels to develop healthily.

Keywords: 17β-estradiol; NADPH oxidase; Oxidative stress; Oxygen-induced retinopathy; Receptor.

Fig 1. Retinal flat mounts and the percentage of retinal avascular area on P9

A. Retinal vascularization was distributed to almost the whole retinas of normoxia control pups. The mature retinal vessels were found in the peripheral region while the immature retinal vessels were found in the central region as showed by the stained vessel wall. B. Large capillary-free areas were found in the retinas of hyperoxia alone exposure pups. C–E. The percentage of capillary-free area was decreased in the retinas of pups treated with hyperoxia plus different concentrations of 17β-E2. F. The capillary-free area was broadened in the retinas of pups treated with hyperoxia plus 10.0 μg of 17β-E2 and 100.0 μg of ICI182780. No neovascular tufts or clusters were found on the border of the avascular region and the vascular region in any of the retinal flat mounts. A′. Magnification of retinal flat mounts of normoxia control pups on P9. The blurred mature vessels in the whole retinal flat mounts of normoxia control pups on P9 obviously appeared under magnification (×4) . G. Statistical analysis of the difference in retinal avascular area represent with percentage of the percentage of the capillary-free area to total retina area between normoxia and hyperoxia groups as well as between hyperoxia alone exposure group and hyperoxia plus different concentrations of 17β-E2 with or without ICI182780 groups (n = 6 in each group) . *P<0.05, *** P<0.01. I is the abbreviation of ICI182780.

Fig 2. MDA concentration in the retina on P9 and P13

A and B. Retinal MDA concentration was significantly increased in hyperoxia alone exposure P9 (A) and P13 (B) pups compared to that in normoxia control pups, and was significantly decreased upon hyperoxia plus different concentrations of 17β-E2 treatment at P9 (A) and P13 (B) pups. No significant difference was found between hyperoxia alone exposure pups and hyperoxia plus 10.0 μg of 17β-E2 and 100.0 μg of ICI182780-treated pups. *P<0.05, ** or *** P<0.01. I is the abbreviation of ICI182780.

Fig 3. NADPH oxidase concentration and activity in the retina on P9 and P13

A and B. Retinal NADPH oxidase concentration was significantly increased in hyperoxia alone exposure P9 (A) and P13 (B) pups compared to that in normoxia control pups, and was significantly decreased upon hyperoxia plus different concentrations of 17β-E2 treatment at P9 (A) and P13 (B) pups. No significant difference was found between hyperoxia alone exposure pups and hyperoxia plus 10.0 μg of 17β-E2 and 100.0 μg of ICI182780-treated pups. *P<0.05, ** or *** P<0.01. I is the abbreviation of ICI182780. C and D. Retinal NADPH oxidase activity was significantly increased in hyperoxia alone exposure P9 (C) and P13 (D) pups compared to that in normoxia control pups, and was significantly decreased upon hyperoxia plus different concentrations of 17β-E2 treatment at P9 (C) and P13 (D) pups. No significant difference was found between hyperoxia alone exposure pups and hyperoxia plus 10.0 μg of 17β-E2 and 100.0 μg of ICI182780-treated pups. *P<0.05, ** or *** P<0.01. I is the abbreviation of ICI182780.

Fig 4. VEGF concentration on P9 and P14

A. Retinal VEGF concentration on P9 was significantly decreased in the hyperoxia alone exposure pups compared to that in the normoxia control pups, and was significantly increased upon hyperoxia plus 10.0 ug of 17β-E2 treatment compared to that in hyperoxia alone exposure pups. *P<0.05, *** P<0.01. I is the abbreviation of ICI182780. B. Retinal VEGF concentration on P14 was significantly increased in the hyperoxia alone exposure pups compared to that in the normoxia control pups, and was markedly decreased in the hyperoxia plus 1.0 and 10.0 ug of 17β-E2 treated pups compared to that in hyperoxia alone exposure pups. No significant difference was found between hyperoxia alone exposure pups and hyperoxia plus 10.0 μg of 17β-E2 and 100.0 μg of ICI182780-treated pups. *P<0.05, ** or *** P<0.01. I is the abbreviation of ICI182780.