Oxygen-Induced Retinopathy from Recurrent Intermittent Hypoxia Is Not Dependent on Resolution with Room Air or Oxygen, in Neonatal Rats

Abstract

Preterm infants often experience intermittent hypoxia (IH) with resolution in room air (RA) or hyperoxia (Hx) between events. Hypoxia is a major inducer of vascular endothelial growth factor, which plays a key role in normal and aberrant retinal angiogenesis. This study tested the hypothesis that neonatal IH which resolved with RA is less injurious to the immature retina than IH resolved by Hx between events. Newborn rats were exposed to: (1) Hx (50% O₂) with brief hypoxia (12% O₂); (2) RA with 12% O₂; (3) Hx with RA; (4) Hx only; or (5) RA only, from P0 to P14. Pups were examined at P14 or placed in RA until P21. Retinal vascular and astrocyte integrity; retinal layer thickness; ocular and systemic biomarkers of angiogenesis; and somatic growth were determined at P14 and P21. All IH paradigms resulted in significant retinal vascular defects, disturbances in retinal astrocyte template, retinal thickening, and photoreceptor damage concurrent with elevations in angiogenesis biomarkers. These data suggest that the susceptibility of the immature retina to changes in oxygen render no differences in the outcomes between RA or O₂ resolution. Interventions and initiatives to curtail O₂ variations should remain a high priority to prevent severe retinopathy.

Keywords: angiogenesis; astrocytes; insulin-like growth factor-I; intermittent hypoxia; oxygen-induced retinopathy; retina; vascular endothelial growth factor.

Figure 1

Representative retinal flatmounts demonstrating the effects of neonatal IH paradigms on retinal vascular integrity at P14 and P21. Retinas were stained with ADPase. Images are 10× magnification and the scale bar is 100 µm. White dotted line boxes indicate higher magnification (20×, scale bar 50 µm) images in the panels below.

Figure 2

Representative retinal flatmounts demonstrating the effects of neonatal IH paradigms on astrocyte and retinal vascular integrity at P21. Astrocytes and Müller cells were stained for glial fibrillary acidic protein (GFAP) immunoreactivity (green), and retinal vasculature was stained with isolectin B4, a biomarker for endothelial cells (red). Merged images are presented in the bottom panel. Images are 10× magnification and the scale bar is 100 µm. White dotted line boxes indicate higher magnification (20×, scale bar 50 µm) images in the panels below the GFAP and isolectin images.

Figure 3

Representative H&E images of retinas showing the effects of neonatal IH on retinal layer integrity and number of endothelial cells in the nerve fiber layer (NFL)/ganglion cell layer (GCL) at P14 (upper panel) and P21 (lower panel). The layers are identified in the RA panel at P14, which is presented as 20× magnification (scale bar is 50 µm) for identification of the retinal layers. All other images are 40× magnification (scale bar is 20 µm). ILM (inner limiting membrane); IPL (inner plexiform layer); INL (inner nuclear layer); ONL (outer nuclear layer); and R&C (rods and cones).

Figure 4

Vitreous fluid vascular endothelial growth factor (VEGF) (A), insulin-like growth factor (IGF)-I (B), and soluble VEGF receptor (sVEGFR)-1 (C) levels in response to variations in oxygen tension. Data are expressed as mean ± SEM (n = 6 samples/group). Groups are compared using ANOVA.

Figure 5

Effects of neonatal IH on retinal (A,B) and choroidal (C,D) VEGF levels at P14 (A,C) and P21 (B,D). Data are expressed as mean ± SEM (n = 6 samples/group). Groups are compared using ANOVA.

Figure 5

Effects of neonatal IH on retinal (A,B) and choroidal (C,D) VEGF levels at P14 (A,C) and P21 (B,D). Data are expressed as mean ± SEM (n = 6 samples/group). Groups are compared using ANOVA.

Figure 6

Effects of neonatal IH on retinal (A,B) and choroidal (C,D) sVEGFR-1 levels at P14 (A,C) and P21 (B,D). Data are expressed as mean ± SEM (n = 6 samples/group). Groups are compared using ANOVA.

Figure 7

Effects of neonatal IH on retinal (A,B) and choroidal (C,D) IGF-I levels at P14 (A,C) and P21 (B,D). Data are expressed as mean ± SEM (n = 6 samples/group). Groups are compared using ANOVA.

Figure 8

Graphic representation of the neonatal IH paradigms used in this study. Hyperoxia (50% O₂) following a brief, repetitive hypoxia (12% O₂) is presented in (A). Hyperoxia (50% O₂) following a brief, repetitive normoxia (21% O₂) is presented in (B). Normoxia (21% O₂ with brief episodes of hypoxia (12% O₂) is presented in (C).