Oxygen-Induced Retinopathy (OIR)

Summary: OIR model mimics retinopathy of prematurity (ROP) and human ischemic retinopathies.

Model Description

Oxygen-induced retinopathy model mimics retinopathy of prematurity (ROP) and human ischemic retinopathies. Two rodent models, mouse and rat, are commonly used to study the pathophysiology of ROP and to test the preclinical efficacy of drug candidates for ROP. However, mouse and rat oxygen-induced retinopathy (OIR) models differ in their mode of induction. In mice, chronic hyperoxia is induced versus alternating hyperoxia/hypoxia in rats. The manifestation of their vascular phenotypes also is different. In mice, vaso-obliteration occurs primarily in the central retina. In contrast, vaso-obliteration is more peripheral in the rat oxygen-induced retinopathy model and in human ROP (Kim, D’Amore and Connor, 2016).  

At Experimentica, we have fully implemented and validated both mouse and rat OIR models with a clinically relevant reference compound and the most comprehensive list of read-outs. More information about both mouse and rat OIR model can be shared on demand. 

Animal speciesMice, Rats
Method of inductionExposure to high oxygen levels during retinal vasculature development
Follow-up periodTypically 5-6 days after the animals are returned into normoxic condition
Route of compound administrationTopical (e.g. eye drops), intravitreal injection, systemic
Read-outsMorphological assessment of retinal vasculature using confocal microscopy:
– Avascular area/vascular obliteration,
– Area of neovascular tufts,
– Microglial activation.

Outcomes and Read-Outs 

In vivo imaging

Fluorescein angiography (rat; Heideleberg Spectralis HRA2). At later timepoints of OIR development, P18-P20, pre-retinal neovascular tufts and avascular areas can be imaged in the periphery (Fig. 1). 

SD-OCT (Envisu R2210 and Envisu R2210, Bioptigen Inc./Leica Microsystems) allows imaging of retinal layers and picks neovascularization in the inner retina. 

Fig. 1. High-resolution imaging using fluorescein angiography (FA).
Fig. 1. High-resolution imaging using fluorescein angiography (FA). FA images from normoxic control animals at P19 and from oxygen-induced retinopathy rats at P18 and P20 are shown. Image in the top row were acquired from the central retina, where vascular tortuosity (arrowheads) is readily observed in oxygen-induced retinopathy retinas. Images in the second row were taken from the periphery of the eye, highlighting the characteristic oxygen-induced retinopathy pathology. Arrows point to pre-retinal neovascular tufts and asterisk to avascular areas. During retinal vascular development, the hyaloid vessels start to regress, which can be captured by FA (bottom row). 

Functional assessment

Flash electroretinography (fERG). Functional deficits are observed both in a-wave and in b-wave amplitudes in the rat OIR model at P21. 

Fig. 2. Abnormal flash electroretinograms in the rat oxygen-induced retinopathy model.
Fig. 2. Abnormal flash electroretinograms in the rat oxygen-induced retinopathy model. A. a-waves are derived from retinal rod and cone photoreceptors. Oxygen-induced retinopathy rats exhibit decreased a-wave amplitudes compared to normoxic control animals at P21. This deficit is particularly evident at stronger light stimuli, suggestive of a defect in cone photoreceptor function. B. Oxygen-induced retinopathy rats also show decreased b-wave amplitudes compared to normoxic controls. Deficits in the b-wave are indicative of dysfunction in ON-bipolar cells and Müller cells. 


Avascular areas and areas of neovascularization are quantified from retinal wholemounts. Fig. 3 depicts the change in avascular area and neovascularization in the rat OIR model. 

Fig. 3. Quantification of vascular parameters in the rat oxygen-induced retinopathy model.
Fig. 3. Quantification of vascular parameters in the rat oxygen-induced retinopathy model. AVAs and areas of NV were quantified at three different time points during oxygen-induced retinopathy. No statistically significant differences were detected between different time points in oxygen-induced retinopathy retinas, but the mean of NV area was slightly higher at P20 compared to P18 and P21, which was concomitant with increased biological variation in the amount of NV. 


Authors' picture