The mouse model of oxygen-induced retinopathy (OIR) is commonly used to investigate various aspects of the pathogenesis of the retinopathy of prematurity (ROP) as well as angiogenesis in general. Retinal astrocytes were suggested to be involved in retinal angiogenesis. This study aimed to describe their localization and cell density during the course of physiological vascularization and pathological revascularization.
Mice expressing H2B-GFP (green fluorescent protein fused to histone 2B) from the endogenous Pdgfra promoter were kept in 75% oxygen from P7 (post natal day 7) to P12 (mouse model of OIR). Retinal flatmounts or cryosections were immunostained for glial fibrillary acidic protein (Gfap), glutamine synthetase (Glul), collagen IV (Col IV), desmin (Des), caspase 3 (Casp3), paired box 2 (Pax2), or Ki67. Astrocytic nuclei were counted with the ImageJ macro AuTOCellQuant. The hypoxic state of the retina was investigated by Hypoxyprobe.
The GFP signal of the Pdgfra reporter mice co-localized with Pax2, a nuclear marker for retinal astrocytes. This bright label was much easier to quantify than Gfap or Pax2 staining. Quantification of the cell density of astrocytes during physiological development specified the spreading of astrocytes in a concentrical wave from the optic nerve head towards the periphery. Astrocyte density was reduced during the remodelling of the primary vascular plexus into a hierarchical vascular tree (maximal astrocyte density at P1: 2800 astrocytes/mm2, final astrocyte density: 800 astrocytes/mm2). In the OIR model, cell density of astrocytes was elevated in the peripheral vascularized zone. In contrast, astrocyte density dropped to a half (400 astrocytes/mm2) of the normal value in the central avascular zone during the hyperoxic phase between P8 and P10 by apoptosis and rose only after P17 as the retinal network normalized. An additional drop of astrocyte density was observed within the angles between the large vessels of the central avascular zone during hypoxia between P12 and P17. Astrocyte density was not altered at vascular tufts.
The hyperoxia effect on astrocytes including the reduced astrocyte density is not the reason for vascular tuft formation. Hypoxia-affected astrocytes in combination with a reduced astrocytic network in the central avascular zone during the hypoxic phase are important determinants in the formation of pathological features during retinal revascularization.