In Vivo Imaging of Retinal Hypoxia in a Model of Oxygen-Induced Retinopathy

Abstract

Ischemia-induced hypoxia elicits retinal neovascularization and is a major component of several blinding retinopathies such as retinopathy of prematurity (ROP), diabetic retinopathy (DR) and retinal vein occlusion (RVO). Currently, noninvasive imaging techniques capable of detecting and monitoring retinal hypoxia in living systems do not exist. Such techniques would greatly clarify the role of hypoxia in experimental and human retinal neovascular pathogenesis. In this study, we developed and characterized HYPOX-4, a fluorescence-imaging probe capable of detecting retinal-hypoxia in living animals. HYPOX-4 dependent in vivo and ex vivo imaging of hypoxia was tested in a mouse model of oxygen-induced retinopathy (OIR). Predicted patterns of retinal hypoxia were imaged by HYPOX-4 dependent fluorescence activity in this animal model. In retinal cells and mouse retinal tissue, pimonidazole-adduct immunostaining confirmed the hypoxia selectivity of HYPOX-4. HYPOX-4 had no effect on retinal cell proliferation as indicated by BrdU assay and exhibited no acute toxicity in retinal tissue as indicated by TUNEL assay and electroretinography (ERG) analysis. Therefore, HYPOX-4 could potentially serve as the basis for in vivo fluorescence-based hypoxia-imaging techniques, providing a tool for investigators to understand the pathogenesis of ischemic retinopathies and for physicians to address unmet clinical needs.

Figure 1. Sensitivity and hypoxia-specificity of HYPOX-4 in retinal cells.

(A) R28 cells were treated with HYPOX-4 (100 μM) and variable oxygen concentrations. HYPOX-4 dependent fluorescence increased with decreasing oxygen concentration. (B) R28, (C) ARPE19 and (D) MIO-M1 cells were treated with concentrations of HYPOX-4 ranging from 10 to 100 μM and 0.1% oxygen concentration; a HYPOX-4 dose-dependent fluorescence was observed in all of these cell types. (E,F) R28 cells were treated with HYPOX-4 (100 μM) and 0.1% oxygen for 4 hours. Hypoxia-specific fluorescence cell imaging was achieved. (G,H) Minimal fluorescence was observed in normoxic cells (n = 8, *p < 0.05).

Figure 2. In vivo imaging of retinal hypoxia in mouse OIR (P13) and age matched room air (RA) pups.

HYPOX-4 was administered systemically to OIR mouse pups 2 hours after return to room air on P12, as well as to age-matched room air pups. In vivo imaging was performed 24 hours post-injection of HYPOX-4. (A) Bright field image of OIR (P13) retina; (B) An image of the same retina in vivo, hypoxia was clearly detected by HYPOX-4 dependent fluorescence within the central avascular retina (green); (C) Bright field image of age-matched RA pup (P13); (D) HYPOX-4 dependent fluorescence was undetectable in aged-matched RA pups; (E) OIR mouse retina showing ex vivo HYPOX-4 dependent fluorescence in the central avascular retina (green); (F) The same retina counterstained with IB4, highlighting the peripheral vascular retina; (G) E and F merged; (H) RA pups showed minimal ex vivo HYPOX-4 dependent fluorescence; IB4 staining of the retinal vasculature (red) from an RA pup; (I) Hypoxia was confirmed in OIR (P12) pups by immunostaining of pimonidazole-adducts (red); blood vessels were counterstained with IB4 (green).

Figure 3. Localization of HYPOX-4 and pimonidazole in transverse retinal sections from OIR mice.

OIR pups (P12) were treated with HYPOX-4 or pimonidazole and the spatial distribution of hypoxia was determined in retinal cross-sections. (A) HYPOX-4 dependent fluorescence activity indicated alternating regions of hypoxia in the inner retina overlapping with retinal avascularity (green); hypoxia was visualized in the inner plexiform and inner nuclear layers. Presumably, oxygen diffusion out of the major vessel indicated by the white arrow, inhibits increased azo/nitroreductase activities and consequently the retention of HYPOX-4. (B) Pimonidazole-adduct immunostaining confirmed retinal hypoxia in the inner plexiform and inner nuclear layers; additionally, this method detected hypoxia in the ganglion cell layer (red). (A,B) retinal nuclei were stained with DAPI (blue). Abbreviations: GCL = ganglion cell layer, IPL = inner plexiform layer, INL = inner nuclear layer, OPL = outer plexiform layer, ONL = outer nuclear layer.

Figure 4. Effect of HYPOX-4 on retinal physiology was assayed in RA-raised mice via electroretinography (ERG) analysis.

(A,B) ERG measurements of dark-adapted mice 7 days post systemic administration of HYPOX-4 revealed no significant changes in mean a-wave and b-wave amplitudes at various flash intensities compared to vehicle (PBS) and sodium fluorescein (control) groups. The TUNEL assay was performed in retinal transverse sections to assess retinal apoptosis and taken as a measure of retinal toxicity. RA mice were treated with 100 μM HYPOX-4 or DNase 1; HYPOX-4 showed no cellular apoptosis. (C,F) DAPI staining of nuclei; (D) DNase 1-treated retinal transverse sections serving as a positive control, fragmented DNA was clearly visible; (G) HYPOX-4 treated retinal transverse sections showed no cellular apoptosis; (E) C and D merged; (H) F and G merged. (I,J) In vitro cellular proliferation was assessed by the BrdU assay using HYPOX-4 treated R28 and MIO-M1 cells. No effect on cellular proliferation was observed.