Pigment epithelium-derived factor (PEDF) alone or in combination with an anti-VEGF drug as a possible treatment of choroidal neovascularization and ischemic ocular diseases

Abstract

Ocular neovascular diseases such as choroidal neovascularization (CNV) and age-related macular degeneration (AMD) are today the most common causes of visual loss and blindness. Although intravitreal anti-vascular endothelial growth factor (VEGF) therapy has achieved major breakthroughs in the treatment of CNV and wet AMD, it does not always successfully suppress CNV, and no efficacious treatment is yet available to prevent photoreceptor degeneration. The formation of CNV is a compensatory response to ocular ischemia. If CNV can be stabilized, it may be helpful for photoreceptor survival and maintaining the normal physiological function of the retina. With a wide use of optical coherence tomography angiography (OCTA) in the clinic, quiescent CNVs without exudation are much more frequently found than previously assumed. In this study, I examined the efficacy of a pigment epithelium derived factor (PEDF) protein alone or combined with a VEGF antagonist (Avastin) in the treatment of rat quiescent CNV in vivo. The rat quiescent CNV was established by subretinal injection of an adeno-associated viral (AAV)-VEGF-A165 vector, which expresses VEGF and leads to vascular growth from the choroid into the subretinal space. The fluorescein angiography (FA), indocyanine green angiography (ICG) and optical coherence tomography (OCT) were performed at the third week, fourth week, sixth week and seventh week after subretinal injection of the AAV-VEGF-A165 vector. The maximal thickness of CNV was measured in the OCT images. Picrosirius red stain was used to quantify mature and immature collagen content in the formalin-fixed, paraffin-embedded quiescent CNV samples. The collagen Ⅳ, VEGF and PEDF were examined in the quiescent CNV samples by immunohistochemistry (IHC), and the percentage of positive staining area to the total CNV area was calculated by ImageJ software. To analyze the effect of PEDF on retinal photoreceptor cells, the outer nuclear layer (ONL) area as a percentage of the corresponding CNV area was quantified. The apoptosis of ONL cells above the CNV area was detected by TUNEL staining. PEDF combined with anti-VEGF therapy significantly inhibits VEGF expression in quiescent CNV. The ONL area as a percentage of the corresponding CNV area is increased in the PEDF treatment group (92.29, 120.27) and decreased in the Avastin treatment group (55.11, 75.59) (P ﹤0.001, Wilcoxon test with Kruskal-Wallis test). The apoptosis of ONL cells above the CNV area is reduced in the PEDF treatment group (0.95 ± 0.60, n/1000μm) compared to the vehicle treatment group (2.26 ± 0.95, n/1000μm) (P ﹤0.001, one-way ANOVA test). Thus, PEDF shows a neuroprotective effect for the retinal photoreceptor cells. In this in vivo experiment, a VEGF antagonist (Avastin) can significantly inhibit the formation of CNV in this rat quiescent CNV model and reduce its thickness. However, quiescent CNV can supply oxygen to the photoreceptor cells and protect photoreceptor cells from degeneration, thus maintaining vision. Although anti-VEGF (Avastin) treatment alone inhibits CNV, it also accelerates photoreceptor cell degeneration in this in vivo quiescent CNV model. In the clinic, development of macular geographic atrophy and fibrosis are found in patients with wet AMD under long term anti-VEGF treatment. PEDF has no effect on mature and immature collagen in this quiescent CNV model. The information on the therapeutic effect is shown in Table 1. PEDF is a potential vascular stabilizing and neuroprotective factor for stabilizing CNV and maintaining vision.

Ischemic/hypoxic retinopathy is a common condition that can cause visual impairment and blindness. However, the changes of the choroidal blood vessels under ischemic/hypoxic conditions are not clearly known. The apoptosis of retinal ganglion cells (RGCs) and the retinal degeneration under ischemic/hypoxic conditions cannot be treated. In this study, I established an ischemic/hypoxic ex vivo eye model by incubating the freshly enucleated rat eyes in Dulbecco's modified eagle medium (DMEM) at 4 °C for 14 hours. After 14 hours, eyes (including control eyes, eyes intravitreally injected with vehicle or PEDF protein) were fixed for electron microscopy (EM) and immunohistochemistry (IHC) respectively. The area of lumen (μm2/μm of Bruch’s membrane) and the area of the whole vessel (μm2/μm of Bruch’s membrane) were analyzed and immunohistochemical staining for VEGF and PEDF in the retina and choroidal vessels was performed. The effects of PEDF on the choriocapillaris and retinal neural cells under ischemia/hypoxia were investigated. In addition, the oxygen concentration within the vitreous was measured by an oxygen-sensitive microsensor in both the ischemic/hypoxic ex vivo eye model and the living rats under anesthesia. TUNEL staining was performed and the apoptosis of retinal neural cells was analyzed. In the living rats, the concentration of oxygen within the vitreous was on average 16.4 % of the oxygen concentration in the air. In the ischemic/hypoxic ex vivo eyes, the oxygen concentration within the vitreous was gradually decreased and the concentration was about 50% of the value in the eyes immediately after enucleation after about 400 minutes of incubation, indicating mild hypoxia during this process. After 14 hours of ischemia/hypoxia, the endothelial cells of the choroidal vessel were ultra-structurally similar with respect to cell organelles compared to the immediately fixed control eyes, but the morphology of the choroidal vessels changed dramatically. In the ischemic/hypoxic eyes, filopodia-like projections filled out the choroidal vessel lumen and appeared identical to the labyrinth-capillaries found in surgically extracted choroidal neovascular membranes from patients with wet AMD. The structural changes within the choriocapillaris in this ischemic/hypoxic eye model can mimic early changes in the process of pathological angiogenesis as observed in patients with CNV or wet AMD. This ex vivo eye model can be used to investigate short term drug effects on the choriocapillaris after ischemia/hypoxia. PEDF inhibited the filopodia-like projection formation and kept the choroidal lumen open as in vivo. The area of lumen is significantly reduced in the ischemic/hypoxic group (0.308±0.087μm2/μm) compared to the PEDF-treated group (1.034±0.077μm2/μm) (P＜0.001, t test). The area of vessel is significantly reduced in the ischemic/hypoxic group (0.675±0.048 μm2/μm) compared to the PEDF-treated group (1.583±0.094μm2/μm) (P＜0.001, t test). The intravitreally injected PEDF protein located in the whole retina and the choroidal vessels, indicating that PEDF protein can penetrate the retina and is transported into the choroid. The apoptosis of RGCs (2.36, 2.89; percentage of positive cells per 100μm) and inner nuclear cells (15.81, 22.89; number of positive cells per 1000μm) was significantly reduced in the PEDF-treated eyes compared to the ischemic/hypoxic eyes [ RGCs (3.14, 3.75; percentage of positive cells per 100μm) (P = 0.015), inner nuclear cells (33.69, 45.22; number of positive cells per 1000μm) (P = 0.001) Wilcoxon test]. The detailed information is shown in Table 2. Thus, PEDF is a promising candidate for treating ischemic/hypoxic retinopathy or human CNV alone or combined with other drugs.

In short, PEDF protected retinal neural cells in both the in vivo quiescent CNV model and the ex vivo ischemic/hypoxic eye model. PEDF inhibits the formation of filopodia-like projections and keeps the choroidal vessel lumen open in this ex vivo ischemic eye model.