A model for choroidal neovascularization (CNV) and gene therapy for age-related macular degeneration (AMD)

Several important angiogenic factors, their receptors and regulatory factors, and anti-angiogenic factors were identified in human fetal retinal pigment epithelium (RPE) by RT-PCR, cloning, Western blot, and immunohistochemistry. They include vascular endothelial growth factor (VEGF) 165aa and 121aa; three receptors: Flt1, KDR and neuropilin-1; receptor for advanced glycation end products (RAGE); soluble adenylyl cyclase (sAC); angiopoietins (Ang1, Ang2) and their receptor Tie2; sFlt1 and pigment epithelium derived factor (PEDF). In collaboration with scientists at Chiron, I have packaged VEGF₁₆₅, Ang1, sFlt1 and PEDF cDNAs into adeno-associated virus (AAV). AAV-VEGF was first tested in cultured human fetal RPE to assess its ability to deliver VEGF to the apical and basal sides. At all concentrations, AAV-induced increases in VEGF were significant on the apical, but not on the basal side. These results were compared to adenovirus mediated overexpression of VEGF in RPE. AAV-VEGF was then injected into subretinal space (SRS) of Sprague Dawley or Long Evans rats. GFP and VEGF were expressed in the retina and RPE for 20 months, the longest time tested. Blood vessel proliferation and leakage were detected using BrdU/lectin double staining and fluorescein angiography, respectively. At five weeks to 20 months after AAV-VEGF injection, histological sections showed degenerating photoreceptors, proliferating RPE, and extensive subretinal neovascularization, while electroretinogram a- and b-wave amplitudes were significantly decreased. 3D reconstruction of serial sections demonstrated that choroidal blood vessels penetrated Bruch's membrane, one of them splitting into three in the SRS. This work established a new model for choroidal neovascularization (CNV), which causes the most devastating form of age-related macular degeneration (AMD), the leading cause of blindness for those over age of 60. This is the only extant model in which the neovascularization is long lasting (over 20 months), which is critical to mimicking a chronic disease like AMD. This animal model of CNV can be extended to primates. The same gene transfer techniques were used to deliver sFlt1 and PEDF genes into RPE. Preliminary results show that AAV mediated overexpression of sFlt1 in RPE can effectively inhibit the development of CNV, indicating that sFlt1 could be used therapeutically to treat AMD.