Can an in vivo imaging system be used to determine localization and biodistribution of AAV5-mediated gene expression following subretinal and intravitreal delivery in mice?

Recombinant adeno associated viruses (AAV) are the most commonly used vectors in animal model studies of gene therapy for retinal diseases. The ability of a vector to localize and remain in the target tissue, and in this manner to avoid off-target effects beyond the site of delivery, is critical to the efficacy and safety of the treatment. The in vivo imaging system (IVIS) is a non-invasive imaging tool used for detection and quantification of bioluminescence activity in rodents. Our aim was to investigate whether IVIS can detect localization and biodistribution of AAV5 vector in mice following subretinal (SR) and intravitreal (IVT) injections. AAV5 carrying firefly luciferase DNA under control of the ubiquitous cytomegalovirus (CMV) promoter was injected unilaterally IVT or SR (in the central or peripheral retina) of forty-one mice. Luciferase activity was tracked for up to 60 weeks in the longest surviving animals, using repeated (up to 12 times) IVIS bioluminescence imaging. Luciferase presence was also confirmed immunohistochemically (IHC) and by PCR in representative animals. In the SR group, IVIS readings demonstrated luciferase activity in all (32/32) eyes, and luciferase presence was confirmed by IHC (4/4 eyes) and PCR (12/12 eyes). In the IVT group, IVIS readings demonstrated luciferase activity in 7/9 eyes, and luciferase presence was confirmed by PCR in 5/5 eyes and by IHC (2/2 eyes). In two SR-injected animals (one each from the central and peripheral injection sites), PCR detected luciferase presence in the ipsilateral optic nerves, a finding that was not detected by IVIS or IHC. Our results show that when evaluating SR delivery, IVIS has a sensitivity and specificity of 100% compared with the gold standard PCR. When evaluating IVT delivery, IVIS has a sensitivity of 78% and specificity of 100%. These finding confirm the ability of IVIS to detect in-vivo localized expression of AAV following SR delivery in the retina up to 60 weeks post-treatment, using repeated imaging for longitudinal evaluation, without fading of the biological signal, thereby replacing the need for post mortem processing in order to confirm vector expression. However, IVIS is probably not sensitive enough, compared with genome detection, to demonstrate biodistribution to the optic nerve, as it could not detect luciferase activity in ipsilateral optic nerves following SR delivery in mice.