Abstract
Purpose :
We have reported amyloid deposits ex vivo in retinas of beagle dogs who suffer from a naturally occurring cognitive dysfunction syndrome, with symptoms and brain pathology similar to Alzheimer’s disease (AD). As in humans, amyloid beta brain load is inversely correlated with cognitive function. Here, we compare several imaging methods which could be used in vivo to track amyloid in this disease.
Methods :
Dogs were categorized by a battery of non-verbal cognitive function tests. Two cognitively impaired dogs and 3 cognitively normal dogs were imaged with a Heidelberg HRA in blue laser auto-fluorescence; and then given IV injections of CRANAD-28, previously used in rodent brain as a two-photon dye to image amyloid. Animals were reimaged 5 minutes post injection. One flat-mounted retina from a cognitively normal dog, positive for CRANAD-28 staining was counter stained with DAPI and histology studies were performed with confocal microcopy and in a fluorescence microscope, custom modified for polarization imaging. One week later, one animal was re-imaged in vivo in blue auto-fluorescence and then with a two-photon adaptive optics scanning light ophthalmoscope with excitation at 920 nm.
Results :
In one-photon excitation, amyloid deposits fluorescently labelled with CRANAD-28 were visible in vivo in the HRA blue auto-fluorescence mode and ex vivo in fluorescence microscopy, close to the anterior retinal surface. As previously reported, ex vivo, deposits were present in the retinas of both cognitively impaired and non-impaired dogs. One week later, in vivo some deposits remained visible. Deposits, visible using CRANAD-28 in one-photon excitation were also visible in two-photon excitation. Ex vivo, deposits visible in crossed-circular polarization correlated with those visible in fluorescence imaging with a sensitivity of 93.6%, and a specificity of 96.1%.
Conclusions :
The beagle dog is a valuable model of retinal pathology in human Alzheimer’s disease. As shown here, multi-modal imaging of retinal deposits in this model could include one and two-photon excitation of CRANAD-28, or non-invasive imaging in polarized light. In locating deposits ex vivo, polarized light imaging showed high sensitivity and specificity compared to fluorescence imaging. Both fluorescent labelling with CRANAD-28 and polarization imaging could potentially allow in vivo diagnosis and tracking of disease and therapies in this animal model of AD.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.