In this paper, Sharma et al. demonstrate the first in vivo two-photon adaptive optics images of extrinsically labeled mouse retinal cells acquired through the eye’s pupil. To enhance signal levels in the images, the authors first labeled retinal ganglion cells using green fluorescent protein (GFP) or G-CaMP3 (a calcium indicator) in normal black C57BL/6J mice. Capitalizing on the large numerical aperture inherent in the mouse eye (~0.50), Sharma et al. used a two-photon adaptive optics scanning laser ophthalmoscope specifically designed for the mouse eye to visualize retinal ganglion cell bodies, dendrites and axons and noted changes in dendrite and axon morphology in different axial planes separated by as few as 5 microns. Despite showing an order of magnitude reduction in mean signal intensity compared with GFP-labeled cell bodies (due to G-CaMP3’s inherently lower fluorescence efficiency), it was still possible to clearly resolve G-CaMP3-labeled cell bodies and visualize a subset of dendrites and axons. While the authors do not present functional imaging data, they do provide exciting calculations which show that the light levels used to acquire their two-photon adaptive optics retinal images are well below the thresholds needed to stimulate a cone or rod photoreceptor in the living mouse eye. This work suggests that the authors’ techniques for performing two-photon adaptive optics imaging in the mouse could allow for simultaneous structural and functional imaging as the infrared light used for structural imaging of the retina would not affect retinal function.
In conclusion, Sharma et al. show that two-photon adaptive optics imaging provides sufficient transverse and axial resolution to clearly visualize microscopic retinal features, from cell bodies to dendrites and axons. While it is encouraging that Sharma et al. did not notice any retinal damage using light levels that were above light safety limits scaled for the mouse eye based on human data, further work must be performed to ensure that thermal damage does not occur and, thereby, better determine the extent to which molecular processes in the mouse eye mimic those in human eyes. Nevertheless, the use of two-photon adaptive optics imaging in the mouse eye could allow for structural and functional assessment of individual retinal cells over time, potentially improving our understanding of visual pathways in normal eyes and the dynamic cellular changes that occur in retinal disease. It will be exciting to see future applications of two-photon adaptive optics imaging for functional assessment of the visual system and its hopeful development for visualizing intrinsic retinal fluorophores non-invasively.
You must log in to add comments.