Abstract
Purpose :
The mouse has been an essential model to characterize mammalian retinal physiology and model human disease. Combined with adaptive optics, previous studies have imaged single retinal cells in the living mouse. To date, studies in mice have required anesthesia to provide optical stability. Here we show the first applications of adaptive optics imaging in the awake, behaving mouse without using anesthesia.
Methods :
A custom adaptive optics scanning light ophthalmoscope (AOSLO) was used to image the retina in confocal, split-detection and fluorescence modalities. Two healthy C57BL6/J mice were fit with head plate that was surgically fixed to the cranium using acrylic and cyanoacrylate. After a 1 week recovery, mice were head-fixed above a running wheel that allowed for voluntary movement. Training was performed for 3 sessions to acclimate. For imaging, eyes were dilated (1% tropicamide) and the apparatus was aligned in the imaging arm of the AOSLO. Widefield SLO was also used to image visible/NIR reflectance, fluorescence, and OCT modalities (Heidelberg HRA).
Results :
Head-fixed mice provided exceptional ocular stability. Pupil translation was <140 µm during active running (<5.6% of the 2.5 mm pupil). The Shack–Hartmann wavefront sensor (904 nm), imaged the central 2.0 mm pupil without eyelid occlusion. Eye blinks, saccade-like eye movements and gaze-shifts (5-33 deg) were sparse providing stable epochs. We also observed high-frequency eye motion (2-6 µm, 24-80 Hz), similar to human ocular tremor. Widefield SLO provided active registration of visible and NIR fluorescence and stabilized OCT.
Using adaptive optics, we were able to resolve the smallest retinal capillaries, nerve fiber bundles and single blood cells moving in vessels. Mean blood cell speed in an 18 µm diameter vessel was >2.8x faster in the awake mouse (25.4 mm/s) than in same sized vessels imaged in mice with 1% isoflurane (9.1±1.9 mm/s). Blood cell pulsatility was 481-555 beats/min in the awake mouse and 181-300 beats/min in anesthetized mice (Fig 1), revealing baseline blood flow changes with anesthesia.
Conclusions :
We show the first demonstration of AOSLO imaging of the awake, behaving mouse. With stable pupil position and AOSLO aberration correction, this preparation opens new possibilities to study the native aspects of the mouse eye and image the retina with high resolution while obviating the need of anesthesia that can alter optics and physiology.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.