Missing Motility: High-resolution in vivo imaging of retinal microglia reveals stationary ramified cells.

Abstract

Purpose : Microglia are believed to be highly dynamic cells that survey their microenvironment for signs of distress and disease. The retina is optically accessible and can be used to non-invasively study microglia and their role in health and neurodegeneration. We sought to characterize distribution and dynamics of microglia in vivo.

Methods : Cx3cr1^GFP/GFP and Cx3cr1^+/GFP mice were imaged with a custom combined scanning laser ophthalmoscope (SLO) and optical coherence tomography (OCT) system capable of capturing fluorescence and reflectance channels concurrently. Both widefield and small areas of the retina were imaged over time to assess microglial density, distribution, motility and mobility in the fluorescence channel. A 7.8mW 860nm laser was focused onto a spot on the retina for 2 minutes to elicit a microglial response. Changes in microglia were monitored with SLO, while changes in retinal layer thickness and scattering were monitored with OCT.

Results : Both Cx3cr1^GFP/GFP and Cx3cr1^+/GFP mice had microglial populations that were spatially homogenous across the retinal surface. In both strains only ~6% of cells were motile over 2-10 minute imaging sessions. In contrast, GFP+ monocytes rapidly coursed through the retinal vasculature, indicating that our imaging system had sufficient resolution to detect small, transient movements. Exposure of a small retinal area to 2 minutes of the IR laser increased scattering of the outer retina in the reflectance channel of the SLO system, allowing us to simultaneously observe the population of microglia in that area. Microglia did not respond to the retinal disruption with increased motility for at least 1 hour after exposure, though ameboid cells clustered at the site of exposure by 1 day. Microglia in close proximity to the cluster remained quiescent and ramified. The ameboid clusters resolved after 1 week, which coincided with lessened light scattering of photoreceptors, RPE, and choroid in OCT images.

Conclusions : Our study has revealed a surprising lack of microglial motility in vivo: 94% of cells show no evidence of primary or secondary branch movement even during extended imaging sessions. Furthermore, IR light-induced local disruption of the outer retina does not immediately increase motility, though ameboid microglia were clearly mobile during the days following disruption.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.