Abstract
Purpose :
Retinal astrocytes migrate into the retina from the optic nerve (ON) during late embryonic development and spread centrifugally in the nerve fiber layer (NFL) to cover the inner retina. Astrocytes guide development of the intrinsic retinal vasculature, but factors that pattern the astrocyte network itself remain elusive. We propose that astrocytes use retinal ganglion cell (RGC) axons as directional cues and will not achieve proper spatial distribution without RGCs. We tested this hypothesis by investigating astrocyte development in mice lacking RGCs through genetic ablation via knockout of the transcription factor Math5.
Methods :
Eyes from Math5 null and littermate control mice were fixed at postnatal day 7 (P7), and retinas removed, immunostained, and imaged by confocal fluorescence microscopy. Antibodies against Pax2, CD31, RBPMS, and neurofilament were used to label astrocytes, blood vessels, RGCs, and RGC axons, respectively, in wholemount retinas. Control and null retinas were analyzed in ImageJ and statistical analyses (t-test, 2-way ANOVA) performed in JMP.
Results :
Math5 null retinas were severely deficient in RBPMS+ RGCs, lacked axon fascicles in the NFL, and were devoid of intrinsic retinal vasculature. By P7, astrocytes in control retinas have reached the ora serrata and achieved a uniform density across the entire retina. In Math5 mutants, however, the area encompassed by the wavefront of astrocyte migration was significantly reduced (control 98.1 ± 0.5%, KO 72.3 ± 2.6%, p < 0.01). Additionally, the spatial distribution of astrocytes was dysregulated with an abnormally high density of cells near the optic nerve and a steeply declining gradient to the periphery (control 1365 ± 101 cells/mm2 central, 1520 ± 43 middle, 1663 ± 58 peripheral; KO 3188 ± 438 central, 1935 ± 187 middle, 100 ± 28 peripheral; interaction of genotype with eccentricity p<0.001).
Conclusions :
RGCs are required for normal migration and spatial organization of retinal astrocytes during development, consistent with a model in which the neuroretina guides the maturation of its own astrocytic and vascular networks. Moreover, defects in astrocytes may underlie the vascular abnormalities in retinas lacking RGCs. Further experiments will describe the trajectory of astrocyte development and investigate cellular and molecular mechanisms underlying RGC-astrocyte-blood vessel interactions.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.