Astrocytes in the mouse brain respond bilaterally to unilateral retinal neurodegeneration

Abstract

Glaucomatous optic neuropathy, or glaucoma, is the world's primary cause of irreversible blindness. Glaucoma is comorbid with other neurodegenerative diseases, but how it might impact the environment of the full central nervous system to increase neurodegenerative vulnerability is unknown. Two neurodegenerative events occur early in the optic nerve, the structural link between the retina and brain: loss of anterograde transport in retinal ganglion cell (RGC) axons and early alterations in astrocyte structure and function. Here, we used whole-mount tissue clearing of full mouse brains to image RGC anterograde transport function and astrocyte responses across retinorecipient regions early in a unilateral microbead occlusion model of glaucoma. Using light sheet imaging, we found that RGC projections terminating specifically in the accessory optic tract are the first to lose transport function. Although degeneration was induced in one retina, astrocytes in both brain hemispheres responded to transport loss in a retinotopic pattern that mirrored the degenerating RGCs. A subpopulation of these astrocytes in contact with large descending blood vessels were immunopositive for LCN2, a marker associated with astrocyte reactivity. Together, these data suggest that even early stages of unilateral glaucoma have broad impacts on the health of astrocytes across both hemispheres of the brain, implying a glial mechanism behind neurodegenerative comorbidity in glaucoma.

Keywords: astrocyte; glaucoma; neurodegeneration; reactivity; visual system.

Fig. 1.

RGC transport efficiency varies across the homeostatic optic projection. (A) CTB conjugated to AlexaFluor-555 was injected into the posterior chamber of the eye. (B) Following injection, CTB is transported anterograde to the brain. (C) Fixed brains (n = 5) were delipidated, then rehydrated and stained (as needed) before clearing and imaging on a light sheet microscope. (D) Dorsal view of the CTB (red) imaged within the entire optic projection in a cleared naïve adult (3-mo-old) mouse brain (white indicates full brain autofluorescence obtained in the 488 channel). A = anterior, P = posterior. (E) Dorsal view of the optic projection labeled with CTB, with optic nerve, lateral geniculate nucleus (LGN) superior colliculus (SC), and ipsilateral projection indicated. (F) The contralateral (magenta) and ipsilateral (cyan) optic projections were segmented from the image for quantification using Imaris. (G) Individual brain regions were also segmented; suprachiasmatic nucleus (SCN, green), contralateral LGN (cLGN, pink), contralateral olivary pretectal nucleus (cOPN, orange), contralateral SC (cSC, red), ipsilateral LGN (iLGN, cyan), ipsilateral OPN (iOPN, blue), and ipsilateral SC (iSC, purple) are all visible from the dorsal view (Left). The frontal view (Right) also shows the accessory optic tract (AOT, yellow) D = dorsal, V = ventral. (H) CTB-positive tissue volume within each analyzed brain region. (I) Mean CTB fluorescence intensity quantified across the full optic projection, normalized to the optic chiasm, and compared to that within individual brain regions as an indicator of RGC anterograde transport efficiency. Regions with CTB intensity significantly different from the average of the full homeostatic projection indicated with P values (determined via repeated measures one-way ANOVA followed by Tukey’s test). A–C created with Biorender.com.

Fig. 2.

Anterograde transport loss begins in specific RGC projections within the contralateral optic tract. (A and B) Dorsal views of complete, CTB labeled optic projections corresponding to sham or microbead-injected, IOP-elevated eyes in distinct mice. (C) IOP in all eyes measured via rebound tonometry. Anterior chamber microbead injection resulted in significant IOP increase over contralateral eye and sham condition. (D) Quantification of CTB+ tissue volume for each projection or in defined brain regions (n = 5 sham, n = 5 microbead). In each, sham (lighter circles on the left of each comparison) and microbead (darker squares on the right) are compared. There was a significant decrease in CTB labeling following IOP elevation in the AOT when compared to sham. (E) Differences in mean intensity (normalized to optic chiasm) across projection regions indicate more widespread deficits in RGC anterograde transport efficiency, particularly in the contralateral tract. The full projection exhibits significant change in CTB fluorescence (gray), but when divided into contralateral and ipsilateral projections (Fig. 1F) only the contralateral projection shows a significant change. This reduction in mean and median intensity in the contralateral projection is significant in SC (F and F′), LGN (G and G′), and OPN (H and H′) as well as the AOT (I and I′). Significance for D and E determined via two-way ANOVA followed by Tukey’s test to compare sham and microbead values in each region analyzed. Adjusted P values for each significant (α ≤ 0.05) comparison indicated on graphs.

Fig. 3.

Astrocyte responses across visual brain regions bilaterally and retinotopically mirror unilateral RGC degenerative patterns. Brains (n = 3) were rehydrated and restained for GFAP to examine astrocyte cytoskeletal elements in degenerating visual streams. (A) Virtual slice in horizontal plane containing SC [merge of CTB (red, B) and GFAP (cyan, C)]. (D) GFAP fluorescence intensity in tissue with intact anterograde transport (red outline) is significantly lower than GFAP fluorescence intensity in the portion of SC with impaired anterograde transport (pink). Tissue contralateral to and retinotopically matching these regions (purple, lavender) exhibits the same pattern of GFAP intensity. This also occurred in LGN (E–H), while the AOT (I–L) exhibited the opposite pattern of GFAP intensity relative to RGC transport impairment. Significance for D, H, and I determined by one-way ANOVA followed by Tukey’s test. Adjusted P values for each significant (α ≤ 0.05) comparison indicated on the graph. Blue diagrams showing brain region locations were generated with Biorender.com.

Fig. 4.

Bilateral astrocyte responses include reactive astrocytes. Tissue sections from sham (A–D; n = 5) and microbead occlusion (E–H; n = 5) samples, imaged for CTB (red), GFAP, (cyan), and LCN2 (white) immunopositivity. (I) Higher-resolution Inset from sham sample indicating immunopositivity around descending vasculature. GFAP is detected surrounding vasculature but not lipocalin-2 (LCN2). (J and K) Ipsilateral and cSC insets corresponding to unilateral microbead occlusion condition. GFAP and LCN2 are detected around descending blood vessels on both sides of the brain. (L) Mean LCN2 fluorescence was significantly upregulated in bilateral SCs when IOP was unilaterally increased; the same was true of (M) the percentage of SC positive for LCN2 signal. There was no significant difference in LCN2 levels between hemispheres in either condition. (N and O) Like LCN2, GFAP is significantly bilaterally upregulated in the unilateral microbead occlusion condition relative to sham when measured by mean fluorescence and binarized signal. Significance was determined by one-way ANOVA followed by Tukey’s test. Adjusted P values for each significant (α ≤ 0.05) comparison between hemispheres [right (R) and left (L)] in the same brain or the same hemisphere in sham/MOM indicated on the graph. MOM = microbead occlusion model.