Retinal ganglion cells downregulate gene expression and lose their axons within the optic nerve head in a mouse glaucoma model

Abstract

Little is known about molecular changes occurring within retinal ganglion cells (RGCs) before their death in glaucoma. Taking advantage of the fact that gamma-synuclein (Sncg) mRNA is expressed specifically and highly in adult mouse RGCs, we show in the DBA/2J mouse model of glaucoma that there is not only a loss of cells expressing this gene, but also a downregulation of gene expression of Sncg and many other genes within large numbers of RGCs. This downregulation of gene expression within RGCs occurs together with reductions in FluoroGold (FG) retrograde transport. Surprisingly, there are also large numbers of Sncg-expressing cells without any FG labeling, and among these many that have a marker previously associated with disconnected RGCs, accumulation of phosphorylated neurofilaments in their somas. These same diseased retinas also have large numbers of RGCs that maintain the intraocular portion while losing the optic nerve portion of their axons, and these disconnected axons terminate within the optic nerve head. Our data support the view that RGC degeneration in glaucoma has two separable stages: the first involves atrophy of RGCs, whereas the second involves an insult to axons, which causes the degeneration of axon portions distal to the optic nerve head but does not cause the immediate degeneration of intraretinal portions of axons or the immediate death of RGCs.

Figure 1.

Sncg mRNA is expressed specifically and abundantly in adult mouse RGCs. A, Low-power view of a retina from a 3 month DBA/2J mouse after in situ hybridization detection of Sncg mRNA shows expression in the GCL. B, Higher-power view of an overdeveloped in situ hybridization shows Sncg mRNA only in the GCL and an occasional cell in the inner nuclear layer. C–H, Expression of Sncg (C, F), GAD67 (D, G) mRNAs and merged views together with nuclei in blue (E, H) in cryostat sections (C–E) and flat mounts (F–H) of retinas from C57BL/6J mice. I, Pie chart showing the percentage of GCL cells that express Sncg mRNA (red), GAD67 mRNA (green), and cells that express neither marker (blue), as determined from retina flat-mount images (n = 2260 cells; n = 3 retinas). J, Colocalization of Sncg mRNA (red) and FG retrograde tracer labeling (green), together with nuclei (blue). K, Pie chart of the percentage of GCL cells with both Sncg expression and FG labeling (yellow), Sncg expression only (red), and cells without either marker (blue), as determined from retina flat-mount images (n = 1465 cells; n = 4 retinas). The arrows (A, B) mark displaced ganglion cells. ONL, Outer nuclear layer; INL, inner nuclear layer; IPL, inner plexiform layer. Scale bars: A, 100 μm; B–E, J, 20 μm; F–H, 50 μm.

Figure 2.

Gene expression is downregulated within RGCs in affected regions of DBA/2J retinas. A, Sncg mRNA expression in retina sections of 3 and 9 month DBA/2J mice. B, Sncg mRNA (red), GFAP immunoreactivity (green), and nuclei staining (DAPI; blue) in retina sections of 3 and 9 month DBA/2J mice. C, Colocalization of Sncg mRNA (red) with RGC-specific genes (green) Nfl, Brn3b, and Hsp90; genes that are enriched in but not specific to RGCs, Nf66, tubulin β-III (Tubb3), AchRB2, TrkB, Thy1, and Tau; and genes that have widespread or ubiquitous expression, FKBP12 and PPIA, in retina sections of 3 and 9 month DBA/2J mice shown together with nuclei staining (DAPI; blue). The arrows (A–C) mark RGCs with reduced expression of Sncg. m, Month; IPL, inner plexiform layer. Scale bars: A–C, 20 μm.

Figure 3.

Decreases in the number of Sncg-expressing cells and the amount of Sncg expression per cell occur in RGCs of old DBA/2J mice but not old C57BL/6J mice. Sncg expression (red), and nuclei (blue) in flat-mounted retinas of 2 month (A) and 12 month (C) C57BL/6J, and 2 month (B) and 15 month (D) DBA/2J mice. Scale bar, 20 μm. The arrows (D) mark RGCs with reduced expression of Sncg mRNA. Number of Sncg-expressing cells (RGCs) (E, G) and total Sncg expression per cell (F, H) in retinas of young and old C57BL/6J (E, F) and DBA/2J (G, H) mice. Retina numbers were as follows: 2 month C57BL/6J (n = 3), 12 month C57BL/6J (n = 3), 2 month DBA/2J (n = 5), and 15 month DBA/2J (n = 3). m, Month. Error bars represent the SD. **p < 0.002.

Figure 4.

Sectorial loss of Sncg-expressing cells and decreased Sncg expression occur together with decreases in FG retrograde labeling. A–F, Binarized images show RGCs identified either by their Sncg expression (red) or their FG labeling (green) in 3 month (A–C) and 9 month (D–F) DBA/2J mice. The asterisks in the retina from the 9 month DBA/2J (D–F) show an area with fewer FG-labeled cells than Sncg-expressing cells. The traces in the binarized images mark the outer edges of retina flat mounts used for analyses as well as position of the ONH. G–L, Sncg expression (G, J) and FG labeling (H, K) shown alone, and together (I, L), show variable intensities in different RGCs of 3 month (G–I) and 9 month (J–L) DBA/2J mice. The arrows in the 9 month retinas (J–L) mark RGCs with minimal expression of Sncg mRNA and little or no FG labeling. M, Probability distribution of mean Sncg expression per cell in retinas from 3 month (blue; n = 183,334 cells; n = 3 retinas) and 9 month (n = 256,837 cells; n = 7 retinas) DBA/2J mice. N, Probability distribution of mean FG labeling per cell (using the same cells shown in M) in the retinas from 3 month (blue) and 9 month (red) DBA/2J mice. O, Positive correlation between Sncg mRNA expression and FG retrograde labeling per cell (using the same cells shown in M and N), expressed as Z-scores (value minus the mean for that retina, divided by the retina SD) in the retinas of 3 month (blue) and 9 month (red) DBA/2J mice. P, Correlation between the retina mean value of total Sncg per cell and the total number of Sncg-expressing cells per retina of 3 month (blue dots) and 9 month (red dots) DBA/2J mice. Q, Correlation between the retina mean value of mean FG fluorescence per cell and the total number of Sncg-expressing cells per retina of 3 month (blue dots) and 9 month (red dots) DBA/2J mice. R, Correlation between the retina mean value of the total Sncg per cell and retina mean value of the mean FG per cell in the retinas of 3 month (blue dots) and 9 month (red dots) DBA/2J mice. m, Month. Scale bars: A–F, 500 μm; G–L, 50 μm.

Figure 5.

RGCs with weak and strong accumulation of pNF (pNF+ RGCs) are found in sectors of RGC degeneration and mark the position of other RGCs that also lack FG transport. A–D, RGCs with faint (arrowhead) and strong (arrow) pNF immunoreactivity in their cell bodies and dendrites (A) express Sncg mRNA (B), but at levels that are relatively low when compared with some nearby RGCs (asterisks). E–G, Traces of retinas from 9 month DBA/2J mice representing three extents of disease progression: unaffected (E), mild (F), and severe (G), categorized as described in the text and supplemental Figure 3 (available at www.jneurosci.org as supplemental material). H, Number of pNF+ RGCs per retina in unaffected (bracket I), mild (bracket J), or severe (bracket K) retinas, relative to the number of remaining Sncg-expressing cells. I–K, Probability distribution of NND for both Sncg-expressing cells without pNF accumulation (gray), as well as pNF+ RGCs with weak (green) and strong (red) pNF accumulation, graphed for retinas that are unaffected (I) (n = 114,019 cells; n = 2 retinas), mild (J) (n = 87,146 cells; n = 2 retinas), or severe (K) (n = 55,672 cells; n = 3 retinas). The dashed line marks the value 3 SDs above the mean value observed in retinas of 3 month DBA/2J mice (n = 183,334 cells; n = 3 retinas). L, Pie chart of the percentage of GCL cells within a ∼35 μm radius of pNF+ RGCs that have both Sncg expression and FG labeling (yellow), Sncg mRNA only (red) with or without pNF accumulation, and cells without either marker (blue) (n = 419 cells; n = 4 retinas). Scale bars: A–D, 20 μm; E–G, 500 μm.

Figure 6.

The loss of Sncg-expressing cells is matched by a reduction in tubulin β-III but not pNF labeling within intraretinal axons. Two retina flat mounts, from a 12 month C57BL/6J mouse (A, C, E) and a 15 month DBA/2J mouse with severe degeneration (B, D, F) showing the relationship between Sncg mRNA expression (red) (A, B), tubulin β-III (Tubb3 or TUJ1) immunoreactivity (blue) (C, D), and pNF immunoreactivity (green) (E, F). m, Month. Scale bar, 500 μm.

Figure 7.

Severely affected DBA/2J mice have accumulation of pNF within intraretinal axons and have loss of optic nerve axons. Peripheral retina axons from 12 month C57BL/6J (A–C) and 15 month DBA/2J mice (D–F) labeled with antibodies to tubulin β-III (Tubb3) (A, D) and pNF (B, E), shown next to the merged views (C, F). G, Retina from a 15 month DBA/2J mouse in which a single pNF-positive axon can be followed from the ONH (asterisk) to the most peripheral part of the retina in which an RGC soma also has pNF labeling. H, I, Higher power views of two regions of the same pNF-labeled axon shown in G are shown magnified so that the axon near the soma (H) and ONH (I) can be compared. J–R, Sncg mRNA expression (J–L), and pNF staining of axons inside the retina (M–O) and in the proximal optic nerve (P–R) for one 3 month DBA/2J mouse (J, M, P) and two 9 month DBA/2J mice with lesser (K, N, Q) and greater (L, O, R) loss of Sncg-expressing cells. Note the correspondence between the number of Sncg-expressing cells within the retina and the number of pNF axons in the proximal optic nerve, but the lack of correspondence of the pNF labeling within the retina. m, Month. Scale bars: A–F, H, I, 20 μm; J–O, 100 μm; G, P–R, 50 μm.

Figure 8.

pNF-positive axons in glaucomatous retinas of DBA/2J mice end within the ONH, and pNF+ RGCs survive disconnected for ∼1 month. A, B, pNF-positive axons are found within the retina in unaffected (A) and highly affected (B) 9 month DBA/2J retinas. The asterisk indicates the missing retina surrounding the ONH, analyzed in longitudinal sections of ONH punches in E and F. C, D, Cross sections of the optic nerves, at ∼2 mm from the retina, from the unaffected (C) and affected (D) retinas presented in A and B, respectively, show uniform labeling of optic nerve axons with pNF (green) and TUJ1 (red) in the unaffected DBA/2J mouse (C) and significant loss of both pNF- and TUJ1-positive axons in the affected DBA/2J mouse (D). E, F, Sum projection images of longitudinal sections of ONH and proximal ON region dissected from the unaffected (E) and affected (F) retinas presented in A and B, respectively. Fluorescence intensity measurements (G) of the unaffected (blue) and affected (red) retinas shown in A and B, respectively, show that the majority of pNF-positive axons from the affected retina end within 500 μm of the retina. The dotted line in E and F indicate the position of the choroid, from which distances were measured in G. H, I, Retina flat mount of a 9 month DBA/2J mouse shows the relationship between FG labeling 1 month after injection in the superior colliculi (magenta) (H) and pNF immunoreactivity (green) (I). J, Cross section of the optic nerve from retina presented in H and I immunolabeled with pNF, showing the near complete loss of axons by the time of analysis. K, L, pNF+ RGCs (green) with (K) and without (L) FG retrograde labeling (magenta) 1 month after FG injection, found in the retina shown in H and I. M, Percentage of pNF+ RGCs that have FG labeling in 9 month DBA/2J 1 week (n = 25 cells in each of 6 retinas) and 1 month (n = 25 cells in each of 5 retinas; n = 19 in 1 retina) after FG injection. Note that, 1 week after FG injection, a very small number of pNF+ RGCs have FG labeling, whereas about one-half of pNF+ RGCs have FG 1 month after FG injection. w, Week; m, month. Scale bars: A, B, H, I, 500 μm; C, D, J, 50 μm; E, F, 200 μm; K, L, 20 μm.