Osteopontin drives retinal ganglion cell resiliency in glaucomatous optic neuropathy

Abstract

Chronic neurodegeneration and acute injuries lead to neuron losses via diverse processes. We compared retinal ganglion cell (RGC) responses between chronic glaucomatous conditions and the acute injury model. Among major RGC subclasses, αRGCs and intrinsically photosensitive RGCs (ipRGCs) preferentially survive glaucomatous conditions, similar to findings in the retina subject to axotomy. Focusing on an αRGC intrinsic factor, Osteopontin (secreted phosphoprotein 1 [Spp1]), we found an ectopic neuronal expression of Osteopontin (Spp1) in other RGCs subject to glaucomatous conditions. This contrasted with the Spp1 downregulation subject to axotomy. αRGC-specific Spp1 elimination led to significant αRGC loss, diminishing their resiliency. Spp1 overexpression led to robust neuroprotection of susceptible RGC subclasses under glaucomatous conditions. In contrast, Spp1 overexpression did not significantly protect RGCs subject to axotomy. Additionally, SPP1 marked adult human RGC subsets with large somata and SPP1 expression in the aqueous humor correlated with glaucoma severity. Our study reveals Spp1's role in mediating neuronal resiliency in glaucoma.

Keywords: CP: Neuroscience; Osteopontin; glaucoma; human retina; neuronal types; neuroprotection; optic nerve crush; retinal ganglion cell.

Figure 1.. Ocular hypertension leads to preferential survival of αRGCs and ipRGCs

(A and B) Illustration of the SOHU model surgery using the mouse eye (A) and the physical mechanisms resulting in elevated IOP after SOHU treatment (B). SO, silicone oil. (C) Whole-mount views of retinas among all RGCs and RGC subclasses. Retinas labeled ‘‘pan RGC’’ are labeled with the antibody RBPMS, which marks allRGCs. The rest of the retinas were from Kcng4-YFP, Cartpt-YFP, TWY3-YFP, Foxp2-YFP, and Opn4-YFP mice, in which αRGCs, ooDSGCs, W3-RGCs, F-RGCs, and Opn4-RGCs, respectively, are labeled genetically. Naive, sham-treated contralateral eye; wpi, weeks post-injection. Scale bar, 50 μm. (D) Fraction of pan RGCs and each subclass that survived SOHU treatment at 1 and 4 wpi. n = 5–8 animals per genotype. Data are presented as means ± SEM. The quantifications were generated by comparing 1 or 4 wpi with the naive baseline (100%). Unpaired two-sided Student’s t test, ns, not significant; ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05.

Figure 2.. Resilient αRGCs and ipRGCs demonstrate elevated Spp1 expression after SOHU treatment

(A) Representative retinal whole-mount images of Kcng4-YFP naive (top) and 4 wpi (bottom), labeled with antibodies to YFP (green), Spp1 (red), and Rbpms (blue). Arrows indicate the overlap of Spp1 and YFP; empty arrowheads indicate ectopic Spp1 expression, which is YFP negative. However, all Spp1 expression is restricted in Rbpms cells, indicating a restricted expression of Spp1 in RGCs. (B) Quantifications of Spp1-positive RGCs numbers in both conditions indicating a significant increase of ectopic Spp1 expression, with the results being normalized to the naive group. n = 5 animals per condition. (C) Quantification of the proportion of non-α-type RGC numbers exhibiting ectopic expression of Spp1 to the number of RGCs positive for Spp1. n = 5 animals per condition. (D) Quantifications of the overlap between Spp1 and other markers for RGC subclasses (representative images shown in Figures S2A–S2E). n = 5–7 animals per condition. (E) Vertical section of the naive retina (top) and 4 wpi (bottom), labeled with antibodies to Spp1 and Opn4 (melanopsin). Arrows indicate the overlap of Spp1 andOpn4 under SO treatment. Green, Spp1; red, Opn4. (F) Vertical section of Kcng4-YFP (αRGCs) naive retina (top) and 4 wpi (bottom), labeled with antibodies to pS6 and Spp1. Arrows indicate the overlap of Spp1 and pS6. Green, Spp1; red, pS6. (G) Fractions of the number of RGCs that have high-pS6-positive levels in both conditions, while the majority of the pS6-positive increase is coupled with Spp1-positive elevation. Scale bars (A, E, and F), 50 μm. n = 5 animals per condition. Unpaired two-sided Student’s t tests; ****p < 0.0001; *p < 0.05.

Figure 3.. Spp1 is essential for driving αRGC resiliency

(A) Schematic of the AAV construct for CRISPR-Cas9-mediated Spp1 knockdown in Kcng4-Cre-positive neurons (left) and timeline of experiment design for knockdown Spp1 at 2 weeks before SO injection and tissue harvest at 4 wpi (right). (B) Control sgRNA/Cas9 Kcng4-YFP retina (top left), retina under SO treatment (top right), sgSPP1/Cas9-infected retina (bottom left), and sgSPP1/Cas9-infectedretina under SO treatment (bottom right), labeled with antibody to YFP. (C) Quantification of normalized αRGC survival. n = 5 animals per condition. (D) Schematic of the AAV construct for Spp1 overexpression in Foxp2-Cre-positive neurons (left) and timeline of experiment design for overexpressing Spp1 at 2 weeks before SO injection and tissue harvest at 4 wpi (right). (E) Control AAV-expression retina (top left), retina under SOHU treatment (top right), AAV-Spp1-infected retina (bottom left), and AAV-Spp1-infected retina under SOHU treatment (bottom right), labeled with antibody to Foxp2. (F) Quantification of normalized F-RGC survival. Scale bars (B and E), 50 μm; n = 5 animals per condition. Paired t test; ns, not significant; **p < 0.01.

Figure 4.. Spp1 is enriched in adult human RGCs with large somata, and SPP1 expression correlates with the glaucoma severity in human patients

(A) Sample images of SPP1 (green) and RBPMS (red) in the retina from an adult donor sample eye with no notable ocular history, including expression in RGC and horizontal cell subsets. (B) Sample images of SPP1 (green), TBR1 (red, a previously characterized transcription factor found in T-RGCs/midget OFF-RGCs), and Nissl (gray) in the retina from a donor in (A). (C) Percentage of SPP1-positive and TBR1-positive RGCs in the adult human retina from the donor in (A) and (B). n = 2 retinas. (D) Quantification of RGC soma size of SPP1-positive, TBR1-positive, and the average Nissl (NeuroTrace)-positive RGCs in the adult human retina. n = 2 retinas. (E) SPP1 concentration in aqueous humor was quantified using ELISA from patients with mild (n = 11) and severe (n = 13) forms of primary open-angle glaucoma (POAG) and age-matched controls (n = 16). Significance was assessed by one-way ANOVA followed byTukey’s Honest Significant Difference (HSD) test for pairwise comparisons, demonstrating a significant difference between patients with severe glaucoma and control patients (p = 0.02). AH, aqueous humor. (F) Model illustrating the differential actions of Spp1 and survival of different RGC subclasses between SOHU and ONC treatments. In the SOHU model, Spp1 expression is upregulated, and ectopic Spp1 expression provides protection for RGCs. In the ONC model, Spp1 expression is downregulated, and restricted Spp1 expression in αRGCs only provides limited protection for αRGCs. Scale bars (A and B), 50 μm.