Changes of Osteopontin in the Aqueous Humor of the DBA2/J Glaucoma Model Correlated with Optic Nerve and RGC Degenerations

Abstract

Purpose.: To identify age-dependent regulated aqueous humor (AH) factors in DBA2/J (D2J) mice and to correlate them with optic nerve degeneration and intraocular pressure (IOP) by population and individual analysis.

Methods.: AH samples of D2J mice aged 2 (n = 3), 7 (n = 5), and 10 months (n = 14) were analyzed by mouse cytokine antibody array. Ten-month samples were classified into eyes with (D2J+) or without (D2J−) optic neuropathy. Ten-month-old C57/Bl6 (B6; n = 13) and DBA2/Rj (D2Rj; n = 15) mice served as controls. IOP was recorded from 2 to 10 months. Individual AH osteopontin (OPN) was determined in 31 D2J eyes (10 months) and was correlated with optic neuropathy and IOP. OPN mRNA was detected by in situ hybridization. OPN blood plasma content of D2J and B6 was monitored from 8 to 10 months. Effect of OPN on cell survival in the ganglion cell layer (GCL) or metabolism was tested in ex vivo–cultured D2Rj eyes and murine neuronal precursors.

Results.: In array analysis, OPN was detected in 10-month-old D2J mice only. They significantly differed between D2J− and D2J+ (P = 0.006). By Western blot analysis, a sevenfold OPN increase in D2J+ was determined compared with B6. Individual analysis confirmed the positive correlation of OPN with optic neuropathy. IOP was not correlated with OPN. OPN blood plasma contents steadily increased with age in D2J. OPN⁺ cells were detected within the ciliary body of D2J, and OPN⁺ RGCs were ≈30% reduced. OPN treatment inhibited cell degeneration within the GCL in ex vivo–cultured D2Rj eyes and increased the metabolic activity of neuronal precursor cells.

Conclusions.: OPN is an age-dependent increased AH factor associated with degeneration of the optic nerve in D2J mice. By modulating the metabolism of neuronal cells, deregulated levels of OPN could be involved in degenerative processes affecting RGCs or optic nerve axons in the D2J model.

The D2J mouse strain develops a hereditary optic neuropathy, accompanied by a reduction of RGC and a progression of IOP. Consequently, D2J mice commonly serve as a model for glaucoma research. ^{1 –5} This strain carries mutations in Gpnmb^R150X and Tyrp^1b, leading to iris pigment dispersion and stroma atrophy. ^6,7 As a consequence, the mice tend to develop a synechiae between iris and/or cornea and lens, which in turn results in an IOP elevation. In a recent individual analysis, we showed that not all D2J mice develop an increase in IOP and, furthermore, not every D2J mouse developing optic neuropathy had an increase in IOP. ⁸ In a follow-up study, we showed in addition that D2J mice display functional deficits of the retina by electroretinogram measurements that were not correlated with IOP. ⁹ Both findings strongly suggested that ocular hypertension alone cannot be postulated as a compulsory factor for the onset of the optic neuropathy or RGC degeneration in D2J. In several forms of human glaucoma—for example, primary open angle glaucoma (POAG)—degeneration of the optic nerve axons or RGCs also develops independently of IOP elevation. ^10,11 Therefore, IOP is not considered to account for the onset of the pathogenesis in these forms of glaucoma but as a diagnostic risk marker. In the past, quantitative and qualitative analysis of human glaucomatous AH samples revealed particular differences in protein composition, with several factors significantly correlated with the disease. ^{12 –29} In POAG, the most prominent factor is TGF-β₂, which not only is increased in approximately 50% of the cases but has also been shown to induce glaucomatous changes in optic nerve astrocytes and trabecular meshwork cells in vitro and to reduce outflow facility in perfused human and porcine anterior eye segments. ^{30 –44} However, human AH is only limitedly available; therefore, correlation studies of AH changes and disease onset and progression to identify early or initiating factors are almost impossible. Hence, the D2J model offers the opportunity to analyze and correlate AH changes with age, IOP, RGC degeneration, and optic neuropathy. Here, we report that the novel AH component osteopontin (OPN) is strongly elevated in 10-month-old D2J eyes and moreover that it is associated with the development of optic neuropathy. Furthermore, we provide the first experimental data that indicate that OPN mediates the metabolic activity of murine neuronal precursor cells in vitro and reduces the decline of cells within the ganglion cell layer (GCL) in ex vivo–cultivated eyes. 

Materials and Methods

Mice

All experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the local animal welfare authorities. D2J and B6 mice were obtained from Jackson Laboratories (Bar Harbor, ME); D2Rj mice were from Janvier Breeding Center, Le Genest St. Isle, France). Animals were housed in cages containing white pine bedding and were maintained in a 12-hour light/12-hour dark cycle with standard rodent chow available ad libitum. At respective ages, animals were killed by sublethal exposure to CO₂ and cervical dislocation. Eyes then were enucleated for AH extraction, histologic processing, or ex vivo cultivation. For population analysis, 13 B6, 15 D2Rj, 19 D2J, and five D2J mice aged 2 months (array control) were used; individual analysis was performed on 31 D2J eyes. 

IOP Measurements

IOPs were measured by noninvasive rebound tonometry ^{45 –47} in 2-month intervals from 2 to 10 months (population analysis) or singularly at D2J aged 10 months (individual analysis). At 7 months, five D2J mice were killed for AH extraction; from then on, the D2J cohort size was 14 from then on. During measurements, animals were restrained in a custom-made device allowing measurements without anesthesia and avoiding increased intrathoracic pressure. Single values resemble mean averages of five measurements per eye. None of the animals examined in this study showed corneal calcifications or cataract. 

Classification of Optic Neuropathy

Optic nerves were fixed in Ito's solution and embedded in an epoxy resin (Epon; Hexion Specialty Chemicals, Inc., Houston, TX). Cross-sections were stained with toluidine blue. To assess the degree of optic neuropathy, areas with gliotic changes and axon degeneration in each individual eye (left [OS]; right [OD]) were encircled, and the percentage was calculated in relation to the complete cross-section area (Fig. 1B, top left/right). All optic nerve sections were viewed in a light microscope (Aristoplan; Leitz, Wetzlar, Germany). All photographs were taken with a digital camera (DC500; Leica, Wetzlar, Germany). D2J eyes of the population analysis (n = 25) were subdivided into eyes with optic neuropathy (≤20%; D2J−; n = 9) or without it (20%; D2J+; n = 16; Fig. 1C). D2J eyes for individual analysis were subdivided into eyes with optic neuropathy (<10%, n = 15; 10% ≤ 70%, n = 5; or ≥80%, n = 11; see Fig. 4B). 

Figure 1.

 

IOP progression of B6 and D2J and optic neuropathy classification in 10-month-old D2J mice. (A) IOP progressions from 2 to 10 months of B6, D2J− and D2J+ mice, measured in 2-month intervals. Horizontal bars: mean IOP. (B) Morphologic classification of the optic neuropathy. Top: methodical determination of the percentage of optic neuropathy. Bottom, left: cross-section shows a B6 optic nerve without optic neuropathy. Bottom, right: cross-section shows a D2J+ optic nerve with optic neuropathy >95%. (C) Tabular summary of the percentage of optic neuropathy determined for each eye of 10-month-old D2J mice. D2J− are highlighted in gray. OS, left eye; OD, right eye; n.e., not evaluable.

AH Sample Preparation

Approximately 3 to 5 μL AH was collected from enucleated eyes in pointed glass capillaries by penetrating the center of the corneas under binocular view, avoiding any contact with the inner tissues of the anterior chamber. Samples of B6 (n = 23), D2Rj (n = 26), and D2J mice aged 2 (n = 10) and 7 (n = 10) months were directly pooled. Samples of 10-month-old D2J were pooled according to the morphologic classification of optic neuropathy. D2J samples for individual ELISA analysis were stored separately. Pools and individual samples were stored at −80°C. 

Cytokine Antibody Array

Thirty microliters of the pooled AH samples were assayed (RayBio Mouse Cytokine Antibody Array 4; RayBiotech; http://www.raybiotech.com/map_all_m.asp#28) according to the manufacturer's instructions. Chemiluminescence signals were visualized by exposure to light-sensitive films (Hyperfilm ECL; GE Healthcare, Piscataway, NJ) for 20 minutes. Films were digitized, and densitometric quantifications were made with analysis software (Lumi-Analyst; Boehringer, Mannheim, Germany). Given RLU values represent the mean ± SEM of both spotted probes on the array. 

Western Blot Analysis

Twenty microliters each of B6 and D2J+ AH pools were supplemented with 25% (vol/vol) SDS-loading buffer (RotiLoad; Roth, Karlsruhe, Germany) and denatured at 60°C for 7 minutes. Twenty-five-microliter aliquots were separated by SDS-PAGE and transferred onto a nitrocellulose membrane (Protran BA83, 0.2 μm; Schleicher & Schüll, Dassel, Germany) by the tank blot method at 70 V for 45 minutes in 1× transfer buffer (10 mM CAPS [pH 11; 3-(cyclohexylamino)-1-propanesulfonic acid], 20% [vol/vol] methanol, 0.1% [wt/vol] SDS). Membranes were blocked in TBST/5% BSA (Tris-buffered saline, 0.1% [vol/vol] Tween-20, 5% [wt/vol] bovine serum albumin, pH 7.2) washed in TBST. Then the rabbit polyclonal anti-mouse OPN antibody (ab8448; Abcam) was added, diluted 1:500 in TBST/1% BSA, and allowed to react for 1 hour at room temperature (RT). After washing twice for 5 minutes with TBST, alkaline phosphatase-conjugated goat anti–rabbit IgG (Promega, Madison, WI) was added, diluted 1:10,000 in TBST/1% BSA for 30 minutes at RT. Blots were washed three times in TBST for 5 minutes and once in detection buffer (100 mM Tris-HCl, 100 mM NaCl, pH 9.5). For detection, 1 mL detection reagent (CDP-Star; Roche, Indianapolis, IN) was added to the membranes and incubated for 5 minutes at RT. Chemiluminescence signals were visualized by exposure to light-sensitive films (Hyperfilm ECL; GE Healthcare) for 1 minute to 10 minutes. Quantification was performed with analysis software (Lumi-Analyst; Boehringer). 

Blood Plasma Collection

Blood plasma was collected weekly for 2 months from five B6 and four D2J mice, aged 8 months at the beginning of the experiment and not part of either the population or the individual analysis. Plasma was extracted from mouse tail blood. Mouse tails were cupped, and whole blood was collected in reaction tubes. To avoid coagulation, one-tenth volume of 500 mM EDTA was added immediately. Plasma was separated by centrifugation for 5 minutes at maximum speed. Plasma samples were stored individually at −80°C until use. 

Enzyme-Linked Immunosorbent Assay

Mouse OPN ELISA kits were purchased from R&D Systems (Wiesbaden-Nordenstadt, Germany) and used according to the manufacturer's instructions. AH samples were diluted 1+49 and blood plasma samples 1+99 in ELISA sample buffer before application. 

In Situ Hybridization

OPN fragments were produced by PCR using primers 5′-gcttttgcctgtttggcattg-3′ (forward; nt242–263) and 5′-gacctcagaagatgaactctc-3′ (reverse; nt1080–1101) and were subcloned into the vector (TopoII; Invitrogen, Carlsbad, CA) in sense and antisense orientation. Digoxigenin (DIG)-labeled mRNA probes were synthesized with a DIG RNA labeling kit (Roche) according to the manufacturer's instructions. Ten-micrometer cryosections of 4% paraformaldehyde (PFA)-fixed D2J and B6 eyes were air dried, postfixed in 4% PFA, washed in DEPC-PBS, and bleached in 6% H₂O₂ for 10 minutes. Sections were washed in DEPC-PBS, acetylated in TAE buffer (100 mM tromethamine, 4% acetic anhydride), and prehybridized in buffer (Prehyb; 40% formamide, 5× SSC, 5× Denhardt's, 250 μg/mL yeast t-RNA) for 2 hours at RT. Then the first buffer was exchanged for another (Hyb; Prehyb-buffer plus 1 μg/mL RNA probe). Hybridization was performed at 70°C overnight (12–14 hours). Slides were washed for 10 minutes in 5× SSC at 65°C, 20 minutes in 2× SSC at 37°C, 20 minutes in 2× SSC plus 0.5 μg/mL RNaseA at 37°C, and twice for 20 minutes in 0.2× SSC at 65°C. After blocking, AP-conjugated antidigoxigenin Fab fragments diluted 1:2000 in blocking solution (Roche) were added for 1 hour at RT. After washing twice in washing buffer (100 mM Tris pH 7.5, 150 mM NaCl) for 20 minutes, the slides were incubated in developing buffer (100 mM Tris pH 9.5, 100 mM NaCl, 50 mM MgCl₂, 0.1% Tween 20, levamisole, 250 μg/mL NBT, 150 μg/mL BCIP; changed every 8 hours) at RT for 48 hours. The reaction was stopped with solution (STOP [1% SDS, 20 mM EDTA, 20 mM Tris pH 7.5, 100 mM NaCl]), and slides were mounted in glycerol. Sections were morphologically analyzed under a light microscope (Aristoplan; Leitz), and photographs were taken with a digital camera (DC500; Leica). 

Ex Vivo–Cultured D2Rj Eyes

Before OPN treatment studies, the progression of retinal disorganization/degeneration caused by ex vivo cultivation was preevaluated in D2Rj eyes. The D2Rj interstrain control was chosen because they do not develop hereditary retinal degenerations, as described for D2J mice. In brief, eyes were enucleated, lenses were not removed to avoid mechanical damage to the retina, and corneas were only carved to allow consistent medium flow. D2Rj eyes were cultivated in DMEM/F-12/10% FBS without supplements in 60-mm Petri dishes under constant agitation in a cell culture incubator at standard cell culture conditions. Morphology of the retinas was examined after 4, 7, and 10 days (three D2Rj eyes per time point). Structural disorganizations were morphologically observable already at day 4 within the nuclear layers (inner nuclear layer, outer nuclear layer; Fig. 2A). The GCL, in contrast, was almost intact at 4 days and showed mild structural disorganization at 7 days (Figs. 2A, 2B). After 10 days, disorganization of GCL was too massive (Fig. 2C), and it was decided to perform OPN treatment experiments up to 7 days at maximum. 

Figure 2.

 

Morphology of ex vivo–cultivated D2Rj eyes. Toluidine blue–stained sections of ex vivo–cultivated D2Rj eyes after (A) 4 days, (B) 7 days, and (C) 10 days. Signs of structural disorganization of the nuclear layers (INL, inner nuclear layer; ONL, outer nuclear layer) were already observable at 4 days and became more evident at 7 and 10 days. Morphology of the GCL appeared almost completely intact up to 7 days, allowing quantification of the cell numbers. After 10 days of cultivation, severe degeneration and structural disorganization were also visible in the GCL (arrowhead) and throughout all other retinal layers, including tissue edema (open arrows). L, lens; V, vitreous cavity. Scale bars, 50 μm.

OPN treatments were performed in triplicate for each time point (4, 7 days) and each concentration (0, 200, 1000 ng/mL OPN), using a total of 18 D2Rj eyes aged 3 months. Cultivation was performed in DMEM/F-12/10% FBS with indicated supplements under standard cell culture conditions. After treatment, serial sagittal sections of the eye globes were prepared. Only midsagittal sections (open iris plus optic nerve visible) were used for quantitative analysis to guarantee a comparably sectioned plane in all investigated eyes. Sections were stained with toluidine blue, and stained cells of the complete GCL next to the vitreous cavity (V) were counted throughout the entire retina and averaged out of a total of nine sections (three sections per eye globe). All sections were viewed with a light microscope (Aristoplan; Leitz). Values given represent the mean ± SEM of nine counted sections per time point and concentration. 

Metabolic Activity Assays

The in vitro effect of OPN on cellular metabolism was tested on a murine neuronal precursor cell line. ⁴⁸ These cells were mistakenly used as RGC in the past, but recharacterization by van Bergen et al. ⁴⁸ (2009) has shown that they do not correspond to the original rat RGC-5 characterized by Krishnamoorthy et al. ⁴⁹ (2001). Cells were seeded to 96-well plates at a density of 2.0 × 10³ cells/well in DMEM/F-12/10% FBS. Plates were incubated overnight under standard cell culture conditions. To mimic oxidative stress, cells were preincubated in DMEM/F-12/10% FBS containing 150 μM H₂O₂ for 6 hours before medium was changed to DMEM/F-12/10% FBS supplemented with 0, 250, 1000, or 2000 ng/mL OPN. Metabolic activity was directly assessed (CellTiter 96 AQueous MTS Assay; Invitrogen) after 36- and 72-hour cultivation, respectively, according to the manufacturer's instructions. Experiments were performed in triplicate, and values given represent the mean ± SEM. 

Statistical Analysis

The statistical significance of the differences between OPN content of 10-month-old D2J− and D2J+ mice and the differences in blood plasma OPN contents in B6 and D2J over time and between these strains were tested using an unpaired two-tailed t-test. Analysis of individual correlations of OPN and degree of optic nerve degeneration and analysis of ex vivo–cultivated D2Rj eyes and neuronal precursor cells was made by one-way ANOVA followed by a Tukey's multiple comparison test. Association of OPN contents and IOP within the three groups of optic nerve degeneration was tested by correlation analysis calculating the correlation coefficients and P values. 

Results

IOP Progression

IOP values in B6 mice displayed strong individual variability, but the mean values remained constant at approximately 8.5 (±1.1) mm Hg from 2 to 8 months (Fig. 1A). A similar progression was observed for D2Rj (data not shown), in accordance with our previous study. ⁸ IOP progression of both D2J+ and D2J− showed an initial decline between 2 and 4 months (6.0–5.0 mm Hg in D2J−; 6.2–4.4 mm Hg in D2J+), but from then on the mean values increased steadily from 5.0 to 10.5 mm Hg (D2J−) and 4.4 to 9.3 mm Hg (D2J+), respectively (Fig. 1A). 

Optic Nerve Morphology

The degree of glaucomatous damage was morphologically determined and subsequently classified in the eyes of 10-month-old B6, D2J, and D2Rj mice. Neither in B6 nor in D2Rj eyes were degenerative changes detected (Fig. 1B). Nine D2J eyes had optic neuropathy ≤20%, and 16 D2J eyes had optic neuropathy >20% (Fig. 1C); 14 of these displayed a damage area of ≥95% (Figs. 1B, 1C). 

Cytokine Antibody Array

Seven candidates were specifically upregulated in the AH of 10-month-old D2J mice (Figs. 3A–C). However, when discriminated against optic neuropathy, a statistically significant difference between D2J+ and D2J− was found only for OPN (P = 0.0063; Fig. 3D). By densitometric quantification, a value of 37 ± 0.7 RLU for D2J− and 49 ± 0.6 RLU for D2J+, respectively, was determined. The increase of OPN within the AH of D2J+ compared with B6 was confirmed by Western blot analysis, revealing an increase of approximately sevenfold (Fig. 3E). 

Figure 3.

 

Mouse cytokine antibody array. (A) Layout of the array (RayBio Mouse Cytokine Antibody Array 4; higher quality available at: http://www.raybiotech.com/map_all_m.asp#28). Reprinted with permission. ©2007–2010 RayBiotech, Inc. Location of OPN is boxed. (B) Array readout of 10-month-old B6 (B6–10) and D2Rj (D2Rj-10) controls. (C) Array readout of 2- and 7-month-old D2J control (D2J−2, D2J−7), 10-month-old D2J− (D2J−10[-]), and D2J+ (D2J−10[+]) mice. (B, C) Spots corresponding to OPN are boxed. (D) Densitometric quantification of (C) demonstrating the elevation of OPN content in D2J samples given as relative light units (RLU) and indicating the significant difference between D2J− and D2J+ mice (*P = 0.0063). (E) Western blot confirming the increase of OPN in D2J+ compared with B6 samples (left); densitometric quantification of the blot revealed a sevenfold increase (right).

Individual Correlation of OPN, Optic Neuropathy, and IOP in D2J Eyes

OPN content in the AH of D2J eyes aged 10 months (n = 31) was individually determined by ELISA (Figs. 4A, 4B). Concomitantly, the degree of optic neuropathy was diagnosed for each eye, and eyes were grouped based on degree: optic neuropathy <10% (n = 15), 10% ≤ 70% (n = 5), and ≥80% (n = 11; Figs. 4A, 4B). D2J mice with optic neuropathy <10% had a mean average OPN concentration of 67 pg/μL, with individual values not exceeding 207 pg/μL, In the group with 10% ≤ 70% optic neuropathy, the mean average was slightly increased to 95 pg/μL; however, the maximum value did not exceed 218 pg/μL. In the group with optic neuropathy ≥80%, a maximum of 619 pg/μL was detected, and the mean average was increased to 281 pg/μL, differing statistically significant from both other groups (*P _{10≤70%/≥80%} < 0.05; **P _<10%/≥80% < 0.001). Notably, 6 of 11 eyes (55%) had values higher than the maxima detected in both other groups (Figs. 4A, 4B). Correlations of individual IOPs with OPN content were not statistically significant in all three groups (P _<10%= 0.632; P _10≤70% = 0.338; P _≥80% = 0.334; Figs. 4B, 4C). 

Figure 4.

 

Correlation of OPN content, optic neuropathy and IOP in individual D2J eyes. (A) Individual OPN concentrations measured in D2J eyes with optic neuropathy of <10%, 10% ≤ 70%, or ≥80%. Horizontal bars: mean OPN concentrations. (B) Tabular summary of OPN concentration, degree of optic neuropathy, and IOP determined for individual eyes. (C) Plots of OPN concentration versus IOP for the three groups.

OPN Blood Plasma Concentration in B6 and D2J Mice

OPN concentrations within the blood plasma of B6 and D2J mice were individually different at all time points analyzed. In B6, the mean average plasma concentration was approximately 2000 pg/mL and remained stable throughout the duration of the study (P _w32/w40 = 0.7556; Fig. 5A). D2J also had a mean average concentration of 2000 pg/mL at week 32; however, the average constantly increased to approximately 3000 pg/mL at week 40. The difference between week 32 and week 40 was statistically significant (*P _w32/w40 = 0.048; Fig. 5B). The difference in OPN contents between B6 and D2J at week 40 was statistically highly significant (**P = 0.0003; Fig. 5). 

Figure 5.

 

Progression of OPN blood plasma concentration in B6 and D2J mice. (A) Individual values of 5 B6 mice over 8 weeks; trendline indicates a constant mean value of 2000 pg/mL. (B) Individual values of four D2J mice over 8 weeks; trendline indicates a constant increase of the mean value from 2000 pg/mL to 3000 pg/mL.

Localization of OPN mRNA in B6 and D2J Eyes

In situ hybridization (ISH) analysis for the detection of OPN-expressing cells was performed in 10 B6 and eight D2J eyes. No cells positively labeled for OPN were detected within the anterior segments of all B6 eyes or in six of the D2J eyes. In the other two D2J eyes, several positively labeled cells were detected in the ciliary body (Fig. 6A). Within the posterior eye, OPN-labeled cells were exclusively detected within the RGC layer in both strains (Fig. 6B). However, quantification of OPN⁺ cells indicated a reduction of approximately 30% in D2J compared with B6. 

Figure 6.

 

OPN expressing cells in the anterior and posterior eye segment of B6 and D2J mice. (A) Ciliary body of a B6 (left) and a D2J (right) eye. OPN⁺ cells (arrows) were detected by ISH in D2J eyes only. (B) OPN⁺ ganglion cells (arrows) in B6 (left) and D2J (right) retinas.

Cell Survival in the GCL of Ex Vivo–Cultivated D2Rj Eyes

After 4 days of ex vivo cultivation, no significant differences in numbers of toluidine blue–stained cells within the GCL were detectable between untreated controls and OPN treated eyes (189 ± 30 vs. 195 ± 11 vs. 192 ± 13; Fig. 7A). After 7 days, the number of cells was statistically significant decreased to 137 ± 29 in untreated controls (***P _4d/7d < 0.0001). Eyes treated with 200 ng/mL OPN, in contrast, had a mean cell count of 199 ± 25, which was comparable to the value after 4 days but significantly higher than in the 7-day controls (***P _co/200 < 0.0001). In eyes cultivated in the presence of 1000 ng/mL OPN, however, the cell number was statistically significant reduced (156 ± 15) compared with eyes treated with 200 ng/mL (**P _1000/200 < 0.001) and comparable to the number determined in the controls (Fig. 7B). 

Figure 7.

 

OPN effects on ex vivo–cultivated D2Rj eyes and murine neuronal precursor cells. (A) Cell survival within the GCL of ex vivo–cultivated D2Rj eyes against different OPN concentrations. Numbers of toluidine-stained cells in the GCL counted on sections of enucleated D2Rj eyes cultivated in medium containing 0, 200, or 1000 ng/mL OPN for 4 days (left) or 7 days (right). Plot indicates a significant decrease of cells in controls (0 ng/mL), whereas numbers remained constant in eyes treated with 200 ng/mL OPN. 1000 ng/mL OPN seemed not to inhibit cell degeneration because cell counts were significantly lower. **P < 0.001; ***P < 0.0001. (B) Metabolic activity of H₂O₂ stressed murine neuronal progenitor cells against different OPN concentrations. Metabolic activity determined by MTS assay in neuronal progenitors treated with 0, 250, 1000, or 2000 ng/mL OPN for 36 (left) or 72 (right) hours. Plot indicates significantly increased activity of all OPN-treated neuronal progenitors after 36 hours; this applied only for neuronal progenitors treated with 1000 ng/mL OPN after 72 hours. *P < 0.05.

Metabolic Activity of H₂O₂-Stressed Murine Neuronal Precursor Cells

Thirty-six hours after H₂O₂-induced oxidative stress, neuronal progenitor cells propagated in the presence of OPN displayed significantly higher metabolic activities than untreated controls, independent of the concentration (*P < 0.05 for all; Fig. 7B). After 72 hours, however, only progenitors cultivated with 1000 ng/mL OPN displayed a significantly higher metabolic activity than untreated controls (*P < 0.05; Fig. 7B). Values obtained in precursor cells treated with 250 ng/mL or 2000 ng/mL appeared higher than those in controls, but the differences were not statistically significant (Fig. 7B). 

Discussion

The intention of this study was to identify age-dependent and optic neuropathy–correlated changes in the AH composition of D2J mice. Here we report that OPN was significantly increased in 10-month-old D2J mice only and was additionally elevated in D2J+ mice. This indicated a distinct age dependency and an association with optic neuropathy. In further investigations, this association was confirmed by individual analysis of D2J eyes. Of the 11 D2J eyes with optic neuropathy ≥80%, more than half (55%) concurrently had OPN concentrations even exceeding the maximum concentration determined in D2J eyes with lesser optic neuropathy. Notably, there was no correlation between OPN concentrations and IOP. 

A potential relevance for OPN in deleterious processes affecting the optic nerve or the D2J retina is suggested from the literature. In several studies, a strong correlation between OPN and different neurodegenerative pathologic conditions, such as Alzheimer's disease, Parkinson's disease, multiple sclerosis, and stroke, has been described. ^{50 –56} In these diseases, the degenerations are accompanied by upregulation of OPN either directly at the lesion sites or within the cerebral or spinal fluid. Therefore, OPN is considered a prognostic marker for these diseases or their severity, respectively. Whether OPN is actively involved in the degenerative process or is upregulated during an elicited protective response is not yet completely clarified. However, data have been collected in rodent models of these diseases that argue for a protective but also a degenerative function of OPN, most likely in a context-dependent manner. Meller et al. ⁵⁶ observed that ventricularly administered OPN significantly reduced the infarct size in a murine model of stroke, emphasizing an active neuroprotective function. In contrast to that, Maetzler et al. ⁵⁷ found that MPTP-induced neurodegenerations were significantly reduced in OPN^−/− mice, from which the authors concluded an active function for neurodegeneration in Parkinson's disease. There are indications for the ambivalent function of OPN in ocular tissues. Chidlow et al. ⁵⁸ detected a distinct transient upregulation of OPN within the plexiform layer of rat retinas after excitotoxic and ischemic insults. ⁵⁸ In patients with Devic's disease, a demyelinating disease affecting the optic nerve, OPN was one of the strongest induced genes, and the authors discussed that it conveys underlying MΦ-mediated inflammation by its chemoattractive capacity. ⁵⁹ Translated to our findings, it is tempting to speculate that OPN is initially upregulated to counteract deleterious processes such as ischemia, inflammation, or increased IOP and to protect RGCs and optic nerve axons in the D2J model as well. However, if these insults persist or recur frequently, as in glaucoma, this could lead to constant overproduction—thus, local accumulation of OPN—which then could negate the protective effect into the opposite, degenerative, effect. The very preliminary experimental data we present here could be interpreted in a way that supports this hypothesis at least to some extent. Based on the experiments with enucleated D2Rj eyes, and with the admission that this experimental setup is technically not fully developed and leaves room for discussion about the causes for observable disorganization and degeneration, we think one can conclude that the supplementation of medium with 200 ng/mL OPN had an inhibitory or a diminishing effect on cell loss within the GCL compared with the control eyes. Unfortunately, toluidine blue staining is not cell type specific, so we cannot discriminate between RGCs and displaced amacrine cells. Given that both cells are of neuronal origin, one can speculate that OPN, at least at a concentration of 200 ng/mL, might mediate a protective effect for neuronal cells. MTS assays with murine neuronal precursor cells showed that OPN is able to stimulate the metabolic activity of these cells, which could be regarded as an indication that such a hypothetical protective effect of OPN might work by regulating the cell metabolism. Notably, in both experiments, increased concentrations of OPN did not have that protective or stimulatory effect, respectively. This could indicate negation of the positive effects when a certain limiting concentration is exceeded. In any case, it will be left to more sophisticated studies extending our preliminary results to prove a potential protective effect of OPN and perhaps a dose dependency. 

Based on the role of OPN as a cytokine and the fact that D2J mice have been reported to exhibit defects of ocular immune privilege and signs of constant mild inflammation, ^7,60 a connection of the immune system with the degenerative processes mediated by OPN is suggestive. This, however, could reveal OPN not as a protectant but as an active mediator of degenerative events, as it is discussed for other neurodegenerative diseases. Studies investigating this aspect of OPN capacities would be of great interest. Further studies focusing on the identification of the origin of the increased OPN in the AH of D2J mice will also be required. Our ISH experiments revealed OPN⁺ cells within the ciliary body of D2J mice only, perhaps indicating the induction of OPN synthesis. Another likely source would be the blood plasma, the basis of AH, because we detected a progressive increase of OPN contents in the plasma of D2J mice over time. Mo et al. ⁶¹ found an impaired integrity of the blood/aqueous barrier in aging D2J mice. 

Taken together, our results identified OPN as an age-dependent increased AH factor associated with degeneration of the optic nerve in D2J mice. By potentially affecting the metabolism of neuronal cells, deregulated levels of OPN could be involved in degenerative processes affecting RGCs or optic nerve axons in the D2J model. Therefore, OPN is a promising candidate for future studies to analyze the actual participation of OPN in this model in detail and eventually to test the relevance of OPN in human glaucoma. 

Footnotes

 Disclosure: M.T. Birke, None; C. Neumann, None; K. Birke, None; J. Kremers, None; M. Scholz, None

The authors thank Anke Fischer and Ekaterina Gedova for expert technical assistance.