Downregulation of SARM1 Protects Retinal Ganglion Cell Axonal and Somal Degeneration Via JNK Activation in a Glaucomatous Model of Ocular Hypertension

Abstract

Purpose: This study aimed to assess the expression of sterile alpha and TIR motif containing protein 1 (SARM1) in both chronic and acute glaucomatous animal models and investigate the underlying SARM1-JNK signaling mechanism responsible for the protective effects of SARM1 downregulation on retinal ganglion cell (RGC) soma and axons in a chronic intraocular hypertension (COH) model.

Methods: The COH model was induced by injecting magnetic microbeads into the anterior chamber, whereas the acute model was created through ischemia–reperfusion (I/R) injury. Immunohistochemistry and Western blot were used to assess SARM1 expression and JNK phosphorylation in the retina and optic nerve. SARM1 downregulation was achieved through the intravitreal injection of adeno-associated virus (AAV)2-shRNA. Quantitative analysis of RGC survival was performed by the counting of Brn3A-positive RGCs, and surviving axons were assessed through optic nerve toluidine blue stain.

Results: The expression of SARM1 increased 1 week after microbead injection in the optic nerve, whereas the retinal SARM1 expression decreased at 3 days post-injection in the COH model. After 24 hours of reperfusion, SARM1 expression increased in both the optic nerves and the retinas in the I/R injury model. SARM1 downregulation led to increased survival of RGC soma and axons in the COH model. In this model, JNK phosphorylation was significantly reduced concomitant with decreased SARM1 expression.

Conclusions: Elevated SARM1 expression was observed in the optic nerves in both the COH and I/R injury models. Downregulation of SARM1 exhibited a protective effect on RGC soma and axons in the COH model, with JNK identified as a downstream regulator of SARM1 in this context.

Glaucoma is a progressive degenerative disease that globally leads to irreversible blindness. The pathological mechanisms of various types of glaucoma involve the progressive death of retinal ganglion cells (RGCs) and optic nerve degeneration. Although elevated intraocular pressure (IOP) is widely recognized as the primary risk factor for glaucoma, current clinical treatments primarily focus on IOP control through medication and surgery. However, there remains a subset of patients with glaucoma who experience continued vision loss despite successful IOP reduction.¹ Therefore, it is crucial to develop therapies independent of IOP reduction that can effectively protect RGCs and delay their progressive death. 

Axonal injury is considered to be the primary early insult that drives glaucomatous neurodegeneration, leading to subsequent loss of RGC soma and dendrites.^2,3 Previous studies have demonstrated that treatments aimed at protecting RGC soma fail to effectively provide long-term protection for the distal RGC axons.^4–6 Consequently, damage occurs in the distal axons, disrupting the connection between the retina and central nervous system, and resulting in an inability to maintain visual function. These findings suggest that axon degeneration involves unique molecular signaling pathways independent of those involved in RGC soma apoptosis. Therefore, targeted protection of axons during the early stages of glaucoma may offer more effective preservation of visual function and delay progressive optic nerve degeneration. 

Sterile alpha and TIR motif containing protein 1 (SARM1), identified as a toll-like receptor adaptor protein, plays a pivotal role in axonal degeneration, instigating the process through nicotinamide adenine dinucleotide (NAD+) depletion.^7–9 Deficiency of SARM1 has been shown to protect both axons and neuronal somas in various models of axonal injury.^10–12 In the visual nervous system, enhanced survival of RGC soma and axons has been observed in SARM1 knockout mice in multiple RGC injury models, including the TNF-alpha-induced neuroinflammation model, the rotenone-induced retinal degeneration model, the silicone oil-induced ocular hypertension (SOHU) glaucoma model, and expressed a pathogenic form of human myocilin that induced the chronic elevated IOP mouse model.^13–16 

C-Jun N-terminal kinases (JNK), members of the mitogen-activated protein kinase (MAPK) family, play a crucial role in axonal degeneration and RGC soma death in glaucoma.² Previous studies have confirmed JNK inhibitors delay RGC soma and axonal degeneration.^17,18 Activation of SARM1 is necessary for initiating this MAPK cascade, involving MKK4 and JNK, in the optic nerve crush (ONC) model.¹⁹ There was contrasting finding that suggested that SARM1 deficiency neither stops DLK/JNK activation in RGC somas after ONC nor halts their degeneration. However, it still confirms the protective effect of deficiency of SARM1 on axons.²⁰ These discrepancies may arise from variations in post-injury time points and neuronal subtypes. Therefore, elucidating the upstream and downstream relationships between SARM1 and JNK is essential to protect RGC somas and axons from degeneration. 

In this study, we investigated the temporal expression patterns of SARM1 and JNK in a chronic intraocular hypertension (COH) model induced by magnetic microbeads and an acute glaucoma model induced by ischemia-reperfusion (I/R) injury. Additionally, we examined the neuroprotective effects of downregulated SARM1 expression on both RGC axons and somas in the COH model. Furthermore, we explored the impact of decreased SARM1 expression on JNK phosphorylation. Our findings may offer insights into potential strategies for delaying glaucomatous neurodegeneration. 

Materials and Methods

Animals

Male Wistar rats (200–250 g) were obtained from the Shanghai Lab Animal Research Center for use in these experiments. The rats were housed under a 12-hour light/12-hour dark cycle and fed with continuous access to food and water. All animal care procedures adhered to the guidelines set by the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care and Use Committee at Fudan University. 

Chronic Intraocular Hypertension Model Induced by Magnetic Microbeads

The chronic experimental glaucoma model was induced using magnetic microbeads (BM547; Bangs Laboratories, Fishers, IN, USA; diameter = 9 µm; concentration = 50 mg/mL), as described in previous studies.^21,22 Rats were anesthetized with a mixture of ketamine/xylazine (80:10 mg/kg) administered via intraperitoneal injection. Prior to microbead injection, topical anesthesia was achieved using 0.4% oxybuprocaine eye drops, followed by mydriasis using 0.5% Tropicamide Phenylephrine Eye Drops. Magnetic microbeads (8–10 µL) were injected into the anterior chamber and evenly distributed around it using a magnet. Subsequently, the anterior chamber angle was blocked to induce gradual elevation of IOP. Post-injection, erythromycin eye ointment was applied for moisturization and prevention of infection. The control eye served as a sham-operated control group and received an equivalent volume of saline injection. 

The IOP of both eyes was measured using the TonoLab tonometer (TonoLab; Icare, Espoo, Finland) in accordance with the manufacturer’s instructions. Prior to and every time point following the injections, IOP measurements were obtained. A successful glaucomatous rat model was defined as those maintaining IOP levels at approximately 1.5 times higher than control rats, whereas rats with IOPs that returned to normal were excluded from the study. 

Rat Ischemia–Reperfusion Injury Model

The I/R injury model was used to simulate acute glaucoma in this study, following the methodology described in our previous publication.²³ Briefly, rats were anesthetized and a 30-gauge needle connected to a standard saline reservoir that was inserted into the anterior chamber of the surgical eye. For sham control eyes, only the needle was inserted. The IOP of the surgical eye was elevated to 90 to 100 millimeters of mercury (mm Hg) measured by a tonometer and maintained for 45 minutes. The rat eyes were evaluated at sequential time points of 6 hours, 24 hours, and 48 hours after reperfusion before euthanization. 

Protein Extraction and Western Blot Analysis

Posterior myelinated optic nerve and retina tissue were lysed in RIPA buffer containing phosphatase and protease inhibitor cocktails (Roche, Mannheim, Germany). Protein concentration was determined using the BCA Protein Assay. A total of 20 to 30 µg of protein samples were loaded for Western blot analysis. 

For Western blot analysis, protein samples were denatured by boiling in SDS-PAGE Sample Loading Buffer (Beyotime Biotechnology, Shanghai, China), separated by SDS-PAGE, and then transferred to PVDF membranes, as previously described.²⁴ The anti-SARM1 antibody (ab226930) provided by Abcam (Cambridge, UK), the SAPK/JNK antibody (9252), and the Phospho-SAPK/JNK (Thr183/Tyr185) Rabbit monoclonal antibody (4668) provided by Cell Signaling Technology (Beverly, MA, USA), and the DLK/MAP3K12 antibody (GTX124127) provided by GeneTex (San Antonio, TX, USA) were used. The peroxidase-conjugated anti-β-actin antibody from Sigma-Aldrich (St. Louis, MO, USA) was used as a loading control. Horseradish peroxidase-conjugated goat anti-rabbit IgG secondary antibody from Santa Cruz Biotechnology (Dallas, TX, USA) was used here. The immunoreactive bands were detected using ECL Ultra Western HRP Substrate (Millipore, MA, USA). ImageJ software was utilized for protein expression analysis. β-actin and the total form of JNK were used for normalization. 

Immunohistofluorescence in the Retina and Optic Nerve

Immunohistofluorescence was performed as previously described.²⁴ Briefly, rat eyeballs with optic nerve were enucleated together and fixed in paraformaldehyde (4%). Subsequently, both the eyeballs and optic nerves were cryoprotected sequentially in 20% and 30% sucrose solutions. The eyeballs and optic nerves were embedded in optimal cutting temperature compound, respectively (Sakura Tissue-Tek, Torrance, CA, USA). The embedded eyeballs and optic nerves were then sectioned into retinal cross-sections of 10 µm thickness for the former, and longitudinal sections of 8 µm thickness for the latter. Standard staining methods were applied to these retinal and optic nerve sections. Primary antibodies used included anti-SARM1 antibody (ab226930; Abcam, Cambridge, UK), Phospho-SAPK/JNK (Thr183/Tyr185; 81E11) Rabbit monoclonal antibody (4668; Cell Signaling Technology, Beverly, MA, USA) and Anti-β3-tubulin Tubulin antibody (ab78078; Abcam, Cambridge, UK). Then, the sections were incubated with following fluorescent secondary antibodies: Alexa Fluor 555-conjugated donkey anti-rabbit IgG, Alexa Fluor 488-conjugated donkey anti-mouse, and Alexa Fluor 647-conjugated donkey anti-mouse (Thermo Fisher Scientific, Fair Lawn, NJ, USA). The nuclei were stained with DAPI. Finally, all the sections were examined by confocal laser-scanning microscope (Leica Microsystems, Heidelberg, Germany). 

Adeno-Associated Virus Administration

The adeno-associated virus (AAV)2-U6-rSarm1 shRNA1-U6-rSarm1 shRNA2-ZsGreen (AAV-SARM1-shRNA) and its negative control virus AAV2-ZsGreen Negative Control (AAV-NC) were obtained from HANBIO (Shanghai, China). The AAV-SARM1-shRNA contained a tandem array of two rSarm1 shRNAs. The specific sequence details of the rSarm1 shRNA were provided as referenced in a previous study²⁵: 

• rSarm1 shRNA1:
• 5′-CAACGACAATACAGCATATGAGCAACGACAATACAGCATATGATTCAAGAGATCATA
• TGCTGTATTGTCGTTGTTTTTT-3′
• rSarm1 shRNA2:
• 5′-CAGATGGAATGGTGCCCTCTAGCAGATGGAATGGTGCCCTCTATTCAAGAGATAGAG
• GGCACCATTCCATCTGTTTTTT-3′

The titers of the 2 virus titers were approximately 1.3 × 10¹² Viral particles (VPs) per milliliter. A total volume of 4 to 5 µL AAV (5 × 10⁹ VP per eye) was intravitreally injected into rat eyes using a 33-gauge needle connected to a 10 µL Hamilton microsyringe, as previously described.²⁶ To prevent fluid outflow due to high IOP, the needle was held in place for at least 60 seconds. Tissue samples were collected 4 weeks after AAV injection for subsequent experiments. 

RNA Isolation and Real-Time PCR Analysis

Total RNA of retinas and optic nerves was extracted using RNAsimple Total RNA kit (TIANGEN, Beijing, China) according to the manufacturer's protocol. The concentration and purity of the RNA samples were assessed using NanoDrop (Thermo Fisher Scientific). Reverse transcription of the total RNA was performed using the PrimeScript RT reagent Kit with gDNA Eraser (RR047A; Takara Biomedical Technology, Beijing, China), following the manufacturer's protocol. Quantitative PCR was conducted on a CFX96 Real-Time System (Bio-Rad, Richmond, CA, USA) using TB GreenPremix Ex Taq Ⅱ (RR820A; Takara Biomedical Technology). The rat SARM1-specific primers for quantitative PCR (qPCR) had the following sequences: forward primer 5′-CGTTTGGAGGCTCAGTGCATAGG-3′ and reverse primer 5′-CAGGCGTTTCAGGCTCTGGATG-3′. To normalize gene expression levels, β-actin mRNA was used as an internal control with forward primer 5′-CTGAGAGGGAAATCGTGCGTGAC-3′ and reverse primer 5′- AGGAAGAGGATGCGGCAGTGG-3′. The relative expression of SARM1 was calculated using the 2^−ΔΔCT method after normalization to β-actin expression.²⁵ 

Immunofluorescence Analysis of Retinal Flat Mounts and Quantification of RGCs

The rat eyeballs dissected at the designated time point were rinsed with phosphate-buffered saline (PBS), and subsequently fixed in 4% paraformaldehyde (PFA) for 1 hour at room temperature. Following another round of PBS washing, the eyeballs were dissected, and the retinas were carefully removed by making four radial incisions to enable flat mounting on glass slides. The retinal flat mounts were fixed again in 4% PFA for an additional 30 minutes before being washed twice with PBS. To block and permeabilize, PBS containing 2% Triton X-100 and 5% donkey serum was used for a duration of 1 hour at room temperature. Subsequently, the flat mounts were incubated for 72 hours at 4°C with anti-Brn3A antibody (ab245230; Abcam) diluted in PBS containing 0.5% Triton X-100 and 2% donkey serum. After 3 rounds of washing with PBS for a duration of 10 minutes each time, the flat mounts were incubated with 555-conjugated donkey anti-Rabbit (1:1000 dilution; Thermo Fisher Scientific) for 90 minutes at room temperature. Following this step, the flat mounts were rinsed with PBS before being sealed using glycerol as mounting medium. Finally, all the retinal flat mounts were examined and photographed by confocal laser-scanning microscope (Leica Microsystems, Heidelberg, Germany). 

To quantify the Brn3A positive RGCs, each retinal flat mount was visually divided into four quadrants, and the central and peripheral regions were partitioned as described in our previous study.²⁶ The central regions were defined as 1.5 mm from the optic disc, and the peripheral regions were defined as 1.5 mm to 3 mm from the optic disc. One field was counted in each central region, whereas two fields were counted in each peripheral region using Image J software. A total of 12 microscopic fields in each retina were counted, representing approximately 20% to 30% of the retinal area. The numbers of Brn3A positive RGCs were utilized to estimate the surviving RGC population.²⁷ 

Optic Nerve Toluidine Blue Stain and Surviving Axon Count

After being fixed with 2.5% glutaraldehyde and 2% PFA in 0.01 M PBS at 4°C overnight, optic nerves were post-fixed with 1% osmium tetroxide (OsO4) in 0.1 M phosphate buffer (PB, pH 7.4) for 2 hours at room temperature. Subsequently, the optic nerves were sequential dehydration using alcohol concentrations of 30%, 50%, 70%, 80%, 95%, and 100% to 100% for 20 minutes each step, and finally, with acetone twice for 15 minutes. 

Embed-812 resin mixture was used to embed the optic nerves. Semi-thin cross sections (1.5 µm thickness) of the optic nerves located approximately 3 to 5 mm distal to the optic nerve head were collected using an ultramicrotome (Leica HistoCore Nanocut R). Semi-thin cross sections were stained with 1% toluidine blue. Observation and photography were performed under a Leica thunder microscope equipped with a 100 × oil immersion objective lens. RGC axons were quantified by blind counting using Image J software along with an axonJ plugin. Approximately 10% of the total area from each optic nerve section was counted, and these results were extrapolated to estimate the entire number of axons. 

Statistical Analysis

The data were expressed as the mean ± standard error of the mean (SEM) with individual values points plotted on the graph. Statistical analysis and graphing were performed using GraphPad Prism software (version 9.0; San Diego, CA, USA). The Student's t-test was used for two groups comparison, whereas 1-way analysis of variance (ANOVA) with the Tukey's multiple comparisons test was used for multiple comparisons. A significance level of P < 0.05 was considered statistically significant, whereas P > 0.05 indicated non-significance (NS). 

Results

The Expression of SARM1 was Increased in the Optic Nerve, Not in the Retina in the COH Model

We established the rat COH model using magnetic microbeads, as described in previous studies.²¹ In order to confirm the successful development of the COH model, IOP measurements were taken before microbead injection and at every time point after injection. As shown in Figure 1A, there was a significant increase in IOP in the experimental eye compared to the control eye post-injection. The peak IOP (29.6 ± 1.8 mm Hg) was observed at 3 days post-injection, and this elevated level of IOP (26.2 ± 1.7 mm Hg) persisted for 2 weeks after injection. 

Figure 1.

 

The expression of SARM1 increased in the ONs in the COH model. (A) Measurements of IOP were taken before and every time point following the injection of magnetic microbeads for a duration of 2 weeks. The IOP exhibited a significant increase 3 days post-injection and remained elevated throughout the entire 2-week period compared to the control eyes. The results are expressed in mm Hg and presented as the mean ± SEM; ***P < 0.001 indicates significant differences from the control eyes (n = 5 for each group). (B) Western blotting analysis of SARM1 expression in the optic nerves and the retinas at 3 days, 1 week, and 2 weeks post-injection. The band with the higher molecular weight is marked with an arrow indicating to SARM1. (C, D) Quantitative analysis of SARM1 expression in the optic nerves and the retinas. The data were normalized to β-actin expression. SARM1 expression increased significantly 1 week after injection in the optic nerves. The results are presented as the mean ± SEM; ***P < 0.001, *P < 0.05 indicates significant differences (n = 5–6 for each group). (E) The optic nerves’ longitudinal sections were immunostained with β3-tubulin Tubulin (green) and SARM1 (red) 3 days, 1 week, and 2 weeks after magnetic microbeads injection. The arrows indicate colocalization of β3-tubulin Tubulin and SARM1. (F) Retinal sections were stained with SARM1 (red) and DAPI (blue) 3 days, 1 week, and 2 weeks after magnetic microbeads injection.

Previous studies have demonstrated that the expression of SARM1 is upregulated in excitotoxicity mouse models, and SARM1-dependent axon degeneration is implicated in neuroinflammatory models of glaucoma and ONC.^13,20,28 Therefore, we investigated alterations in SARM1 expression in our COH model. The band with higher molecular weight is representative of SARM1 expression, according to a previous study.²⁹ The level of SARM1 significantly increased 1 week after injection and remained elevated for 2 weeks compared to control eyes in the optic nerve (Figs. 1B, 1C). Additionally, we observed a slight decrease in retinal SARM1 expression at 3 days post-injection; however, it was restored by 1 week, as shown in Figures 1B and 1D. Furthermore, we examined the localization of SARM1 in the optic nerve and the retina. Immunofluorescent double labeling of SARM1 with a neuron cell marker (β3-tubulin) was performed on optic nerves. We found that SARM1 expression increased at both 1 and 2 weeks post-injection in the optic nerve compared to the control eyes. Partially colocalization with β3-tubulin confirmed that SARM1 was present in RGC axons (Fig. 1E). In the retina, SARM1, mainly localized in the ganglion cell layer (GCL). Three days post-injection in the COH model, SARM1 levels were reduced, as shown in Figure 1F. 

Phosphorylation of JNK Increased Early in the COH Model

To assess the extent of JNK phosphorylation in the COH model, we conducted an analysis of phosphorylated JNK (p-JNK) and total JNK expression in both the optic nerves and the retinas through Western blot assay. The phosphorylation of JNK expression in the optic nerves significantly increased 3 days after injection, which was earlier than the timing of SARM1 expression upregulation (Figs. 2A, 2B). Immunofluorescent analysis in the optic nerves further confirmed a significant increase in p-JNK expression compared to control eyes at 3 days post-injection, and the increased expression was maintained at 1 week post-injection. P-JNK was also identified partially co-labeling with β3-tubulin (Fig. 2C). Additionally, we observed an early elevation of p-JNK expression in the retinas 3 days after injection in the COH model. However, there were no significant differences in p-JNK levels between COH and control eyes at 1 and 2 weeks post-injection (Figs. 2D, 2E). Immunofluorescent analysis also confirmed the level of p-JNK increased significantly 3 days after injection mainly in the GCL (Fig. 2F). Notably, this temporal pattern of increased p-JNK expression coincided with decreased SARM1 protein levels. These findings suggest that JNK activation and changes in SARM1 protein expression occur concurrently during both optic nerves and retinal degeneration in the COH model. 

Figure 2.

 

Increased p-JNK expression in the COH model. (A) Western blot analysis of p-JNK expression in the optic nerves 3 days, 1 week, and 2 weeks after injection, with normalization to total JNK expression. (B) Quantitative analysis of p-JNK expression in the optic nerves. The two bands of p-JNK and total JNK were analyzed in combination. The level of p-JNK increased 3 days post-injection in the optic nerves and maintained for 1 week. The results are presented as the mean ± SEM; ***P < 0.001, **P < 0.01 indicate significant differences (n = 5–6 for each group). (C) The optic nerves’ longitudinal sections were immunostained with β3-tubulin Tubulin (green) and p-JNK (red) 3 days, 1 week, and 2 weeks after magnetic microbeads injection. The arrows indicated colocalization of β3-tubulin Tubulin and p-JNK. (D) Western blot analysis of p-JNK expression in the retinas at 3 days, 1 week, and 2 weeks post-injection, with normalization against total JNK expression. (E) Quantitative analysis of p-JNK expression in the retinas. The two bands of p-JNK and total JNK were analyzed in combination. The level of p-JNK increased 3 days post-injection in the retinas. The results are presented as the mean ± SEM; **P < 0.01 indicate significant differences (n = 5–6 for each group). (F) Retinal sections were stained with p-JNK (red) and DAPI (blue) 3 days, 1 week, and 2 weeks 3 days after magnetic microbeads injection. The level of p-JNK increased significantly 3 days after injection mainly in the GCL.

The Expression of SARM1 and p-JNK Increased in the I/R Injury Model

We further examined the expression profiles of SARM1 and p-JNK in the I/R injury model, utilized as an acute glaucomatous animal model. As depicted in Figures 3A and 3B, SARM1 expression exhibited a significant increase compared to control eyes after 24 hours of reperfusion in both the optic nerves and the retinas. Subsequently, there was a dramatic decrease in SARM1 levels after 48 hours of reperfusion due to severe damage caused by I/R injury. Elevated expression of p-JNK was observed in the optic nerves after 24 hours of reperfusion and in the retinas after 6 hours of reperfusion (Figs. 3C, 3D). Interestingly, a temporal trend was noted in retinal p-JNK expression, characterized by an initial increase followed by a subsequent decrease. Interestingly, the level of p-JNK after 48 hours of reperfusion showed no significant difference compared to that after 24 hours of reperfusion in the optic nerves. These findings suggest that excessive protein loss of total JNK in the optic nerves does not significantly alter the proportion of p-JNK within the total JNK. 

Figure 3.

 

Increased SARM1 and p-JNK expression in the I/R injury model. (A) Western blot analysis of SARM1 expression at 6, 24, and 48 hours post-reperfusion in the optic nerves and the retinas. (B) Quantitative analysis of SARM1 expression in the optic nerves and the retinas. The data were normalized to β-actin expression. SARM1 expression increased significantly both in the optic nerves and the retinas. The results are presented as the mean ± SEM; ***P < 0.001, **P < 0.01, and *P < 0.05 indicate significant differences (n = 4–5 for each group). (C) Western blot analysis of p-JNK expression at 6, 24, and 48 hours post-reperfusion in the optic nerves and the retinas. (D) Quantitative analysis of p-JNK expression in the optic nerves and the retinas. The data were normalized to total JNK expression. The two bands of p-JNK and total JNK were analyzed in combination. The p-JNK expression increased significantly both in the optic nerves and the retinas. The results are presented as the mean ± SEM; ***P < 0.001 indicate significant differences (n = 3–5 for each group).

Downregulation of SARM1 in the Optic Nerve and Retina was Achieved by Recombinant Adeno-Associated Virus - SARM1 shRNA – ZsGreen

To investigate the neuroprotective effects of reduced SARM1 expression on the survival of RGCs and optic nerves, an adeno-associated virus vector (AAV2-SARM1 shRNA-ZsGreen; AAV-SARM1-shRNA) was designed to downregulate SARM1 expression. Two weeks prior to magnetic microbeads injection, either AAV2-SARM1 or AAV2-ZsGreen Negative Control (AAV-NC) was injected, as depicted in Figure 4A. The transfection efficiency in rat retina was assessed by observing ZsGreen fluorescent protein expression in retinal flat mounts for each viral construct (Fig. 4B). At 4 weeks post-injection, both the optic nerves and the retinas exhibited a significant decrease in SARM1 mRNA levels compared to eyes injected with AAV-NC (Fig. 4C). Specifically, SARM1 mRNA levels were reduced to 56.7% in the axon and 60.1% in the retina. Consistent with the PCR findings, Western blot analysis revealed a corresponding reduction in SARM1 protein expression in the control group, with levels decreasing to 66.8% in the axon and 74.9% in the retina, relative to eyes injected with AAV-NC. In the COH group, following the administration of AAV-SARM1-shRNA, the protein expression further decreased to 49.5% in the axon and 58.1% in the retina compared to the COH eyes injected with AAV-NC (Figs. 4D, 4E). Immunofluorescence data also suggested decreased SARM1 expression in both the optic nerves and the retinas with AAV-SARM1-shRNA treatment (Figs. 4F, 4G). 

Figure 4.

 

Downregulation of SARM1 in the optic nerves and the retinas was achieved by AAV2- SARM1 shRNA – ZsGreen. (A) Timeline of the major procedures for the study. (B) Transduction profiles of AAV encoding ZsGreen expression throughout the entire retina four weeks post-AAV injection. (C) The RT-PCR analysis was performed to evaluate the expression of SARM1 mRNA in the optic nerves and the retinas following AAV-SARM1-shRNA or AAV-NC injection, with β-actin as a reference gene for normalization. The level of SARM1 mRNA significantly decreased with AAV-SARM1-shRNA treatment 4 weeks after AAV injection. The results are presented as the mean ± SEM; *P < 0.05 indicate significant differences (n = 3–5 for each group). (D) Western blot analysis of SARM1 expression in retinal and axon samples obtained from both the control group and the COH group. (E) Quantitative analysis of SARM1 expression in the optic nerves and the retinas. The results were normalized to β-actin expression. SARM1 expression decreased significantly both in the control group and the COH group after AAV-SARM1-shRNA injection. The results are presented as the mean ± SEM; ***P < 0.001, **P < 0.01, and *P < 0.05 indicate significant differences (n = 5 for each group). (F) The optic nerves’ longitudinal sections were immunostained with β3-tubulin Tubulin (gray) and SARM1 (red) treated with AAV-SARM1-shRNA or AAV-NC in the COH model. The arrows indicated colocalization of β3-tubulin Tubulin and SARM1. The level of SARM1 expression was significantly reduced by AAV-SARM1-shRNA. (G) Retinal sections were stained with SARM1 (red) and DAPI (blue) treated with AAV-SARM1-shRNA or AAV-NC in the COH model. The expression of SARM1 exhibited a significant reduction following treatment with AAV-SARM1-shRNA.

Downregulation of SARM1 Expression Conferred Significant Neuroprotection to Both Retinal Ganglion Cell Soma and Axons in the COH Model

The total number of surviving axons was quantitatively analyzed to assess optic nerve injury using toluidine blue stain. As shown in Figure 5A, the COH eyes injected with control AAV-NC exhibited numerous instances of axon loss, glial scarring, and myelin debris. Notably, there was a significant reduction in total axon counts in the COH eyes injected with AAV-NC (14.50 ± 1.23 axons/100 µm²) compared to the control eyes solely injected with AAV-NC (23.13 ± 0.62 axons/100 µm²). However, treatment with AAV-SARM1-shRNA led to a significant increase in total axon counts (20.72 ± 0.59 axons/100 µm²) in the COH eyes along with reduced gliosis and myelin debris formation (Fig. 5B). 

Figure 5.

 

Toluidine blue-stained optic nerve cross-sections showed axon degeneration. (A) The 100 × light microscopy images of optic nerve semi-thin cross-sections from the control group and 2 weeks after injection of magnetic microbeads with AAV-SARM1-shRNA or AAV-NC. The upper row scale bar = 50 µm and the bottom panel scale bar = 10 µm while showing the magnification of the upper row. The areas of glial scarring and myelin debris are indicated by red arrows. (B) Quantification of surviving axons in optic nerves of the COH model. Total axon counts increased significantly in the COH eyes injected with AAV-SARM1-shRNA. The results are presented as the mean ± SEM; ***P < 0.001 indicate significant differences (n = 5–6 for each group).

To evaluate RGC soma survival, RGC counts were performed on rat retinal wholemounts using Brn3A staining, as shown in Figure 6. In the central region of the eye, RGC density significantly increased to 2133 ± 86/mm² whereas it reached 1766 ± 86/mm² in the peripheral region of COH eyes treated with AAV-SARM1-shRNA; and the RGC density decreased to 1568 ± 84/mm² and 1383 ± 75/mm², respectively for the central and peripheral regions of COH eyes injected only with AAV-NC. 

Figure 6.

 

RGC soma surviving was performed by Brn3A staining in retina flat mount. (A) Representative retina flat mount images of central and peripheral regions exhibited Brn3A positive RGC soma in both the control group and the COH group treated with AAV-SARM1-shRNA or AAV-NC. Scale bar = 100 µm. (B) Quantification of surviving RGC soma in retina flat mount. Twelve microscopic fields in each retina were counted to calculate RGC density. RGC density increased significantly in the COH eyes with AAV-SARM1-shRNA treatment. The results are presented as the mean ± SEM; ***P < 0.001, **P < 0.01, and *P < 0.05 indicate significant differences (n = 5–6 for each group).

AAV-SARM1-shRNA treatment attenuated the damage to RGC axon and soma induced by elevated IOP, although there was no significant difference in RGC axon and soma between control eyes injected with only AAV-SARM1-shRNA or AAV-NC. These findings suggest that downregulation of SARM1 expression can mitigate injury to RGC axons and soma in a COH model. 

JNK Acted Downstream of SARM1 in the COH Model

In different axonal injury models, the upstream and downstream relationships between SARM1 and JNK were different.^19,20 The precise relationship between these two signaling molecules remains unclear in ocular hypertensive models. Therefore, we aimed to investigate the upstream and downstream interactions within our COH model. Western blot analysis of the optic nerves and the retinas revealed that AAV-SARM1-shRNA-induced downregulation of SARM1 significantly decreased p-JNK expression both in the control group and the COH group. However, knockdown of SARM1 did not affect dual leucine zipper kinase (DLK) expression, an upstream protein of JNK (Fig. 7). We also found that decreased p-JNK expression in both the optic nerves and the retinas following treatment with AAV-SARM1-shRNA by immunofluorescence analysis (see Figs. 7C, 7D). The optic nerves and the retinas harvested 3 days post-microbead injection were subjected to Western blot analysis for protein expression profiling. The level of p-JNK decreased significantly 3 days post-injection consistent with the reduction in SARM1 induced by AAV-SARM1-shRNA (Supplementary Fig. S1). These findings suggest that JNK acts as a downstream effector regulated by SARM1 in our COH model. 

Figure 7.

 

The p-JNK acted downstream of SARM1 in the COH model. (A) Western blot analysis of DLK and p-JNK expressions in retinal and axon samples treated with AAV-SARM1-shRNA or AAV-NC treatment in the COH model. (B) Quantitative analysis of DLK and p-JNK expressions in the optic nerves and the retinas. DLK expressions were normalized to β-actin expression, and p-JNK were normalized to total JNK expression. The two bands of p-JNK and total JNK were analyzed in combination. The p-JNK expression decreased significantly both in the control group and in the COH group consistent with SARM1 which decreased. The results are presented as the mean ± SEM; ***P < 0.001, **P < 0.01, and *P < 0.05 indicate significant differences (n = 5 for each group). (C) The optic nerves’ longitudinal sections were immunostained with β3-tubulin Tubulin (gray) and p-JNK (red) treated with AAV-SARM1-shRNA or AAV-NC in the COH model. The arrows indicate colocalization of β3-tubulin Tubulin and p-JNK. The modest level of p-JNK expression was significantly reduced by AAV-SARM1-shRNA. (D) Retinal sections were stained with p-JNK (red) and DAPI (blue) treated with AAV-SARM1-shRNA or AAV-NC in the COH model. The expression of p-JNK exhibited a significant reduction following treatment with AAV-SARM1-shRNA.

Discussion

SARM1 serves as a metabolic change sensor during the early stages of axon degeneration, and its deficiency has been confirmed to have neuroprotective effects in various neurodegenerative disorders.³⁰ Considering that axonal injury is believed to be an initial event in glaucoma progression,² it is crucial to investigate the role of SARM1 in different glaucoma models for potential intervention at early disease stages. Whereas previous studies have primarily focused on the protective effects of SARM1 deficiency in ONC models, further investigations should be conducted using intraocular hypertension models.^19,20 

In our study, we investigated the changes in levels of SARM1 and JNK in both chronic and acute glaucoma models. As depicted in Figures 1B and 1C, there was a significant increase in SARM1 expression within the optic nerves 1 week after microbead injection, which persisted for 2 weeks in the COH model. Conversely, 3 days post-injection, a decrease in SARM1 expression was observed within the retina. These findings suggest that following injury, SARM1 is activated and functions primarily within axons, leading to its elevated expression therein. Additionally, it appears that SARM1 expressed within the retina is transported to axons to exert its effects in the COH model; hence reduced retinal expression of SARM1 can be detected during early stages of injury, as shown in Figure 1D. Subsequently, over time, SARM1 levels gradually return to equilibrium during late-stage injury. In our acute model, SARM1 expression was found to be significantly upregulated at the 24-hour time point, as shown in Figures 3A and 3B, unlike in the COH model where there was no significant change in the early stage of axonal expression. The reason may be that in the acute model, IOP elevation is more intense, and then the stimulation is relatively large. This intense stimulus requires a rapid mobilization of substantial SARM1 expression. However, the transport of SARM1 from the retina to the optic nerve may be too slow to keep pace with the acute response. Consequently, the elevated expression of SARM1 in the retina and optic nerve was detected at the 24-hour time point. Notably though, no significant changes were observed in glaucomatous RGC soma with regard to SARM1 levels when using an ocular hypertension glaucoma model induced by silicone oil.¹⁵ These inconsistent results may stem from variations between different injury models and tissue collection time points. 

The level of JNK phosphorylation was investigated in our COH model (see Fig. 2). JNK signaling was found to be activated both in the optic nerves and the retinas, consistent with previous findings in various glaucoma-relevant injury models.^31–33 In our previous study, no significant changes were observed in the level of p-JNK at all time points after surgery.²⁴ The activation of JNK should be detected during the initial phase of the experimental model, such as 3 days post-injection, as demonstrated in this study. In addition, we also found that the timing of p-JNK expression upregulation in the optic nerves was earlier than SARM1 expression upregulation (see Figs. 2A, 2B). Elevated IOP leaded to the response of several upstream factors of p-JNK, resulting in the upregulation of p-JNK expression in our COH model. 

The protective effects of SARM1 deficiency on RGCs have been confirmed in various models of RGC injury.^14,15 However, the extent of protection provided by SARM1 inhibition varies across different models. In the silicone oil-induced ocular hypertension model and rotenone-induced retinal degeneration model, SARM1 knockout provides neuroprotection to both RGC somas and axons, whereas, in the ONC model, it only offers neuroprotection to axons.^14,15,20 Our findings also demonstrate that decreased expression of SARM1 induced by AAV-SARM1-shRNA injection has protective effects on both RGC somas and axons in a magnetic microbeads-induced COH model (see Figs. 5, 6). Recent studies have identified Ser-548 phosphorylation as a key regulatory site for SARM1’s NAD cleavage activity,³⁴ which is involved in mitochondrial transport along axons after ischemia/reperfusion injury.³⁵ These results suggest that phosphorylation of SARM1 at Ser-548 may play a role in axon degeneration. Further investigations are warranted to explore whether SARM1 phosphorylation is implicated in RGC death across different glaucoma models. 

The MAPK family plays an indispensable role in glaucomatous axonal injury signaling, particularly the JNK pathway and its upstream regulator DLK.^36,37 Previous studies have explored the connections among SARM1, DLK, and JNK in the ONC model.^19,20 In this study, we observed that JNK functions as a downstream mediator of SARM1. As depicted in Figure 7 and Supplementary Figure S1, there was a significant decrease in p-JNK expression following AAV-SARM1-shRNA injection. In the control group, p-JNK expression was not significantly different after AAV-SARM1-shRNA treatment, only the mean value was slightly downregulated (see Supplementary Fig. S1D). This could be due to the low basal expression of p-JNK in the control group and the brief period of AAV following injection. Meanwhile, there were no notable differences in DLK expression among the groups treated with AAV-SARM1-shRNA or AAV-NC. These findings suggest that DLK may serve as an upstream regulator of SARM1 in our COH model. 

In conclusion, we have demonstrated a significant increase in SARM1 levels in the optic nerves of both the COH model and the I/R injury model. Notably, downregulation of SARM1 expression provided substantial protection to RGC soma and axons in the COH model. Furthermore, our findings indicate that activation of MAPK signaling member JNK is dependent on SARM1 in the COH model. These results offer valuable insights into the potential therapeutic targeting of SARM1 for glaucoma treatment in clinical settings. 

Acknowledgments

Supported by the National Key Research and Development Program of China (Nos. 2020YFA0112703 and 2020YFA0112700), supported by grants from the National Natural Science Foundation of China (Nos. 82171055 and 82271085), sponsored by Program of Shanghai Municipal Commission of Science and Technology (21S11905900), and Xuhui Hospital and Regional Cooperation Project (23XHYD-28). 

Author Contributions: Jihong Wu, Xuejin Zhang, and Kaidi Wang conceived and designed the experiments. Xuejin Zhang, Ting Li, Junfeng Li, and Rong Zhang performed the experiments and provided the reagents/materials/analysis tools. Xuejin Zhang wrote the paper. All contributing authors read and approved the final version of the manuscript. 

Disclosure: X. Zhang, None; T. Li, None; R. Zhang, None; J. Li, None; K. Wang, None; J. Wu, None