Introduction
Glaucoma is a chronic neurodegenerative disease that is characterized by progressive vision loss and changes to the optic nerve head [
50]. It affects more than 70 million individuals globally, making it one of the leading causes of irreversible blindness worldwide [
50]. High intraocular pressure (IOP) is a main risk factor for the development of the disease and the progression of vision loss, however there are indications that glaucoma is a multi-factorial disease. Factors such as immune responses, mitochondrial dysfunction, and age contribute to the loss of retinal ganglion cells (RGC), and it is likely that their respective contributions to RGC damage vary during glaucoma pathogenesis [
40,
50].
A number of recent studies have indicated that metabolic crisis and oxidative stress can cause RGC dysfunction in glaucoma and that cellular NAD+ levels may be a critical determinant of the cells’ survival [
21,
28,
36,
40,
44]. Low levels of NAD+ can be caused by several factors, including retinal or systemic metabolic dysfunction, but axonal injury or RGC stress independently cause downregulation of the enzyme NMNAT2, which functions to convert nicotinamide mononucleotide (NMN) to NAD+ [
10,
14,
21,
26,
29]. Conversely, providing NAD or its precursors has been shown to attenuate RGC loss in several animal models of glaucoma [
43,
51].
Low levels of NAD+ also activate the enzyme sterile alpha and TIR motif-containing 1 (
Sarm1) that cleaves NAD+ into ADP-ribose, cyclic ADPR, and nicotinamide through dimerization of its TIR domain [
9,
42]. Upon SARM1 activation, remaining axonal NAD+ is quickly degraded leading to axonal degradation through Wallerian degeneration and subsequent neuronal death [
22,
39,
48]. Furthermore, SARM1 activation can exacerbate axonal degeneration by promoting neuroinflammation through NFκB signaling [
19,
25,
30,
54]. Consistent with its role in axonal degradation, studies have shown that inhibition of SARM1 activity can be neuroprotective in rodent models of neurodegenerative conditions, including traumatic brain injury and experimental autoimmune encephalomyelitis (EAE) [
2,
4,
13,
23,
31,
34,
45,
54].
The role of SARM1 in injuries affecting ocular neurons is less well established. Deletion of
Sarm1 has been shown to be protective to neurons following optic nerve crush in some studies, but not in others [
13,
31]. A protective effect has also been reported following intravitreal injection of TNF-alpha and acute elevation of IOP [
25,
31]. However, the chronic nature of glaucoma is a defining characteristic of the disease and a challenge for its management. In this study we sought to determine if loss of
Sarm1 protects RGC morphologically and functionally from glaucomatous damage due to chronic IOP elevation. Our data demonstrate that deletion of the
Sarm1 gene provides significant reduces RGC loss in mice experiencing 16 weeks of moderately elevated IOP. Finally, we demonstrate that naive
Sarm1 knock out (KO) mice display moderate deficiencies in visual acuity, indicating a role of this gene during ocular development.
Materials and methods
Animals
Mice used in this study were either normal C57BL/6J (B6) or homozygote C57BL/6J-Sarm1em1Agsa/J (Sarm1 KO, Jackson Laboratory, stock #034399, Bar Harbor, ME, USA). Mice were housed in the animal facility of the Iowa City Veterans Affair Healthcare System in a 12 h light–dark cycle and were fed chow and water ad libitum. Both male and female mice were used for this study. All studies were approved by the University of Iowa and Iowa City VA Committees for Animal Care and Use and conducted according to the ARVO statement for the use of Animals in Ophthalmology and Vision Research.
Western blot analysis of endogenous SARM1
Mouse brains from transgenic animals were lysed by dounce homogenization in cold RIPA buffer (50 mM Tris–HCl pH7.4, 1 mM EDTA, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, and protease inhibitor cocktail from Prometheus). Brain extracts were sonicated for 20 s then centrifuged at 5000×g for 5 min at 4 °C. Supernatants were collected and subjected to a second round of centrifugation to remove excess cell debris. The supernatant was mixed with laemmli buffer and 15 µg protein separated by SDS-PAGE. Endogenous Sarm1 and β3-tubulin were detected by western immunoblotting (anti-Sarm1, 1:500, Biolegend # 696602; anti-β3-tubulin, 1:5000, BioLegend #657409). Primary antibodies were detected with a DyLight 680-conjugated goat anti-mouse secondary (1:5000, Invitrogen #35519) and visualized with a LiCor Odyssey Fc imaging system.
Immunohistochemistry
Optic nerves were dissected immediately following euthanasia and fixed in 4% paraformaldehyde. Nerves were embedded in paraffin and longitudinal sections of 5 micron thickness were prepared. Sections underwent 3 min of heat-induced epitope retrieval (Decloaking Chamber, NxGen) in sodium citrate buffer (10 mM Sodium Citrate, 0.05% Tween 20, pH 6.0). Slides were then blocked with 1% BSA in TBS for 2 h at room temperature and incubated overnight at 4 °C with a 1:100 dilution of the primary antibody (rabbit anti-SARM1 IgG, Cell Signaling #13022, Danvers, MA). Sections were then rinsed and incubated with a 1:1000 dilution of the secondary antibody for two hours (Donkey anti-Rabbit IgG Alexa Fluor® 555, Abcam, Waltham, MA). DAPI was included in all stains to facilitate orientation.
Induction of elevated IOP
To induce elevated IOP in mice, an adenoviral vector was used that expresses a pathogenic form of human myocilin Ad5RSVmyocilin
Tyr437HisFlag (Ad5myoc, University of Iowa Viral Vector Core, Iowa City, IA, USA), as described previously [
18,
53]. Newborn (P2–P5) B6 and
Sarm1 KO mice were subcutaneously injected with Ad5myoc (3 µL of 3 × 10
6 PfU) to promote tolerance to the vector and limit potential ocular inflammation. At eight weeks of age, mice were anesthetized using 4% isoflurane for induction and 2.5% isoflurane for maintenance with a flow rate of 1L oxygen/min, and 3 × 10
8 PfU virus in 1 µL of PBS was deposited in the anterior chamber of both eyes via transcorneal injection. 0.5% proparacaine hydrochloride (Bausch + Lomb, Bridgewater, NJ, USA) and 1% tropicamide (Sandoz, Princeton, NJ, USA) eye drops were administered during this process for local anesthesia and dilation. After injection the needle was kept in place for 30 s in order to prevent washout. Both coloration and pulsation of the eye vasculature were monitored throughout the injection to ensure that no ischemic injury occurred. Control mice received an injection of an equivalent amount of sterile PBS.
IOP in anesthetized mice was determined using a rebound tonometer (Tonolab, Colonial Medical Supply, Windham, NH, USA) described previously [
24]. IOP measurements were taken between 10 AM and 1 PM by an investigator blinded to the animals’ status.
Imaging of the optic nerve and measurement of axonal size
Immediately following euthanasia, the optic nerve was removed and fixed in 2% glutaraldehyde and 2% paraformaldehyde. The nerves were then post-fixed in osmium tetraoxide and embedded into Eponate 12 resin. 1 μm thick sections of the optic nerve were taken perpendicular to the length of the nerve. Sections were then stained using 1% p-phenylenediamine (PPD) and images were taken at 100× magnification using a Zeiss Axioscope 5 microscope.
Measurement of optokinetic reflex (OKR)
Visual acuity in mice was measured using an OptoDrum (StriaTech, Tübingen, Germany). The animal is placed inside on a platform and computer screens surrounding the platform will display a rotating stripe pattern. The reflexive head movements of the mice were recorded by a camera and the system’s software determined if the response of the animal tracked with the pattern it was shown. The spatial frequency of the displayed pattern is either increased or decreased in small increments to determine the level of visual acuity in cycles/degree (c/d) at 99.8% contrast. If the animal is unable to track the rotating stripe pattern by turning their head, it indicates that the pattern is not perceived by the animal and helps to define the level of visual acuity in said animal. Multiple trials are run in each session to ensure the proper analysis of visual acuity by the system. The researcher conducting each test was blinded to the background and status of the mice tested.
Quantification of RGC density
To analyze RGC density, mice were first euthanized, enucleated, and eyes were fixed in 4% paraformaldehyde for two hours. After fixation, retinas were dissected and placed in 0.3% Triton-X100 in PBS for 6 h. Following three freeze and thaw cycles, retinas were blocked using 1% BSA/0.3% Triton-X100 in PBS for one hour at room temperature. Retinas were then incubated in goat anti-Brn3a primary antibody (Santa Crux, TX, diluted 1:200 in 1% BSA/0.3% Triton-X100/1% DMSO/PBS) at 4 °C for 48 h on a rocker platform. Retinas were washed thoroughly in PBS and binding was visualized after incubation in a donkey anti-goat Alexa Fluor 546 secondary antibody solution (Invitrogen, Carlsbad, CA, USA, 1:200 in PBS) for 3 h at room temperature in the dark. Retinas were again washed in PBS and then whole-mounted using Vectashield (Vector Laboratories, Burlingame, CA, USA). 24 images were taken of each retina at predetermined locations peripherally, mid-peripherally, and centrally at a magnification of 20× using an Olympus BX41 microscope (Olympus, Center Valley, PA, USA). Each image is 330 × 438 μm or 0.1445 mm2, thus all 24 images taken together account for 2.89 mm2 or approximately 25–30% of the total mouse retinal area. Brn3a+ cells were counted manually by a blinded investigator by using the cell counter plugin for ImageJ software (National Institutes of Health, Bethesda, MD, USA). RGC density per mm2 was then calculated by averaging the cell counts of each area of the retina.
Statistical analysis of data
All calculations were computed using GraphPad Prism 10.0.2 (GraphPad Software, San Diego, CA, USA). P values of < 0.01 were considered statistically significant for this study. Comparisons between more than two groups were carried out using ANOVA and Tukey’s multiple comparison test. All data is presented as a mean of animals tested.
Discussion
Therapeutic options for glaucoma are currently limited to reducing IOP, but recent data suggests that modulation of intracellular NAD+ levels can provide neuroprotection in the disease.
Sarm1 is a significant regulator of NAD+ levels, suggesting that it may be a viable target for pharmaceuticals exploration. Since studies in other model systems have indicated a role of SARM1 during embryonic development, our initial efforts were directed to assess the functional and morphologic effects of
Sarm1 ablation on the visual system [
5,
48]. During the development of the visual system as much as half of the RGC die after failing to establish connections to the superior colliculus [
3]. While it is conceivable that blocking a pathway that contributes to axonal degradation could result in more surviving neurons, we did not observe increased numbers of RGC or optic nerve axons in
Sarm1 KO mice, indicating that this enzyme is not required for this process. We did, however, notice a reduction in the average axon diameter in knockout animals. Smaller axonal size is generally correlated with slower conduction velocity, but experimental verification of this notion was outside the scope of this study [
27]. We also noticed reduced optokinetic responses in knockout animals when compared to wild-type mice. Whether this is related to the reduced caliber of optic nerve axons remains to be determined. Axon diameter can influence the timing of signal conduction, whereas the optokinetic response involves a complex integration of visual input, central processing, and motor output making the relationship between axon diameter and the optokinetic response not straightforward [
38]. One possible explanation for reduced visual function may be that RGC dendritic arbors in
Sarm1 KO mice appear to exhibit a reduced complexity compared to wildtype counterparts [
6]. This could potentially lead to changes in their receptive fields and account for the observed attenuation in optokinetic responses.
Our data further indicate that, within the optic nerve, SARM1 is primarily expressed by oligodendrocytes. Non-neuronal expression of SARM1 has been reported in astroglia [
23,
30], and in both cultured and in vivo oligodendrocytes [
11]. Using identical experimental procedures, we did not observe SARM1 immunoreactivity in the neural retina, which does not contain oligodendrocytes. Considering the established involvement of SARM1 in axonal degeneration, this finding was unexpected. Within the retina, prior investigations that have either indirectly indicated SARM1 expression by RGCs [
31,
49,
52] or have directly demonstrated its presence within retinal homogenates [
33]. Furthermore, single cell transcriptomic data from several studies indicate that
Sarm1 is expressed at low levels by RGC and photoreceptor cells [
7,
46,
47]. Thus the absence of distinct immunoreactivity in our experiments may be indicative of low Sarm1 levels in RGC, but could also be related to the shortcomings inherent to immunohistochemistry, including epitope availability within the retina.
A main finding of our study is that Sarm1 ablation confers significant neuroprotection in chronic glaucoma. In our glaucoma model, control mice lose between 19 and 14% of RGC after 4 months of moderately elevated IOP and display a significant decline of visual acuity. In contrast, Sarm1 KO mice exposed to the same level and duration of IOP elevation, loose only approximately half the number of RGC and maintain visual function similar to baseline levels.
These findings agree with the emerging concept that NAD+ metabolism is a crucial determinant of RGC survival. It is likely that NAD+ levels are decreased in the glaucoma retina and that the activity of the NAD+ synthase NMANT2 decreases in stressed RGC. The physiologic role of SARM1 is to further deplete NAD+ cellular stores to cause rapid degradation and removal of damaged axons. Removal of
Sarm1 most likely confers protection of RGC due to the persistence of higher NAD+ levels due to the absence of the NADase activity of the enzyme [
48]. The data further imply that in the absence of SARM1 NAD+ levels in the glaucomatous optic nerve remain sufficiently high to prevent the activation of axon destructive cascades.
Alternatively, it is possible that the observed reduction of RGC in eyes with glaucoma is the result of decreased neuroinflammation in
Sarm1 KO mice. It has been firmly established that retinal inflammation is a significant contributor to RGC damage and axonal loss [
40].
Sarm1 is not only activated by neuroinflammatory signals, but also further promotes inflammation through NF-κB pathway signaling [
19,
30,
54]. Deletion of
Sarm1 should lessen NF-κB signaling, resulting in reduced neuroinflammation and enhanced neuronal survival [
30,
54].
Although
Sarm1 KO mice clearly experienced less RGC loss than control mice in this glaucoma model, they are not completely protected, underscoring the fact that glaucoma is a multi-factorial disease [
50]. Factors impacting the onset and progression of vision loss involve diverse pathomechanisms, including RGC mitochondrial dysfunction and oxidative stress [
28,
36], vascular factors [
1], and autoimmune processes [
17,
32,
35]. Many of these likely persist and will cause damage to RGC even if
Sarm1 disruption prevents the initiation of Wallerian degeneration.
Our data indicating that loss of
Sarm1 affords protection of RGC in glaucoma agree with earlier studies that employed acute injury models to cause RGC death [
13,
25,
31]. Importantly, the findings of our research indicate that the protective benefit can be sustained over several months despite unabated IOP elevation, a critical aspect for therapeutic efficacy given the chronic nature of glaucoma [
37]. Thus
Sarm1 deletion or inhibition may effectively reduce RGC loss in glaucoma and, given the scarcity of treatments for glaucoma, the potential of
Sarm1 as a therapeutic target is intriguing. Potent inhibitors of
Sarm1 have recently been introduced and may provide viable pharmacologic options [
4,
12,
20]. However, chronic systemic suppression of
Sarm1 activity may give rise to undesirable side effects in patients and it is conceivable that gene therapy targeted to the eye may be a more effective strategy [
8]. Gene therapy designed to reduce
Sarm1 expression has been successful in some experimental models and may be aided by the fact that even a partial reduction of transcript levels is sufficient to delay axonal degeneration [
10,
15,
16,
49].
On the other hand, inhibition of
Sarm1 does not address the metabolic dysfunction that causes its activation and subsequent axon loss. It is possible that supplementing patients with NAD+ or its precursors will restore metabolic balance and prevent activation of
Sarm1 [
14,
41,
43,
51]. Alternatively, it may be possible to increase cellular NAD+ levels through gene therapy with NMNAT2, to compensate for its reduced expression levels [
10].
Taken together, our data demonstrate that
Sarm1 ablation reduces RGC loss in a rodent glaucoma model with chronic elevated IOP. The profound and sustained effect observed in the model paves the way for further exploration of the potential benefits for the treatment of human glaucoma. Concurrently, our results underscore the potential of NAD+ supplementation as a viable therapeutic strategy to preserve NAD+ homeostasis and prevent
Sarm1-mediated axonal and RGC degradation [
43,
51]. Our data elucidate a pivotal role for
Sarm1 in RGC vulnerability in glaucoma and enhances our comprehension of its long-term effects within a chronic rodent model reflective of human pathology.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.