The Role of Neuroinflammation in Glaucoma: An Update on Molecular Mechanisms and New Therapeutic Options

Glaucoma is a multifactorial optic neuropathy characterized by the continuous loss of retinal ganglion cells, leading to progressive and irreversible visual impairment. In this minireview, we report the results of the most recent experimental studies concerning cells, molecular mechanisms, genes, and microbiome involved in neuroinflammation processes correlated to glaucoma neurodegeneration. The identification of cellular mechanisms and molecular pathways related to retinal ganglion cell death is the first step toward the discovery of new therapeutic strategies. Recent experimental studies identified the following possible targets: adenosine A2A receptor, sterile alpha and TIR motif containing 1 (neurofilament light chain), toll-like receptors (TLRs) 2 and 4, phosphodiesterase type 4 (PDE4), and FasL-Fas signaling (in particular ONL1204, a small peptide antagonist of Fas receptors), and therapies directed against them. The continuous progress in knowledge provides interesting data, although the total lack of human studies remains an important limitation. Further research is required to better define the role of neuroinflammation in the neurodegeneration processes that occur in glaucomatous disease and to discover neuroprotective treatments amenable to clinical trials. The hereinafter reviewed studies are reported and evaluated according to their translational relevance.


INTRODUCTION
Glaucoma is a chronic and progressive optic neuropathy characterized by death of retinal ganglion cells (RGCs) with consequent localized or diffuse thinning of the nerve fiber layer and increased cupping of the optic nerve head (ONH) (1). It is a social disease; its prevalence is believed to grow strongly, and about 80 million glaucomatous patients are expected in 2020 and 112 million in 2040 (2,3). More than 7 million people worldwide are blind due to glaucoma, and the prevalence of bilateral blindness caused by this pathology varies from 6 to 16% in western countries (4).
The precise mechanisms leading to RGCs loss are not fully understood. Several studies have suggested that glaucoma has important analogies with other neurodegenerative pathologies correlated to inflammatory responses like amyotrophic lateral sclerosis, Alzheimer's disease, Parkinson's disease, Huntington's disease, and frontotemporal dementia (5)(6)(7)(8).
The purpose of this minireview is to discuss the consolidated knowledge of neuroinflammation in glaucoma and to focus on new experimental evidences about possible therapeutic targets and latest treatment proposals.
Recently, in a study in DBA/2J mice with experimental glaucoma, the triggering receptor expressed on myeloid cells/TYROsine kinase binding protein (TREM/TYROBP) signaling network has been identified as the principal regulator mechanism of microglial responses to elevated IOP; infiltrating monocyte-like cells are likely responsible for the early proinflammatory signals (33).
Microglial cells communicate with astroglia via signaling proteins. Depending on this interaction, astroglia are differentiated into two types of reactive astrocytes, A1 and A2 (34)(35)(36). A1 has a detrimental effect (37); on the other hand, A2 has a neuroprotective function (38). Indeed, microglia and astroglia collaborate to regulate the inflammation process.
Different genes are implicated in the inflammatory pathways and are upregulated in the retina and ONH (39,40). The first to be upregulated are TLR signaling pathways: for example, HSPs increase the expression of major histocompatibility complex (MHC) II and cytokine production (21). The second pathway is represented by nuclear factor-kappa B (NF-κB), which causes an increased expression of IL-1 cytokine family that promotes the cascade of inflammatory cytokines (TNF-α and IL-6). TNFα is found in the optic nerve in glaucoma patients (41)(42)(43) and also Fas ligand (FasL), a proapoptotic protein, both being implicated in glaucoma pathogenesis (44). Recently, Oikawa et al. (45) demonstrated in feline's early glaucoma the upregulation of genes related to cell proliferation and immune responses (linked with the TLR and NF-κB signaling pathway), and they observed that proliferating cell types are different in ONH sub-regions. Microglia-macrophages were found in the prelaminar region and in the lamina cribrosa, while oligodendrocytes were more numerous in the retrolaminar region.

MOLECULAR MECHANISMS
After reporting the fundamental concepts related to the neuroinflammation cells and the genes involved, we now refer to the latest studies on the main molecular mechanisms selected according to their presence in neurodegenerative diseases and their translational relevance.
The exosomes have a demonstrated role in neurodegenerative diseases. In an experimental glaucoma model (46), the exosomes, produced by naive microglia (BV-2 cells) in a condition of elevated hydrostatic pressure (BV-Exo-EHP, e.g., high IOP), induced the activation of retina microglia, production of cytokines, hypermotility, proliferation, and increase of phagocytosis. They promoted increase of reactive oxygen species and cell death, causing a reduction of retina ganglion cell number. The translation relevance of this study is that the exosomes own an autocrine function in the neuroinflammation pathway correlated to neurodegeneration.
Necroptosis, a recently discovered genetic form of cell death, has a fundamental role in neurodegenerative diseases (47). It is similar to necrosis with cell swelling, granular cytoplasm, chromatin fragmentation, and cellular lysis. Necroptosis differs from apoptosis because the cell content moves into the extracellular matrix in a passive way through the altered cell membrane. Necroptosis is induced by TNF-α and also by Fas and TNF-related apoptosis-inducing ligand (TRAIL), interferons (IFNs), TLR signaling, and viral infection via DAI (DNA sensor DNA-dependent activator of IFN regulatory factor). Ko et al. (47) in a neuroinflammation model of experimental glaucoma proved that the axon degeneration is sterile alpha and TIR motif1 (SARM1)-dependent and is induced by TNF-α with a consequent oligodendrocyte loss and RGC death. The necroptosis perpetrator is mixed lineage kinase domain-like pseudokinase (MLKL) through the reduction of the axon survival factors nicotinamide mononucleotide adenylyltransferase 2 (NMNAT2) and stathmin 2 (STMN2) that inhibit SARM1 NADase action. TNF-α also activates SARM1dependent axon degeneration in sensory nerve cells through a different necroptotic signaling mechanism.
Several recent studies have dealt with the topic of SARM1 axon degeneration pathway. The axon health is preserved by the balance between the pro-survival NMNAT2 and STMN2 and pro-degenerative molecules dual leucine zipper kinase (DLK) and SARM1 (48). Activated DLK reduces the concentration of protective factors (NMNAT2 and SCG10) and exposes axons to traumatic and metabolic damages; this event is also related to mitochondrial dysfunction that in an independent way reduces the NMNAT2 and SCG10 concentrations in the axons (49). The equilibrium between axon survival and selfdestruction is strictly related to axonal NAD+ metabolism. Sasaki et al. (50) studied in cell cultures and in vivo the biomarker cADPR, which controls NAD+ levels via SARM1 and mobilizes calcium. SARM1 has a basal activity in normal conditions. After injury, the axon degeneration is preceded by SARM1-dependent rise in the amount of axonal cADPR, but the contribution of cADPR in degenerative mechanisms has not been proven. The mitochondrial dysfunction can produce an incomplete activation of SARM1 and could predispose the axons to neurodegeneration.
In the context of neuroinflammation, it is mandatory to report the interesting developments that have taken place in recent years regarding the relationship between microbiome and glaucoma. Chen et al. conducted a study in germ-free mice, demonstrating that the absence of gastrointestinal (GI) bacteria abolished the development of glaucoma (51). The immune mechanism involved in the pathogenesis of the neural damage is related to CD4+ T-lymphocytes that recognize HSPs. Bacterial HSPs crossreacted with both mouse and human HSPs. High IOP values induced the passage of CD4+ T-cells in the retina, and these Tcells were to blame for the extended phase of neurodegeneration in glaucomatous disease. They recognized specifically both bacterial and human HSPs. The induction of neurodegeneration in glaucoma requires pre-exposure to microbial flora, either the GI and oral one (51). In the past century, the presence of elevated titers of antibodies directed against small HSPs and cross-reacting with human HSPs was demonstrated in glaucoma patients  (52). It was reported that these antibodies exert cytotoxicity when directly applied to human retina (53). It was proposed that immunomodulation should become the basis of glaucoma therapy. However, the immunosuppressive therapy could be dangerous and sometimes burdened by serious side effects. One solution could be the identification of antigens triggering or enhancing the autoimmune mechanism(s) and their elimination, aiming at reducing the active aggression by autoimmune T-cells.
Since the development of glaucoma was shown to be linked to the microbiome, some of its components are potentially curable, such as Helicobacter pylori, a Gram-negative flagellar bacterium present worldwide, in 20-90% of individuals. Only active H. pylori infection induces cellular immune responses against the nervous system, due to molecular mimicry and cross-reactivity with components of host nerves. A large meta-analysis was recently published by Doulberis et al. (54). When the diagnosis of H. pylori infection was performed by gastric biopsy, the odds ratio (OR) for an association was very high (5.4), with confidence interval of 3.17-9.2 (highly significant); the strong association diminished to an OR of 2.08 when only serum antibodies were measured. Similarly, also in glaucoma, the antigens recognized by the immune system were reported in quite a large number. Geyer and Levo (55) suggested that the therapy should be directed not only against the IOP but also against autoreactive lymphocytes, on the basis of the lack of neurodegeneration in experimental mice models after depletion of either B-cells and in particular T-cells directed against HSP-derived peptides (55). However, this might not be sufficient. Beutgen

NEW POSSIBLE TARGETS AND PROPOSALS ON PHARMACOLOGICAL THERAPIES
Hereinafter, the possible new therapeutic targets to reduce or block the neuroinflammation, which are more likely to be used in future clinical practice, are outlined. They are listed in alphabetical order. In Table 1, we provided more detail on new targets, their neuroinflammatory and neurodegenerative effects, and, where provided, the therapeutic options.

Astroglial Nuclear Factor-Kappa B
The astroglial NF-κB was evaluated as a possible treatment target for the modulation of immune response in an experimental transgenic glaucoma mouse with deletion of IkappaB kinase beta (IκKβ) in astroglial cells (59). The study showed a reduction of the increase in proinflammatory cytokines (in particular  Frontiers in Neurology | www.frontiersin.org TNF-α) with consequent protection of axon from degeneration and RGCs from apoptosis, as demonstrated also by pattern electroretinogram (PERG) data.

Endothelin-1
Since 2012 (71), it was clarified that the increment of endothelin-1 (ET-1) in glaucomatous patients is linked to sub-clinical inflammation. Nor Arfuzir et al. (60), using Sprague Dawley rats that underwent intravitreal injection of ET-1, proved that the pretreatment with magnesium acetyltaurate (MgAT) was protective against the increase induced by ET-1 in the retinal expression of IL-1β, IL-6, and TNF-α and against the activation of NF-κB and c-JUN induced by ET-1. MgAT promoted a higher RGC survival.

FasL-Fas Signaling
FasL, a type II transmembrane protein of the TNF group, promotes apoptosis after binding to the Fas receptor and is involved in the pathogenesis of glaucoma also through inflammatory pathways. Krishnan

Monocyte-Like Cells
Recently, in a mouse model of ocular hypertension (DBA/2J eyes), it was proved that monocyte-like cells enter the ONH and that monocyte-platelet interactions occur in glaucomatous tissue (62). Using these monocyte-like cells as therapeutic target, the authors demonstrated that the treatment with DS-SILY, a peptidoglycan that hinders the platelet adhesion to the vessel endothelium and to monocytes, and the treatment with genetic targeting of Itgam (CD11b), an immune cell receptor that blocks effusion of the monocytes from the vessels, can both be neuroprotective by reducing neuroinflammation.

Phosphodiesterase Type 4 Signaling
Cueva Vargas et al. (64) evaluated the anti-neuroinflammatory activity of ibudilast, a clinically approved cAMP phosphodiesterase (PDE) inhibitor with special affinity for PDE type 4 (PDE4). In an OHT rat model, the intraocular administration of ibudilast decreased the production of proinflammatory cytokines and reduced macroglia and microglial reactivity in the retina and optic nerve. Ibudilast had a positive effect on RGC soma survival, avoided axonal degeneration, and enhanced anterograde axonal transport in POAG eyes through activation of the cAMP/PKA pathway.

Sterile Alpha and TIR Motif Containing 1
Since SARM1 induces Wallerian degeneration, this pathway could be a therapeutic target (65). The SARM1 inhibition prevents axonal degeneration in traumatic injuries and neurodegenerative disorders. As reported before, SARM1 is involved in innate immune response, and it is possible that the relationship between immune regulation and neurodegenerative disorders will play an important role for new therapeutic possibilities. The discovery of a reliable biomarker of axonal damage, i.e., the neurofilament light chain (NfL) and the possibility of testing NfL in plasma or serum, could be an important step to develop SARM1 inhibitors that protect axons from degeneration (66).

Toll-Like Receptors 2 and 4
As reported, TLR signaling is involved in homeostasis and in pathology of CNS. TLR2 and TLR4 expressed by microglia take part in the glial response and in the neuroinflammation. Yang et al. (67) demonstrated that in C57BL/6J mice, the alpha 1-antitrypsin (AAT), a serine protease inhibitor, blocks microglial activation in chronic ocular hypertension model. It was shown that AAT inhibits leukocyte migration and has antithrombotic, antiapoptotic, and anti-inflammatory properties (68). In OHT rats, Ji et al. (69) demonstrated that the human umbilical cord mesenchymal stem cell (hUC-MSC) transplantation blocks TLR4-related microglial activation and neuroinflammatory pathways.

NEW PROPOSAL HOLISTIC APPROACHES
In this section, we reported studies proposing holistic medicinal treatments to modulate/reduce the neuroinflammation in an experimental model of glaucoma. Table 2 illustrates new therapies with details on modulations of inflammatory mechanisms.

Antioxidants
Oxidative stress can alter the immune function of the glia and promotes the neuroinflammation in glaucoma. The effects of the antioxidant Tempol were tested on ocular hypertensive retina and optic nerve samples and on NF-κB, a redox-sensitive transcriptional regulator of neuroinflammation. The analysis of markers of oxidative stress (proinflammatory cytokines, including IL-1, IL-2, IFN-γ, and TNF-α) demonstrated that the treatment was able to decrease the neuroinflammation markers (72).

Ketogenic Diet
A recent research study (73) evidenced that the ketogenic diet reduces inflammation through inhibition of AMPK activation and HCAR1-mediated inhibition of the NLRP3 inflammasome. The way by which ketogenic diet works in neuroprotection is not completely known up to now. Researches supposed that ketogenic diet reduces oxidative stress, inhibits class I histone deacetylases, promotes Nrf2 activation to upregulate antioxidants, and inhibits NF-κB to reduce inflammation (74,75).

Coriolus and Hericium Nutritional Mushrooms
Trovato Salinaro et al. (76) reported the potential effect of Coriolus and Hericium, nutritional mushrooms, in the therapy of neurodegenerative diseases including Alzheimer's disease and glaucoma; they act as modulators of mechanisms of cellular protection (antioxidant properties) from mitochondrial dysfunction and neuroinflammation.

Hydrophilic Saffron Extract
Studies conducted (77) in a mouse model of laser-induced unilateral OHT on a hydrophilic saffron extract (crocin 3%) demonstrated the reduction of morphological features of microglial activation in OHT and in contralateral eyes. The treatment with saffron extract in part inverted the downregulation of P2RY12. The saffron extract administrated orally protected RGCs from death by reducing neuroinflammation correlated with increased IOP.

DISCUSSION
The co-participation of the innate immune response and the inflammation in the pathogenesis of optic nerve degeneration in glaucomatous disease is now an established and proven event (78)(79)(80)(81)(82). The purpose of this minireview was to outline potential targets for future therapies, some of which are already being studied, and to describe possible treatments related to alternative medicine. These recent studies performed on animal models have improved our knowledge on cellular mechanisms, partly already known, but mainly on the signaling pathways of neuroinflammation. The studies analyzed openly the way to new possibilities of modulation of neuroinflammation (such as Frontiers in Neurology | www.frontiersin.org antagonist of A 2A R, stem cells, MgAT, resveratrol, and alpha 1-antitrypsin) with a view to achieve effective neuroprotective treatments for glaucomatous disease. An important limitation is the almost total lack of researches in human. On the basis of the data reported, the possible ways of developing future research are illustrated in Figure 1. Further studies are needed to transfer the current experimental evidence into clinical research in order to slow down or to prevent the progressive glaucoma neurodegeneration.

AUTHOR CONTRIBUTIONS
All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.