Therapeutic strategies for glaucoma and optic neuropathies

Abstract

Glaucoma is a neurodegenerative eye disease that causes permanent vision impairment. The main pathological characteristics of glaucoma are retinal ganglion cell (RGC) loss and optic nerve degeneration. Glaucoma can be caused by elevated intraocular pressure (IOP), although some cases are congenital or occur in patients with normal IOP. Current glaucoma treatments rely on medicine and surgery to lower IOP, which only delays disease progression. First-line glaucoma medicines are supported by pharmacotherapy advancements such as Rho kinase inhibitors and innovative drug delivery systems. Glaucoma surgery has shifted to safer minimally invasive (or microinvasive) glaucoma surgery, but further trials are needed to validate long-term efficacy. Further, growing evidence shows that adeno-associated virus gene transduction and stem cell-based RGC replacement therapy hold potential to treat optic nerve fiber degeneration and glaucoma. However, better understanding of the regulatory mechanisms of RGC development is needed to provide insight into RGC differentiation from stem cells and help choose target genes for viral therapy. In this review, we overview current progress in RGC development research, optic nerve fiber regeneration, and human stem cell-derived RGC differentiation and transplantation. We also provide an outlook on perspectives and challenges in the field.

1. Introduction

Glaucoma, a group of progressive optic neuropathies, is the second most prevalent eye disease leading to blindness globally, followed in impact only by cataracts (Kingman, 2004). Approximately 80 million people worldwide have glaucoma (Tham et al., 2014). Glaucoma is primarily categorized into two subtypes: open-angle glaucoma, and angle-closure glaucoma. Both open-angle and angle-closure glaucoma can be further divided into primary and secondary glaucoma (Allison et al., 2020). Primary open-angle glaucoma (POAG) is the most common glaucoma type, particularly affecting aging individuals (Quigley, 1996; Shastry, 2013) and accounting for ~90% of glaucoma cases in the United States and ~74% of cases globally (Quigley and Broman, 2006; Reeder et al., 2008). Dysfunction in the ocular drainage system is not the only characteristic of POAG - while an elevated IOP is the major risk factor for POAG, vascular impairment of ocular blood has also been shown to play a role in the pathogenesis of POAG in multiple studies. Elevated IOP is a major risk factor for POAG progression, and current medical therapy is limited to lowering IOP (Pang and Clark, 2020). Although glaucoma is mostly caused by IOP elevation, some cases are congenital (Kaur and Gurnani, 2023; Nutt et al., 2023) or occur in patients with normal IOP (Fox and Fingert, 2023; Leung and Tham, 2022). Despite significant clinical and basic scientific advancements, glaucoma remains a major vision-threatening disease. Further understanding of the disease pathobiology is urgently needed to develop more efficient drugs to prevent vision loss.

Death of retinal ganglion cells (RGCs) is a primary factor in glaucoma development and is considered the main underlying cause of visual loss in glaucoma patients (Davis et al., 2016). RGCs are the sole transmitters of visual information from the retina to the brain. Therefore, death of RGCs disrupts transmission of visual information, impairing visual function. However, lowering IOP may be insufficient to prevent glaucoma progression or RGC loss in some patients (Kass et al., 2002). Therefore, it is crucial to incorporate neuroprotective agents as adjunctive therapy in combination with existing glaucoma treatment options. Further, a research focus on promoting RGC survival and function could help develop novel effective strategies to preserve vision.

This review provides an overview of developmental aspects of RGCs, focusing on key transcription factors involved in their regulation. Subsequently, we summarize the experimental models used to investigate prophylactic and interventional therapeutic approaches to preserve RGC integrity. Moreover, we explore current advancements in therapeutic approaches, including potential application of replenishment cell therapy. Because stem cells can differentiate into neuronal cells including RGCs, we further overview current literature on human stem cell-derived RGC transplantation. Lastly, we discuss the clinical aspects of this research, highlighting both promising aspects and challenges that need to be addressed.

2. RGC development

Retinogenesis is a complex process of coordinated steps that ultimately give rise to six neuronal cell types and one glial cell type in a conserved overlapping temporal order (Cepko et al., 1996; Livesey and Cepko, 2001). The first differentiation wave occurs during embryonic development, leading to formation of RGCs, followed by horizontal cells, amacrine cells, and cone photoreceptors. Retinal development continues after birth, resulting in genesis of rod photoreceptors, bipolar cells, and finally Muller glia.

The gene regulatory network underlying RGC development has been intensively studied, revealing numerous transcription factors that are essential for RGC genesis and optic nerve growth (Fig. 1). At the center of this network is the proneural atonal BHLH transcription factor 7 (ATOH7), which is expressed in retinal precursor cells around embryonic day 11 (E11)–E14.5 of the mouse retina and gradually decreases expression afterward (Brown et al., 1998; Brzezinski et al., 2012; Pan et al., 2008). ATOH7 is essential for RGC differentiation and optic nerve formation in mice, zebrafish, and humans (Brown et al., 2001; Kay et al., 2001; Keser et al., 2017; Kondo et al., 2018; Wang et al., 2001). As RGC differentiation proceeds, a cascade of downstream transcription factors is expressed including POU4F1/2, ISL1, SOX4, and SOX11. Given the central role of ATOH7 in RGC genesis, we next discuss relevant recent work.

Figure 1.

Schematic diagram of gene regulatory cascade involved in retinal ganglion cell (RGC) development. Various transcription factors specify RGCs from early retinal progenitor cells (RPCs) to mitotic ganglion cells. RGCs initially differentiate under the control of the retinal progenitor transcription factor PAX6. This leads to expression of the bHLH transcription factor ATOH7 in retinal precursors. ATOH7 triggers an RGC gene regulatory program that includes expression of transcription factors SOX4, POU4F2, ISL1 and SOX11, all required for RGC terminal differentiation.

2.1. Role of ATOH7 in RGC development

ATOH7 has been widely considered essential for establishing the RGC lineage during retinal development. However, a single-cell transcriptomics study using Atoh7-knockout mice showed that ATOH7-expressing cells are not restricted to RGC genesis, as ATOH7 is usually expressed in a cluster of retinal progenitor cells expressing transcription factors that promote other lineages such as NeuroD1 and Neurog2 (Wu et al., 2021a). Further, single-cell RNA sequencing indicates that the RGC genetic program is normally initiated upon Atoh7 loss but fails to proceed in later stages, suggesting an independent pathway can promote RGC formation (Wu et al., 2021b). This hypothesis is further supported by a study demonstrating that inhibiting apoptosis in mice lacking Atoh7 largely rescues RGC numbers during development (Brodie-Kommit et al., 2021). Over time, RGCs in Atoh7/Bax-deficient mice mature, exhibit action potentials, and become integrated into retinal circuits, but they experience significant abnormalities in axonal guidance. These results highlight the critical role of ATOH7 in ensuring RGC survival and demonstrate there are both ATOH7-dependent and -independent mechanisms involved in RGC specification. Whether this is true for human retinogenesis remains to be investigated.

How ATOH7 expression is regulated also has been studied intensively. The early progenitor gene Pax6 is required for Atoh7 expression, likely acting directly to regulate its expression by binding cis-regulatory elements near Atoh7 (Riesenberg et al., 2009). Interest in the Atoh7 enhancer landscape was further intensified by discovery of a deletion in a non-coding DNA element upstream of ATOH7 associated with nonsyndromic congenital retinal nonattachment disorder, characterized by optic nerve atrophy and complete blindness (Miesfeld et al., 2020). Interestingly, deletion of the comparable region in mice leads to significant loss of Atoh7 expression but only a slight decrease in RGCs (Miesfeld et al., 2020). These results point toward differences in requirement of human and mouse regulatory elements of ATOH7, warranting further investigation.

An encouraging method to potentially restore vision involves replenishment of lost RGCs either endogenously, through stimulation of cell regeneration, or from exogenous sources using in vitro stem cell differentiation and cell transplantation (Vetter et al., 2017). A recent study endeavored to generate RGCs by inducing expression of specific transcription factors crucial for RGC development within Muller glial cells (Todd et al., 2022). These cells are recognized for their ability to generate various types of retinal cells following injury in zebrafish (Lahne et al., 2020; Martins et al., 2022). The authors genetically modified mice to express different combinations of ASCL1, POU4F2, ISL1, and ATOH1 within Muller glial cells. Interestingly, their findings suggest that POU4F2 and ISL1, together with ASCL1, control the gene activity in mature mouse Muller glial cells toward a cellular state resembling RGCs, a step towards regenerating functional RGCs.

Overall, RGC development is a complex process involving differentiation, maturation, and functional specialization. These recent in vivo findings shed light on important aspects of transcriptional regulation of ATOH7 during RGC development and highlight the crucial role of ATOH7 non-coding regulatory elements in making this process robust and precise. Nevertheless, RGC development is accompanied by several limitations and challenges. These include complexity of development, heterogeneity of RGC subtypes, limited regenerative potential, complex interactions with surrounding neurons, and other genetic and environmental factors. Overcoming these limitations and challenges requires interdisciplinary efforts and application of advanced techniques (e.g., genomics, imaging, functional studies) to gain comprehensive understanding of the intricate processes governing RGC development.

3. Experimental models of glaucoma and therapeutic approaches

Various models have been developed to investigate glaucoma, each with advantages and disadvantages (A. Bouhenni et al., 2012). Models can be categorized by the mechanism in which they induce glaucoma, establishing a framework to evaluate the unique strengths and limitations of each model as well as possible avenues for improvement (Table 1).

Table 1.

Experimental models of glaucoma

Glaucoma type	Approach	Model	Animals	References
POAG	Elevation of IOP	Microbead	Rodents Primates Rabbits	(Astafurov et al., 2014; Ito et al., 2016; Morgan and Tribble, 2015; Ngumah et al., 2006)
		Silicon oil	Mice Primates	(Fang et al., 2021; Moshiri et al., 2022; Zhang et al., 2019a; Zhang et al., 2019b)
		TGF-β transduction	Mice	(Patil et al., 2022)
	Genetic	MYOC mutation	Mice	(Senatorov et al., 2006; Zhou et al., 2008)
		Type I collagen mutation	Mice	(Aihara et al., 2003; Mabuchi et al., 2004)
PACG	Elevation of IOP	Episcleral vein cautery, laser photocoagulation	Rodents Primates Pig	(Danias et al., 2006; Fu and Sretavan, 2010; Ollivier et al., 2004; Ruiz-Ederra et al., 2005; Ruiz-Ederra and Verkman, 2006; Urcola et al., 2006)
		Laser photocoagulation of the trabecular meshwork	Rodents Rabbit Primates	(Biermann et al., 2012; Gherezghiher et al., 1986; Levkovitch-Verbin et al., 2002; Wang et al., 1998)
		Morison’s model	Rodents	(Jia et al., 2000; Morrison et al., 2015; Morrison et al., 2018)
Normotensive Glaucoma	Chemical induction	NMDA injection	Rodents	(Honda et al., 2019; Kimura et al., 2015; Liu et al., 2022; Namekata et al., 2013)
	Genetic	GLAST knockout	Mice	(Harada et al., 2007b)
Pigmentary Glaucoma	Genetic	DBA/2J	Mice	(Anderson et al., 2002; Jakobs et al., 2005; Turner et al., 2017)
Other	Traumatic optic neuropathy	Optic nerve crush	Rodents	(Cameron et al., 2020; Tang et al., 2011; Templeton and Geisert, 2012)

PACG: primary angle closure glaucoma; POAG: primary open angle glaucoma; GLAST: glutamate/aspartate transporter; MYOC: myocilin; NMDA: N-methyl-D-aspartate; TGF-β: transforming growth factor beta

3.1. Experimental models of glaucoma

3.1.1. Induction of elevated IOP

While the overarching pathophysiology is not well understood, elevated IOP appears to be the most reliable risk factor for glaucoma development and may play an important role in its pathogenesis (Weinreb et al., 2014). Thus, many animal models use elevated IOP to induce glaucoma. One such method is Morrison’s model, in which injection of hypertonic saline into the episcleral veins of rodents results in sclerosis of the anterior chamber angle and obstruction of aqueous outflow (Morrison et al., 2018). A single injection leads to sustained IOP elevation, leading to axonal nerve injury that correlates with the degree of IOP elevation (Jia et al., 2000) (Morrison et al., 2015) (Feng et al., 2013). In addition, laser photocoagulation targeting the trabecular meshwork can elevate IOP in experimental animal models (Biermann et al., 2012; Gherezghiher et al., 1986; Lanzetta et al., 1999; Levkovitch-Verbin et al., 2002; Wang et al., 1998).

One of the most popular glaucoma animal models is the microbead induction method, in which microbeads are injected into the anterior chamber of the eye to obstruct aqueous outflow and elevate IOP, with or without use of a magnet to guide the beads into the iridocorneal angle (Ito et al., 2016) (Ito et al., 2016). Multiple injections are required to maintain prolonged IOP elevation, with use of magnetic microbeads possibly allowing for fewer injections. The model has been validated in mice, rats, and primates (Morgan and Tribble, 2015). Varying the parameters of microbead injections such as bead size and bead number can tune the degree of IOP elevation and number of injections needed to maintain long-term IOP elevation (Ito et al., 2016; Morgan and Tribble, 2015).

Silicon oil injection into the anterior chamber also can lead to IOP elevation of almost double the contralateral control eye following a single injection (Zhang et al., 2019a) (Zhang et al., 2019b). This model in rhesus macaques results in 50%–60% RGC loss in the mid-peripheral region after three months (Moshiri et al., 2022). However, IOP levels in some primates are significantly decreased in silicon oil-injected eyes compared to contralateral controls, thought to be secondary to ciliary body atrophy. This indicates that further optimization is required before more widespread use of this model (Moshiri et al., 2022). However, the silicon oil model uniquely allows for removal of the silicon oil at any point, resulting in quick and stable return of IOP to normal levels (Zhang et al., 2019b). Recent advancements with this technique involve modulating how often the pupil remains dilated and thus reconnecting the anterior and posterior chambers, altering dynamics of the resulting IOP elevation (Fang et al., 2021).

3.1.2. Chemical induction

While elevated IOP seems to play a large role in glaucoma, a significant subset of patients have normal IOP throughout their disease (Killer and Pircher, 2018). Thus, animal models that do not rely on elevated IOP are needed to better understand disease in normotensive glaucomatous patients. Glutamate toxicity and overactivation of N-methyl-D-aspartate (NMDA) receptors are implicated in the pathogenesis of many central nervous system disorders including glaucoma (Harada et al., 2007b). Thus, researchers have proposed use of NMDA injections as an animal model of glaucoma (Liu et al., 2022). A single intravitreal NMDA injection in mice significantly reduces RGC number after only one day and results in >50% reduction in as little as five days (Liu et al., 2022). A 20% reduction in thickness of the ganglion cell complex also has been observed 14 days post-injection (Kimura et al., 2015).

3.1.3. Inflammatory models

Increasing evidence implicates TNF-α as a mediator of RGC death during glaucoma, via binding to the TNF-R1 death receptor (Tezel, 2008). Systemic administration of etanercept, a TNF-α inhibitor, to mice with IOP elevated secondary to episcleral vein cauterization ameliorates RGC loss, axonal degeneration, and microglial activation (Roh et al., 2012). Thus, there has been some interest in using the inflammatory mediator to induce glaucomatous neurodegeneration in animal models. One such study in Sprague Dawley rats demonstrated a 39% decrease in RGCs eight weeks after a single intravitreal injection of TNF-α, a result that closely mimics that obtained with laser photocoagulation (Nakazawa et al., 2006). A single intravitreal injection of TNF-α in rats also dramatically increases the number of apoptotic RGCs, and this effect is significantly ameliorated by co-treatment with a TNF-R1 blocker (Cheng et al., 2021).

3.1.4. Genetic models

Genetic models have long provided valuable insights into the pathogenesis of many diseases including glaucoma. The DBA/2J mouse strain, homozygous for a cadherin mutation (Cdh23ahl), is perhaps the most classic example, in which mice develop anterior segment anomalies, iris atrophy, peripheral anterior synechiae, and pigment dispersion, leading to elevated IOP and glaucoma at 7–9 months in most animals (Jakobs et al., 2005; Turner et al., 2017).

Manipulation of the myocilin (MYOC) gene, mutated in 3%–4% of patients with POAG, also has been used to induce glaucoma in mice (Senatorov et al., 2006). An induced Myoc Y345H mutation leads to a statistically significant IOP elevation of 2 mmHg in 12–18-month-old mice compared to wild-type mice, as well as 20% fewer RGCs in the peripheral retina and morphological changes characteristic of apoptosis (Senatorov et al., 2006; Zhou et al., 2008).

Finally, there has been growing interest in using glutamate/aspartate transporter (GLAST) knockout to induce glaucoma in mice (Honda et al., 2019). GLAST knockout mice exhibit RGC degeneration starting at 2 weeks of age and demonstrate 50% RGC loss by 8 months (Harada et al., 2007a). Since these changes occur in the absence of significant IOP elevation, this system could serve as a powerful model for normotensive glaucoma (Harada et al., 2007a).

3.1.5. Other models

While many models can contribute to understanding of neuroprotective strategies, the optic nerve crush (ONC) model is a uniquely valuable tool to study axonal regrowth in the optic nerve (Tang et al., 2011). This method involves dissection of tissue to reveal the optic nerve followed by mechanical crushing with forceps, and it serves as a useful model for both traumatic neuropathies and glaucoma (Cameron et al., 2020). While ONC does not mimic any known pathophysiologic process linked to glaucoma, it does induce patterns of RGC loss consistent with glaucoma (Cameron et al., 2020). While the ONC model is reproducible across various mouse strains and leads to consistent RGC death with a single procedure, significant surgical skill is required to not damage the ophthalmic artery or other optical blood vessels, which can lead to undesired ischemic changes in the retina (Tang et al., 2011).

3.2. Therapies for experimental glaucoma

Under glaucomatous stress, pathological changes encompass several factors, such as oxidative stress, deprivation of neurotrophic factors, apoptosis, glutamate excitotoxicity, and immune responses. Multiple strategies have been used to modulate these pathways to try to improve effectiveness of glaucoma treatment. In particular, Food and Drug Administration (FDA) approval of Luxturna, an adeno-associated virus (AAV) vector-based gene therapy used to treat RPE65-positive retinal dystrophy (Mendell et al., 2021), marks a crucial milestone. This breakthrough has not only provided a successful treatment option for Leber congenital amaurosis but also significantly impacted development of gene therapies for other eye-related conditions including glaucoma because several animal studies show encouraging gene therapies for experimental glaucoma (Hakim et al., 2023). The approval of using AAVs in humans allows us to apply gene therapy to clinical trials.

3.2.1. IOP-lowering treatments

Elevated IOP is a major risk factor for POAG progression, so current therapy seeks to lower IOP. Several groups have developed gene therapies to modulate expression of genes related to enhancing aqueous humor outflow through the trabecular meshwork or uveoscleral pathway to lower IOP and mitigate glaucomatous damage (Borras, 2017; Chern et al., 2022; Hakim et al., 2023). Induction of MMP1 expression in the trabecular meshwork reduces IOP in a sheep model of steroid-induced ocular hypertension (Borras et al., 2016; Gerometta et al., 2010). Further, AAV-mediated MMP-3 expression increases outflow in dexamethasone-induced ocular hypertension in transgenic MYOC mouse models and nonhuman primates (O’Callaghan et al., 2023).

3.2.2. Neuroprotective approaches

Neuroprotective therapeutics demonstrate potential to directly enhance RGC survival and inhibit apoptotic pathways that lead to cell death. Numerous studies demonstrate that neurotrophic factors such as brain-derived neurotrophic factor (BDNF), basic fibroblast growth factor, and ciliary neurotrophic factor (CNTF) significantly regulate RGC survival, regeneration, and neuron plasticity, influencing processes like synapse formation, dendritic branching, and neurotransmitter release (Foldvari and Chen, 2016; Johnson et al., 2011; Mansour-Robaey et al., 1994; Mo et al., 2002; Mysona et al., 2017; Sawai et al., 1996; van Adel et al., 2003). Consequently, several studies have transduced BDNF within the retina (Martin et al., 2003; Ren et al., 2012; Schuettauf et al., 2004). While BDNF gene therapy in glaucoma models provides some protection, overexpression of BDNF alone may be insufficient to promote RGC survival because of rapid downregulation of its receptor, tropomyosin-related kinase B (TRKB), after injury. Accordingly, AAV-mediated upregulation of TRKB significantly enhances RGC survival in axotomized rats by treating them with BDNF (Cheng et al., 2002). In addition, AAV2-mediated overexpression of both BDNF and TRKB enhances RGC survival in a mouse model of optic nerve injury by activating pro-survival pathways (Osborne et al., 2018). Studies further support that systemic administration of 7,8-dihydroxyflavone can alleviate RGC loss in rats after optic nerve injury through activation of the TRKB/AKT/ERK pathway (Chen et al., 2023; Galindo-Romero et al., 2021).

CNTF also has demonstrated its potency as a neuroprotective agent in multiple animal models of glaucoma by increasing RGC survival and promoting axon regeneration (Cui et al., 1999; Do Rhee et al., 2022; LeVaillant et al., 2016; van Adel et al., 2003). CNTF gene therapy exerts a greater impact on axonal regrowth compared to AAV2-BDNF transduction following optic nerve injury in rats (Leaver et al., 2006). Consistently, a single injection of AAV-CNTF provides sustained and global alterations in endogenous genes associated with cell growth and differentiation after one year (LeVaillant et al., 2016). Further, AAV-CNTF effectively attenuates RGC loss in a rat model of laser-induced glaucoma (Pease et al., 2009). Importantly, Goldberg et al. recently reported the safety of CNTF treatment in human patients with glaucoma using an intravitreal implant of CNTF-encapsulated cell therapy (Goldberg et al., 2023). All the evidence warrants the feasibility of CNTF for glaucoma treatment in patients.

Many preclinical studies have explored strategies to inhibit cell apoptosis as a means to prevent RGC loss. Nicotinamide adenine dinucleotide (NAD⁺) is a coenzyme involved in cellular energy metabolism and various cellular processes, including DNA repair, mitochondrial function, and cellular stress response (Covarrubias et al., 2021; Williams et al., 2018). Emerging research suggests that age-related decline in NAD levels can compromise cellular resilience and increase vulnerability to RGC damage. Oral administration of vitamin B3, an NAD⁺ precursor, as well as Nmnat1 gene therapy, which encodes a NAD⁺-producing enzyme, protect both RGC soma and axon loss and preserve electrical activity in RGCs during ocular hypertension in DBA/2J mice (Williams et al., 2017). In addition, delivering the NMNAT2 enzyme to RGCs protects both RGC somas and axons and effectively preserves RGC function, as determined by pattern electroretinography in mice subjected to ONC and elevated IOP (Fang et al., 2022).

Heat shock proteins (HSPs) are chaperon proteins that enhance cell survival under stresses (Piri et al., 2016). In addition, small HSPs protect RGCs in experimental glaucoma models. Peptain-1 and −3a, derived from α-crystallin domains of αB-crystallin and HSP20, respectively, effectively protect both RGC somas and axons during glaucomatous stress in the mouse model (Nam et al., 2022b; Stankowska et al., 2019). In particular, delivery of Hsp27 in the mouse eye results in greater neuroprotection during ocular hypertension compared to other small HSPs (αA-crystallin, αB-crystallin, Hsp20) (Nam et al., 2022a).

3.2.3. Anti-oxidative and anti-inflammatory approaches

Multiple hypotheses have been suggested for the underlying axon degeneration in glaucoma. These include hypoxia and oxidative stress resulting from reduced blood flow and impaired oxygen supply to the optic nerve. Anti-oxidant agents can neutralize free radicals and protect cellular components from oxidative stress. In addition, the neuroinflammatory response characterized by activation of microglia and astrocytes can lead to release of pro-inflammatory cytokines and other factors that contribute to axonal degeneration. Therefore, anti-oxidative or anti-inflammatory agents could ultimately promote RGC survival. For example, intravitreal injection of rapamycin inhibits glial activation by blocking the mTOR/ROCK pathway, reducing RGC death (Wang et al., 2023). Another study revealed that use of a GLP-1R agonist to inhibit transformation of A1 astrocytes prevents production of inflammatory cytokines, thereby rescuing RGCs (Sterling et al., 2020).Thus, the development of antioxidant or anti-inflammation reagents may be a therapeutic strategy for treating glaucoma.

3.3. Limitations and challenges

While there are several preclinical research models to choose from, each model has unique strengths and weaknesses. Some models require significant surgical skill, such as Morrison’s model and episcleral vein cauterization. Other models, while reliable, suffer from a high dropout rate, such as the DBA/2J model. Other models are applicable for only one subtype of glaucoma, such as NMDA injection, while others are more versatile, such as the silicon oil method. The ultimate decision of which model to choose lies with the investigator and their specific needs, expertise, and availability of equipment. No one model will unveil all the secrets glaucoma has to offer. Further research is still needed to both refine these models and use them to their fullest potential.

With respect to therapeutic development, a major challenge in axonal regenerative research is the phenomenon known as “stalled growth.” Despite efforts to develop experimental treatments that promote axonal regrowth, regenerating axons fail to reach their intended post-synaptic targets, resulting in limited functional recovery (de Lima et al., 2012). Neuroregeneration research faces obstacles due to inhibitory factors in the surrounding environment, such as extracellular inhibitors of axon growth, myelin-associated proteins, and formation of glial scar tissue following injury (Silver et al., 2014; Yiu and He, 2006). Additionally, lack of directional guidance cues after injury can result in random or failed axonal growth. Insufficient supply of neurotrophic factors in the regenerative environment also limits axonal growth potential. Consequently, additional investigations are needed to address this stalled growth and discover effective approaches to overcome this problem in axonal regeneration research.

4. Stem cell replacement

4.1. Stem cell-derived RGC differentiation

Since human RGCs do not regenerate or are not replaced after injury, stem cell-derived RGCs provide an opportunity for cell replacement therapy. Protocols for stem cell-to-RGC differentiation in two-dimensional (2D) (Ji and Tang, 2019; Yuan et al., 2021) and three-dimensional (3D) (Hua et al., 2020) culture are well-established and summarized in previous reviews. Thus, here we only discuss human stem cell studies in the past decade (Table 2).

Table 2.

Human stem cell-derived RGC differentiation

Chemical-based protocol (2D culture)
Differentiation base	Stem cell type	Time	Special molecules	Yield	Lab	References
Gene/ chemical	hiPSCs	8 days	NGN2, GDNF, DAPT	14%	Goldberg	(Luo et al., 2022)
	hESCs			20%
Chemical	hiPSCs	35 days	DAPT, Rock inhibitor (Y27632), forskolin, cAMP, XAV939, SB431542	22–50%	Mills	(Chavali et al., 2020)
Chemical/growth factor	hiPSCs	36 days	SB431542, XAV939, LDN193189, IGF1, bFGF, SHH, DAPT, Y27632, Follistatin 300, Foskolin, cAMP,	85–93%	Chavali	(Vrathasha et al., 2022)
Chemical/ growth factor	hiPSCs	60 days	FGF2	N/A	Takahashi	(Osakada et al., 2009)
Chemical/ growth factor	hiPSCs	15 days	SHH, FGF8, DAPT, Follistatin, Cyclopamine	54%	Ahmad	(Teotia et al., 2017)
Chemical	hESCs	30 days	SB431542, Dorsomorphin, IWP2	N/A	Goldberg	(Zhang et al., 2020)
Chemical	hESCs	30 days	Dorsomorphin, IDE2, Nicotinamide, Forskolin, DAPT, SB431542, LDN193189, Noggin	20–50%	Zack	(Sluch et al., 2017)
Chemical	hESCs	N/A	Noggin, Dickkopf1(DKK1), IGF1	N/A	Reh	(Lamba et al., 2006)
Chemical/growth factor	hESCs+iPSCs	40 days	FGF2, DAPT	N/A	Scholer	(Greber et al., 2011)
Chemical/growth factor	hESCs	60 days	IGF1, DKK1, Noggin, DAPT	N/A	Reh	(Chao et al., 2017)
Chemical/growth factor	hESCs	36 days	IGF1, DKK1, Noggin	N/A	Hewitt	(Daniszewski et al., 2018)
Retinal organoid protocol (3D culture)
EB attachment method	Stem cell type	Observe time	Special molecules/approaches	Yield	Lab	References
Gelatin	hiPSCs	56 days	FGF2	N/A	Orieux	(Rabesandratana et al., 2020)
Matrigel	hiPSCs	25–50 days	DKK1	N/A	Ge	(Luo et al., 2021a; Luo et al., 2018)
FBS	hiPSCs	40 days	N/A	N/A	Meyer	(Fligor et al., 2018)
MEF	hiPSCs	90 days	Noggin, DKK1, IGF1, FGF2	13%	Meyer	(Sridhar et al., 2013)
Matrigel	hiPSCs	35 days	RA	N/A	Canto-Soler	(Vergara et al., 2017; Zhong et al., 2014)
Matrigel	hESCs	30 days	Reducing the conc. of knockout serum replacement additive	N/A	Sasai	(Nakano et al., 2012)
Matrigel	hESCs	30 days	BMP4	N/A	Gamm	(Ludwig et al., 2023)
Laminin	hESCs	50 days	Noggin, DKK1, IGF1, FGF9	15%	Gamm	(Meyer et al., 2011)
Matrigel	hESCs	21 days	BMP4, knockout serum replacement	N/A	Takahashi	(Kobayashi et al., 2018)
Matrigel	hESCs	27 days	IWR1e	28%	Lee	(Aparicio et al., 2017)
PDL/Laminin	hESCs	30 days	SEMA3A, SLIT1, KSR	N/A	Azuma	(Yokoi et al., 2017)

BMP: bone morphogenetic protein; DAPT: γ-secretase inhibitor; DKK1: Wnt inhibitor; EB: embryonic body; FGF: fibroblast growth factor; GDNF: glial derived neurotrophic factor; RA: retinoic acid; hESC: human embryonic stem cell; hiPSC: human induced pluripotent stem cell; IGF: insulin-like growth factor; N/A: not applied; SB431542: Smad2 inhibitor; XAV939: Wnt inhibitor

Chemical-based protocols were first introduced to 2D culture for RGC differentiation from human induced pluripotent stem cells (hiPSCs) or human embryonic stem cells (hESCs). There is no significant difference in RGC differentiation yield between hiPSCs and hESCs induced by chemical-based protocols. Inhibitors of signaling pathways including SMAD2 (SB431542), WNT (XAV939, DKK1), AMPK (Dorsomorphin), ALK2/3 (LDN193189), BMP (Noggin), and γ-secretase (DAPT) have been used to promote RGC differentiation. It usually takes 4–8 weeks to observe RGC features in cultures differentiated by chemical-based protocol. However, a group developed a rapid protocol by expressing neural prompter gene NGN2 in the stem cell-to-RGC culture, requiring only 8 days for RGC differentiation (Luo et al., 2022). The rapid protocol shows that 14% of hiPSCs develop into RGCs, and 20% of hESCs develop into RGCs. Although the rapid protocol reduces the time for RGC differentiation, the yield remains lower than traditional chemical-based protocols (20%–50%) (Chavali et al., 2020; Sluch et al., 2017).

Addition of growth factors such as fibroblast growth factor (FGF) and insulin-like growth factor (IGF) to 2D differentiation cultures can increase RGC yield. Remarkably, adding FGF8 to differentiation cultures significantly boosts the RGC yield to 54% (Teotia et al., 2017). A study even showed that adding FGF and IGF1 to hiPSC cultures increases RGC differentiation yield to 93% (Vrathasha et al., 2022).

A 2011 study reported that adding basement-membrane matrix components to differentiating mouse embryonic stem cells results in formation of self-organized optic-cup structures (Eiraku et al., 2011). At the same time, another group developed a protocol by plating stem cell aggregates (embryoid bodies) on laminin-precoated dishes, which subsequently formed optic vesicle-like structures [later called retinal organoids (ROs)] from hiPSCs (Meyer et al., 2011). RGCs start to appear at 30–50 days in human ROs (Table 2) (Fligor et al., 2018), recapitulating development of fetal retinas (~6–8 weeks) (Luo et al., 2019). The yield of RO-derived RGCs is unclear in most RO protocols, but a few reports show 13%–28% (Aparicio et al., 2017; Meyer et al., 2011; Sridhar et al., 2013). Due to the 3D structure of ROs and their content of various neuronal cells, RGCs harvested from ROs show similar transcriptional profiles as human fetal RGCs (Sridhar et al., 2020). Thus, increasing studies are applying RO-derived RGCs to cell replacement therapy.

4.2. Stem cell-derived RGC transplantation

Here we overview transplantation of human stem cell-derived RGCs in normal and injured eyes (Table 3). Transplantation of donor human RGCs into host mouse retinas results in human cells located in the ganglion cell layer. Intravitreal or subretinal injection of 10,000–500,000 donor cells into normal or ONC rodent eyes shows that donor cells remain 5 months after transplant, with a 0.134% survival rate (Vrathasha et al., 2022). Another donor cell transplant study used cyclosporine for immunosuppression, although the donor cell survival rate was not recorded (La Torre et al., 2012). Interestingly, an additional study showed that ONC reduces survival of donor cells after transplantation (Luo et al., 2022), perhaps due to severe inflammatory responses that lead to cell death. Further, a study in a larger animal model showed that one million donor RGCs subretinally injected into a monkey eye survive for 3 months without immunosuppression treatment (Chao et al., 2017).

Table 3.

Human stem cell-derived RGC transplantation

2D-derived donor cells
Host/age	Health condition	Source of donor cells	Duration	Transpla nted cells	ISx	Outcome	Injection method	References
Rat/adult	Normal	hESC-derived RGCs	1 week	50,000	N/A	Observation of donor cells in the host GCL	IVT	(Zhang et al., 2020)
Mouse/4 mns	Normal	hiPSC-derived RGC	2 and 5 months	500,000	N/A	Observation of donor cells in the host GCL (survival efficiency = 0.134%)	IVT	(Vrathasha et al., 2022)
Mouse/adult	Normal	hESC-derived RGCs	4 weeks	100,000	Cyclosporine	Observation of donor cells in the host retina	Subretinal	(La Torre et al., 2012)
Monkey/unknown	Normal	hESC-derived RGCs	4 and 12 weeks	1000,000	N/A	Donor RGCs survive in the host eye for 3 months without immunosuppression	Subretinal	(Chao et al., 2017)
Mouse/adult	Normal	hESC-derived RGCs	1 week, 1 month	10,000	N/A	ONC reduces the survival and/or integration of donor cells after transplantation. Donor stem cells may protect endogenous RGCs from death	IVT	(Luo et al., 2022)
	ONC
RO-derived donor cells
Host/age	Health condition	Source of donor cells	Duration	Transpla nted cells	ISx	Outcome	Injection method	References
Rabbit, monkey/adult	Normal	hiPSC-derived RGCs	1 week–3 months	100,000	N/A	Scaffold and donor cells were detected in both rabbit and monkey eyes after transplantation	Subretinal Epiretinal	(Li et al., 2017)
Cat/adult	Normal	hESC-derived organoids	28–66 days	5–9 ROs per eye	Cyclosporine Prednisolone	Immunosuppression regimen improved donor cell survival, reduced immune response; promoted donor cells’ synaptic connectivity	Subretinal	(Singh et al., 2019)
Monkey/adult	Normal	hiPSC-derived organoids	4 weeks-1 year	2–3 ROs per eye	Ozurdex	Donor cells were survived up to one year, and long neurite outgrowth was observed	Epiretinal	(Luo et al., 2021b)
	High IOP
Mouse	NMDA	TiPSC-EBs	2 weeks	25,000	N/A	TiPSC-EBs were differentiated into RGC-like cells with distinct neuronal structure	IVT	(Zhou et al., 2021)
Mouse/8–10 wks	NMDA	hESC-derived optic cup-like structure	5 weeks	20,000	N/A	Donor cell integration was observed in the host GCL	IVT	(Wang et al., 2019)
Mouse/11–13 wks	ONC	hiPSC-derived organoids	4 weeks	200,000	Cyclosporine	Donor cells (RBPMS+) were identified in vivo and no neurite outgrowth was observed	IVT	(Rabesandr atana et al., 2020)

GCL: retinal ganglion cell layer; hiPSC: human induced pluripotent stem cell; hESC: human embryonic stem cell; ONC: optic nerve crush; ROs: retinal organoids; IVT: intravitreal injection; ISx: immunosuppressant; N/A: Not applied; NMDA: NMDA-induced ganglion cell injury model; IOP: intraocular pressure; TiPSC-EBs: T cell-derived iPSC embryoid bodies.

Because RGCs derived from human ROs show similar transcriptional profiles to human fetal RGCs (Sridhar et al., 2020), studies have recently used RO-derived RGCs for cell replacement. In addition to intravitreal and subretinal injection, epiretinal injection also has been applied to RO transplantation (Li et al., 2017; Luo et al., 2021b). In a rodent model, intravitreal injection of 20,000–200,000 donor RGCs into mouse eyes with NMDA-induced or ONC injury results in donor RGCs in the host retina 2–4 weeks after transplantation (Rabesandratana et al., 2020; Wang et al., 2019; Zhou et al., 2021). A cat model showed that immunosuppression drugs cyclosporine and prednisolone improve RO transplantation, with donor cells surviving for 66 days (Singh et al., 2019). Further, in a non-human primate model, scaffolds facilitate RO-derived RGC survival in the monkey eye for 3 months. Another study showed that RO-derived RGCs remain in the monkey eye one year after transplantation when treated with immunosuppression drug Ozurdex, with extended neurites from donor RGCs observed in the host optic nerve (Luo et al., 2021b), suggesting that human-derived RGCs integrate into the monkey visual system.

Yet more transplanted cells do not guarantee longer survival in the host eye(Venugopalan et al., 2016). Most studies used donor cells with peak RGC marker expression in ROs. However, investigation of donor cells from different developmental stages of ROs remains interesting for future RGC replacement therapy. One group has applied CRISPR/Casp9 technology to generate a BRN3B-P2A-tdTomato-P2A-THY1.2 reporter line (Sluch et al., 2017), allowing donor RGC enrichment by immunopanning or fluorescence-activated cell sorting against THY1 protein (Wang et al., 2019). Another study showed that attaching dissociated RO cells to a plate for one week significantly promotes the expression of several RGC markers and RGC axonal outgrowth, indicating further differentiation and maturation (Rabesandratana et al., 2020). Although one study indicates that RO-derived donor RGCs (Luo et al., 2021b) survive longer in host eyes than cells generated by chemical-based methods (Vrathasha et al., 2022), no studies have directly compared cell survival from chemical-based or RO protocols—this remains important for future work.

4.3. Neuroprotection

Stem cells are thought to secrete trophic factors to support endogenous neuronal survival. Consistently, transplantation of human stem cells promotes donor mouse RGC survival in a co-transplantation model in rodent eyes (Wu et al., 2018). Several other studies have observed profound neuroprotective effects on host RGCs with neural progenitor cells (Rettinger et al., 2021), retinal neurons (Luo et al., 2022), or even non-retinal neurons (Wu et al., 2010). Further, hiPSC-derived RGC transplantation alleviates ONC-induced RGC death in a mouse model (Luo et al., 2022). Thus, providing continuous secretion of trophic factors could be a strategy to increase donor cell survival after transplantation.

4.4. Limitations and challenges

Because RGCs are the only neurons in the eye that connect to the brain, regeneration of their axons in the optic nerve is critical to restore vision. Despite the encouraging progress of stem cell therapy in animal models, it remains clinically challenging (Kuriyan et al., 2017). Although a study of mouse primary RGC transplantation showed that extended axons are observed in the brain (Venugopalan et al., 2016), long-distance regenerative axons of human stem cell-derived donor RGCs are limited in the host’s optic nerve (Luo et al., 2021b).

In addition, donor RGC integration into the host visual circuit is another critical point. The internal limiting membrane of the retina is the main barrier that prevents donor cells from migrating into the ganglion cell layer and connecting to endogenous neurons (Zhang et al., 2021). Enzymatic or mechanical removal of the internal limiting membrane may facilitate donor cell survival and integration in the host eye. A study has successfully recorded the voltage reading of transplanted hiPSC-RGCs on mouse retinal tissue (Vrathasha et al., 2022), indicating the electrophysiologic function of donor neurons. In addition, multielectrode array (MEA) recording, a functional measurement for multiple neurons, has been successfully applied in photoreceptor transplantation (Tu et al., 2019). In RGC, MEA is able to classify multiple RGCs by their electrophysiologic responses at the same time (Hilgen et al., 2022). However, little has been done in RGC transplantation. Thus, applying MEA in RGC replacement studies remains a promising future test to rule out whether donor RGCs functionally connect to endogenous circuits.

5. Clinical aspects

5.1. Current advances in clinical medicine and surgery for glaucoma treatment

IOP is the major modifiable risk factor in preventing glaucomatous vision loss, although its exact cause remains elusive (Stein et al., 2021). Elevated IOP leads to progressive ocular neuropathy characterized by RGC apoptosis (Kang and Tanna, 2021). In managing adult glaucoma, medical therapy generally stands as the first-line treatment (Mohan et al., 2022). Recent years have seen remarkable strides in novel pharmacotherapies and innovative drug delivery systems for glaucoma to mitigate the challenges of chronic medication use (Mohan et al., 2022; Wagner et al., 2022). In addition, laser and surgical interventions can significantly reduce IOP over extended periods of time and can be more cost-effective than prolonged medication use (Wagner et al., 2022). However, these interventions come with elevated procedural risks and a higher potential for failure (Lim, 2022). Traditional glaucoma surgery with incisional procedures has been gradually replaced by minimally invasive (or microinvasive) glaucoma surgery (MIGS), which offers improved safety with reasonable efficacy (Wagner et al., 2022). Nevertheless, more large-scale randomized trials are necessary to evaluate the long-term efficacy of MIGS (Pereira et al., 2021; Wagner et al., 2022).

5.2. Medical therapy

According to the 2022 Preferred Practice Pattern guidelines from the American Academy of Ophthalmology, the initial goal of therapy is to achieve at least 25% reduction in IOP, even in cases of normotensive glaucoma (2022). Over the past few decades medications for glaucoma have evolved significantly, with improved effectiveness and safety of alpha agonists, beta blockers, topical carbonic anhydrase inhibitors, and prostaglandin analogs compared to older options (e.g., pilocarpine, oral carbonic anhydrase inhibitors) (Ostler et al., 2021).

Rho kinase inhibitors represent a recently introduced pharmacotherapy class that increases conventional outflow and decreases episcleral venous pressure by promoting formation of pores in the Schlemm’s canal of endothelial cells and facilitating relaxation of smooth muscle fibers in the trabecular meshwork (Moshirfar et al., 2018; Tanna and Johnson, 2018; Thieme et al., 2000). As one Rho kinase inhibitor, netarsudil 0.02% also acts as a norepinephrine transporter inhibitor and exhibits comparable IOP-lowering efficacy as timolol 0.5%, but higher incidence of adverse effects including conjunctival hyperemia, subconjunctival hemorrhage, and cornea verticillate (Asrani et al., 2020; Serle et al., 2018). The value of Rho kinase inhibitors lies in their different mechanism of action compared to currently used medications, although adverse effects may affect long-term compliance (Asrani et al., 2020; Moshirfar et al., 2018; Serle et al., 2018).

Latanoprostene bunod 0.024% (LBN) presents another notable advancement (Hoy, 2018). It metabolizes into latanoprost acid and butanediol mononitrate, which metabolizes into 1,4 butanediol and then nitric oxide (Hoy, 2018). As latanoprost acid binds to the prostaglandin F receptor and enhances uveoscleral outflow by remodeling the extracellular matrix of the ciliary muscle via matrix metalloproteinases, nitric oxide leads to vasodilation, relaxation of smooth muscle cells, and increased trabecular outflow (Hoy, 2018). LBN has demonstrated comparable IOP-lowering efficacy to timolol 0.5% and substantially reduces both diurnal and nocturnal IOP levels (Liu et al., 2016; Medeiros et al., 2016; Weinreb et al., 2018; Weinreb et al., 2016).

Omidenepag isopropyl (OMDI) is another novel topical ocular hypotensive agent (Matsuo et al., 2022). Upon instillation, OMDI converts to its active form, omidenepag, offering a distinctive feature as a nonprostaglandin-selective prostanoid EP2 receptor agonist (Matsuo et al., 2022; Schlötzer-Schrehardt et al., 2002). By activating the EP2 receptor in the ciliary body and trabecular meshwork, OMDI initiates G-protein-mediated signaling cascades, resulting in increased intracellular adenosine 3’,5-cyclic monophosphate (cAMP) levels and enhanced aqueous humor outflow through trabecular and uveoscleral pathways (Fuwa et al., 2018). Demonstrating both IOP-lowering efficacy and acceptable safety, OMDI’s potential is underscored by a four-week study revealing non-inferiority to latanoprost 0.005% in IOP reduction (Aihara et al., 2020), and a 52-week study indicating sustained IOP reduction when used alone or with timolol 0.5% (Aihara et al., 2021). Notably, OMDI stands out for avoiding prostaglandin-associated periorbitopathy complications, distinguishing it from prostaglandin F receptor agonists (Matsuo et al., 2022). However, it is recommended that OMDI should be avoided in aphakic or pseudophakic eyes to prevent macular edema (Aihara et al., 2021; Matsuo et al., 2022).

New drug delivery systems have been developed to address the challenges of chronic medication use, including adherence, side effects, costs, and IOP target attainment. One such system is the Durysta bimatoprost implant, a biodegradable sustained-release device utilizing the NOVADUR system within the anterior eye chamber (Mohan et al., 2022). Comprising biodegradable polymers that hydrolyze into water and carbon dioxide, Durysta exhibits non-inferiority to timolol 0.5%, though there are concerns about reduced corneal endothelial cell density (Craven et al., 2020; Medeiros et al., 2020). Additionally, the potential of bimatoprost ocular rings and travoprost punctum plugs as sustained-release alternatives is being explored, pending phase 3 trials and subsequent clinical approval (Brandt et al., 2017; Mohan et al., 2022; Perera et al., 2016).

5.3. Laser therapy

When medications cannot reduce IOP and prevent vision loss, laser and surgical treatments are indicated. Laser therapy effectively reduces IOP and minimizes long-term costs of multiple anti-glaucomatous medications (Stein et al., 2021). The choice of laser procedure depends on underlying etiology of the glaucoma (Wagner et al., 2022). For example, laser trabeculoplasty is indicated for open-angle glaucoma that poorly responds to pharmacotherapy (Wagner et al., 2022). This procedure involves applying a thermal laser directly to the trabecular meshwork, leading to structural changes that enhance aqueous humor outflow (Samples et al., 2011). Recent advancements in laser trabeculoplasty, including pattern scanning trabeculoplasty and titanium-sapphire laser trabeculoplasty, have demonstrated short-term efficacy and safety profiles comparable to selective laser trabeculoplasty (Tsang et al., 2016).

5.4. Surgical therapy

Surgical interventions are typically considered when pharmacotherapy and laser procedures fail to achieve desirable IOP reduction (Wagner et al., 2022). Traditional bleb-based, incisional IOP-lowering procedures such as trabeculectomy effectively lower IOP but are associated with complications including endophthalmitis, bleb leakage, and blebitis (Jampel et al., 2012; Lim, 2022). These procedures also have high reoperation rates (Jampel et al., 2012).

To address these issues, MIGS procedures have been developed (Jabłońska et al., 2022). These approaches offer minimally invasive solutions with relatively favorable safety profiles and moderate efficacy. MIGS has revolutionized glaucoma surgery, although large-scale randomized trials are necessary to evaluate long-term efficacy (Pereira et al., 2021; Wagner et al., 2022). MIGS devices primarily center on three pressure-lowering mechanisms: the trabecular outflow pathway, subconjunctival space, and suprachoroidal space (Jabłońska et al., 2022; Wagner et al., 2022). One such device targeting the trabecular outflow pathway is the Hydrus microstent, evolving from the iStent and approved by the US FDA in 2018 (Samet et al., 2019). With its unique design, this microstent enhances aqueous flow by maintaining Schlemm’s canal patency (Samet et al., 2019). Another noteworthy MIGS option is the Xen gel stent, which is made of porcine collagen-derived gelatin cross-linked with glutaraldehyde and diverts aqueous humor to the subconjunctival space (Pereira et al., 2021). The device hydrates within 1–2 minutes upon contact with aqueous humor during implantation and then swells in place, bending and conforming to tissue (Pereira et al., 2021). In a prospective multicenter study, this stent achieved a 20% IOP reduction in 65.8% of cases after 24 months, demonstrating a favorable safety profile (Reitsamer et al., 2019).

5.5. Limitations and challenges

Reducing IOP is considered the most important factor in treating glaucoma. Optimal selection among medical, laser, and surgical options presents a clinical management challenge. Despite the range of pharmacological interventions, failure of medical treatment persists due to factors including ineffectiveness, intolerance, poor compliance, and low adherence, collectively promoting progression of glaucoma (Butt et al., 2016).

While ocular hypotensive agents and lasers have diminished the need for anti-glaucoma surgery, a subset of patients still need surgical intervention. This could arise either from inadequate IOP reduction with maximal medical therapy or from intolerance to ocular hypotensive agents due to accumulated exposure to topical medications and preservatives (Bettin and Di Matteo, 2013). While trabeculectomy has long been regarded as the surgical gold standard, it is association with postoperative complications necessitating additional interventions (Chen et al., 1997; Wagner et al., 2022). MIGS emerges as a revolutionary surgical approach, offering rapid patient recovery (Pereira et al., 2021; Wagner et al., 2022). However, studies with extended follow-up are needed to confirm the sustained efficacy of these novel techniques.

6. Conclusion

Glaucoma management has witnessed significant advancements in preclinical animal studies, clinical medicine, and surgery. Although animal models do not perfectly mimic human disease, these studies still have translational values in regenerative medicine. Understanding the regulatory mechanism of RGC development provides insight for gene therapy or human stem cell-to-RGC differentiation for cell replacement therapy. Although advanced experimental studies of gene therapy and cell replacement for axon regeneration and RGC restoration are encouraging, a long road remains ahead before these approaches fully reach clinical utility. Nonetheless, therapy continues to evolve with the introduction of novel medications, including Rho kinase inhibitors, and new drug delivery systems. In clinics, laser and surgical interventions effectively reduce IOP, with traditional incisional surgeries being gradually replaced by MIGS. MIGS offers improved safety, though large-scale randomized trials are needed to assess its long-term efficacy. Here, we overview the role of ATOH7 in RGC development and suggest multidisciplinary approaches to restore RGC loss by leveraging insights from work in animal models of experimental glaucoma, basic study of RGC development, molecular and cellular therapies, stem cell differentiation and cell replacement therapies, and current clinical procedures. Although many challenges remain, growing encouraging studies give hope in restoring vision in patients suffering from glaucoma.