The concept of gene therapy for glaucoma: the dream that has not come true yet

Abstract

Gene therapies, despite of being a relatively new therapeutic approach, have a potential to become an important alternative to current treatment strategies in glaucoma. Since glaucoma is not considered a single gene disease, the identified goals of gene therapy would be rather to provide neuroprotection of retinal ganglion cells, especially, in intraocular-pressure-independent manner. The most commonly reported type of vector for gene delivery in glaucoma studies is adeno-associated virus serotype 2 that has a high tropism to retinal ganglion cells, resulting in long-term expression and low immunogenic profile. The gene therapy studies recruit inducible and genetic animal models of optic neuropathy, like DBA/2J mice model of high-tension glaucoma and the optic nerve crush-model. Reported gene therapy-based neuroprotection of retinal ganglion cells is targeting specific genes translating to growth factors (i.e., brain derived neurotrophic factor, and its receptor TrkB), regulation of apoptosis and neurodegeneration (i.e., Bcl-xl, Xiap, FAS system, nicotinamide mononucleotide adenylyl transferase 2, Digit3 and Sarm1), immunomodulation (i.e., Crry, C3 complement), modulation of neuroinflammation (i.e., erythropoietin), reduction of excitotoxicity (i.e., CamKIIα) and transcription regulation (i.e., Max, Nrf2). On the other hand, some of gene therapy studies focus on lowering intraocular pressure, by impacting genes involved in both, decreasing aqueous humor production (i.e., aquaporin 1), and increasing outflow facility (i.e., COX2, prostaglandin F2α receptor, RhoA/RhoA kinase signaling pathway, MMP1, Myocilin). The goal of this review is to summarize the current state-of-art and the direction of development of gene therapy strategies for glaucomatous neuropathy.

Introduction

Glaucoma is a heterogenous group of eye conditions that can lead to vision loss, by irreversible damage to retinal ganglion cells (RGCs). There are different types of glaucoma, and they are generally classified based on the clinical features of the condition (Weinreb et al., 2014; Wiggs and Pasquale, 2017; Schuster et al., 2020). Different types of glaucoma include, primary open-angle glaucoma (POAG), the most common type, linked with gradual loss of balance between production and drainage of aqueous humor in the anterior segment of the eye, leading to increased intraocular pressure (IOP) and subsequent, perpetuating degeneration of the optic nerve axons; normal-tension glaucoma (NTG), classified as a subtype of POAG, occurs when the optic nerve shows clinical signs of damage, despite low values of IOP, the exact mechanisms of this type of glaucoma is still poorly understood; angle-closure glaucoma associated with the narrowing of the iridocorneal angle, leading to a critical increase of IOP that very often is sudden and results in rapid vision loss; secondary glaucoma, caused by various associated conditions or underlying health problems, such as ocular trauma, ocular inflammatory diseases, steroid intake, diabetes, pigment dispersion syndrome, pseudoexfoliation syndrome; congenital glaucoma that is present at birth and is caused by developmental abnormalities of the anterior segment of the eye, this type of glaucoma is exclusively and always associated with increased IOP; juvenile glaucoma represents subtype of the best-described genetic background, with myocilin gene mutation as the most important association. The above-named types of glaucoma have numerous subtypes and variants, proving its heterogeneity.

Glaucoma, as a progressive optic neuropathy, is thought to be the leading cause of severe visual impairment or permanent vision loss worldwide. The exact mechanism of RGC death in glaucoma is still unknown, although apoptosis has been suggested to be the final common pathway for cell death. RGC death in glaucoma is most commonly associated with elevated IOP, which has made lowering the IOP, up to now, the only and the most effective treatment method (Daliri et al., 2017; He et al., 2018). Since current therapeutic strategies, i.e., pharmacological, and surgical approaches targeting increased IOP, are not sufficient enough to prevent glaucoma-related blindness, new effective therapeutic strategies focused on RGC neuroprotection, and their regeneration are expected to be developed, to prevent or at least delay the death of RGC (Chang and Goldberg, 2012; Doozandeh and Yazdani, 2016; Gauthier and Liu, 2016; Sena and Lindsley, 2017; Boia et al., 2020; Xuejiao and Junwei, 2022).

Retrieval Strategy

In this review, we made an effort to search two databases, PubMed and ClinicalTrials.gov, for the following terms, glaucoma and “gene therapy”, glaucoma and neuroprotection, glaucoma and regeneration, we did not apply any time frame of the search (until February 1, 2023), nor did we apply any filters. To calculate the translation coefficient, we searched both databases for the following phrases: glaucoma, “gene therapy”, neuroprotection, neuroregeneration, regeneration, glaucoma and “gene therapy”, glaucoma and neuroprotection, glaucoma and regeneration, “optic neuropathy”, “optic neuropathy” and “gene therapy”, “optic neuropathy” and neuroprotection, “optic neuropathy” and regeneration, cancer, stroke, Alzheimer’s disease. In this search strategy, we did not apply any time frame to the search, but we filtered the search results by excluding reviews, systematic reviews, and meta-analyses in PubMed and selecting only interventional trials in ClinicalTrials.gov.

Gene Therapy

Gene therapy is a type of treatment that involves replacing, inactivating, or introducing a new gene into cells to treat a disease or condition with a proven genetic background (Patil et al., 2019; Wang et al., 2019). There are several different methods that can be used to deliver the therapeutic genetic material to target cells in vivo. Viral vector-based gene therapies are the most common method of gene delivery and use a viral vector to deliver the therapeutic genetic material to the target cells. The virus itself is modified so that it has low immunogenicity and lacks pathogenic properties, but it still has the ability to bind to the target cell to deliver the genetic material (Sudhakar and Richardson, 2019; Wang et al., 2019). Among the known viral vectors, adeno-associated viruses (AAV) have become the most widely used in both preclinical and clinical studies (Bordet and Behar-Cohen, 2019). AV have also become the vectors of choice for RGC-targeted gene delivery. AAV serotype 2 (AAV2) shows high tropism to RGC due to its high expression of heparin sulfate proteoglycan, which mediates attachment of the viral particles (Summerford and Samulski, 1998). Transfection of RGCs by intravitreal administration, the most reasonable route, results in an efficiency of approximately 85% according to available data (Hellström et al., 2009; Wilson and Di Polo, 2012). A single intravitreal injection transfects RGC bodies and axons for more than 7 months, and AAV-delivered genes begin to be expressed 2–3 weeks after injection (Dudus et al., 1999). In non-viral gene therapy, the exogenous genetic material is delivered to target cells using a non-viral vector, such as a lipid nanoparticle or synthetic polymer. This method is less efficient than viral vector gene delivery, but is considered less likely to trigger an immune response (Bordet and Behar-Cohen, 2019; Patil et al., 2019; Huang and Chau, 2021). Gene editing methods use restrictive enzymes, such as clustered regularly interspaced short palindromic repeats (CRISPR)-Cas9, to make precise changes in the DNA of target cells (He et al., 2022). Gene editing can be used to correct genetic defects or to insert fragments of new genetic material into the DNA of target cells (Wang et al., 2020; Wu et al., 2020; Hu et al., 2023). Gene transfer is the least precise method, in which genetic material is transferred directly into a living organism by systemic administration, i.e., intravenous injection. This method is not widely used because it is difficult to control the exact location of the introduced genetic material.

Gene therapy in ocular diseases

Although gene therapy is a relatively new field undergoing continuous and intense study, it holds great promise as a potential treatment for many ocular diseases, including glaucoma.

A growing number of studies have reported the use of gene therapy in inherited retinal and optic nerve degenerations, including Leber congenital amaurosis, retinitis pigmentosa, choroideremia, Usher syndrome, Stargardt disease, Leber’s hereditary optic neuropathy, achromatopsia, and X-linked retinoschisis, as well as wet age-related macular degeneration, diabetic macular edema, and retinal vein occlusion (Naik et al., 2021; Wasnik and Thool, 2022).

Luxturna^TM (voretygene neparvovec), by Spark Therapeutics, became the first viral ocular gene therapy among 22 gene therapies approved worldwide through 2019 (Ma et al., 2020). Luxturna^TM was approved by the United States Food and Drug Administration Agency in December 2017 for the treatment of Leber’s congenital amaurosis type 2, an inherited retinal dystrophy with different genetic backgrounds. The goal of this treatment is to introduce the gene encoding the isomerohydrolase enzyme involved in the production of 11-cis retinol using an AAV vector in the biallelic mutation of the Rpe65 gene (Wang et al., 2019; Kang and Scott, 2020; Britten-Jones et al., 2022; Girach et al., 2022; Kiser, 2022).

Glaucoma is a complex disease, and the exact genetic background that contributes to the development of this disease is not fully understood. Genome-wide association studies have identified 127 genetic loci associated with an increased risk of developing POAG (Aboobakar and Wiggs, 2022). These studies have helped to improve our understanding of the genetic factors that contribute to the development of glaucoma and may contribute to the development of new treatments for the disease. Gene therapy for glaucoma involves the delivery of a therapeutic gene to ocular cells, primarily RGCs for direct neuroprotection, or to trabecular meshwork and ciliary body cells to improve their function and reduce IOP. Therefore, gene therapy research in glaucoma focuses on two main aspects, neuroprotection of RGC or reduction of IOP (Borrás, 2017; Rhee and Shih, 2021).

Viral vectors used in glaucoma gene therapy studies

Viral vectors are the most commonly used vectors in glaucoma gene therapy studies. Both DNA and RNA viruses are used and are mainly adenoviruses (AVs), AAVs, and lentiviruses (LVs) (Daliri et al., 2017). AV contain double-stranded DNA that is highly immunogenic but relatively susceptible to manipulation. AV do not integrate their genome into the host DNA, which can lead to rapid loss of expression of the introduced gene, especially in proliferating cells. They have the largest transgene capacity of up to 37 kb and very high transfection efficiency (Sudhakar and Richardson, 2019). AAV viruses are parvoviruses. They contain single-stranded DNA without the ability to integrate into the host cell genome. They are characterized by very low immunogenicity and no pathogenicity. Despite all these positive features, the small capacity of ~3 kb of transgene DNA limits their usefulness. Among the different AAV serotypes, AAV2 shows the highest transfection efficiency of RGCs (Hellström et al., 2009; Mowat et al., 2014; Ratican et al., 2018; Rodriguez-Estevez et al., 2020). Horizontal and amacrine cells can also be transfected with AAV2 (Donahue et al., 2021). Recombinant AAV (rAAV) is another viral vector commonly used in glaucoma gene therapy research. It contains a reduced genome of coding sequences and a deletion of the Rep gene, resulting in a reduced ability to integrate into the host genome and low immunogenicity (Wang et al., 2019). There are several subtypes of rAAV. Self-complementary AAV (scAAV) and AAV with a triple tyrosine mutation (tyrosine substitution by phenylalanine) in the capsid are characterized by a high degree of RGC transduction efficiency (Mowat et al., 2014). LVs are retroviruses containing single-stranded RNA that integrate their genome into the host genome, which is associated with a risk of mutagenesis (Arsenijevic et al., 2022). Unlike AV and AAV, the transgene inserted by LV can be passed on to subsequent generations of cells, making it a useful vector for transfection of proliferating cells. LV can accommodate transgenes up to 10 kb (Moore et al., 2018; Poletti and Mavilio, 2021; Arsenijevic et al., 2022). Typically, all the viral vectors can be delivered to the retina by intravitreal and subretinal injections (Petrs-Silva et al., 2009; Mowat et al., 2014).

The right vector ensures efficient transfection of target cells, but the right promoter to drive gene expression allows for effective transgene expression. CMV, CAG, SYN1, Nefh, Thy1, and Mcp1 are promoters that are commonly used in ocular research, in combination with AAV. Despite of relatively low specificity for RGC, CMV and CAG promoters are characterized by a high degree of transduction, reaching up to 85%. In contrast, SYN1, Nefh, Thy1, and Mcp1 show higher specificity, but result in lower efficiency of expression in RGC (Ratican et al., 2018). Wang et al. (2020) evaluated the efficacy of different mouse and human promoters (CAG, Sncg, Thy1, Tubb3, GAP43, Isl2, CaMKIIa, RBPMS) and examined the expression of the reporter gene, enhanced green fluorescent protein (EGFP), in human pluripotent stem cells differentiated into RGC (hPSC-derived RGC), as well as in mouse RGC, after intravitreal injection. Sncg was found to be the promoter with the highest efficiency and specificity for RGC transfection. Furthermore, the mouse type of promoter (AAV2-mSncg-EGFP) was more efficient than the human type (AAV2-hSncg-EGFP) (Wang et al., 2020).

The Intraocular Pressure-Dependent and -Independent Animal Models of Glaucoma

In order to test experimental constructs in the form of gene therapy, various animal models of glaucoma have been used in different studies. Glaucoma was induced by IOP elevation by topical steroid application (Borrás et al., 2016), episcleral vein cauterization (Pacwa et al., 2023), polystyrene microbeads (Smedowski et al., 2014; Krishnan et al., 2016; Wójcik-Gryciuk et al., 2020), paramagnetic microbeads (Wu et al., 2020). The retinal ischemia/reperfusion model has been used in several studies as a faster alternative to induce optic nerve injury by transiently elevating IOP up to 110 mmHg for 60 or 90 minutes with intracameral saline injection (Igarashi et al., 2016; Tan et al., 2020b). Other potential methods to induce high-pressure glaucoma include laser coagulation of the trabecular meshwork (Pease et al., 2009) or silicone oil-induced ocular hypertension/glaucoma model (Fang et al., 2022). One of the most widely used glaucoma animal models is the DBA/2J mouse, in which mutations in two melanosomal protein genes, Tyrp1 and Gpnmb, result in pigment dispersion syndrome with spontaneous age-related elevation of IOP (Scholz et al., 2008; Nair et al., 2014; Hines-Beard et al., 2016; Donahue et al., 2021). IOP is elevated in most DBA/2J mice from 6 to 16 months of life, with significantly increased IOP at 8–9 months of age, with significantly elevated IOP by 12 months of age (Libby et al., 2005). Because glaucoma is not always associated with elevated IOP, i.e., NTG, studies have been conducted on low-tension glaucoma models based on induction of excitotoxic insult with N-methyl-D-aspartate (NMDA) (Shiozawa et al., 2020) or genetic manipulation of glutamate transporter genes (Harada et al., 2007; Kimura et al., 2020). Finally, optic nerve crush serves as a useful model of acute optic neuropathy and has also been used in some studies as a surrogate model for glaucomatous optic nerve injury because it induces RGC apoptosis (Osborne et al., 2018a; Cameron et al., 2020).

Augmentation and Silencing Gene Therapies for Glaucoma

The augmentation approach involves the introduction of additional DNA containing a functional or modified/enhanced version of the lost or malfunctioning gene into the cell. The exogenous gene produces a functional product to replace or enhance the action of the originally altered protein. Silencing gene therapy induces the downregulation of cellular processes to control disease by silencing specific genes thought to be associated with specific symptoms (Tsai et al., 2022). Among silencing strategies, the administration of small interfering RNA (siRNA) or short hairpin RNA (shRNA) are the leading methods. siRNA are double-stranded RNA molecules of typically 20–24 bp, while shRNA contain a tight hairpin turn. Both approaches act via the RNA-induced silencing complex, which becomes part of a machinery that leads to the synthesis of complementary mRNA fragments, resulting in cleavage of the target mRNA. Delivery of shRNA usually results in a longer-lasting silencing effect than siRNA, which is particularly important in rapidly proliferating cells (Martineau and Pyrah, 2007; Tsai et al., 2022; Figure 1).

Figure 1.

Overview of experimental gene therapies for glaucoma.

The IOP lowering gene therapies target the ciliary body and trabecular meshwork of the iridocorneal angle, resulting in IOP lowering. RGC neuroprotection therapies that directly target RGCs and provide IOP-independent neuroprotection. Created with Illustrator (Adobe.com). IOP: Intraocular pressure; RGC: retinal ganglion cell.

IOP-targeting gene therapy for glaucoma

RGC death is most commonly associated with elevated IOP, making IOP lowering the most effective and only treatment option. However, high IOP is not the only pathological factor for glaucoma progression. One third of glaucoma patients with normal IOP will develop typical signs and symptoms of glaucoma (i.e., NTG), and this ratio is even higher in the Asian population (Iwase et al., 2004). On the other hand, ocular hypertension without optic nerve damage is approximately eight times more common than glaucoma, suggesting that glaucoma may be a primary optic nerve disease that makes RGCs more susceptible to risk factors such as elevated IOP (Pfeiffer et al., 2013). IOP reduction has been the target of gene therapy using the AAV-shh10 construct carrying the CRISPR-Cas9 system, resulting in disruption of the aquaporin 1 gene (AAV-shh10-AQP1), which plays a critical role in aqueous humor production (Wu et al., 2020). In this study, intraocular injections of AAV vectors were performed in a mouse model of corticosteroid- and microbead-induced ocular hypertension and resulted in a reduction of IOP through selective editing of the AQP1 gene in the ciliary epithelium. This therapy may provide a potential opportunity to administer a single injection to achieve a sustained therapeutic effect of lowering IOP.

Clinical observation of POAG patients with reduced aqueous levels of cyclooxygenase-2 (COX2), the key enzyme for prostaglandin synthesis, and the efficacy of POAG treatment with topical prostaglandin analogues led to attempts to conduct a gene therapy study using a transgene for the prostaglandin F2α receptor and COX2. The feline immunodeficiency virus - LV vector was used to deliver transgenes to the ciliary epithelium and trabecular meshwork in a normotensive feline glaucoma model (LV-pCOX2^co-G). Similarly, Chern et al. used a transgene carried by a rAAV vector (rAAV2/2[MAX]-smCBA-TC40-COX2-P2A-PTGFR-TC45) and obtained a 12-month dose-dependent IOP reduction of 12.6–43.2% in normotensive Brown Norway rats (Chern et al., 2022).

RhoA is a GTPase that activates RhoA kinase and is responsible for the dynamic reorganization of actin filaments in cells. It is also involved in apoptosis and cell proliferation mechanisms. RhoA kinase inhibitors block the polymerization of the actin cytoskeleton in trabecular meshwork cells and improve aqueous humor outflow, resulting in a reduction in IOP. RhoA kinase inhibitors are currently used as topical treatments for glaucoma and ocular hypertension (e.g., netarsudil, ripasudil) (Tanna and Johnson, 2018). Borrás et al. (2015) introduced an scAAV2 containing a transgene that results in the expression of a dominant-negative mutant RhoA protein lacking the green fluorescent protein (GFP) binding site (scAAV2-dnRhoA2). Transfection of trabecular meshwork cells by intracameral injection of this AAV construct in a rat model prevented RhoA activation and resulted in significant nocturnal IOP reduction (Borrás et al., 2015).

The exoenzyme C3 transferase (C3), an ADP-ribosyl transferase, has the potential to selectively ribosylate RhoA, B and C proteins and can inhibit RhoA. By affecting the outflow in Schlemm’s canal, it leads to the regulation of aqueous humor outflow (Arsenijevic et al., 2022). Tan et al. (2020a) used scAAV2-C3 to transfect the cultured trabecular meshwork cells and confirmed the remodeling of the cytoskeletal structure of the trabecular meshwork. In addition, in vivo scAAV-C3 intravitreal injection in mouse and monkey models resulted in IOP reduction with concomitant corneal edema due to unintended transfer of AAV2-C3 vector to corneal endothelial cells (Tan et al., 2020a). Similar results were obtained with the LV-C3 construct in cultured trabecular meshwork cells and in a monkey model (Tan et al., 2019).

Exogenous steroids, i.e. dexamethasone, decrease the phagocytic ability of trabecular meshwork cells, resulting in increased extracellular matrix deposition, increased resistance to aqueous humor outflow, and subsequent elevation of IOP. Metalloproteinase 1 (MMP1) belongs to the group of collagenases that are down-regulated in trabecular meshwork cells under steroid treatment, which further leads to the accumulation of extracellular matrix proteins, further disturbances in aqueous humor outflow and ocular hypertension (Zhang et al., 2007; Borrás, 2017). Based on these considerations, gene therapy using AAV-MMP1 gene delivery was attempted in steroid-induced ocular hypertension. Intracameral injection of scAAV2-GRE-MMP1 resulted in increased MMP1 content in trabecular meshwork cells and subsequent reduction of steroid-induced IOP elevation in the sheep model (Borrás et al., 2016).

Myocilin is a structural protein that is highly expressed in the trabecular meshwork. Mutations in the MYOC gene are responsible for approximately 4% of glaucoma cases and lead to the development of juvenile glaucoma (Tamm, 2002). The pathogenic effects of MYOC gene mutations include protein misfolding, ER stress, and cytotoxicity in trabecular meshwork cells, leading to increased aqueous humor outflow resistance and elevated IOP. The gene therapy constructed of adenovirus vector with CRISPR assembly (AV5-crMYOC) has been tested in vitro and in vivo, in NTM5 (normal trabecular meshwork 5) cell line expressing mutant MYOC containing the most common, Y437H mutation in MYOC gene, and in high-tension mouse glaucoma model (Tg-MYOC^Y437H) carrying similar mutation (Y437H), respectively (Jain et al., 2017). Application of AV5-crMYOC in NTM5 mutant cells improved clearance of MYOC protein aggregates, decreased markers of oxidative cell stress, and improved cell survival. In an animal model, intravitreal injection of Ad5-crMYOC prevented age-related IOP elevation and lowered IOP in animals with elevated IOP levels.

RGC-neuroprotection-targeting gene therapy for glaucoma

It is believed that factors other than elevated IOP may be involved in the pathogenesis of glaucoma. To date, various risk factors for glaucoma development have been identified, such as axonal transport failure, neurotrophic factor deprivation, activation of intrinsic and extrinsic apoptotic signaling, mitochondrial dysfunction, excitotoxic insult, oxidative stress, reactive gliosis, and synaptic connectivity breakdown (Kimura et al., 2016). Considering that the currently available treatment options, i.e. IOP lowering, are not sufficient to prevent terminal glaucoma progression, addressing the aspect of IOP-independent RGC neuroprotection provides new hope for a more sufficient treatment of this disease.

Neurotrophic factors

Decreased expression of genes involved in cellular metabolism and axonal transport, resulting in neurotrophic factor deprivation, has been described in animal models of glaucoma (Dias et al., 2022). There have been several attempts to utilize the neurotrophic factors, mainly brain-derived neurotrophic factor (BDNF), which plays a key role in the survival, plasticity, and growth of neurons, including RGCs (Dekeyster et al., 2015; Kimura et al., 2016; Kuo and Liu, 2022). According to the concept of disruption of retrograde transport of neurotrophic factors in glaucoma, decreased levels of BDNF and upregulation of its receptor, TrkB, have been found in glaucomatous RGC (Domenici et al., 2014; Kimura et al., 2016). As a result of direct BDNF protein supplementation, RGC apoptosis was reduced in an animal model of glaucoma (Fu et al., 2009; Johnson et al., 2011; Mysona et al., 2017), allowing the development of further studies on gene therapies based on the use of BDNF transgenes. Wójcik-Gryciuk et al. (2020) applied the intravitreal augmentation gene therapy AAV2-BDNF in a rat model of glaucoma. A unilateral model of high tension glaucoma was created by intracameral injection of polystyrene microbeads (Smedowski et al., 2014), followed by intravitreal injection of AAV2-BDNF. The authors concluded that overexpression of BDNF in the glaucoma model allowed to normalize the retinal concentration of TrkB, primarily elevated, as a hypersensitivity RGC response to reduced BDNF levels, providing RGC neuroprotection (Wójcik-Gryciuk et al., 2020). Similarly, Osborne et al. (2018a) based their gene therapy study on different AAV2-BDNF constructs. The research was conducted in HEK293 and SH-SY5Y neuroblastoma cells undergoing hydrogen peroxide-induced oxidative insult, as well as in vivo in normotensive mice. In addition, the authors used a combined construct consisting of modified BDNF (without proBDNF coding sequence) and TrkB (TrkB-2A-mBDNF). The combined approach, AAV2-TrkB-2A-mBDNF, resulted in a reduction in cell apoptosis in vitro that was comparable to AAV2-BDNF or AAV2-TrkB alone, as analyzed by terminal deoxynucleotidyl transferase dUTP nick end labeling staining (Osborne et al., 2018a). Analogous therapy, performed in an animal model of optic nerve crush in mice and in a high-pressure glaucoma model in rats, showed that the combined AAV2-TrkB-2A-mBDNF therapy was more neuroprotective than either AAV2-BDNF or AAV2-TrkB alone. In electroretinography (ERG), there was an improvement in the positive scotopic threshold response, with no change in the amplitude of the a- and b-waves (Osborne et al., 2018b). In contrast, Nishijima et al. (2023) tested an intravitreal gene therapy encoding only the intracellular domain of TrkB (iTrkB), which results in more constitutive activation and allows for more efficient packaging into the AAV vector due to the reduced size of the transgene (AAV-F-iTrkB). The authors tested the AAV construct in several animal models of optic nerve damage. In the NTG model of glutamate/aspartate transporter knockout mice, intravitreal treatment with AAV-F-iTrkB alleviated RGC loss by histology, limited thinning of the ganglion cell complex by optical coherence tomography, and preserved visual responses in the multifocal ERG. In the high-voltage silicone oil-induced glaucoma model, the applied gene therapy prevented RGC loss by histology. AAV-F-iTrkB treatment was also effective in protecting RGCs in an acute optic nerve crush model. In the latter, it not only prevented loss of RGC bodies, but also protected synaptic connections between cells, provided protection of the RGC dendritic tree, and induced axonal outgrowth (Nishijima et al., 2023).

Shiozawa et al. (2020) used tyrosine triple mutan AAV2 as a vector for BDNF gene transduction (tm-scAAV2-BDNF), using a mouse model of excitotoxic insult (Zhou and Wollmuth, 2017; Armada-Moreira et al., 2020) induced by an intravitreal injection of NMDA. This gene therapy reduced the loss of RGCs (positive for Brn3a) and limited the thinning of the inner retina, as well as improved the amplitude of the b-wave in the ERG, which was accompanied by a reduction in reactive gliosis in the retina (Shiozawa et al., 2020). Similar results were reported using the viral vector tm-scAAV2-BDNF in a mouse model of retinal ischemia/reperfusion induced by transient elevation of IOP with intracameral saline injection (Igarashi et al., 2016).

In addition to BDNF, which is the most studied neurotrophic factor in experimental glaucoma therapies, there are also reports demonstrating the efficacy of augmentation gene therapy with other growth factors and neurotrophins, such as ciliary neurotrophic factor (Pease et al., 2009) or vascular endothelial growth factor (Shen et al., 2018). AAV2-PEDF (pigment epithelium-derived factor) gene therapy promoted RGC survival (positive for Tuj1) and altered fibroblast growth factor-2, interleukin (IL)-1β, ionized calcium-binding adaptor molecule 1, and glial fibrillary acidic protein immunostaining in the inner retina of mice after optic nerve crush, although it did not promote axonal outgrowth. The effects of the therapy were significantly enhanced when applied in combination with human mesenchymal stem cells and were additionally associated with pronounced axonal outgrowth (Nascimento-Dos-Santos et al., 2020).

Regulation of apoptosis and neurodegeneration

The exact mechanism of RGC death in glaucoma is still unknown, although apoptosis has been suggested to be the final common pathway for cell death (You et al., 2013; Levkovitch-Verbin, 2015; Pietrucha-Dutczak et al., 2018). There are several stages of RGC damage occurring during the development of glaucoma and RGC death is considered to be a biphasic process. First, one or more trigger factors (i.e., increased IOP, ischemia or simply aging of the organism) evokes the primary, but limited damage to RGC. Subsequently, oxidative stress and excitotoxicity, due to excess of glutamate release from dying neurons, induce secondary damage to surrounding neurons. The chronic neurodegeneration occurs, and intracellular glutamate is released from the dying cells and dispersed among other neighboring cell populations, triggering a cascade of apoptotic events leading to further cell death. Due to the undoubtful role of apoptosis in RGC loss in glaucoma, attempts have been made to use anti-apoptotic gene delivery as a therapeutic target. During the activation of the intrinsic apoptotic pathway, the Bax protein plays a pivotal role, leading to the formation of the major outer membrane protein complex in the outer mitochondrial membrane. This complex is a pore through which cytochrome c passes into the cytoplasm, activating further stages of cell death (Almasieh et al., 2012). Bcl-2 and Bcl-xL delay the RGC death due to the anti-apoptotic mechanism (Bax antagonize and inhibit the major outer membrane protein complex formation) (Harder et al., 2012; Voss and Strasser, 2020). Donahue et al. (2021) applied intravitreal injection of AAV2-Bcl-xL-mCherry in 5-month-old DBA/2J mice. Expression of Bcl-xL did not prevent IOP elevation but attenuated the RGC soma pathology and axonal degeneration in the optic nerve in IOP-independent manner.

Inhibition of RGC apoptosis has also been attempted via viral-mediated ocular delivery of the X-linked inhibitor of apoptosis, Xiap gene (also known as Birc4), a caspase 3 inhibitor. McKinnon et al. (2002) performed intravitreal injection of AAV-Birc4 driven by the CBA promoter in a high-tension rat glaucoma model obtained via hypertonic saline injection to the limbal vessels, which resulted in greater than 30% improvement in survival of optic nerve axons. A similar study, using the AAV-Xiap, applied in a glaucoma magnetic microbead mouse model, by Visuvanathan et al. (2022) demonstrated functional (pattern ERG) and structural neuroprotection of the RGC bodies and optic nerve axons, accompanied by reduction of reactive gliosis in the optic nerve.

Fas ligand (FasL) is a transmembrane protein belonging to the TNF family. FasL interaction with Fas receptor on cell surface participates in apoptosis induction. FasL is represented by two forms, membrane (mFasL, mCD95L) which shows pro-apoptotic activity and soluble (sFasL, sCD95L) which attenuates the activity of mFasL (Devel et al., 2022). Krishnan et al. (2016) used AAV-mediated gene delivery of sFasL (AAV2-sFasL) to achieve RGC neuroprotection in two mouse models of optic neuropathy, acute (microbead high-pressure glaucoma model) and chronic (DBA/2J glaucomatous mice). As a result of the applied gene therapy, preservation of RGC and axon density, inhibition of glial activation and reduction of TNFα levels were observed. This neuroprotective effect was independent of IOP. There was also a significant change in the levels of apoptosis regulatory proteins, a decrease in TNFα and pro-apoptotic mediators Fas, FADD and Bax, and an increase in the expression of anti-apoptotic factors, cCLIP (cellular FLICE-like inhibitory protein), Bcl2, cIAP2 (cellular inhibitor of apoptosis 2) (Krishnan et al., 2016).

Nicotinamide mononucleotide adenylyltransferase 2 (Nmnat2) is the enzyme involved in the NAD⁺ biosynthesis process that restores the proper level of NAD⁺. Nmnat2 is highly expressed in the brain and nervous system and has been implicated in axonal survival (Hicks et al., 2012). The beneficial effect of Nmnat2 gene therapy on RGC survival has been demonstrated. Overexpression of Nmnat2 by intravitreal injection of an AAV2-Nmnat2 construct driven by the mSncg promoter, enhanced the RGC neuroprotection and limited optic nerve axon degeneration. This therapy demonstrated the preservation of visual function in two models of optic neuropathy, the traumatic optic neuropathy model (optic nerve crush) and the silicone oil-induced ocular hypertension/glaucoma model glaucoma model (Fang et al., 2022). The use of the mSncg promoter allowed precise manipulation of RGC gene expression without the risk of affecting other retinal cell populations, and increased the specificity and potency of transgene expression in RGC.

Wang et al. (2020) conducted studies using the AAV viral vector driven by the mSncg promoter with CRISPR-CAS9 gene editing, resulting in inhibition of the RGC-specific pro-degenerative genes Digit3 and Sarm1. In glaucoma, the phenomenon of downregulation of Sncg during RGC degeneration has been reported previously (Surgucheva et al., 2008). Mice with acute optic neuropathy (optic nerve crush), treated with an intravitreal injection of AAV-mSncg-CRISPR-CAS9, using guide RNA genes Digit3 and Sarm1, showed reduced expression of these genes, preservation of ganglion cell complex thickness, and prevention of RGC soma shrinkage and optic nerve axon loss (Wang et al., 2020).

In addition to its direct effect on IOP, the exoenzyme C3 transferase may also exert neuroprotective effects on RGCs in an IOP-independent manner. Tan et al. performed a gene therapy study with bilateral intravitreal injection of scAAV2 vectors encoding C3 protein (scAAV2-C3) in rats, followed by a short-term unilateral ischemia/reperfusion model on day 7 after viral particle injection. Terminal deoxynucleotidyl transferase dUTP nick end labeling and cleaved caspase 3 immunostaining were used to measure the number of apoptotic cells. This therapy showed a significant reduction in the expression of RhoA, caspase-3 protein, and the number of apoptotic RGCs, as well as a significant attenuation of retinal thickness loss in the ischemia/reperfusion model (Tan et al., 2020b).

Heat shock proteins (Hsps) are stress response factors that may be involved in neurodegeneration processes by regulating the cell cycle and cellular response to proteotoxic and genotoxic insults (Pietrucha-Dutczak et al., 2018). Proteotoxic stress is often associated with protein misfolding and cellular aggregation as a result of impaired clearance systems. Small Hsps (sHsps) are ATP-independent molecular chaperones with the potential to prevent the formation of protein aggregates. The efficacy of sHsp as a gene therapy approach for RGC protection has been tested in the acute optic neuropathy model of ischemia/reperfusion and in the high tension microbead glaucoma model (Nam et al., 2022). Of the constructs tested, AAV2-HspB1, -HspB4, -HspB5, or -HspB6, containing the mini-RGC-specific promoter Ple345 (neurofilament light, NefL), the AAV2-HspB1 showed the most superior efficacy in preventing RGC loss, axonal transport disturbances, and glial activation in the animal models tested.

Immunomodulation

Dysregulation of the complement system, primarily increased expression of C1q and C3 components, has been reported in glaucoma and ocular hypertension (Harder et al., 2017). However, it is not entirely clear whether this phenomenon is a primary cause of glaucomatous damage or a secondary consequence of retinal changes due to elevated IOP. Bosco et al. (2018) used intravitreal injections of the viral vector AAV2-CR2-Crry carrying the Crry gene for the C3 complement receptor 2 (CR2) inhibitor protein. As a result, Crry protein was overexpressed in the RGC and inner retina of the DBA/2J mouse model of glaucoma, leading to a decrease in C3 complement component levels. The authors reported a neuroprotective effect of the applied gene therapy on RGC bodies and axons, and a significant delay in disease progression despite persistently elevated IOP (Bosco et al., 2018). In parallel to the pro-degenerative activity of complement component C3, it may also have a protective function towards RGCs. The study by Harder et al. (2017) indicates the neuroprotective role of astrocyte-derived C3 complement component in the early stages of glaucoma in response to elevated IOP in DBA/2J.Wlds mice, which may be related to epidermal growth factor receptor signaling.

Modulation of neuroinflammation

Erythropoietin (EPO) exerts neuroprotective effects on neurons by modulating inflammation in neurons, astrocytes, and microglia (Bond and Rex, 2014). EPO reduces cellular damage caused by oxidative/nitrative stress, limits immune cell recruitment and infiltration, reduces microglial proliferation and reactivity, and prevents apoptosis. Hines-Beard et al. (2016) used a gene therapy approach to transfer the gene encoding erythropoietin with attenuated erythropoietic activity (EpoR76E) by treating 5-month-old DBA/2J with the rAAV-EpoR76E vector. Reduced expression of proinflammatory cytokines, such as IL-1, IL-12, IL-13, IL-17, and chemokine (C-C motif) ligand 4 and 5, and increased levels of antioxidant proteins were observed in retinas from 8-month-old mice. The authors also reported preservation of axonal anterograde transport and a significant reduction in microglial numbers, but not proliferation. Functional measures, flash visual evoked potentials, showed preservation of N1 and P1 amplitudes compared to control animals.

Targeting NMDA excitotoxicity

Glutamate binding to NMDA receptors leads to the Ca²⁺ influx into the RGC cytosol, resulting in the excitotoxic damage and downregulation of Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) signaling and subsequent induction of apoptosis (Hartwick et al., 2008; Guo et al., 2021). CaMKII activation exerts a neuroprotective effect on RGCs under both healthy and degenerative conditions in the retina. Guo et al. (2021) showed that intravitreal application of AAV-CaMKIIα in two animal models, the microbead high-pressure glaucoma model and the glutamate/aspartate transporter-deficient mouse NTG model, protects RGC bodies and their long-range projections, thus preserving functional vision.

Regulation of transcription factors

MYC-associated protein X (Max) is a ubiquitous transcription factor involved in cell proliferation, differentiation, and apoptosis (Grandori et al., 2000). Since RGC degeneration and apoptosis are associated with decreased Max expression (Petrs-Silva et al., 2004), Lani-Louzada et al. (2022) used intravitreal gene therapy with rAAV2-Max in the rat high-pressure glaucoma model (limbal vessel cauterization) and in the rat optic nerve crush model. Cultures of rat retinal explants isolated from eyes of postnatal day 14 rats that received intravitreal injections of rAAV2-Max at birth were also analyzed. Increased expression of Max was detected compared to controls (rAAV2-GFP), which prevented RGC loss and improved RGC function as assessed by pattern ERG (Lani-Louzada et al., 2022).

Nuclear factor-E2-related factor 2 (Nrf2), a transcription factor activated by oxidative stress and electrophilic molecules shows relevance in many ocular diseases, including glaucoma (Pietrucha-Dutczak et al., 2018). In the field of gene therapy research on RGC neuroprotection, there are also scientific reports on beneficial effects achieved by increasing the expression of Nrf2. Fujita et al. (2017) used the AAV2-Nrf2 vector with the Mcp1 or CMV promoter in a mouse optic nerve crush model. As a result of the treatment, there was a significant histologic neuroprotection detected as a decrease in the number of Sytox-positive cells, indicating less RGC death, and a better number of Brn3a-positive RGCs compared to the control. Despite better RGC survival, there was no evidence of functional neuroprotection using visual acuity measurements, regardless of the type of promoter used (Mcp1 or CMV), but there was an improvement in contrast sensitivity (Fujita et al., 2017).

Superoxide dismutase (SOD) is one of the most important and active Nrf2-regulated antioxidant defense enzymes in almost all living cells exposed to oxygen (Pietrucha-Dutczak et al., 2018). Since the oxidative stress is one of the common mechanisms of neurodegeneration (Himori et al., 2013; Lin and Kuang, 2014; Naguib et al., 2021; Sanz-Morello et al., 2021), the enhancement of antioxidants and antioxidant enzymes has become a target of interest in optic neuropathies. Jiang et al. (2016) aimed to analyze the effects of AAV2-SOD2 gene therapy on RGC survival and mitochondrial dysfunction in the rat high-tension glaucoma model (trabecular meshwork and episcleral vein cauterization). The authors showed that high-pressure glaucoma is associated with decreased SOD and catalase activities, mitochondrial dysfunction and increased malondialdehyde presence. Preconditioning with AAV2-SOD2 attenuated RGC damage and mitochondrial disintegration (Jiang et al., 2016).

Silencing gene therapies for glaucoma

Similar to augmentation gene therapies, experimental silencing therapies target either the IOP level itself or IOP-independent RGC neuroprotection (Naik et al., 2021).

The IκB kinase (IkappaB kinase or IKK) is part of the enzymatic complex that regulates the cellular response to inflammatory stimuli. Binding of IκB to NF-κB inactivates the latter, and phosphorylation of IκB allows release of NF-κB, which further regulates MMP expression in the trabecular meshwork and ciliary body. It has been documented that MMP content in these structures plays an important role in IOP regulation (Weinreb et al., 2020). The changes in MMPs, via effects on IκB, have become a therapeutic target for IOP-lowering silencing therapies in rodent and non-primate models. Zeng et al. (2020) investigated hyperbranched cationic glycogen derivative-mediated IκBα gene silencing using intracameral injection of IκBα-siRNA-loaded DMAPA-Glyp complex in normotensive rats, which resulted in a short-lasting but significant reduction in IOP levels related to MMP2 levels. In a high-pressure glaucoma model in rhesus monkeys (laser coagulation of the trabecular meshwork), intracameral injection of IκBα-siRNA resulted in a significant reduction of elevated IOP that lasted for 28 days. This effect was correlated with increased expression of MMP2 and MMP9 in the trabecular meshwork and ciliary body (Sun et al., 2022).

SYL040012 is a siRNA drug candidate for IOP reduction in glaucoma and ocular hypertension that selectively interferes with the β2 adrenergic receptor in the ciliary body, thereby reducing elevated IOP (Martínez et al., 2014). The in vivo efficacy of the silencing was tested in normotensive New Zealand White rabbits, with a 20% reduction in IOP achieved and lasting up to 190 hours (Martínez et al., 2014).

SH2 domain containing tyrosine phosphatase 2 (Shp2) participates in transmembrane signaling by receptor tyrosine kinases and regulates various intracellular pathways, but most importantly decreases the expression of several growth factor receptors. In the case of RGC, Shp2 interacts with Caveolin1 (Cav1) and through this interaction, it impairs BDNF/TrkB signaling (Abbasi et al., 2021). The effects of gene silencing therapy (AAV2-shRNA-Shp2) were tested in wild-type and Cav1^–/– mouse models of high pressure glaucoma (using polystyrene microbeads) and ischemia/reperfusion model (intracameral saline injection). Shp2 knockdown attenuated RGC apoptosis and provided histological evidence of neuroprotection (Brn3a and terminal deoxynucleotidyl transferase dUTP nick end labeling staining) as well as functional neuroprotection (preservation of pSTR in ERG) in the wild-type glaucoma mouse model, but not in Cav1^–/– mice, further demonstrating that Shp2 acts through interaction with caveolin1 (Abbasi et al., 2021).

The long non-coding RNA growth arrest-specific transcript 5 (GAS5) regulates proliferation, apoptosis, and cell migration, and alters insulin signaling, neuronal survival, and neuroinflammation (Goustin et al., 2019; Patel et al., 2023). Knockdown of GAS5 using shRNA resulted in increased EZH2 expression and decreased ABCA1 expression in RGCs of a rat model of high pressure glaucoma (limbal laser coagulation). As a result of upregulation of EZH2 and suppression of ABCA1, a neuroprotective effect towards RGC was observed (Zhou et al., 2019).

The membrane protein Nogo-A and its associated receptor NogoR-1 (NogoR-66) are expressed in many neuronal cells and oligodendrocytes and play an inhibitory role in neuroregeneration. Upon insult, downregulation of Nogo-A promotes neurite outgrowth and functional recovery (Schwab, 2010). Antagonizing NogoR-1 or inactivating the Nogo-A protein provides neuroprotection, synaptic protection of RGC and prevents visual deterioration, as almost all RGC express NogoR-1 (Fu et al., 2011; Wang et al., 2015; Mdzomba et al., 2018; Solomon et al., 2018). Oncomodulin is a calcium-binding protein belonging to the parvalbumin family that is derived from macrophages and released into the vitreous and retina, and can promote optic nerve regeneration after insult (de Lima et al., 2012a, b; Williams et al., 2020; Wong and Benowitz, 2022). Application of gene therapy consisting of siRNA-NogoR-1 delivered by a non-viral approach with an oncomodulin-truncated protein with high affinity for its RGC receptors (siRNA-OM/tp-Nogo-1) to primary rat RGC culture resulted in promotion of axonal outgrowth, increased intracellular cyclic adenosine monophosphate levels, and decreased RhoA activity (Cui et al., 2014).

Perspectives and Conclusions

There is increasing evidence for the efficacy of various experimental gene therapies in glaucoma, but they are still in the early stages of research and there is a lack of human clinical trials using these sophisticated treatment modalities. According to the analysis of the glaucoma patient cohort presented by Gruzei et al. (2022), the eligibility of glaucoma patients for gene therapy, defined as rapid disease progression despite controlled IOP (visual field progression, MD slope < –1 dB/year), was calculated to be 10% of the 374 patients in the cohort. An analysis of the available PubMed reports (excluding reviews, systematic reviews and meta-analyses) and clinical trials database (only for interventional studies) shows a clear lack of translational studies that would lead to human trials for gene therapies in glaucoma, thus there is an emerging need to intensify research in this niche area of ophthalmology (Figure 2).

Figure 2.

Analysis of PubMed and ClinicalTrial.gov search results for specific search terms.

PubMed search results were filtered to exclude reviews, systematic reviews, and meta-analyses to analyze only research articles. In ClinicalTrial.gov, only interventional studies were included in the analysis. The translational coefficient was calculated as the percentage ratio of search results for the same phrases in ClinicalTrial.gov and the PubMed database. The last three search phrases, cancer, stroke, and Alzheimer’s disease have been used for comparison purpose. Out of analyzed phrases, the best translational coefficient was calculated for “optic neuropathy” and “gene therapy” search phrase, which could be explained by very intense research progress achieved in the topic of gene therapies for Leber’s Hereditary Optic Neuropathy, out of which many is currently tested in clinical trials. “Gene therapy”, in overall, showed second-best translational coefficient, which reflects on all gene therapy clinical trials conducted for variety of diseases. The worst translational coefficients were calculated for glaucoma and “gene therapy” (0%), neuroregeneration, neuroprotection, glaucoma and regeneration, regeneration phrases. Non-glaucomatous optic neuropathies showed noticeable better coefficients than glaucoma and they were higher than overall glaucoma translational coefficient. Glaucoma and neuroprotection studies in ClinicalTrials.gov search represent NT-501 CNTF implant, near to infrared and green light retinal stimulation, metformin, GlaucoT Glaucoma Treatment Device, nicotinamide, CBS eyedrop, Brimonidine tartrate 0.2% ophthalmic solution, GlaucoCetin, DNB-001, Gingko Biloba, and α-tocopherol trials. Glaucoma and regeneration studies in ClinicalTrials.gov represent topical insulin, and NT-501 CNTF implant trials. All trials representing “Optic neuropathy” and “gene therapy” were conducted in LHON. “Optic neuropathy” and neuroprotection search in ClinicalTrials.gov represent one gene silencing therapy with QPI-1007 in non-glaucomatous neuropathies, citicoline, NT-501 CNTF implant, near to infrared and green light retinal stimulation, ACTHAR, phenytoin, and minocycline trials. “Optic neuropathy” and regeneration represent NT-501 CNTF implant, and citicoline trials. Green colored circles represent search results used to calculate translation coefficients for glaucoma and optic neuropathy studies. Other colors refer to comparison groups (cancer, stroke, Alzheimer’s disease) representing non-ocular pathologies. Created with Illustrator (Adobe.com).

In particular, with recent technological advances and the elucidation of disease pathways, there is an increasing likelihood that future glaucoma treatment will focus more on direct, IOP-independent neuroprotection of RGCs and regeneration of the optic nerve to reverse glaucomatous damage and alleviate the socioeconomic burden of this disease. This potentially brings gene therapy and genome editing to the forefront of the glaucoma field, as they can be tailored to target specific disease pathways and have long-term and beneficial effects, providing new hope for glaucoma patients and a new tool for glaucoma specialists.

Limitations

This review provides a comprehensive overview of the topic of gene therapy for glaucoma. For the purpose of the review, we performed a very strict selection of the articles included in the analysis, which could be considered as a strength of the study. We performed a search based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline (Haddaway et al., 2022; Figure 3). However, there are several potential limitations to this report. The literature search was conducted using PubMed and ClinicalTrials.gov only, and we did not analyze other databases, so some studies may have been missed. In the analysis, we included only original studies from PubMed and interventional studies from ClinicalTrials.gov with available full text and published in English. We limited the search to the term “gene therapy,” but there are other names for this particular intervention, such as “gene delivery,” etc. We believe these represent a minority of the search and some were also identified using the terms “neuroprotection” and “regeneration”.

Figure 3.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart.

Additional file: Open peer review report 1 ^{(75.3KB, pdf)} .