Figure 1|Experimental models of RGC degeneration and methods to evaluate gene function in RGC survival.
Experimental models of RGC degeneration: (1) Excitotoxicity model: Historically, excitotoxicity was first demonstrated to kill the inner retinal neurons after systematic glutamate injections. In adult mice, a single injection of high dose of N-methyl-D-aspartate (NMDA) into the vitreous chamber will induce massive RGC death (Nakamura et al., 2021) (Figure 1A). In our hands, an intravitreal administration of toxic levels of NMDA solution (1.5 μL, 20 mM) will lead to more than 80% of RGC loss one week after injection in the adult mouse retina. (2) Optic nerve crush model: The optic nerve in mice can be easily accessed intraorbitally and crushed by fine forceps. Optic nerve crush is a widely used model to investigate RGC survival and axon regeneration (Figure 1B). Axotomy caused by optic nerve crush results in delayed death of RGCs, with about 75% loss of RGCs 2 weeks after injury (Guo et al., 2016). (3) Glaucoma model: Several induced and genetic models have been developed to mimick elevated intraocular pressure (IOP) of glaucoma patients (Figure 1C). Injection of microbeads into the anterior chamber to occlude aqueous outflow is a popular way to induce ocular hypertension in mice (Calkins et al., 2018). The RGC loss is about 30% at 8 weeks after microbead injection (Yang et al., 2016). (4) Ischemia/reperfusion model: Retinal ischemia/reperfusion could be induced by raising the IOP to ultra-high level (such as 120 mm Hg) for 60 minutes by placing a needle connecting an elevated saline reservoir to the anterior chamber (Figure 1D). Ischemia results in a significant loss of RGCs. In the meantime, ischemia also leads to loss of other inner retinal neurons.
Methods to evaluate gene function in RGC survival: To evaluate the role of a gene in RGC survival, initial examination often involves probing gene expression levels or its enzymatic activities before and after injury. Mechanistic investigation requires gain-of-function and loss-of-function assays. Although chemical activators and inhibitors can be used to alter gene function or enzymatic activities, more convincing results require direct manipulation of gene expression or enzymatic activities. Intravitreal injection of AAVs (adeno-associated viruses) is an efficient way to deliver genes into RGCs (Figure 1E). When combined with molecular and genetic techniques such as site-directed mutagenesis, Cre-LoxP, and CRISPR-Cas9, AAV-mediated gene transfer could achieve various purposes to manipulate a gene. Quantification of the RGC numbers from the gain-of-function and loss-of-function studies of a gene will help determine its role in RGC survival.
Clear and definitive labeling of RGCs is a prerequisite for RGC quantification. The well-established method to label all RGCs is to perform immunohistochemistry using antibodies recognizing neuron-specific class III beta-tubulin (Tuj1) (Guo et al., 2016) or RNA-binding protein with multiple splicing (RBPMS) (Masin et al., 2021). Tuj1 and RBPMS are more reliable markers for RGC quantification compared to others such as Brn3a/3b as their expression levels may have been downregulated at the early phase of RGC degeneration after injury onset.
RGC density in mice varies in different retinal quadrants as well as different distances from the retinal center. To achieve objective comparison of RGC numbers among experimental and control groups, sampling multiple regions in retinal flat-mounts from all quadrants at a fixed radius is a more reliable method than sampling retinal sections. Based on our experience, following Tuj1 immunostaining, sampling square regions located ~500 μm from the edge of retinal flat-mounts provides consistent quantification of RGC numbers as these regions contain less dense RGC axons allowing clearly discernible RGC somas to be scored (Guo et al., 2016) (Figure 1F).