Biology and therapy of inherited retinal degenerative disease: insights from mouse models

Abstract

Retinal neurodegeneration associated with the dysfunction or death of photoreceptors is a major cause of incurable vision loss. Tremendous progress has been made over the last two decades in discovering genes and genetic defects that lead to retinal diseases. The primary focus has now shifted to uncovering disease mechanisms and designing treatment strategies, especially inspired by the successful application of gene therapy in some forms of congenital blindness in humans. Both spontaneous and laboratory-generated mouse mutants have been valuable for providing fundamental insights into normal retinal development and for deciphering disease pathology. Here, we provide a review of mouse models of human retinal degeneration, with a primary focus on diseases affecting photoreceptor function. We also describe models associated with retinal pigment epithelium dysfunction or synaptic abnormalities. Furthermore, we highlight the crucial role of mouse models in elucidating retinal and photoreceptor biology in health and disease, and in the assessment of novel therapeutic modalities, including gene- and stem-cell-based therapies, for retinal degenerative diseases.

Keywords: Mouse mutants; Photoreceptor; Retinal development; Retinal disease.

Fig. 1.

Structure of the human and mouse eye. (A) Schematic cross-sections of the human and mouse eye. Light is focused by optical elements (such as cornea and lens) on the neural retina at the back of the eye. The central cone-only region of the human retina is called the fovea and is responsible for high resolution. The region surrounding the fovea is termed macula and contains higher density of cones compared with the peripheral retina. The area of human retina is ~1094 mm², with the macula and fovea being ~3 and 1.5 mm², respectively (http://webvision.med.utah.edu). The total number of rods and cones in the human retina are 120 million and 6 million, respectively. The highest density of cones is at the center of the fovea (~161,900/mm²), which has no rods. The mouse retina lacks a distinct fovea and/or macula. The retinal pigment epithelium (RPE) monolayer separates the choroidal blood supply from the photoreceptors and is crucial for visual function. The lens is much larger in mouse than humans relative to the eye size. (B) Photograph of a mouse retinal section stained with hematoxylin and eosin, indicating different cellular layers. The outer nuclear layer (ONL) contains photoreceptor cell bodies, from which the inner segment (IS) and outer segment (OS) extend towards the RPE. The inner nuclear layer (INL) includes amacrine, bipolar and horizontal neurons, whereas ganglion cells, axons of which form the optic nerve, reside in the ganglion cell layer (GCL). Outer and inner plexiform layers (OPL and IPL, respectively) contain synaptic regions. (C) Schematic representation of the rod and cone photoreceptors, which have distinct compartmentalized morphology. The outer segment includes hundreds of membranous discs that contain visual pigment and other phototransduction components. The metabolic machinery is present in the inner segment. The visual proteins are transported to the outer segment via a connecting cilium. The nucleus is contained in the cell body, and the presynaptic region includes one or more ribbon-like structures for docking of synaptic vesicles.

Fig. 2.

A broad classification of proteins associated with retinal diseases according to their localization or function in photoreceptors and retinal pigment epithelium (RPE). As illustrated, RPE65, LRAT and MERTK, which are associated with LCA and arRP, are RPE proteins, whereas CRX, NRL and NR2E3 are photoreceptor-specific transcription factors. The remaining disease-associated proteins that are listed localize to the outer segment (OS), connecting cilium (CC) and/or basal body (BB) of the photoreceptor (here a rod is represented). Abbreviations: adRP, autosomal dominant retinitis pigmentosa; arRP, autosomal recessive retinitis pigmentosa; CSNB, congenital stationary night blindness; ESCS, enhanced S-cone syndrome; LCA, Leber congenital amaurosis; RP, retinitis pigmentosa.

Fig. 3.

Characterization of retinal degeneration in human patients and mouse mutants. (A) Human ocular fundus photographs, optical coherence tomograms (OCT) and electroretinograms (ERG; see Box 1). (i) Wide-field color fundus image in an adult normal subject shows preserved macula and peripheral retina, with normal coloration of the underlying retinal pigment epithelium (RPE) and choroid. (ii) In an adult patient with retinitis pigmentosa [Affected (RP)], areas of atrophy accompanied by pigmentary changes indicate underlying photoreceptor degeneration. OCT imaging allows ‘histological-like’ assessment of retinal structure in vivo, including identification of different retinal layers. Whereas, in a normal subject, the photoreceptor layer (outer nuclear layer; see rectangular areas marked by the broken red line) is well preserved, marked thinning is evident in a patient with RP, with some sparing only in the area of the fovea, which contains only cone photoreceptors. This thinning reflects loss of photoreceptors as part of the progressive degeneration. The black arrow in the fundus images shows the location of the OCT scan across the macula, and the area in the red rectangle is magnified in the image to the right. (iii) ERG testing allows measurement of retinal function in response to light stimulation. Under dark-adapted conditions (scotopic), stimulation of the normal eye with a dim or bright white flash elicits a well-formed rod response (black traces, upper left panel) or mixed rod/cone response (upper right panel), respectively. In light-adapted conditions (photopic), single flash stimulation of the eye results in a normal cone response (lower left panel) whereas rapid stimulation (30 Hz) results in flicker waveform (lower right panel). By contrast, in RP patients, severe attenuation of these electrophysiological responses of the retina is evident (red traces). (B) Mouse ocular fundus photographs, OCT and ERG. (i) The normal [wild type (WT), C57BL/6J] mouse retina fundus has a uniform color and the blood vessels are visible. (ii) The rd1 mutant mouse retina shows large areas of atrophy and discoloration, where the photoreceptors and possibly also the RPE have degenerated. The blood vessels are not visible in the degenerating retina. OCT imaging in a 3-month-old rd1 mouse shows a striking difference in retinal thickness compared with the normal control (rectangular areas marked by broken red line). The OCT scan position is indicated by a green line in each fundus image. (iii) The dark-adapted (scotopic; indicating rod function) and light-adapted (photopic; indicating cone function) ERG responses are robust in the normal mouse (WT, black traces, at 3 weeks of age) and are practically non-detectable in the rd1 mutant mouse (red traces). Anatomical and functional studies in mouse rd mutants are thus similar to what is generally observed in RP patients.

Fig. 4.

Schematic view of major proteins involved in phototransduction. The phototransduction events are broadly similar in rod and cone photoreceptors, and, given their complexity, we show here only the key proteins associated with rod phototransduction. During phototransduction (black arrows), the capture of photon(s) results in activation of rhodopsin, leading to dissociation of transducin (G protein) subunits βγ from Gα, which in turn activates cGMP-phosphodiesterase (PDE). PDE catalyzes the hydrolysis of cGMP to GMP, thereby causing closure of cyclic-nucleotide-gated (CNG) channels in the photoreceptor outer segment membrane. The closure of CNG channels results in photoreceptor hyperpolarization and transmission of the electrochemical signal(s) to second-order neurons in the inner retina via modulation of neurotransmitter release (not shown here). Channel closure also blocks Ca²⁺ entry, resulting in reduced intracellular Ca²⁺ (not shown here) and transmission of a feedback signal for recovery by engaging guanylyl cyclase activating proteins (GCAP). At low Ca²⁺ levels, GCAP activates guanylate cyclase (GC) and stimulates cGMP synthesis, thereby restoring cGMP levels and leading to re-opening of CNG channels. Termination of phototransduction (red arrows and T bars) also requires the inactivation of rhodopsin, which is initiated by its phosphorylation by rhodopsin kinase [G-protein receptor kinase (GRK)], facilitating the binding of arrestin to rhodopsin. In the dark and at high intracellular Ca²⁺ levels, recoverin inhibits GRK and controls the life time of activated rhodopsin. The transducin-PDE complex is inactivated by the hydrolysis of bound GTP that is greatly accelerated by the RGS9 complex (not shown here). The latter consists of regulators of G protein signaling member 9 (RGS9), G protein β5 and RGS9 associated protein (R9AP) (not illustrated in the figure). The negative feedback loop associated with Ca²⁺ concentration is critical for maintaining phototransduction.

Fig. 5.

From gene discovery to therapy of retinal degenerative diseases. (A) Schematic representation of the discovery of a gene associated with retinal disease. To identify the genetic defect associated with a diseases phenotype, DNA sequences of affected individuals (black squares) are compared with those of healthy individuals (white squares and circles) in the family. A hypothetical genetic difference in an individual with dominant disease is shown in the sequence (T/A). (B) Paradigms for retinal disease modeling and drug discovery. Drosophila (fly) and zebrafish embryos are useful for high-throughput large-scale drug screening; mouse models are excellent for elucidating disease mechanisms and for testing therapies; small-molecule (drug) screenings are often performed using cell culture systems derived from either mouse embryonic fibroblasts (MEFs) or human induced pluripotent stem cells (iPSCs). (C) Development of therapies using mouse models. (i) The accessibility of the eye and retina allows for delivery of appropriate drugs (e.g. neurotropic factors), gene therapy vectors (viral vectors used for specific gene replacement in recessive disease) and even cells (i′,i″) using intravitreal or subretinal injections/ transplantation in mice as well as in larger animal models. Surgical manipulation of the mouse retina is complicated by the small size of the eye and large lens. (i′) In vivo images of cell transplantation into an albino mouse eye. Human embryonic-stem-cell-derived retinal pigment epithelium (hESC-RPE) cells can be transplanted into the mouse eye via transvitreal (left panel; green arrow) or subretinal (right panel; green arrow indicates the location of cell grafts) transplantation. (i″) Histological (left panel) and immunofluorescence (right panel) images of an albino mouse eye after cell transplantation, demonstrating survival and integration of the hESC-RPE cells in the mouse retina. (Left panel) Transplanted pigmented hESC-RPE cells integrate as a monolayer (black staining) in the mouse retina. Inset and arrow indicate transition from host albino cells to grafted cells in the subretinal space. (Right panel) hESC-RPE cells are positive for both the human nuclear antigen (HuNu; green) and premelanosome 17 (PMEL17; red; a typical marker for RPE). (ii) Delivery of gene therapy vectors into photoreceptors or RPE cells will lead to the expression of the appropriate protein and facilitates rescue of function and phenotype.