ReviewRetinal remodeling during retinal degeneration
Introduction
Many forms of blindness, particularly those arising from photoreceptor degenerations, have no satisfactory resolution, which has inspired prosthetic and biological schemes to rescue or reconstruct retinal tissues. But does the neural retina remain receptive to such intervention? Though the neural retina grossly appears to survive the loss of its photoreceptors in retinitis pigmentosa and age-related macular degeneration, most brain pathways are clearly not so forgiving. Is the retina really different? The answer is: no.
Of the retinal degenerative diseases, retinitis pigmentosa (RP) is the best characterized (Bird, 1995) with an incidence of approximately 1 in 3500–4000 (Bunker et al., 1984) and involves more than 40 known genes, with many mutations in the rhodopsin gene (http://www.sph.uth.tmc.edu/RetNet/). Over 160 gene loci have been associated with retinal degenerations (Farrar et al., 2002). RP and its allied diseases (e.g. the cone and cone/rod dystrophies), age-related macular degeneration and environmental challenges to the retina, all stress and kill photoreceptors. Roughly speaking, most retinal degenerations fall into three forms: rod-degenerative forms, the mixed rod/cone-degenerative forms or debris-associated forms (e.g. mertk defects and light damage). Though these forms have differing specific mechanisms and dependencies, and may trigger diverse modalities of cell death, all share a common outcome: photoreceptor loss and neuronal remodeling.
Our laboratory employs a fusion of molecular and computational technologies (Computational Molecular Phenotyping: CMP) to concurrently visualize sets of small molecules with subcellular resolution, allowing a multi-dimensional segmentation of cell classes in normal retinal tissues with the aim of cataloging cell classes and elucidating their circuitries. We have used CMP to screen samples of human RP (Fig. 1), as well as natural and engineered rodent models of retinal disease, examining all cell types in 18 human cases of RP and greater than 200 cases of rodent retinal degenerations encompassing 18 different models. We concurrently visualized glial transformations, neuronal translocations, neuronal loss and the emergence of ectopic neurite complexes. The fusion of phenotyping and ultrastructure also enabled the visualization of novel synaptic assemblies, illustrating that the degenerating retina produces new synapses with vigor. Though these seeming plasticities might be exploited to rescue vision, the new circuitry is more likely corruptive of visual processing and reflects, we believe, attempts by neurons to find synaptic excitation. Even minor rewiring may corrupt signal processing in retinal pathways, leaving many current approaches to bionic and biological retinal rescue unsustainable. The ultimate conclusion is that the sequelae of retinal degenerative disease are far more complex than previously believed, and schemes to rescue vision via bionic implants or stem/engineered cells are based on presumed preservation of normal wiring and cell population patterning after photoreceptor death. Those beliefs are incorrect: retinal neurons die, migrate, and create new circuitries. Vision rescue strategies need to be refined.
Section snippets
Why was retinal remodeling overlooked despite years of work in retinal degenerations?
Certainly, the impetus to analyze remnant tissues held little motivation after photoreceptor death, but the progression of strategies to rescue vision in these retinas has brought issues related to the status of the neural retina into the forefront. Most papers concerned with the survival of the neural retina after photoreceptor loss have focused on simple cell counts to document numbers of cells surviving as opposed to documenting specific changes occurring in the surviving cohorts. Another
Retinal remodeling: phases and patterns of expression
We began our work in retinal degeneration after CMP-profiling samples of aged human retina from patients with RP, which revealed confusing and profound alterations in neural retinal structure (Fig. 1). Did animal models of retinal degeneration maintain their neural structure and circuitry, as implied in the literature? Was human remodeling a spurious conflation of disease and long-life? The answers were: no and no. CMP-profiling of numerous retinal disease models now show that all known retinal
Specific models and their defects
Remodeling is a general phenomenon but displays variations in speed and coherence that likely derive from the nature and duration of photoreceptor stress and death. Over twenty specific models of retinal degeneration have been analyzed (Table 1) in addition to tissues from RP human patients. Specific systems include natural (RCS rat, rd1 mouse, rd2 mouse, orJ mouse, pcd mouse, and nrJ mouse), transgenic (Tg GHL mouse, Tg rhoΔCTA mouse, Tg TG9N mouse, Tg rdcl mouse, Tg P23H rat [3 lines], Tg
Corrupt visual circuitry
The RCS rat retina shows panretinal defects at a density of about 790 major defects/mm2, equivalent to 40 000/eye, a quarter of which are microneuromas that include bipolar, amacrine and ganglion cell processes with abundant synapses. Is microneuroma circuitry normal? To explore the nature and scope of anomalous rewiring in microneuromas formed during retinal remodeling in late-stage retinal degenerations, we used serial section ultrastructural CMP (Jones et al., 2003a, Jones et al., 2003b) in
Self signaling
The technology of excitation mapping can be used to ‘ask’ blind retinas about their excitatory environment. While electrophysiology is hampered by the lack of a simple stimulus and difficult sampling conditions, the glutamate-gated channel permeant probe AGB (1-amino-4-guanidobutane) enables concurrent sampling of the integrated excitation histories of all retinal neurons in vitro (Marc, 1999, Marc and Jones, 2002) and can be used in vivo by intravitreal injection, followed by CMP profiling. We
Discussion of implications for rescue
Much previous literature argued that retinal degenerations such as RP affect only the sensory retina. Many approaches to retinal rescue are based on this clearly incorrect assumption. What does this mean for the current state of vision rescue research? At first it may appear to be discouraging and even insurmountable given that the field of vision rescue has been laboring for 20 years to overcome the loss of photoreceptors alone, only recently recognizing the scope of neural remodeling in the
Acknowledgements
This research was supported by NEI EY 002576 and EY 015128, and Research to Prevent Blindness.
References (99)
- et al.
Augmented rod bipolar cell function in partial receptor loss: an ERG study in P23H rhodopsin transgenic and aging normal rats
Vis. Res.
(2001) - et al.
Retinal rod photoreceptor-specific gene mutation perturbs cone pathway development
Neuron
(1999) Retinal photoreceptor dystrophies
Am. J. Ophthalmol.
(1995)- et al.
Prevalence of retinitis pigmentosa in Maine
Am. J. Ophthalmol.
(1984) - et al.
Chromosomal localization of the neurological mouse mutations tottering (tg), Purkinje cell degeneration (pcd), and nervous (nr)
Brain Res. Mol. Brain Res.
(1996) - et al.
Subretinal electrical stimulation of the rabbit retina
Neurosci. Lett.
(1997) - et al.
Horizontal cells of the normal and dystrophic rat retina: a wholemount study using immunolabelling for the 28-kDa calcium-binding protein
Exp. Eye Res.
(1993) - et al.
Abnormalities in rod photoreceptors, amacrine cells, and horizontal cells in human retinas with retinitis pigmentosa
Am. J. Ophthalmol.
(2000) - et al.
Muller cell and neuronal remodeling in retinal detachment and reattachment and their potential consequences for visual recovery: a review and reconsideration of recent data
Vis. Res.
(2003) - et al.
Regulation of hippocampal synaptic plasticity by the tyrosine kinase receptor, REK7/EphA5, and its ligand, AL-1/Ephrin-A5
Mol. Cell Neurosci.
(1998)