Neural remodeling in retinal degeneration

https://doi.org/10.1016/S1350-9462(03)00039-9Get rights and content

Abstract

Mammalian retinal degenerations initiated by gene defects in rods, cones or the retinal pigmented epithelium (RPE) often trigger loss of the sensory retina, effectively leaving the neural retina deafferented. The neural retina responds to this challenge by remodeling, first by subtle changes in neuronal structure and later by large-scale reorganization. Retinal degenerations in the mammalian retina generally progress through three phases. Phase 1 initiates with expression of a primary insult, followed by phase 2 photoreceptor death that ablates the sensory retina via initial photoreceptor stress, phenotype deconstruction, irreversible stress and cell death, including bystander effects or loss of trophic support. The loss of cones heralds phase 3: a protracted period of global remodeling of the remnant neural retina. Remodeling resembles the responses of many CNS assemblies to deafferentation or trauma, and includes neuronal cell death, neuronal and glial migration, elaboration of new neurites and synapses, rewiring of retinal circuits, glial hypertrophy and the evolution of a fibrotic glial seal that isolates the remnant neural retina from the surviving RPE and choroid.

In early phase 2, stressed photoreceptors sprout anomalous neurites that often reach the inner plexiform and ganglion cell layers. As death of rods and cones progresses, bipolar and horizontal cells are deafferented and retract most of their dendrites. Horizontal cells develop anomalous axonal processes and dendritic stalks that enter the inner plexiform layer. Dendrite truncation in rod bipolar cells is accompanied by revision of their macromolecular phenotype, including the loss of functioning mGluR6 transduction. After ablation of the sensory retina, Müller cells increase intermediate filament synthesis, forming a dense fibrotic layer in the remnant subretinal space. This layer invests the remnant retina and seals it from access via the choroidal route. Evidence of bipolar cell death begins in phase 1 or 2 in some animal models, but depletion of all neuronal classes is evident in phase 3. As remodeling progresses over months and years, more neurons are lost and patches of the ganglion cell layer can become depleted. Some survivor neurons of all classes elaborate new neurites, many of which form fascicles that travel hundreds of microns through the retina, often beneath the distal glial seal. These and other processes form new synaptic microneuromas in the remnant inner nuclear layer as well as cryptic connections throughout the retina. Remodeling activity peaks at mid-phase 3, where neuronal somas actively migrate on glial surfaces. Some amacrine and bipolar cells move into the former ganglion cell layer while other amacrine cells are everted through the inner nuclear layer to the glial seal. Remodeled retinas engage in anomalous self-signaling via rewired circuits that might not support vision even if they could be driven anew by cellular or bionic agents. We propose that survivor neurons actively seek excitation as sources of homeostatic Ca2+ fluxes. In late phase 3, neuron loss continues and the retina becomes increasingly glial in composition.

Retinal remodeling is not plasticity, but represents the invocation of mechanisms resembling developmental and CNS plasticities. Together, neuronal remodeling and the formation of the glial seal may abrogate many cellular and bionic rescue strategies. However, survivor neurons appear to be stable, healthy, active cells and given the evidence of their reactivity to deafferentation, it may be possible to influence their emergent rewiring and migration habits.

Introduction

The vertebrate retina is a tandem device: the photoreceptive, glutamatergic sensory retina drives the signal processing neural retina. Most retinal degenerations initially afflict the sensory retina and/or the RPE and many forms produce photoreceptor-depleted foci that expand and fuse, leading to extreme visual impairment or total blindness. Diverse sources of emerging data reveal negative plasticity in the remnant neural retina following loss of the sensory retina. The goal of this review is to draw together the lines of evidence; to detail the essential features of remodeling in retinal degenerations; to evaluate these processes in the context of known CNS plasticity and remodeling in response to challenges such as deafferentation. As a prototypical CNS assembly, the retina should be expected to remodel extensively in response to pathological stimuli. Alterations should be expected in the neural retina when photoreceptors degenerate. We will endeavour to show these expectations valid.

Though the notion of negative retinal remodeling induced by retinal degenerations is not new, it has not received much prior attention and certainly has not been integrated into schemata for retinal rescues. It can be argued that neural remodeling has been passively and actively neglected. A recent Medline search on “retinal degeneration” yielded ≈300 papers containing the words bipolar, horizontal, amacrine or Müller cells out of >13,000 articles retrieved. Less than 3% of relevant literature addresses the destinies of retinal neurons and glia after the death of their photoreceptor partners. This is partly due to the essential transparency of remodeling. In the absence of photoreceptor drive, remodeling is undetectable in vivo and histological methods such as routine light microscopy are insensitive to the early responses of neurons to photoreceptor stress and death. The vast majority of scientific papers dealing with the fate of the neural retina during and after photoreceptor degeneration have focused on numerical cell survival, largely neglecting classical and advanced methods for neuronal visualization. Given our current knowledge regarding the molecular deconstruction of photoreceptors during challenge or degeneration (John et al., 2000; Rex et al., 2002), we might profitably inquire of the morphological and biochemical integrities of the ≈55–60 classes of neurons that form a mammalian retina. Confocal microscopy, advanced immunocytochemical techniques, and quantitative molecular phenotyping have opened new windows on remodeling. Its pervasiveness emerges as irrefutable. Even so, various recent media still present the view that the neural retina is either refractory to photoreceptor loss or that changes may be minimal (Scarlatis, 2000; Chow et al., 2002; Gekeler and Zrenner, 2002; Margalit et al., 2002; Zrenner, 2002).

Degeneration of the sensory retina is often accompanied by early changes in circuitry of the neural retina (Wong, 1997; Banin et al., 1999; Peng et al., 2000; Strettoi and Pignatelli, 2000; Aleman et al., 2001; Strettoi et al (2002), Strettoi et al (2003)). These changes do not abate, but rather devolve to a protracted phase of negative remodeling of neuronal and glial elements in the remnant neural retina (Marc et al., 2001; Ren et al., 2001; Jones et al (2002), Jones et al (2003a); Sullivan et al., 2003). The key trigger seems to be the loss of cone photoreceptors. Remodeling involves three levels of restructuring in the neural retina: (1) neuronal death; (2) migration of cells; (3) rewiring. These are not trivial or collectively rare events, though visualizing them requires selective detection technologies. Remodeling potentially corrupts spatial processing and may even transform retinas into self-signaling neuronal assemblies with little ability to provide “proper” signal throughput as gauged by normal assemblies of visual circuits. Such changes may render many cell and bionic implant rescue strategies unfeasible. Beyond this, remodeling may constrain windows of opportunity for genetic rescues or molecular retinal maintenance. The mechanisms triggering remodeling are unknown, but it is a general feature of mammalian neural retinas challenged by sensory deafferentation effected by a range of photoreceptor degenerations. Our theme is that the mammalian retina is similar to other areas of the central nervous system (CNS) in its response to deafferentation on molecular, cellular and systems levels. Both retina and CNS display atrophy, apoptosis, reshaping of axonal and dendritic fields, revision of synaptic connectivity and efficacy that alter the signal processing attributes of entire pathways, alterations in gene expression, and glial transformations. Remodeling in the CNS is often guided by remnant synaptic inputs, while the retina becomes a system unto itself. Retinal remodeling has three implications we wish to emphasize. 1. Remodeling exposes the cryptic, incipient plasticity of the normal mammalian retina. 2. Remodeling challenges cellular and implant rescue strategies. 3. Remodeling may enable experimental revision of the remnant retina.

Neural remodeling in the degenerating mammalian retina follows a stereotyped sequence of events after photoreceptor degeneration. As this review will examine many of these events in some detail, it is profitable to start with an overview of remodeling. The remodeling scenario is outlined in Fig. 1, using the human retina as an exemplar, recalling that mammalian retinas used as experimental surrogates are virtually identical to >90% the human retina in terms of cone density, circuitry, and the molecular signatures of individual cell types. Retinal degenerations are of three basic forms: rod-initiated, cone-initiated, and debris initiated. We will use the common rod-initiated RP-like process for an outline and note key deviations in certain models where necessary.

Phase 0: The normal retina. The retina represents separate transport compartments bounded by cellular seals. The sensory retina is composed of photoreceptor cells and is internally compartmentalized by high-resistance tight junction arrays formed distally at the basolateral RPE margins and high-resistance intermediate junctions arrays formed proximally by Müller cell apical microvilli. The latter constitutes a diffusion barrier perforated by photoreceptor inner limbs: the perinuclear soma, nucleus, axon and synaptic terminal. The RPE is the blood retinal barrier for the outer limbs of photoreceptors: myoids, ellipsoids and outer segments. The neural retina is a bounded entirely by high-resistance intermediate junction seals of the Müller cells: distally at the apical microvilli, proximally at the Müller cell foot pieces and internally at vessels invested by Müller cell processes. This scheme functionally isolates the extracellular sensory transduction compartment of photoreceptors from their extracellular neural synaptic compartment, similar to hair cells and other sensory cell arrays. Within the sensory retina, the RPE and Müller cells regulate ionic transport, metabolite transport, fluid transport, retinoid transport, and outer segment turnover. Within the neural retina, Müller cells specifically regulate extracellular K+, GABA and glutamate levels; provide glutamine for glutamate synthesis; regulate extracellular volume (likely via anion flux and taurine transport); and provide for carbon skeleton recycling and redistribution.

Phase 1: Rod degeneration. Whether triggered by malfunction or insufficiency of specific proteins, degenerating rods display progressive decreases their ability to produce outer segments. The rate of synthesis lags RPE phagocytosis and outer segments shorten. Though the mechanisms of photoreceptor stress and forms of apoptosis may be debated, rods with shortened outer segments are at risk. Most rods are fated to die and move through stressed, deconstructed, irreversibly stressed, and apoptotic stages. Some rods may escape stress and enter the neural retina as survivor rods, but this appears to be rare. The details of cone stress during this phase are not clear, but the subretinal space is compressed and cone outer segments are truncated. Initial changes in the neural retina are dramatic but transparent to simple histological examinations. During initial stress and rod deconstruction, rod neurites sprout and enter the inner retina, extending to the ganglion cell layer in some cases and expressing rhodopsin. In late phase 1 or early phase 2, rods may retract neurites and remnant presynaptic terminals prior to death.

Phase 2: Cone degeneration. When rod outer segments are lost, plasma membrane rhodopsin levels in rod inner segments rise and the rate of rod death increases. Collapse of the subretinal space parallels the ablation of cone outer segments and remodeling of the inner segment. Cone deconstruction includes redistribution of cone opsins to inner segments, changes in expression of key proteins, and perhaps transient extension of axons into the neural retina. Cones begin to die, likely due to multiple stressors, but death may be accelerated by migration of activated microglia into the sensory retina. Elaboration of a fibrotic distal seal by Müller cells, some of which show nuclear translocation into the outer nuclear layer, entombs the remaining cones. Initial remodeling of the neural retina begins as rod and cone bipolar cells retract their dendrites and revise their expression of synaptic signaling receptor proteins. The rod-targeted axon terminal fields of horizontal cells undergo retraction and the cone-targeted horizontal cell somas hypertrophy and extend new neurites into the inner plexiform layer. Every bipolar and horizontal cell appears to be remodeled. Though a fraction of bipolar cells dies at this stage, the changes are quantitatively invisible by routine histological analysis.

Early phase 3: Progressive neurite remodeling. The loss of the sensory retina completes the deafferentation of the neural retina and the Müller cell seal becomes compacted. The seal is irreversibly bound to and invests the neural retina. Müller cell hypertrophy, in part due to up-regulation of intermediate filament synthesis, forms large columns that segment the retina. Neurons of the inner plexiform layer begin to remodel, with new neurites from all cell types appearing in the inner nuclear layer, forming patchy foci in the remnant outer plexiform layer. Global neuronal cell death is statistically detectable.

Middle phase 3: Global remodeling. Global remodeling includes concurrent neuronal death, cell migration, and rewiring via ectopic fascicle evolution and formation of ectopic synaptic microneuromas. Neuronal death shows no bias; all cell types are vulnerable. The rate is also variable and appears partially dependent on the speed and coherence of cone loss in phase 2. Neuronal migration includes eversion of amacrine cells to the distal margin of the retina and inversion of bipolar and amacrine cells to the ganglion cell layer, perhaps guided by Müller cell surfaces. Bundles of mixed neurites from all cell types course through the retina at all levels, most graphically in the remnant outer plexiform layer just proximal to the glial seal. Microneuromas contain synaptic terminals of all types of neurons, including bipolar cell ribbon synapses and conventional synapses of GABAergic and glycinergic amacrine cells. All neurons appear to preserve their normal basic molecular signatures. New synaptic connections are abundant.

Late phase 3: Plateau remodeling. Basic remodeling processes persist. Neuronal death continues and can significantly deplete the inner nuclear and ganglion cell layers, with Müller cells partially filling the space. Neurite fascicles and microneuromas may regress as neuronal cell death depletes the retina. The inner plexiform layer becomes thinner. The optic fiber layer can become thinned and ganglion cell layer vessels migrate into the retina. Preretinal membranes and new extracellular matrices are formed. In some diseases, progressive RPE alterations are evident, including cell death, elaboration of columns of apical processes extending deep into the remnant neural retina, and migration of some survivor RPE cells into the retina, lining retinal vessels.

Different initial scenarios for cone- and debris-initiated retinal degenerations. Cone-initiated diseases may simply reverse the order of photoreceptor loss. Phase 1 would encompass cone stress and apoptosis, followed by phase 2 ablation of the outer nuclear layer via bystander-effect loss of rods. This excludes those cone dystrophies in which rods are spared. Cone–rod dystrophies and debris-initiated diseases may display overlapping phases 1 and 2. Debris-initiated retinal degenerations are largely those in which RPE function is directly or secondarily compromised, resulting in an inability to internalize outer segments and accumulation of debris in the subretinal space. Cumulative stresses such as lipid oxidation, metabolite deprivation, microglia/macrophage activation may lead to a relatively coherent loss of rods and cones. However, after sensory deafferentation, neural remodeling appears to follow the same phase 3 profiles.

Section snippets

Human retinal degenerations initiated by rod, cone and RPE defects

Human retinal degenerations largely originate in the sensory retina and its adjacent transport tissues, the RPE and choroid. They are diversely sorted by clinical phenotypes, affected gene, protein function and malfunction, and cellular association (Daiger and Sullivan, 2003). With respect to ultimate photoreceptor loss, inherited disorders form a continuum (Deutman and Hoyng, 2001; Rivolta et al., 2002). Some cone dystrophies (COD) exist as isolated photopic dysfunction, with preservation of

Rod and cone death patterns in retinal degenerations

Significant effort is now being expended to determine why and how photoreceptors die in retinal degenerations and what environmental factors might impact the speed and scope of cell death (Remé et al., 1998; Hao et al., 2002). Many of these details clearly lie outside the scope of this review, but the initiators, secondary stressors, and durations of stress prior to photoreceptor death likely shape the natures and extents of early remodeling defects.

Developmental plasticity

Prior to exploring reactions of the mature neural retina to sensory deafferentation, it is useful to review evidences of plasticity in the retina. Circuit assembly in the mammalian retina involves significant postnatal refinement, including improved high spatial frequency cutoffs of retinal ganglion cells at about pnd 30 in cats (Rusoff and Dubin, 1977). There is increasing evidence that light history impacts maturation. In a remarkable study, Ikeda and Wright (1976) demonstrated that the

Deprivation and deafferentation-induced remodeling in the CNS

As we are about to consider remodeling in the retina, as brief review of CNS remodeling is in order. The mammalian CNS remodels during development and learning, as well as in response to injury or disease. Plastic responses produced by CNS neurons in association with learning and memory, e.g. changes in the numbers of dendritic spines (Sorra and Harris, 2000), are compelling examples of the flexibility of neuronal circuits. Altered visual experience can induce compensatory axonal sprouting

Systems expressing early remodeling of neural retina

Only a few model systems have been carefully examined for evidence of early remodeling, but the data are unequivocal: even at early stages of photoreceptor stress and deconstruction, the neural retina reacts with bipolar and horizontal cell remodeling. The most extensive analysis has been carried out in the rd1 mouse, but similar phenomena have been observed in the rd10 mouse, Crx−/− mouse, FVB/N mouse, RCS rat, and P347L transgenic pig. Early remodeling is transparent to conventional

Anatomical evidence for late remodeling in human retinal degenerations

The earliest data suggesting that advanced retinal degenerations triggered more severe defects than merely the loss of part or all of the sensory retina arose from standard histological surveys of retinitis pigmentosa tissue. The advent of animal models has driven many of these studies from recent memory, leaving many with the impression that loss of the sensory retina is the only significant defect arising in retinal degenerations. The true status of the neural retina has not been better

Basic mammalian pathways

Even in rod-rich mammalians, cone circuitry dominates (Strettoi et al., 1994). Of the ≈55–60 classes of retinal elements that constitute a generalized mammalian retina (MacNeil et al (1999a), MacNeil et al (1999b); Masland (2001a), Masland (2001b); Marc and Jones, 2002; Rockhill et al., 2002), all but five are predominantly or exclusively driven by cones. The mammalian rod pathway is composed of but two purely rod-driven neurons (rod bipolar cells and horizontal cell axon terminals functioning

Implications of remodeling for rescue strategies

The clinical relevance of remodeling in retinal blinding diseases is substantial. The fundamental strategies of subretinal and epiretinal prosthetic implants, photoreceptor transplants, stem cell therapies, gene rescue therapies, survival factor treatments and combinations thereof are based upon the hope that the neural retina remains viable after extensive photoreceptor degeneration; that the neural retina should act as a platform to accept donor cells and quantitatively respond to

The future: retinal plasticity and remodeling

The scope of true physiological plasticity in the adult mammalian retina is unknown, but given our knowledge of the extent of network adaptation and structural revision displayed by non-mammalian retinas, extensive functional plasticity in terms of synaptic strengths or even structural changes may be expected. Further, the processes invoked in early development of retinal lamination (Johnson et al., 1999), spatial patterning (Reese and Galli-Resta, 2002) and functional activity (Tian and

Acknowledgements

The preparation of this review was funded in part by NEI R01 EY02576 (RM), TeleThon Project E0833 (ES) and NEI R01 EY12654 (ES). Our thanks to many colleagues who advised us on revisions. The remaining errors are ours alone.

References (276)

  • Y Chu et al.

    Horizontal cells of the normal and dystrophic rat retinaa wholemount study using immunolabeling for the 28 kDa calcium binding protein

    Exp. Eye Res.

    (1993)
  • M.H Chun et al.

    The synaptic complex of cones in the fovea and in the periphery of the macaque monkey retina

    Vis. Res.

    (1996)
  • F.E Coblentz et al.

    Evidence that ganglion cells react to retinal detachment

    Exp. Eye Res.

    (2003)
  • J.E Cook et al.

    Retinal mosaicsnew insights into an old concept

    Trends Neurosci.

    (2000)
  • A Cowey et al.

    Transneuronal retrograde degeneration of retinal ganglion cells after damage to striate cortex in macaque monkeys. Selective loss of P-beta cells

    Neuroscience

    (1989)
  • R.N Fariss et al.

    Abnormalities in rod photoreceptors, amacrine cells, and horizontal cells in human retinas with retinitis pigmentosa

    Am. J. Ophthalmol.

    (2000)
  • J.C Fiala et al.

    Dendritic spine pathologycause or consequence of neurological disorders?

    Brain Res. Rev.

    (2002)
  • S.K Fisher et al.

    Müller cell and neuronal remodeling in retinal detachment and reattachment and their potential consequences for visual recoverya review and reconsideration of recent data

    Vis. Res.

    (2003)
  • J Freigang et al.

    The crystal structure of the ligand binding module of axonin-1/tag-1 suggests a zipper mechanism for neural cell adhesion

    Cell

    (2000)
  • C.C Garner et al.

    Molecular mechanisms of CNS synaptogenesis

    Trends Neurosci.

    (2002)
  • K Gregory-Evans et al.

    Abnormal cone synapses in human cone-rod dystrophy

    Ophthalmology

    (1998)
  • N Gupta et al.

    Activated microglia in human retinitis pigmentosa, late-onset retinal degeneration, and age-related macular degeneration

    Exp. Eye Res.

    (2003)
  • J.R Heckenlively et al.

    Clinical findings and common symptoms in retinitis pigmentosa

    Am. J. Ophthalmol.

    (1988)
  • R Adler et al.

    Cell death in age-related macular degeneration

    Mol. Vis.

    (1999)
  • Anderton, B.H., Callahan, L., Coleman, P., Davies, P., Flood, D., Jicha, G.A., Ohm, T., Weaver, C., 1998. Dendritic...
  • E Apfelstedt-Sylla et al.

    Extensive intrafamilial and interfamilial phenotype variation among patients with autosomal dominant retinal dystrophy and mutations in the human RDS/peripherin gene

    Br. J. Ophthalmol.

    (1995)
  • K Arikawa et al.

    Localization of peripherin/rds in the disk membranes of cone and rod photoreceptorsrelationship to disk membrane morphogenesis and retinal degeneration

    J. Cell Biol.

    (1992)
  • J.L Badano et al.

    Beyond Mendelan evolving view of human genetic disease transmission

    Nat. Rev. Genet.

    (2002)
  • Baehr, W., Frederick, J.M., 2000. Inherited retina diseases: vertebrate animal models. In: Encyclopedia of Life...
  • V Baekelandt et al.

    Alterations in GAP-43 and synapsin immunoreactivity provide evidence for synaptic reorganization in adult cat dorsal lateral geniculate nucleus following retinal lesions

    Eur. J. Neurosci.

    (1994)
  • A Bansal et al.

    Mice lacking specific nicotinic acetylcholine receptor subunits exhibit dramatically altered spontaneous activity patterns and reveal a limited role for retinal waves in forming ON and OFF circuits in the inner retina

    J. Neurosci.

    (2000)
  • E.L Berson et al.

    Ocular findings in patients with autosomal dominant retinitis pigmentosa and a rhodopsin gene defect (pro23-to-his)

    Arch. Ophthalmol.

    (1991)
  • R Biran et al.

    Surfactant-immobilized fibronectin enhances bioactivity and regulates sensory neurite outgrowth

    J. Biomed. Mater. Res.

    (2001)
  • R Blanco et al.

    Ionotropic glutamate receptors in isolated horizontal cells of the rabbit retina

    Eur. J. Neurosci.

    (1999)
  • J.C Blanks et al.

    Photoreceptor degeneration and synaptogenesis in retinal-degenerative (rd) mice

    J. Comp. Neurol.

    (1974)
  • S.A Bloomfield et al.

    Light-induced modulation of coupling between AII amacrine cells in the rabbit retina

    Vis. Neurosci.

    (1997)
  • T.A Blute et al.

    Direct imaging of NMDA-stimulated nitric oxide production in the retina

    Vis. Neurosci.

    (2000)
  • C Bowes et al.

    Retinal degeneration in the rd mouse is caused by a defect in the beta subunit of rod cGMP-phosphodiesterase

    Nature

    (1990)
  • R.B Brem et al.

    Immunolocalization of integrins in the human retina

    Invest. Ophthalmol. Vis. Sci.

    (1994)
  • D.V Buonomano et al.

    Cortical plasticityfrom synapses to maps

    Annu. Rev. Neurosci.

    (1998)
  • D.A Butts et al.

    Retinal waves are governed by collective network properties

    J. Neurosci.

    (1999)
  • S.A Capurso et al.

    Deafferentation causes apoptosis in cortical sensory neurons in the adult rat

    J. Neurosci.

    (1997)
  • L.D Carter-Dawson et al.

    Differential effect of the rd mutation on rods and cones in the mouse retina

    Invest. Ophthalmol. Vis. Sci.

    (1978)
  • J Chen et al.

    Mechanisms of rhodopsin inactivation in vivo as revealed by a COOH terminal truncation mutant

    Science

    (1995)
  • M.S Chen et al.

    Role of the integrin-associated proteini CD9 in binding between sperm ADAM 2 and the egg integrin a6b1implications for murine fertilization

    Proc. Natl. Acad. Sci.

    (1999)
  • D.T Chiu et al.

    Patterned deposition of cells and proteins onto surfaces by using three-dimensional microfluidic systems

    Proc. Natl. Acad. Sci.

    (2000)
  • A.Y Chow et al.

    Subretinal implantation of semiconductor-based photodiodesdurability of novel implant designs

    J. Rehab. Res. Dev.

    (2002)
  • C.M Cicerone

    Cones survive rods in the light-damaged eye of the albino rat

    Science

    (1976)
  • M.L Condic

    Adult neuronal regeneration induced by transgenic integrin expression

    J. Neurosci.

    (2001)
  • M.L Condic et al.

    Ligand-induced changes in integrin expression regulate neuronal adhesion and neurite outgrowth

    Nature

    (1997)
  • Cited by (716)

    • The role of hydrogen sulfide in the retina

      2023, Experimental Eye Research
    View all citing articles on Scopus
    View full text