Ординатура / Офтальмология / Английские материалы / The Retina and its Disorders_Besharse, Bok_2011

B, complement component 2) indicate that their AMD risk-conferring variants can together account for nearly 75% of AMD cases. There is additional evidence that environmental risk factors, such as smoking and obesity, can combine with genetic risk to increase AMD susceptibility even further. For example, the risk of AMD is 8–10 times higher in individuals who are both homozygous for the risk-conferring C allele in the CFH gene and smoke. Additional variants in the CFH and related genes have also been linked to AMD risk. For example, deletion of two CFH-related genes (CFHR1 and CFHR3) has been shown to have protective effects and decrease the likelihood of AMD.

ARMS2/HTRA1

Both the ARMS2 and HTRA1 genes contain polymorphisms that have been associated with increased risk for AMD and they exhibit a high degree of linkage disequilibrium (i.e., they are essentially always transmitted together). Because of this latter property it has been difficult to determine which of the polymorphisms is responsible for the associated AMD risk. In addition, biological evidence supporting a role for both has been presented. One ARMS2 polymorphism (rs10490924) leads to a change in a region of the gene that is predicted to result in an amino acid change from alanine to serine at position 69 (A69S) of the ARMS2 protein. Some data suggest that the ARMS2 protein is located in mitochondria and that it is involved in modulating oxidative stress and associated apoptosis, but there is no direct evidence yet linking ARMS2 protein dysfunction and the AMD disease process. An additional ARMS2 polymorphism that involves both deletion and insertion of genetic material has been shown to lead to messenger RNA (mRNA) instability and marked reduction in the levels of ARMS2 transcripts, and is significantly associated with AMD risk. ARMS2 protein is highly expressed in the retina and it has been proposed that dysfunctional ARMS2 and/or reduced levels of ARMS2 in mitochondria-rich retinal photoreceptor cells may enhance susceptibility to age-associated mitochondrial dysfunction that leads to photoreceptor compromise and visual loss.

The AMD-linked polymorphism in the HTRA1 gene (rs11200638) lies in the gene’s promoter region and some data have suggested that the risk-conferring allele leads to altered binding of transcription factors, increased transcription of the gene, and elevated HTRA1 protein levels, but other studies have not confirmed this observation.

Treatments

Most therapeutic approaches for AMD have historically focused on the neovascular form of the disease, as it is the

one which is responsible for the most severe and precipitous vision loss. Therapies for neovascular AMD have included thermal laser photocoagulation of leaky neovessels and photodynamic therapy that utilizes laser activation of an intravenously administered compound that damages the endothelial cells of neovessels, leading to thrombosis. These approaches were applicable to less than half the patients with neovascular AMD and thus, to less than 5% of all AMD patients. More recently, molecular inhibitors of vascular growth factors (primarily vascular endothelial growth factor, VEGF) have been exploited to limit neovascularization in a variety of ocular neovascular processes. Such VEGF inhibitors are generally administered in multiple intravitreal injections and act by inhibiting the interaction of VEGF with its receptor on the surfaces of vascular endothelial cells. The binding of VEGF to its receptor promotes endothelial cell proliferation, vessel growth, and increased vascular permeability, leading to the development of CNV in affected eyes. The most striking results to date have been provided by anti- body-based compounds (Lucentis and Avastin) that are directed against VEGF and inhibit its function. These compounds have provided unprecedented benefits in terms of visual stabilization in most patients with CNV and visual improvement in many. It is anticipated that advances in the specificity, efficacy, durability, and deliverability of antiVEGF compounds, as well as compounds directed against additional vascular growth factors, will continue to advance the physician’s ability to combat neovascular AMD. However, similar therapeutic advances have not been made for the dry or atrophic forms of AMD that represent 85–90% of disease cases.

See also: Developmental Anatomy of the Retinal and Choroidal Vasculature; Physiological Anatomy of the Retinal Vasculature; Primary Photoreceptor Degenerations: Retinitis Pigmentosa; Primary Photoreceptor Degenerations: Terminology; Retinal Pigment Epithelium: Cytokine Modulation of Epithelial Physiology; Secondary Photoreceptor Degenerations.

Further Reading

Baird, P. N., Robman, L. D., Richardson, A. J., et al. (2008). Gene–environment interactions in progression of AMD – the CFH gene, smoking and exposure to chronic infection. Human Molecular Genetics 17: 1299–1305.

Berger, J. W., Fine, S. L., and Maguire, M. G. (1999). Age-Related Macular Degeneration. St. Louis, MS: Mosby.

Chong, E. W., Kreis, A. J., Wong, T. Y., Simpson, J. A., and Guymer, R. H. (2008). Dietary omega-3 fatty acid and fish intake in the primary prevention of age-related macular degeneration:

A systematic review and meta-analysis. Archives of Ophthalmology 126: 826–833.

Cong, R., Zhou, B., Sun, Q., Gu, H., Tang, N., and Wang, B. (2008). Smoking and the risk of age-related macular degeneration: A metaanalysis. Annals of Epidemiology 18: 647–656.

Gehrs, K. M., Anderson, D. H., Johnson, L. V., and Hageman, G. S. (2006). Age-related macular degeneration – emerging pathogenic and therapeutic concepts. Annals of Medicine 38: 450–471.

Grisanti, S. and Tatar, O. (2008). The role of vascular endothelial growth factor and other endogenous interplayers in age-related macular degeneration. Progress in Retinal and Eye Research

27: 372–390.

Jager, R. D., Mieler, W. F., and Miller, J. W. (2008). Age-related macular degeneration. The New England Journal of Medicine

358: 2606–2617.

Johnson, E. J. (2005). Obesity, lutein metabolism and age-related macular degeneration: A web of connections. Nutrition Reviews 63: 9–15.

Klein, R. (2007). Overview of progress in the epidemiology of age-related macular degeneration. Ophthalmic Epidemiology 14: 184–187.

Lotery, A. and Trump, D. (2007). Progress in defining the molecular biology of age related macular degeneration. Human Genetics 122: 219–236.

Montezuma, S. R., Sobrin, L., and Seddon, J. M. (2007). Review of genetics in age related macular degeneration. Seminars in Ophthalmology 22: 229–240.

Pieramici, D. J. and Rabena, M. D. (2008). Anti-VEGF therapy: Comparison of current and future agents. Eye 20: 1330–1336.

Provis, J. M., Penfold, P. L., Cornish, E. E., Sandercoe, T. M., and Madigan, M. C. (2005). Anatomy and development of the macula:

Specialization and the vulnerability to macular degeneration. Clinical and Experimental Optometry 88: 269–281.

Rattner, A. and Nathans, J. (2006). Macular degeneration: Recent advances and therapeutic opportunities. Nature Reviews Neuroscience 7: 860–872.

Scholl, H. P. N., Fleckenstein, M., Issa, P. C., et al. (2007). An update on the genetics of age-related macular degeneration. Molecular Vision 13: 196–205.

Sunness, J. S. (1999). The natural history of geographic atrophy, the advanced atrophic form of age-related macular degeneration.

Molecular Vision 5: 25.

Relevant Websites

http://www.ahaf.org – American Health Assistance Foundation. http://www.macular.org – American Macular Degeneration Foundation

(AMDF).

http://www.eyesight.org – Macular Degeneration Foundation (MDF). http://www.mayoclinic.com – Mayo Clinic.

http://www.nei.nih.gov – National Eye Institute (NEI). http://www.nlm.nih.gov – National Library of Medicine, National

Institutes of Health.

Glossary

Epigenetic – The heritable modifications of gene expression that are not the result of changes in the DNA sequence. This can include methylation of DNA that results in inactivation of gene transcription or factors that modify how RNA transcripts are spliced to form the final transcripts that are translated into peptide sequences.

Extrinsic factor – An agent external to the organism that contributes to or is causative of a disease state. This can include drugs, foods, normal nutrients (excess or deficiency), toxins, inhaled chemicals, infectious agents, and exposures to radiation such as light, sound, and high-energy particles.

Intrinsic factor – An agent that is inherent to the organism that contributes to or is causative of a disease state. While commonly these factors are genetic variants in the organism’s DNA that may predispose (or be protective) of specific conditions, other intrinsic factors include epigenetic changes, aging changes, and the effects of the biology of symbiotic bacteria in the skin or gut. Another intrinsic agent is an organism’s immunologic response behavior and memory (though obviously the immunologic memory is heavily affected by the exposure to extrinsic agents, such as viral infections).

Microbiomics – The genetic information expressed by the microbes that are indigenous to a host organism (e.g., bacteria colonized to the skin or intestinal tract).

Introduction

Secondary retinal degeneration occurs when cells in the retina die by a process triggered by factors not inherent to retina. Secondary retinal degeneration can be caused by trauma, infection, inflammation, toxins, anti-retinal antibodies, or as an adverse effect of medications.

* All of the genes that are mentioned in this article are described in Table 2 of Chapter 210 (for the retinal degenerations), RetNet (www.sph.uth.tmc. edu/retnet/), and/or Online Mendelian Inheritance of Man (OMIM) (www.ncbi.nlm.nih.gov/Omim/).

In the past, clinicians have tended to view genetic and nongenetic etiologies of retinal degeneration as easily separated categories. The molecular studies of hereditary retinal degenerations have shown that, while some retinal conditions are caused by mutations in genes with photoreceptor-specific expression, many retinal conditions are the results of mutations in genes that are widely expressed in the body as well as from the secondary effects of metabolic changes caused by the expression of mutated genes in ocular cell types other than photoreceptors as well as from other organs and tissues distant from the eye. Based on our understanding of complex genetic disorders, we now realize that there can be interplay of genetic and nongenetic factors that run the entire spectrum of possibilities. For example, rhegmatogenous retinal detachments, which can lead to secondary photoreceptor degeneration, may be influenced or caused by genetic variants (e.g., COL11A1, VCAN, COL9A1, and COL2A1) that are expressed in nonretinal cells, and whose expression may be limited to a particular period in ocular development. Thus, we have to consider this continuum of causality as we attempt to make useful classifications that can guide diagnostics and therapy. In light of these complexities, we offer the following operational distinctions among primary and secondary photoreceptor and retinal degenerations that may be relevant to therapeutic approaches.

. If the genetic defect is such that it would require actual alteration of the gene expression in the photoreceptors to correct the abnormality and arrest the degeneration, then this can be considered a primary photoreceptor degeneration. The genetic alteration is necessary and sufficient to cause photoreceptor degeneration. The gene that is mutated may (e.g., opsin, peripherin/rds, cone transducin, AIPL1, and GUCY2D) or may not (e.g., splicing factors PRPF8, PRPF3, and PRPF31, IMPDH1, and CA4) be photoreceptor specific. For a primary photoreceptor degeneration, one would expect that the correction of the genetic alteration outside of the photoreceptors would not be sufficient to prevent photoreceptor degeneration. However, a secondary photoreceptor degeneration that results from loss of expression or expression of a mutated protein in either other retinal cells or the retinal pigment epithelium (RPE) (e.g., RPE-65, RGR, and LRAT) might be corrected by gene therapy to the key nonphotoreceptor cells in the retina or RPE.

. If one reviews the genes attributed to primary photoreceptor degenerations, it is clear that many of these causative genes are not limited to photoreceptor-specific

expression. Mutations in these genes have been attributed to nonsyndromic primary photoreceptor degenerations (such as retinitis pigmentosa (RP), Leber congenital amaurosis (LCA), cone dystrophy, and cone–rod dystrophy) as well as syndromic forms (e.g., Usher syndrome, Bardet–Biedl syndrome, Alstrom disease, and Cohen syndrome). Most of these conditions are further described and discussed elsewhere in the encyclopedia. In some instances, the mechanisms of action of these genes may not be solely mediated through their direct effects on the photoreceptors, thus raising the possibility that, in some cases, the photoreceptor degeneration is mediated through a mixed primary and secondary photoreceptor degeneration model (see below). Two unique examples of this potential ambiguity are the ABCA4 (Stargardt disease, cone–rod dystrophy) and RS1 (X-linked retinoschisis) genes. Both genes are specifically expressed in the photoreceptors, but their mechanism of action appears to be mediated through other retinal/RPE cells that lead to a secondary photoreceptor degeneration (see below).

. If the photoreceptors degenerate as the result of an alteration in a gene whose expression is primarily in other retinal or RPE cells, then this would be a primary retinal degeneration with secondary photoreceptor degeneration. Correction of the genetic defect would require modification of the effects of those retinal/RPE cells. A primary retinal degeneration without photoreceptor degeneration can occur such as with optic neuropathies that lead to retinal ganglion cell loss without significant loss of photoreceptors (Table 1).

. A mixed model of primary and secondary photoreceptor degeneration can be considered in two different modes. One is when a genetic alteration within the photoreceptors themselves would not be sufficient to cause photoreceptor degeneration by itself but would

Table 1 Secondary photoreceptor degenerations associated with primary retinal/RPE degeneration/dystrophy

Gene involved, site of cell/tissue expression related to retinal degeneration (RPE-retinal pigment epithelium, RVE-retinal vascular endothelium, RVP-retinal vascular pericytes, MGC-Muller glial cells), and phenotype (LCA-Leber congenital amaurosis, RP-retinitis pigmentosa) (from RetNet)

predispose to degeneration in the presence of an extrinsic or intrinsic agent. The genetic alteration within the photoreceptors could be necessary, conditional, or probabilistic but not sufficient. An intrinsic agent could be a genetic alteration in nonphotoreceptor retinal cells or due to expression elsewhere in the body. As noted above, a number of genes that are expressed in photoreceptors and for whom there are mutations that are known to be responsible for photoreceptor degeneration also have expression in other retinal cells as well as in other tissues. In some of these cases, it is not always clear if the expression in the photoreceptors alone is sufficient to cause cell death or whether or not there is a component of photoreceptor degeneration that is secondary to the effects on other cells and tissues. The only way to distinguish a secondary effect from a primary one would be to create animal models in which the genetic alteration is limited to specific cell populations and to determine if the photoreceptors are spared when their gene expression is normal. This is especially true for the forms of RP that are associated with mutations in genes that affect metabolic processes throughout the body. Examples of these conditions include gyrate atrophy, Bietti crystalline retinopathy, abetalipoproteinemia, and Refsum disease. At this time, we simply cannot establish if the effects of these genetic mutations are mediated by a primary effect on the photoreceptors or by secondary mechanisms. In the case of gyrate atrophy and Refsum disease, there is evidence that nutritional therapy can ameliorate the progression of the condition, which suggests an interplay of a person’s intrinsic genetic makeup and diet (an extrinsic agent), but we still do not know if the effect is due to the systemic reduction of toxic metabolites or a photorecep- tor-specific mechanism is also involved. Similarly, with Bietti crystalline retinopathy, the defect in CYP4V2 has multitissue consequences but it is not known if a systemic correction of the metabolic defect would be sufficient to overcome the enzyme deficiency in photoreceptor cells. Only future studies will be sufficient to distinguish if these conditions are representative of a mixed model of photoreceptor degeneration or secondary photoreceptor degenerations of an intrinsic type (see below).

An extrinsic agent can be a drug or environmental exposure (including something in the diet). There are relatively few established human examples of this model for retinal degenerations, though retinal degeneration-B (rdgB) mutants in Drosophila show light-dependent photoreceptor degeneration. This mixed model could possibly account for some of the cases of photoreceptor degenerations with incomplete penetrance (individuals who have the diseasecausing mutation but show no clinical evidence of retinal degeneration).

. If one could prevent the photoreceptor degeneration by preventing an individual’s exposure to an extrinsic agent

or condition (e.g., toxin, drug, infectious agent, light, and trauma), then this is secondary photoreceptor degeneration of the extrinsic type (even if the body converts that agent to a toxic form as part of a normal metabolic pathway – such as methanol to formaldehyde). Clearly, the primary method of management is to avoid exposure to the extrinsic conditions that would induce the degeneration. This form of degeneration can be due to exposure to an external agent as well as deprivation of a mandatory nutrient (such as vitamin A). The deficiency can be the result of a lack of intake or synthesis of the key nutrient (vitamin-A- or zinc-deficient diet) or due to the inability to process or use such a metabolite/nutrient. Examples would be malabsorption of vitamin A and zinc due to intestinal disorders or drugs which block utilization, such as fenretinide or accutane (Table 2).

. A second mode of a combined primary and secondary photoreceptor degeneration is when one group of photoreceptors, such as the rod photoreceptors, undergoes a primary degenerative process due to a mutation in a gene that is expressed in those photoreceptors that precipitates apoptosis. At the same time, there is a second group of photoreceptors, the cone photoreceptors, which undergoes a secondary degenerative process due to alterations in the cellular environment induced by the death of neighboring cells. This situation is actually very common among patients with retinal dystrophies such as rod–cone (e.g., RP) or cone–rod forms. Recent studies of several mouse models of RP due to rod-photoreceptor specific genes have showed that the nonautonomous death of the cone photoreceptors is influenced by activation of the rapamycin pathway that can be modified by exogenous

Table 2 Retinotoxic drugs and agents, nutrient deficiencies, infectious agents, light injury, and trauma

Drugs

Ethambutol, aminoglycosides, epinephrine, desferroximine, antimalarials (hydroxychloroquine, chloroquine, quinine), vigabatrin, phenothiazines (e.g., fluphenazine, mellaril, and stellazine).

Nutrient deficiencies

Zinc, vitamin A, omega-3 fatty acids.

Infectious

Toxoplasmosis, cytomegalovirus, herpes simplex, varicella zoster, HIV, DUSN (nematode), rubella, syphilis, prion, corona virus, others.

Toxins

Cadmium, iron (siderosis), lead, mercury (suspected), copper (intraocular chalcosis), cobalt, iodoacetic acid (IAA), methanol.

Light

Solar, laser chronic exposure.

Trauma

Commotio, retinal detachment.

Vascular

Occlusive disease, embolic, inflammatory, retinopathy of prematurity (ROP), Coats disease.

insulin, suggesting a possible intrinsic mechanism that could be influenced by a systemic therapeutic approach. The importance of this mechanism cannot be overemphasized since preservation of cone photoreceptor cells and function in a patient with RP would have a dramatic impact on maintaining useful visual function and it does not necessarily require the correction of the primary photoreceptor degeneration mechanism in the rod photoreceptors.

. If one can prevent photoreceptor degeneration by correcting or reversing a systemic or ocular metabolic or immune process, then it is a secondary photoreceptor degeneration of the intrinsic type. A number of these conditions are driven or influenced by genetic etiologies (necessary and sufficient in the case of metabolic syndromes, but often conditional or probabilistic in immune-related conditions), but the retinal degeneration is still secondary. Intrinsic causes are not exclusively genetic, one may have to consider epigenetic factors as well as immunologic memory and the microbiomics of the natural flora. Clearly, one would primarily direct therapy to correcting the primary metabolic or immune disturbance rather than focusing on modifying the behavior of the photoreceptors. Therapy might be directed specifically to the affected eye(s), (such as periocular or intraocular steroid therapy) rather than systemically, but it would be intended to primarily modify effector cells in the tissue, rather than the photoreceptors themselves (Table 3).

. A mixed intrinsic and extrinsic etiology for a secondary photoreceptor degeneration would be when a person has a genetic variant that creates a toxic metabolite in the presence of an extrinsic molecule that would normally

Table 3 Intrinsic factors: genes, phenotypes (e.g., RP nonsyndromic, RP syndromic, and macular degeneration), mechanism (e.g. metabolic, immune, inflammatory (inflamm))

not be encountered. A normal person would not experience a retinal degeneration under the same exposure conditions. This set of conditions has overlap with the purely extrinsic and intrinsic etiologies if the genetic variation simply shifts the dose–response characteristics of the host. For example, a person is genetically predisposed to react to an extrinsic molecule at levels in the normal environment, while another person would experience similar photoreceptor degeneration only when the exposure is at levels that would exceed normal exposures. Reduction of the extrinsic exposure below the normal levels could be beneficial for these individuals (such as Refsum disease or gyrate atrophy). Alternatively, correction of the genetic variant would allow the person to cope with normal exposure levels. An animal model of the mixed intrinsic and extrinsic secondary photoreceptor degeneration would be the RPE65-MET450 mutants (intrinsic) that have varying reduced sensitivity to light-induced (extrinsic) photoreceptor degeneration as compared to animals that have the LEU450 variant in the RP65 gene.

. We are only beginning to understand these types of situations, although it is likely that many of the idiosyncratic reactions that some patients experience to certain situations or medications are the result of genetic variations that affect drug bioavailability, mechanism of action, and elimination. One such example would be the patient who develops cystoid macular edema (CME) after uncomplicated surgery. The surgical intervention would be considered an extrinsic agent. While CME is common in cases of complicated surgery and postsurgical inflammation, it is relatively uncommon (but not rare) in individuals whose surgery and postoperative care are uneventful and have no predisposing clinical conditions. Yet, this is most likely due to intrinsic (nonphotoreceptor-specific) genetic factors that govern inflammation. Persistent CME can lead to secondary photoreceptor degeneration.

If the extrinsic exposure cannot be manipulated, then essentially, one is forced to treat the mixed etiology as a purely intrinsic issue. For example, if a person had a genetic condition from an intrinsic metabolic defect that is light sensitizing such that normal ambient light would trigger photoreceptor degeneration, the distinction between an intrinsic etiology and a mixed intrinsic/extrinsic etiology becomes almost meaningless, since having a person avoid all light exposure to prevent photoreceptor degeneration is neither feasible nor desirable. However, reduction of the light exposure might alter the rate of disease progression, but therapy directed toward the intrinsic factor(s) would be essential to preserve vision under normal exposure circumstances. This situation is comparable to the mixed primary and secondary photoreceptor degeneration category (such as a mutation in an photoreceptor-specific gene that is

responsible for light-dependent degeneration) except that, instead of the intrinsic etiology being disconnected from the photoreceptors themselves, the photoreceptors are directly affected by a genetic variant that renders the photoreceptors vulnerable to the extrinsic factor (e.g., light). While the reduction of the extrinsic exposure would be desirable, it may not be realistic and thus therapy would also have to be directed to the photoreceptors themselves.

The combination of extrinsic and intrinsic factors that affect photoreceptor degeneration is comparable to the genetic and environmental interactions that are often discussed in the context of complex genetic diseases such as age-related macular degeneration. At this time, our understanding of these interactions is very limited, but there are some examples for simpler retinal conditions. Mice that are heterozygous for a deletion in the PDE6B subunit (the rd mouse), a dose of sildenafil citrate (Viagra) that would normally have no effect in the normal mouse, will show a major change in the electroretinogram. It is likely that some of the individuals who experience visual side effects from this medication may have a genetic variant that reduces the overall level of phosphodiesterase activity in their retinas, thus conferring sensitivity. Another mixed etiology of secondary photoreceptor dysfunction (which can ultimately lead to degeneration) can be seen in an individual with a normally adequate intake of vitamin A, who becomes vitamin A deficient due to an acquired or hereditary malabsorption syndrome (including postsurgical bowel resection or remodeling). The etiology may be intrinsic or iatrogenic, but the treatment is directed toward the extrinsic agent by increasing the dose or mode of absorption of the vitamin A.

Mechanisms of Secondary

Photoreceptor Death

Photoreceptors can die from several mechanisms, including physical lysis, destruction by thermal denaturation (such as by laser), or by triggering the apoptotic pathways. Apoptosis can be triggered by a number of disruptions, including loss of key trophic factors such as vascular endothelial growth factor (VEGF), energy depletion through mitochondrial failure, oxidative damage of proteins and lipids, release of calcium by shifts in membrane permeability, which can be caused by deregulation of ionic channels, or from the fixation of complement to the membrane surface. Light levels below those that cause thermal denaturation can lead to direct activation of caspases, calpain 2, and cathepsin D. In addition, mitochondrial-dependent apoptotic pathways also appear to be activated.

In a number of cases, signaling of the apoptotic pathway appears to be governed by at least two pathways: the Wnt pathway and the Jak–STAT pathway. A number of research groups are attempting to identify nonspecific

therapies that can block or inhibit the pathways that result in photoreceptor death. The use of ciliary neurotrophic factor (CNTF) as a trophic factor to inhibit activation of the apoptosis pathway is currently in clinical trials to treat primary and secondary photoreceptor degenerations.

As one considers these multiple mechanisms of photoreceptor death, it becomes clear that a major value of experimental animal models for these conditions is to specifically determine the extent to which photoreceptor death is a primary event and if genetic defects within the photoreceptors themselves are necessary and sufficient to initiate apoptosis. At the same time, this does not negate the importance of understanding the intrinsic (both genetic and nongenetic) factors and extrinsic factors that either trigger or modify cell death that may be amenable to therapeutic intervention at a systemic level. Finally, even in the presence of a combination of factors that lead to photoreceptor death, there is the possibility of interrupting or inhibiting the common apoptosis signaling pathways within the photoreceptor cells and other retinal neurons in order to preserve function and vision.

See also: Primary Photoreceptor Degenerations: Retinitis Pigmentosa; Primary Photoreceptor Degenerations: Terminology; Retinal Ganglion Cell Apoptosis and Neuroprotection.

Hackam, A. S. (2005). The Wnt signaling pathway in retinal degenerations. IUBMB Life 57(6): 381–388.

Ling, C. P. and Pavesio, C. (2003). Paraneoplastic syndromes associated with visual loss. Current Opinion in Ophthalmology

14(6): 426–432.

Poll-The, B. T., Maillette de Buy Wenniger-Prick, L. J., Barth, P. G., and Duran, M. (2003). The eye as a window to inborn errors of metabolism. Journal of Inherited Metabolic Disease 26(2–3): 229–244.

Punzo, C., Kornacker, K., and Cepko, C. L. (2009). Stimulation of the insulin/mTOR pathway delays cone death in a mouse model of retinitis pigmentosa. Nature Neuroscience 12(1): 44–52.

Rattner, A. and Nathans, J. (2006). An evolutionary perspective on the photoreceptor damage response. American Journal of Ophthalmology 141(3): 558–562.

Samardzija, M., Wenzel, A., Aufenberg, S., et al. (2006). Differential role of Jak–STAT signaling in retinal degenerations. FASEB Journal 20(13): 2411–2413.

Siu, T. L., Morley, J. W., and Coroneo, M. T. (2008). Toxicology of the retina: Advances in understanding the defence mechanisms and pathogenesis of drugand light-induced retinopathy.

Clinical and Experimental Ophthalmology 36(2): 176–185. Stone, J., Maslim, K., Valter-Kosci, K., et al. (1999). Mechanisms of

photoreceptor death and survival in mammalian retina. Progress in Retinal and Eye Research 18(6): 689–735.

Wenzel, A., Grimm, C., Samardzija, M., and Reme´, C. E. (2005). Molecular mechanisms of light-induced photoreceptor apoptosis and neuroprotection for retinal degeneration. Progress in Retinal and Eye Research 24(2): 275–306.

Wu, J., Seregard, S., and Algvere, P. V. (2006). Photochemical damage of the retina. Survey of Ophthalmology 51(5): 461–481.

Yang, L. P., Zhu, X. A., and Tso, M. O. (2007). A possible mechanism of microglia–photoreceptor crosstalk. Molecular Vision 13: 2048–2057.

Glossary

Dynamic range – The full range of light levels over which retinal cells or pathways can dynamically modify their function to encode.

Mesopic vision – Intermediate light level vision mediated by both rod and cone photoreceptors under conditions where both are responsive, allowing a seamless transition from rod to cone vision. Photopic vision – High light level vision mediated by cone photoreceptors.

Ribbon synapse – Synapses where a ribbon-like structure is present that optimizes the continuous release of neurotransmitter.

Rod spherule – The specialized synaptic terminal of the rod photoreceptor.

Scotopic vision – Low light vision mediated by rod photoreceptors.

Introduction

Nearly all sensory systems must find a way to represent a wide range of input signals and translate them into meaningful neural responses. The human visual system is able to operate effectively from starlight to bright sunlight, a range that spans about 12 orders of magnitude of light intensity. The pupil serves as the first stage of sensitivity control by changing the amount of the light reaching the photoreceptors. The diameter of the pupil can change by a factor of 4, allowing light intensity to change by a factor of 16. However, this alone cannot account for the complete range of light to which the mammalian retina is sensitive. To reliably transmit changes in light stimuli over this range, the mammalian retina has evolved several specializations to report changes in the light environment that include: (1) the evolution of two photoreceptor types, the rods and cones, that operate at different light levels; (2) several neural pathways with which to encode the output of these photoreceptors; and (3) adaptive mechanisms at all levels of retinal processing to modulate light sensitivity based on light history

Rods versus Cones

The evolution of the duplex retina in vertebrates with two classes of photoreceptors, the rods and cones, marks a

departure from the single type of photoreceptor present in many invertebrates. The use of two photoreceptor subtypes with different sensitivities to light allows the vertebrate retina to respond over a greater range of light intensities. By switching between rods and cones, the vertebrate retina is thus able to maximize visual sensitivity depending on the ambient light level. It is now believed that such an arrangement allows the vertebrate retina to reduce energy consumption in daylight, when the rods are not responsive.

Rods mediate vision when photons are scarce; their design and cytoarchitecture are optimized for maximal sensitivity to incoming photons and are capable of generating a reproducible response to a single absorbed photon, which is critical for setting the sensitivity of scotopic vision when the retinal circuitry pools thousands of rods (see the section titled ’The pathways concept’ below). As the mean background light level increases, rods themselves are able to adapt, which allows them to signal light intensities up to 1000 R* or more per second.

Cone photoreceptors are 100-fold less sensitive than rods, and are critical for our daytime vision under conditions when the exquisitely sensitive rod photoreceptors are saturated. The reduced sensitivity of cones arises from reduced amplification within cone phototransduction, and mechanisms designed to shut off phototransduction more quickly than rods. Furthermore, even in the brightest light the cone photocurrent does not remain saturated, which leaves open the ability to adapt and signal changes in light intensity even under conditions where a majority of the photopigment is bleached. The optimizations of cones to function in bright light with virtually little overlap with rod light levels allows for a smooth transition from rod to cone vision, in what is referred to as the mesopic range.

The Pathways Concept

The adaptive features of the rod and cone light response allow these photoreceptors to remain responsive over a larger range of light intensities by preventing response saturation. However, the dynamic range of the rods and cones themselves cannot account for the 12 orders of magnitude in light intensity we experience. Another strategy used by the mammalian retina to extend further the dynamic range of vision is to utilize multiple neural pathways to carry light-evoked signals. The functional

properties of these pathways (e.g., convergence, gain, and adaptation) can then be adjusted to maximize the visual system’s ability to remain responsive over the largest range of light intensities. To date, the most-studied retinal pathways in mammals are those that carry signals from rod photoreceptors to ganglion cells. These include the circuits that carry light information near scotopic threshold (rod bipolar pathway or classical rod pathway), and those that operate at higher rod light levels that may provide a seamless mesopic transition (rod–cone and rod–OFF pathways) to photopic vision. Rather less is known about the functional properties of the cone pathways that maximize dynamic range.

Rod Bipolar Pathway

Psychophysical studies have demonstrated at the limits of scotopic vision that the human visual system is capable of detecting the absorption of a few photons of light. This remarkable sensitivity arises from the fact that individual rods can reliably signal the absorption of a single photon, and that a specialized retinal circuitry referred to as the rod bipolar pathway can combine these signals such that a ganglion cell projecting centrally can pool thousands of rod signals. The discovery of a dedicated depolarizing bipolar cell and an amacrine cell that carries rod signals in the rabbit retina leads to the finding that this circuit appears conserved across all mammalian species. A hallmark of this pathway is the convergence of rod signals at many stages of processing that is critical for our scotopic sensitivity. For instance, as many as 20–100 rods converge on a rod (ON) bipolar cell, and 20–30 rod bipolar cells converge on an AII amacrine cell. Thus, a ganglion cell that sums the output of many AII amacine cells can pool upward of 10 000 rods. Ultimately, by pooling rod signals and eliminating rod noise to preserve best the single-photon response from individual rods, the rod bipolar pathway can extend the dynamic range of vision down to light levels where a small fraction of the rods absorb a photon.

Signal transfer from rods to rod bipolar cells

At scotopic threshold, vision relies on a sparse number of photons at the retina that produce few photon absorptions per thousands of rods within the 0.2-s integration time of the rod photoresponse. Under these conditions, the transmission of a small, graded hyperpolarization upon photon absorption requires that rod synapse is appropriately optimized. The transmission of small, graded singlephoton responses at the rod synaptic terminal is aided by two specializations. First, the resting dark membrane potential, or voltage, sits at approximately –40 mV, near the steepest point in the relationship between voltage and L-type Ca2þ channel opening (Figure 2). Thus, small changes in membrane potential produce substantial changes in the number of open channels, thereby altering

glutamate release. Second, if the rod bipolar cell is sensing reductions in glutamate release due to photon absorption, then statistical lapses of glutamate release in darkness would mimic light absorption. Thus, the high rate of glutamate release generated in darkness by the specialized synaptic ribbon in the rod spherule reduces the probability of these lapses. Together, these synaptic properties allow the small, light-evoked signals from rods to be reproducibly transferred to downstream neurons.

Despite the rod synaptic specializations for the transmission of single-photon absorptions, the depolarization in darkness due to open cyclic guanosine monophosphate (cGMP)-gated channels is also a complicating factor in the detection of these sparse signals. Open cGMP-gated channels in turn will report internal fluctuations in cGMP, produced by the phototransduction mechanism, which are commonly referred to as dark noise. Since rods generate a small, graded hyperpolarization upon photon absorption, the downstream convergence of thousands of rod signals would cause the light-evoked response from a single rod to be overwhelmed by the dark noise of the majority. Given the magnitude of dark noise in individual rods, it has been proposed that some type of nonlinear combination of rod signals would be required to increase the detection of the single-photon responses in downstream cells. Since rod photoreceptors are relatively depolarized in darkness, the steady release of glutamate from the synapse provides some insights into potential mechanisms. Postsynaptic saturation at the rod-to-rod bipolar synapse would allow noise generated by open cGMP-gated channels in the rod outer segment to be eliminated. It was proposed that the saturation of postsynaptic glutamate receptors would provide a nonlinear way to eliminate the rod noise, since the synapse would not be able to relay small changes in membrane potential that reflect rod noise. Later work suggested that such thresholding is critical for maximizing the detection of the single-photon response in retinal neurons downstream of the rods (Figure 1). In particular, the extent of nonlinear signaling appears to be set to separate optimally the rod single-photon response from rod noise, allowing scotopic vision to reach the highest possible sensitivity.

The mechanism that underlies the nonlinear threshold at the rod synapse has been studied to some extent, but is hindered by a lack of identification of the components of the signaling pathway. Light-evoked signaling between rod photoreceptors and rod (ON) bipolar cells results in a membrane depolarization, effectively inverting the sign of the rod’s hyperpolarizing light response. The postsynaptic mechanism underlying this sign inversion is a G-protein signaling pathway initiated by the metabotropic glutamate receptor, mGluR6. mGluR6, in turn, activates a guanine nucleotide-binding protein, Goa, which leads to a series of unidentified events that close a cationic transduction channel of unknown identity. Thus, upon

light-absorption, glutamate release from rods is reduced, thereby reducing the activity of the mGluR6 signaling pathway and allowing transduction channels to open and depolarize the cell (Figure 2).

In the context of the mGluR6 signaling cascade, it now appears that the nonlinear threshold that eliminates rod noise is due to saturation within the signaling cascade, and not at the level of the glutamate receptors. Furthermore, evidence from axotomized rod bipolar cells indicates that nonlinear signal transfer does not arise due to feedback in the inner plexiform layer. Saturation of the mGluR6 signaling cascade allows the elimination of noise by making the rod bipolar cell insensitive to small fluctuations in glutamate, driven by noise in the rod photoreceptor. Only when the rod’s membrane potential is hyperpolarized sufficiently does the glutamate concentration in the synaptic cleft reduce enough to relieve the synapse from saturation. Such an operation thus allows larger hyperpolarizations due to light absorption to cross the rod synapse, while masking smaller fluctuations that are more likely due to noise in the rod photocurrent or synaptic transmission. Near absolute visual threshold, such synaptic processing is necessary to maximize the detectability of rod signals.

Signal transfer from rod bipolar cells to AII amacrine cells

The convergence of the rod bipolar pathway moving from rods to rod bipolar cells and, finally, AII amacrine cells, requires the further accentuation of the single-photon

response. Two main specializations between these cells appear well tuned to further improve the detection of the single-photon response, and thus push the dynamic range of vision to lower light intensities. First, a specialized ribbon synapse between the rod bipolar cell and the AII amacrine cells allows the coordinated release of multiple vesicles upon stimulation. Such multivesicular release increases the amplitude of the AII amacrine cell response, allowing it to be distinguished from vesicular release due to noise in the rod bipolar cell. Second, the electrical coupling of AII amacrine cells by connexin 36 appears to reduce noise in the network, allowing an improved signal-to-noise ratio.

Ganglion cell sensitivity

Retinal ganglion cells can also relay single-photon responses to higher visual centers, a requirement for the high sensitivity of rod vision. Through the process of neural convergence, as well as the mechanisms described above, ganglion cells and AII amacrine cells have about the same flash sensitivity. Recordings of many groups from dark-adapted cat retinal ganglion cells indicate bursts of2–3 action potentials occurred with a frequency consistent with an upstream origin that may be the rods. The frequency of these bursts increased with background light, suggesting that the bursts were related to photon absorption. Similar conclusions have been drawn using a crosscorrelation analysis from paired ganglion cell recordings in the presence and absence of background lights.