Advertisement
Commentary Free access | 10.1172/JCI33513
Department of Ophthalmology, Children’s Hospital Boston, Harvard Medical School, Boston, Massachusetts, USA.
Address correspondence to: Lois E.H. Smith, Department of Ophthalmology, Fegan 4, Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115, USA. Phone: (617) 919-2529; Fax: (617) 730-0392; E-mail: lois.smith@childrens.harvard.edu.
Find articles by Chen, J. in: JCI | PubMed | Google Scholar
Department of Ophthalmology, Children’s Hospital Boston, Harvard Medical School, Boston, Massachusetts, USA.
Address correspondence to: Lois E.H. Smith, Department of Ophthalmology, Fegan 4, Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115, USA. Phone: (617) 919-2529; Fax: (617) 730-0392; E-mail: lois.smith@childrens.harvard.edu.
Find articles by Connor, K. in: JCI | PubMed | Google Scholar
Department of Ophthalmology, Children’s Hospital Boston, Harvard Medical School, Boston, Massachusetts, USA.
Address correspondence to: Lois E.H. Smith, Department of Ophthalmology, Fegan 4, Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115, USA. Phone: (617) 919-2529; Fax: (617) 730-0392; E-mail: lois.smith@childrens.harvard.edu.
Find articles by Smith, L. in: JCI | PubMed | Google Scholar
Published October 1, 2007 - More info
The role of retinal microglial cells (MCs) in age-related macular degeneration (AMD) is unclear. Here we demonstrated that all retinal MCs express CX3C chemokine receptor 1 (CX3CR1) and that homozygosity for the CX3CR1 M280 allele, which is associated with impaired cell migration, increases the risk of AMD. In humans with AMD, MCs accumulated in the subretinal space at sites of retinal degeneration and choroidal neovascularization (CNV). In CX3CR1-deficient mice, MCs accumulated subretinally with age and albino background and after laser impact preceding retinal degeneration. Raising the albino mice in the dark prevented both events. The appearance of lipid-bloated subretinal MCs was drusen-like on funduscopy of senescent mice, and CX3CR1-dependent MC accumulation was associated with an exacerbation of experimental CNV. These results show that CX3CR1-dependent accumulation of subretinal MCs evokes cardinal features of AMD. These findings reveal what we believe to be a novel pathogenic process with important implications for the development of new therapies for AMD.
Christophe Combadière, Charles Feumi, William Raoul, Nicole Keller, Mathieu Rodéro, Adeline Pézard, Sophie Lavalette, Marianne Houssier, Laurent Jonet, Emilie Picard, Patrice Debré, Mirna Sirinyan, Philippe Deterre, Tania Ferroukhi, Salomon-Yves Cohen, Dominique Chauvaud, Jean-Claude Jeanny, Sylvain Chemtob, Francine Behar-Cohen, Florian Sennlaub
Age-related macular degeneration (AMD), the most common cause of blindness in the elderly, is characterized by degeneration of the macula and can lead to loss of fine color vision. Alterations in inflammatory and immune system pathways, which arise from genetic differences, predispose individuals to AMD. Yet the mechanism of disease progression with respect to inflammation is not fully understood. In this issue of the JCI, the study by Combadière and colleagues shows that CX3C chemokine receptor 1–deficient (CX3CR1-deficient) mice have abnormal microglia that accumulate beneath the retina and contribute to the progression of AMD (see the related article beginning on page 2920).
Age-related macular degeneration (AMD), a degenerative disease of the retina, is the most common cause of visual impairment in the elderly in the developed world (1). The prevalence of AMD is expected to increase as the population ages. AMD is characterized by degeneration of the macula, an area in the central retina with the highest concentration of cones that is responsible for high-acuity and color vision (Figure 1). Damage to the macula can cause profound loss of fine vision (Figure 1A). Genetic factors identified for AMD risk involve regulation of the inflammatory, complement, and chemokine pathways, including the fractalkine receptor CX3C chemokine receptor 1 (CX3CR1) (2). CX3CR1 is expressed in microglia, the resident macrophages in the CNS, and the retina (3, 4) and mediates migration and adhesion of these cells in response to CX3C chemokine ligand 1 (CX3CL1) (5).
Loss of central vision in AMD patients. (A) Representation of the visual field seen by a person with normal vision and by an AMD patient with a loss of central color vision. (B) Schematic of a normal human eye compared with an AMD eye depicting deposits of drusen (yellow spots) and choroidal neovascularization with subretinal hemorrhage in the macula (red). (C) Schematic cross-section of a normal eye through the macula shows retinal neuronal layers, RPE, Bruch’s membrane, and choroid vessels. In AMD, the intimate relation among photoreceptors, RPE, and choroid is disrupted by drusen (lipid and cellular debris–containing deposits). The formation of drusen separates RPE from Bruch’s membrane and the underlying choroidal vessels in association with RPE atrophy and photoreceptor degeneration (hallmarks of dry AMD). In wet AMD with choroidal neovascularization, abnormal leaky choroidal vessels proliferate and penetrate the altered Bruch’s membrane protruding into the subretinal space, causing hemorrhage and rapid loss of vision.
In this issue of the JCI, a new study by Combadière et al. (6) confirmed that in humans, the CX3CR1 M280 allele increases the risk of human AMD and that microglia isolated from these individuals migrate defectively. The authors also showed that in CX3CR1-deficient mice, microglia accumulate in the subretinal space, evoking morphological and pathological features similar to those observed in human AMD. The data suggest an important role of CX3CR1 and of microglia in the pathogenesis of AMD.
The retina is composed of several transparent neuronal layers. Light focused through the cornea and lens passes across these layers to interact with the pigments in the photoreceptor outer segments, causing a chemical transformation. This ultimately propagates a series of action potentials across the retinal and brain neurons, and we perceive an image. The photoreceptors are dependent upon a well-functioning adjoining retinal pigment epithelium (RPE), which phagocytoses the spent photoreceptor outer segments that are shed daily to recycle photopigments (such as rhodopsin) and membrane lipids. The RPE also transports nutrients and breakdown products between the photoreceptors and the adjoining choroidal vascular plexus, separated from the RPE by Bruch’s membrane (Figures 1 and 2). Therefore the photoreceptors, choroidal vessels, and RPE have an interdependent relationship, and loss of any one component causes dysfunction of the others. This dysfunction is often seen first in the macula because the densely packed photoreceptors here have the highest oxygen consumption and metabolic rate of any tissue in the body (7).
Proposed role of CX3CR1 and microglia in AMD. (A) In the normal retina, rod and cone photoreceptor outer segments are composed of thousands of photopigment-containing membrane discs that are continuously shed from the tips of the cells and recycled. The discarded discs are removed by RPE phagocytosis and reprocessed. (B) Additionally, microglia expressing CX3CR1 phagocytize cellular debris caused by daily insults (such as aging and cellular stress) and help maintain a healthy eye. (C) AMD resulting from loss of CX3CR1 expression is initiated by the same daily insults. However, as reported by Combadière et al. (6), when microglia lacking CX3CR1 are recruited to the area of damage they remain there, in part because of abnormal migration. CX3CR1-deficient microglia are proinflammatory (versus antiinflammatory CX3CR1-replete microglia; ref. 16) and recruit other proinflammatory cells. This response exacerbates the cellular damage that occurs with normal degeneration. (D) The increased accumulation of proinflammatory, CX3CR1-negative (16) microglia in the subretinal space appears to add to the formation of drusen (6), which contribute to loss of RPE function and photoreceptor degeneration (dry AMD). (E) Finally, as drusen build up and separate the RPE from the choroidal vasculature, the resulting hypoxia triggers proangiogenic signals, which foster choroidal neovascularization (wet AMD).
Clinically, AMD begins with the asymptomatic appearance of drusen, seen as white or yellow spots beneath the retina. What triggers the formation of drusen is not completely understood, and not all drusen formation leads to AMD. Drusen are deposits of lipid and cellular debris that are found beneath the RPE on Bruch’s membrane (Figures 1 and 2). These deposits increasingly separate the RPE from the underlying choroidal vascular bed, interrupting the function of the RPE and leading to photoreceptor degeneration, a hallmark of dry AMD (Figures 1 and 2). In wet AMD, its most devastating form, abnormal leaky blood vessels grow from the choroidal vasculature through the disrupted Bruch’s membrane and RPE into the subretinal space beneath the photoreceptors, causing profound visual loss (Figures 1 and 2).
Many risk factors for AMD are related to inflammation. These include environmental factors, such as smoking and low omega-3 fatty acid intake (8, 9), and genetic factors, such as polymorphisms complement factor H (10) and CX3CR1 (2), the chemokine CX3CL1 receptor. Chemokines are small proteins that induce directed chemotaxis in nearby responsive immune cells. Macrophage chemoattractants as well as elevated levels of inflammatory mediators and activated complement components are found in drusen samples from AMD patients (11, 12).
However, the mechanism of AMD with respect to inflammation and chemoattractants is not well defined. In their current study, Combadière and coworkers (6) expand our understanding of the role of chemokines in AMD. They suggest that defective microglia with abnormal CX3CR1 function contribute to the formation of drusen in AMD and subsequently promote photoreceptor loss and choroidal neovascularization. The authors show that, compared with mice with normal microglia, mice deficient in CX3CR1 had subretinal accumulation of microglia and drusen-like deposits after insults to the retina including senescence, light-induced retinal degeneration, and laser damage to rupture Bruch’s membrane. The drusen-like material contained photoreceptor pigment (rhodopsin), indicating contribution from shed rod photoreceptor outer segments, as well as markers indicating contribution from accumulated defective microglia lacking CX3CR1 expression. Because these deposits precede photoreceptor degeneration and choroidal neovascularization (Figure 2), the authors suggest that these drusen could cause these AMD features. This study provides new insight into the putative role of microglia in drusen formation and into the role of inflammation in the pathogenesis of AMD, but raises many questions as well. How and why do these defective microglia accumulate — through decreased egress alone or increased attraction to the area? How do CX3CR1-defective microglia cause photoreceptor degeneration and choroidal neovascularization? How do these results reconcile with studies showing that decreased macrophage recruitment increases aspects of AMD?
In aged transgenic mice lacking CC chemokine ligand 2 (Ccl2) or its cognate receptor CC chemokine receptor 2 (Ccr2), both molecules involved in chemoattraction of macrophages and/or microglia, some features of AMD (drusen-like deposits and choroidal neovascularization) have been reported, suggesting that macrophage recruitment by Ccl2 or Ccr2 confers protection against the formation of AMD (13). Mice deficient in IL-10 — a potent inhibitor of cytokine and chemokine production — with inhibited recruitment of macrophages in retina have increased choroidal neovascularization (14). These results suggest that macrophage recruitment into the subretinal space is necessary to remove extracellular deposits and prevent the formation of drusen. Yet global depletion of macrophages and microglia in mice reduces laser-induced choroidal neovascularization (15). In contrast, the results reported here by Combadière and coworkers (6) suggest that accumulated CX3CR1-deficient microglia secondary to increased recruitment or impaired egress from the retina can form drusen and cause photoreceptor degeneration and choroidal neovascularization, all aspects of AMD.
CX3CR1 is involved in leukocyte recruitment, and previous studies have shown that its loss leads to inhibition of leukocyte migration (5). It is likely that an alternate chemoattractant molecule, like that of Ccr2, is involved in recruitment of these cells into areas of retinal damage. But how does accumulation of CX3CR1-deficient microglia cause AMD-like symptoms?
Perhaps it is because macrophages can have many different and even opposite functions depending on their activation state. Given this possibility, the results of the present study fit with the recent finding in muscle that decreased expression of CX3CR1 results in macrophages with a proinflammatory phenotype compared with macrophages with high levels of CX3CR1, which have an antiinflammatory phenotype (16). It is reasonable to speculate that the accumulation in the subretinal space of resident microglia with reduced CX3CR1 function might therefore contribute to low-grade inflammation, leading to the recruitment of other inflammatory cells including proangiogenic, bone marrow–derived macrophages, which in turn induce choroidal neovascularization and the pathology of AMD. Prolonged presence of these cells in the subretinal space may also spur phagocytosis of lipids normally processed by the RPE. These bloated microglia trapped subretinally might become dysfunctional and degenerate, contributing to the accumulation of extracellular cellular debris and to the early sign of drusen formation. Although the present study (6) raises interesting possibilities regarding the mechanism of drusen formation, the exact contribution of microglia and circulating macrophages to retinal degeneration and choroid neovascularization remains undefined.
The study by Combadière and coworkers (6) is an important contribution to the current inflammatory theory in AMD. A clear understanding of the cellular and molecular mechanisms of AMD could bring a major shift in our approach to disease treatment and prevention. Currently the only treatment to slow the progression of dry AMD to wet AMD is dietary supplementation with vitamins and antioxidants (17), which is consistent with the role of oxidation stress in this pathological process (18). Recently, higher intake of dietary lipids thought to be antiinflammatory (omega-3 long-chain polyunsaturated fatty acids and fish) (9) have been found to be associated with decreased likelihood of developing wet AMD (8). For wet AMD, the treatments focus on suppression of choroidal neovascularization, with laser ablation, photodynamic therapy, or, increasingly, anti-VEGF angiogenesis inhibitors. However, these treatments do not address the underlying cause of AMD, and preventive and nondestructive therapies are much more desirable.
A twin study by Seddon et al. found that genetic factors explain 46%–71% of the variation in AMD disease severity, while environment explains 19%–42% (19), so manipulation of some pathways involved in genetic risk could have significant benefit. But the possibility of manipulating the CX3CR1/CX3CL1 pathway specifically to prevent the development of AMD is uncertain at the moment. CX3CL1-deficient mice are protected against development of atherosclerosis (20), and CX3CL1 mediates antitumor responses (21), so manipulation of the CX3CR1/CX3CL1 pathway in AMD is complex and might disrupt the physiological process of microglia and/or macrophage recruitment and clearing of subretinal debris. Nevertheless, the study by Combadière and coworkers (6) suggests a new direction. Further understanding of the molecular basis of the pathogenesis of AMD might lead to new treatment based on inhibition of complement factors that promote choroidal neovascularization as well as pharmacological targeting of chemoattractant cytokines.
We thank Christopher M. Aderman for excellent technical help and for preparing figures for this manuscript. The authors’ research is generously supported by the V. Kann Rasmussen Foundation as well as by NIH grants EY008670, EY017017, and EY14811 (L.E.H. Smith); by NIH grants 1F32 EY017789 and 5 T32 EY07145 (K.M. Connor); and by Children’s Hospital Boston Mental Retardation and Developmental Disabilities Research Center (NIH grant P01 HD18655). We thank the Juvenile Diabetes Research Foundation for fellowship support (J. Chen). This work was also supported by the Research to Prevent Blindness Lew Wasserman Merit Award (L.E.H. Smith).
Address correspondence to: Lois E.H. Smith, Department of Ophthalmology, Fegan 4, Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115, USA. Phone: (617) 919-2529; Fax: (617) 730-0392; E-mail: lois.smith@childrens.harvard.edu.
Nonstandard abbreviations used: AMD, age-related macular degeneration; Ccl2, CC chemokine ligand 2; Ccr2, CC chemokine receptor 2; CX3CL1, CX3C chemokine ligand 1; CX3CR1, CX3C chemokine receptor 1; RPE, retinal pigment epithelium.
Conflict of interest: The authors have declared that no conflict of interest exists.
Reference information: J. Clin. Invest.117:2758–2762 (2007). doi:10.1172/JCI33513.
See the related article at CX3CR1-dependent subretinal microglia cell accumulation is associated with cardinal features of age-related macular degeneration.