Central Role of Oxidative Stress in Age-Related Macular Degeneration: Evidence from a Review of the Molecular Mechanisms and Animal Models.

Age-related macular degeneration (AMD) is a common cause of visual impairment in the elderly. There are very limited therapeutic options for AMD with the predominant therapies targeting vascular endothelial growth factor (VEGF) in the retina of patients afflicted with wet AMD. Hence, it is important to remind readers, especially those interested in AMD, about current studies that may help to develop novel therapies for other stages of AMD. This study, therefore, provides a comprehensive review of studies on human specimens as well as rodent models of the disease, to identify and analyze the molecular mechanisms behind AMD development and progression. The evaluation of this information highlights the central role that oxidative damage in the retina plays in contributing to major pathways, including inflammation and angiogenesis, found in the AMD phenotype. Following on the debate of oxidative stress as the earliest injury in the AMD pathogenesis, we demonstrated how the targeting of oxidative stress-associated pathways, such as autophagy and nuclear factor erythroid 2-related factor 2 (Nrf2) signaling, might be the futuristic direction to explore in the search of an effective treatment for AMD, as the dysregulation of these mechanisms is crucial to oxidative injury in the retina. In addition, animal models of AMD have been discussed in great detail, with their strengths and pitfalls included, to assist inform in the selection of suitable models for investigating any of the molecular mechanisms.

Axonal and cell body protection by nicotinamide adenine dinucleotide in tumor necrosis factor-induced optic neuropathy.

Axonal degeneration often leads to the death of neuronal cell bodies. Previous studies have demonstrated the crucial role of nicotinamide adenine dinucleotide (NAD) biosynthesis in axonal protection of motor neurons, but the role of nicotinamide mononucleotide adenylyltransferase 1 and NAD in optic nerve degeneration is unclear. Intravitreal injection of tumor necrosis factor (TNF) induces optic nerve degeneration and subsequent loss of retinal ganglion cells. We found that the levels of nicotinamide mononucleotide adenylyltransferase 1 mRNA and protein and of NAD were significantly decreased in the optic nerve after intravitreal injection of TNF in rats. The concomitant disorganization of microtubules with vacuoles and neurofilament accumulations in the axons were blocked by exogenous NAD treatment. Nicotinamide adenine dinucleotide also prevented TNF-induced axonal loss and delayed retinal ganglion cell loss 2 months after TNF injection. Microglia identified by immunohistochemistry were increased in the optic nerves after TNF injection; this increase was inhibited by NAD treatment. These results suggest that axonal nicotinamide mononucleotide adenylyltransferase 1 and NAD declines are associated with TNF-induced optic nerve axonal degeneration and that axonal protection of NAD may be related to its inhibitory effect on microglial activation.

Neuroprotective effects of the beta-carboline abecarnil studied in cultured cortical neurons and organotypic retinal cultures.

Presently there is no neuroprotective pharmacological treatment of proven clinical safety and efficacy available. The purpose of this study was to investigate whether the beta-carboline, abecarnil (Abe), which has already passed clinical phase III trials in patients with anxiety disorders, is neuroprotective in in vitro models of cerebral ischemia or excitotoxicity. Abe (100 nM) protected cultured cortical neurons when applied 20 min before or 20 min after combined oxygen glucose deprivation (OGD). Furthermore, cultured cortical neurons were protected from NMDA excitotoxicity when Abe (100 nM) was administered 20 min before or concurrent with 100 microM NMDA. In contrast, in adult rat organotypic retinal cultures, Abe failed to protect retinal ganglion cells (RGCs) against glutamate (Glu) excitotoxicity. Thus, although our data demonstrate that Abe is a potential neuroprotectant in cultured neurons, the lack of effect in an organotypical model of Glu toxicity indicates that further study is required before Abe might be considered for human neuroprotection trials.

TNF-alpha-induced optic nerve degeneration and nuclear factor-kappaB p65.

PURPOSE: To characterize a model of optic nerve axonal degeneration induced by tumor necrosis factor (TNF)-alpha and to determine the role of nuclear factor (NF)-kappaB p65 in axonal degeneration. METHODS: Groups of rats were euthanatized at 1 day, 1 or 2 weeks, or 1 or 2 months after intravitreal injection of TNF-alpha. Morphometric analyses of neurofilament- or Thy-1-positive cells, retinal ganglion cells (flat preparations stained with cresyl violet or retrograde labeling with a neurotracer), the number of axons, immunostaining for myelin basic protein, and TUNEL assays were performed. Levels of NF-kappaB p65 protein in retina and optic nerve were determined by Western blot analysis and immunohistochemistry. The effects of antisense oligodeoxynucleotide (AS ODN) against NF-kappaB p65 and helenalin, an inhibitor of NF-kappaB p65 activation, on TNF-alpha-induced optic nerve degeneration were determined by counting the number of axons. RESULTS: Intravitreal injections of TNF-alpha induced obvious axonal loss and extensive degeneration of the axons from 2 weeks to 2 months after injection, whereas significant retinal ganglion cell loss was noted only at 2 months after injection. NF-kappaB p65 was increased in the optic nerve but not in the retina and was found to colocalize with ED-1 and Iba1, markers of microglia. Inhibition of NF-kappaB p65 with AS ODN or helenalin significantly ameliorated the effects of TNF-alpha-mediated axonal loss. CONCLUSIONS: TNF-alpha causes axonal degeneration with probable delayed loss of retinal ganglion cell bodies. NF-kappaB p65 may play a pivotal role in axonal degeneration, with the possible involvement of microglial cells.

Nuclear factor-kappaB p65 and upregulation of interleukin-6 in retinal ischemia/reperfusion injury in rats.

We previously demonstrated that endogenous interleukin-6 (IL-6) is upregulated and may be neuroprotective after retinal ischemia. The purpose of this study is to investigate the role of nuclear factor kappa-B (NF-kappaB) in regulating IL-6 expression after ischemia. NF-kappaB p65 mRNA levels were significantly elevated between 2 and 12 h after the insult. A high number of NF-kappaB p65 positive cells were detected in the inner retina at 12 h after ischemia. Activated nuclear NF-kappaB p65 and IL-6 were colocalized in cells, which were also marked by a microglial/phagocytic cell marker (ED1) in the inner retina. Carbobenzoxy-L-leucyl-L-leucyl-L-leucinal (MG-132, a proteasome inhibitor, which inhibits IkappaB degradation and hence prevents the activation and translocation of NF-kappaB into the nucleus) abolished the increase in NF-kappaB p65 mRNA levels after the insult, while there was no effect by helenalin (an inhibitor which inhibits NF-kappaB activity by alkylation of the p65 subunit, thereby blocking its binding to the target DNA). However, MG-132 and/or helenalin significantly diminished the increase in IL-6 mRNA levels after the insult. Small interfering RNAs (siRNAs, inhibit target gene expression through the sequence-specific destruction of the target messenger RNA) against NF-kappaB p65 significantly reduced the increase in NF-kappaB p65 mRNA levels as well as IL-6 mRNA levels after ischemia. The number of retinal ganglion cells (RGCs) was also significantly decreased using the inhibitors of NF-kappaB compared with those of the controls after ischemia. These findings support the hypothesis that upregulation of endogenous retinal IL-6 in retinal I/R injury in microglial/phagocytic cells is controlled predominantly by NF-kappaB p65.

Activation of microglia and chemokines in light-induced retinal degeneration.

PURPOSE: Microglial cells, which are activated and recruited by chemokines, have been shown to play crucial roles in neuronal degenerations of the central nervous system (CNS). This study investigated the activation and migration of retinal microglial cells and expression of chemokines in retinas in light-induced photoreceptor degeneration in mice. METHODS: Ninety-five Balb/cJ mice were kept in cyclic light for 1 week followed by dark adaptation for 48 h prior to light exposure of 3 h at 3.5 Klux. Animals were enthuanized at various times after light exposure. Terminal deoxynucleotidyl transferase-mediated dUTP nick end label (TUNEL) assay, rat-anti-mouse CD11b and 5D4 antibodies, isolectin-B4, and a chemokine-specific gene array were used to detect DNA fragmentation during retinal degeneration, to label retinal microglial cells, and to determine the expression of retinal chemokines and chemokine receptors, respectively. Reverse-transcriptase coupled polymerase chain reactions (RT-PCRs) were conducted on selected chemokine mRNAs to confirm the gene array findings. RESULTS: After intense light exposure, TUNEL-positive cells were noted in the outer nuclear layer (ONL) of the retina at 3 h, and their presence were noticeably increased at 1 day but declined at 3 days and 7 days after light exposure. In contrast, CD11b- or isolectin-B4-positive cells were seen in the ONL as early as 6 h and their presence increased significantly at 1 day and 3 days after light exposure. These cells displayed a round or ovoid morphology at 6 h and 1 day but assumed a more ameboid configuration at 3 days. By 7 day, the number of the microglial cells declined in the ONL and they became ramified, and were present mostly in the subretinal space. 5D4-positive cells with large cell bodies were only noted at 3 day and 7 day but not earlier. With chemokine-specific gene array analysis, we identified four chemokines and two chemokine receptors showing significant increases in their gene expressions. Among them, monocyte chemoattractant protein-3 (MCP-3), showed a remarkable 4.4 fold increase in its gene expression. RT-PCR confirmed a marked increase of MCP-3 expression in retinas at 3 h to 1 day, and a return to normal at 3 days following light injury. CONCLUSIONS: Retinal chemokines such as MCP-3 and their receptors are involved in the activation and migration of retinal microglia in light-induced retinal degeneration, which in turn modulate the apoptotic loss of photoreceptor cells in the outer retina.

Early glial responses after acute elevated intraocular pressure in rats.

PURPOSE: To study the responses of glial cells to a short-term elevation in intraocular pressure (IOP) in rats. METHODS: Adult Sprague-Dawley albino rats, 45 to 55 days old, were given India ink intracamerally. After 7 days, 200 spots of laser burn over 360 degrees were delivered by an Argon laser (620-637 nm; 200 mW; 200 mm; 0.2 seconds) aimed at the ink deposits in the trabecular meshwork. IOP was recorded and eye tissues at 12 hours and 1, 3, 5, 7, or 14 days after laser were examined by immunohistochemistry with antibodies against glial fibrillary acidic protein (GFAP), vimentin, S-100, ED1, and OX42. To evaluate neuronal loss, the number of cells in the retinal ganglion cell layer (RGCL) was counted on flat preparations of retinas at various times after elevation of IOP. RESULTS: Significant elevation of IOP from 1 to 7 days and loss of cells in the RGCL from 3 days onward were noted after trabecular laser photocoagulation. In the inner retina, there was a gradual and sustained increase in GFAP and S-100 immunoreactivity, but only a transient increase in vimentin immunoreactivity. No remarkable changes in GFAP, vimentin, and S-100 immunoreactivity were noted at the optic nerve head (ONH). ED1- and OX42-labeled cells were noted in the choroidal plexus, the parapapillary region of the optic nerve, and the ONH from 3 days onward, whereas expression in the retina was unremarkable. CONCLUSIONS: There is differential expression of glial cell markers in the retina and the ONH, with early loss of cells in the RGCL in response to the elevation of IOP. Macroglia such as astrocytes and Müller cells may be involved in the pathophysiology of retinal ganglion cell death or retinal repair, and activated microglial/phagocytic cells may play an important role in modulating the changes in the ONH that occur with the elevation of IOP.

Neuroprotection of photoreceptors by minocycline in light-induced retinal degeneration.

PURPOSE: Microglial cells have been found to play pivotal roles in various neuronal degenerative diseases such as Parkinson's and Alzheimer's diseases. Minocycline, a microglial inhibitor, has recently been shown to be neuroprotective in various models of cerebral ischemia and degenerative diseases of the brain. This study was conducted to evaluate the neuroprotective effect of minocycline and the role of microglia in light-induced retinal degeneration. METHODS: BALB/cJ mice were exposed to intense green light for 3 hours and observed during 1, 3, or 7 days of dark recovery. The animals received intraperitoneal injections of minocycline or vehicle 1 day before exposure to light for 2, 4, or 8 days, depending on the periods of survival. Morphologic, morphometric, immunohistochemical, and electrophysiological studies were performed to evaluate the efficacy of minocycline in the amelioration of light-induced retinal degeneration and the possible involvement of microglial cells. RESULTS: Minocycline treatment provided marked amelioration in the loss of photoreceptors in light-induced retinal degeneration, as evidenced by morphologic, morphometric, and electrophysiologic criteria. Morphologically, the minocycline-treated group showed markedly better preservation of the outer retina after exposure to light. Morphometrically, at 7 days after exposure to light, in the minocycline-treated animals, 89.1% of the normal-appearing photoreceptor nuclei remained, but in the retinas of the vehicle-control group only 38.0% of these nuclei remained. This difference was statistically significant (P < 0.001). At 7 days after exposure to light electroretinography (ERG) showed that minocycline significantly preserved the amplitudes of dark-adapted a- and b-wave and light-adapted b-wave, which were all significantly reduced after exposure to light. Concomitant with this protective effect, at 3 days after exposure to light, the CD11b(+) microglial cells in the outer nuclear layer (ONL) and subretinal space in the minocycline-treated group were significantly decreased (by 63.5%) when compared with those in the light-exposed, vehicle-treated control group (P < 0.01). CONCLUSIONS: Minocycline is neuroprotective against light-induced loss of photoreceptors, possibly through the inhibition of retinal microglial activation.

Interleukin-6 in retinal ischemia reperfusion injury in rats.

PURPOSE: To study the role of interleukin (IL)-6 after retinal ischemia-reperfusion (I/R) injury in rats. METHODS: Intraocular pressure of adult male Lewis albino rats was raised to create retinal ischemia for 1 hour. Retinal reperfusion was reestablished, and the animals were killed at various time points after the injury. Their eyes were enucleated and processed for immunohistochemistry to detect IL-6 and ED-1 (a marker of microglial/phagocytic cells), enzyme-linked immunosorbent assay (ELISA) of IL-6 protein, and semiquantitative real-time RT-PCR for IL-6 mRNA. The neuroprotective effect of IL-6 was evaluated by giving intravitreal injections of 150 or 300 ng rat recombinant IL-6 to eyes immediately after I/R injury and counting cresyl violet-stained retinal ganglion cell layer cells (RGCLCs) and fluorochrome-labeled retinal ganglion cells (RGCs) on flat preparations of retinas at 7 days. RESULTS: IL-6-positive cells appeared after I/R injury in the inner plexiform layer (IPL) and the inner nuclear layer (INL). Their numbers were significantly higher 18 hours after the injury, and most of these cells were also ED-1 positive. ELISA showed noticeable increases in endogenous retinal IL-6 protein levels 8 hours after I/R injury. Semiquantitative real-time RT-PCR showed significant increases in endogenous retinal IL-6 mRNA levels between 2 and 18 hours. Exogenously added IL-6 prevented between 50% and 70% of RGC loss after I/R injury. CONCLUSIONS: IL-6 is upregulated after retinal I/R injury, and its expression by microglia/phagocytic cells may protect RGC layer neurons from I/R injury. Exogenously added IL-6 protects the inner retina after I/R injury.

Nuclear factor-kappa B p65 in NMDA-induced retinal neurotoxicity.

Transcription factors of the nuclear factor-kappa B (NF-kappaB)/Rel family may be involved in neuronal cell death or survival. We examined the role of NF-kappaB p65 in N-methyl-D-aspartate (NMDA)-induced neurotoxicity in the rat retina. Western blot analysis showed that elevated levels of retinal NF-kappaB p65 protein at days 1 and 5 after intravitreal NMDA injection. Immunohistochemistry localized increased NF-kappaB p65 immunoreactivity in the ganglion cell layer (GCL) and the inner nuclear layer (INL) after NMDA injection especially in retinal ganglion cells (RGCs), displaced amacrine cells, and amacrine cells. Concomitant with the early increase in NF-kappaB p65 protein levels, there was an increase in NF-kappaB DNA binding activity after NMDA injection as shown by electrophoretic mobility shift assay (EMSA). These increases in NF-kappaB p65 protein levels and NF-kappaB DNA binding activity were totally abolished by simultaneous injection of NF-kappaB p65 antisense oligodeoxynucleotide (AS ODN). A partial but significant protective effect on the inner retina was noted when the AS ODN was given together with NMDA as shown by morphological analysis, morphometry of cells in the GCL and morphometry of inner plexiform layer thickness as well as quantitative real-time PCR of Thy-1 mRNA levels. These results suggest that activated NF-kappaB p65 may participate in NMDA-induced retinal neuronal cell death and that inhibition of NF-kappaB activation such as the use of AS ODN may be a viable neuroprotective strategy for protective RGCs and other inner retinal neurons.

Hyperthermic pre-conditioning protects retinal neurons from N-methyl-D-aspartate (NMDA)-induced apoptosis in rat.

Glutamate-induced excitotoxicity is associated with a selective loss of retinal neurons after retinal ischemia and possibly in glaucoma. Since heat shock protein (HSP) 70 is known to play a protective role against ischemic neuronal injury, which is also linked to excitotoxicity, we studied the expression of inducible (HSP72) and constitutive (HSC70) forms of HSP70 in apoptosis of retinal ganglion cells (RGCs) after intravitreal injection of 8 nmoles N-methyl-D-aspartate (NMDA), a glutamate receptor agonist. Approximately 18 h after NMDA injection, there were increased numbers of TUNEL-positive cells and cells with elevated HSP72 immunoreactivity in the retinal ganglion cell layer (RGCL), but there were no noticeable changes in HSC70 immunoreactivity. These HSPs positive cells were also Thy-1 positive, a marker for RGCs. Hyperthermic pre-conditioning, which is known to induce HSPs, given 6 or 12 h prior to NMDA injection ameliorated neuronal loss in the RGCL as counted 7 days after NMDA injection but pre-conditioning at 18 h prior to NMDA injection did not have any ameliorative effect. Quercetin, an inhibitor of HSP synthesis, abolished the ameliorative effect of hyperthermic pre-conditioning. Pre-conditioning elevated HSP72 but not HSC70 immunoreactivity and reduced the number of TUNEL-positive cells in the RGCL at 18 h. Our results suggest that intravitreal injection of NMDA induces an up-regulation of HSP72 in a time-dependent manner but not HSC70 in RGCs, indicating a stress response of HSP72 in RGCs and other inner retinal neurons after exposure to NMDA. Hyperthermic pre-conditioning given within a therapeutic window is neuroprotective to the retina against NMDA-induced excitotoxicity, likely by inhibiting apoptosis through the modulation of HSP72 expression.

Involvement of RhoA and possible neuroprotective effect of fasudil, a Rho kinase inhibitor, in NMDA-induced neurotoxicity in the rat retina.

RhoA, a key protein involved in cytoskeleton regulation modulating neurogenesis and neural plasticity, has been implicated in a variety of cellular functions including the modulation of N-methyl-D-aspartate (NMDA) receptor activity. We examined its possible involvement in NMDA-induced excitotoxicity in the retina, and evaluated the neuroprotective effect of fasudil, a Rho kinase inhibitor, in this model of neurotoxicity. RhoA protein levels in NMDA-treated retinas were assessed by Western blot analysis and localized by immunohistochemistry. Fasudil (10(-6)-10(-4) M together with 4 x 10(-2) M NMDA) was given intravitreally and its effect was evaluated by counting the number of cells in the ganglion cell layer (GCL), measuring the thickness of the inner plexiform layer (IPL), and measuring retinal Thy-1 mRNA levels at 5 days after injection. Western blot analysis showed a transient increase in the level of retinal RhoA and ROCKII proteins at 1 day after NMDA injection, and that this increment was significantly prevented by simultaneous injection of fasudil. Immunohistochemistry showed that NMDA induced a substantial increase in RhoA immunoreactivity in the GCL and the IPL. Fasudil injection reduced cell loss in the GCL and the reduction in IPL thickness after NMDA injection. The reduction in Thy-1 mRNA levels by NMDA was also significantly attenuated by concomitant injection of fasudil. These results suggest that RhoA and ROCKII are upregulated and may be involved in NMDA-induced retinal neurotoxicity, and that fasudil is neuroprotective against glutamate-related excitotoxicity.

Relative contribution of VEGF and TNF-alpha in the cynomolgus laser-induced CNV model: comparing the efficacy of bevacizumab, adalimumab, and ESBA105.

PURPOSE: To compare the relative contribution of VEGF and TNF-alpha in the development of laser-induced choroidal neovascularization (CNV) in monkeys and to exploit the feasibility of topical use of suitable antibody fragments for the prevention of experimental CNV. METHODS: To induce experimental CNV, small high-energy laser spots were used to treat several areas of the macula in the retinas of cynomolgus monkeys according to previously published protocols. To prevent abnormalities, bevacizumab (a potent VEGF inhibitor) and adalimumab or ESBA105 (potent TNF-alpha inhibitors) were given by intravitreal injection 1 week before and 1 week and 3 weeks after laser treatment. ESBA105 was also applied topically in a separate group. Control animals were treated with either intravitreal or topical saline. Eyes were monitored by ophthalmic examination, color photography, and fluorescein angiography. RESULTS: Inhibition of VEGF by bevacizumab completely blocked the formation of CNV. Both TNF-alpha inhibitors also significantly reduced laser-induced CNV abnormalities after intravitreal administration. Most important, topical use of the anti-TNF-alpha single-chain antibody fragment ESBA105 also reduced the formation of CNV. CONCLUSIONS: TNF-alpha contributes to laser-induced CNV formation, and its inhibition can be a new therapeutic target for CNV. This study suggests TNF-alpha as another therapeutic target for the prevention and treatment of CNV and adds to the emerging clinical data suggesting the therapeutic value of TNF-alpha inhibitors in age-related macular degeneration (AMD). Further, this study shows that topical therapy with suitable antibody fragments has the potential of being introduced to retinal disease treatment regimens.

Heterogeneous populations of microglia/macrophages in the retina and their activation after retinal ischemia and reperfusion injury.

Activation of Microglia/macrophages has been observed in ischemia-reperfusion injury of the brain. This study was undertaken to investigate the different subpopulations of microglia/macrophages in the normal rat retina and their activation after retinal ischemia. Retinal ischemia was induced by elevation of intraocular pressure to 120 mmHg for 60 min. Microglia/macrophages were identified on frozen retinal sections by four antibodies, namely OX42, 5D4, OX6 and ED1. In the normal retina, there were heterogeneous populations of resident microglia/macrophages as characterized by their differences in morphology, antigen expression and distribution. OX42+ cells had delicate processes and were located in the inner layers of the retina, while 5D4+ cells were highly ramified and mostly scattered in the inner plexiform layer (IPL) and the outer plexiform layer. Few amoeboid ED1+ cells were also seen in the ganglion cell layer and IPL. OX6+ (MHC-II antigen presenting) cells were not detected in the normal retinas. Double labeling with OX42 and 5D4 antibodies on normal retinal sections showed few microglia exhibited positive labeling with both OX42 and 5D4, while the majority of the microglia were labeled with either OX42 or 5D4 antibodies. After retinal ischemia single labeling with these antibodies showed increased number of these antigen-expressing cells, disappearance of normal cellular processes, and rounding or amoeboid like appearance of the cell bodies. At 1 day after ischemia, there was a significant infiltration of round OX42+, ED1+ and OX6+ cells with loss of the cellular processes in the inner retina. From 3 to 14 days, all subpopulations of microglia/macrophages differentiated cellular processes and became dendritic again. Double labeling on retinas after 1 day of recovery showed OX42+ cells were co-labeled with ED1+ or OX6+ cells, but not with 5D4+ cells. Scattered amoeboid OX42+, 5D4+, and ED1+ cells were noted in the subretinal space 3-14 days after ischemia. In summary, there were heterogeneous populations of resident microglia/macrophages in the normal inner retina and they were activated early after ischemia-reperfusion injury and exhibited different antigenic expression which were further altered in the recovery phase.