Three million individuals in the United States are projected to have end-stage manifestations of age-related macular degeneration (AMD) by the year 2020.1 More than one-third of these individuals with end-stage AMD will have geographic atrophy (GA), a major cause of blindness in the elderly population.2,3 Although anti-vascular endothelial growth factor (anti-VEGF) therapy has been successful in managing wet AMD, long-term follow-up data from the CATT and SEVEN-UP trials report that a significant proportion patients, up to 90%, develop fovea-involving GA during the course of therapy.4,5 At present, there are no effective therapies for GA, although multiple agents are actively being investigated through clinical trials.6
Amyloid-beta 1–42 (Aβ42), considered the most pathogenic isoform of Aβ peptide, has been implicated in both neurodegenerative disease and AMD.7,8 Aβ has been found in drusen, particularly drusen at the margins of GA.9 Colocalization of activated complement and amyloid vesicles within drusen links the complement cascade in the pathogenesis of GA.10
Brimonidine, considered a neuroprotective agent in models of glaucoma and retinal ischemia, was shown to reduce the progression of GA in a subset of patients from a phase 2 clinical trial.11,12 Brimonidine has been shown to protect adult human retinal pigment epithelial (RPE) and Müller cells against oxidative stress in vitro.13 It is not known whether the neuroprotective effects of brimonidine translate to these cells when exposed to a neurotoxic agent, Aβ42. This investigation examines the in vitro cyto-metabolic effects of brimonidine on adult human RPE cells, the main cell population affected in AMD, and human Müller glial (MIO) cells when exposed to Aβ42.
Materials and Methods
Adult human RPE cells (ARPE-19; ATCC, Manassas, VA) were obtained and cultured in a 1:1 mixture of Dulbecco's Modified Eagle's Medium and Ham's F-12 Medium (Corning–Cellgro, Mediatech, Manassas, VA). Human retinal Müller cells (MIO-M1), obtained from the Department of Cell Biology of the University College, London, were grown in Dulbecco's modified medium with high glucose (Gibco, Carlsbad, CA). Cells were cultured at a concentration of 10,000 cells per well.
Brimonidine tartrate 0.1% was obtained from an ophthalmic solution (Alphagan P; Allergan, Irvine, CA) that contained Purite as a vehicle. Dilutions were made to reach three different concentrations, one time (1×), two times (2×), and five times (5×). The 1× concentration was considered to be equivalent to 50μL/4mL solution.
Aβ42 (AnaSpec Protein, Fremont, CA) was reconstituted and diluted to achieve a final concentration of 10μM in culture media. This concentration of Aβ42 was determined by titration to cause a 30% reduction in cell viability. Cells were pretreated with brimonidine in their respective concentrations (1×, 2×, and 5×) for 6 hours, then exposed to Aβ42 in culture media and incubated at 37°C for 24 hours. A parallel experiment was run with scrambled Aβ peptide to assess for toxicity due to addition of peptide alone. The plates were taken for the following assays after the completion of the treatment phase.
Reactive oxygen species (ROS) assay. ROS dye powder (2′,7′-Dichlorofluorescein diacetate) (Thermo Fisher Scientific, Waltham, MA) was added to each well and allowed to incubate before being replaced with phosphate buffer solution. The plates were immediately placed into the microplate reader at filter settings of 492 nm excitation/520 nm emission. This assay detects ROS products such as hydrogen peroxide and peroxyl radicals.
Cell viability assay. MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) reagent (Biotium, Fremont, CA) is reduced by healthy mitochondria to an insoluble product that has a purple color; therefore, colorimetry can be performed to quantify the degree of cell viability in each well. Each plate was placed into a spectrophotometer (ELX 808 Absorbance Reader; Biotek, Winooski, VT) set to measure absorbance at a wavelength of 570 nm (signal) and reference wavelength (background) of 630 nm.
Mitochondrial membrane potential (ΔΨm) assay. JC-1 iodide salt (5,5′,6,6′-Tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine, iodide) (Biotium, Fremont, CA) is a cationic dye that accumulates within healthy mitochondria and acquires red fluorescence. In unhealthy mitochondria, the dye remains in the cytoplasm and acquires green fluorescence. The plates were then analyzed by a microplate reader (SpectraMax Gemini XPS; Molecular Devices, Sunnyvale, CA) using the following excitation/emission settings: 550 nm excitation/600 nm emission (red) and 485 nm excitation/535 nm emission (green). The ratio of red to green fluorescence is calculated and used to reflect the ΔΨm. A reduction in the ΔΨm indicates metabolic dysfunction and is considered an early marker for cellular apoptosis.
All assays were run in quadruplets. Untreated (control) wells were normalized to 100% signal for the purpose of comparison to the treated samples. Statistical analysis was performed using GraphPad Prism, Version 5.0 (GraphPad Software, San Diego, CA). Student's t tests were used for pairwise comparisons. A P value of less than .05 was considered statistically significant. All statistical data and error bars are presented to represent standard error of the mean.
Brimonidine at higher doses (2× and 5×) successfully reduced ROS production in RPE cells after exposure to Aβ42 (Figure 1A). Normalized relative fluorescent signal detected in RPE cultures treated with Aβ42 alone was 0.87 ± 0.08, compared to a reduced value of 0.58 ± 0.05 (P = .031) and 0.54 ± 0.08 (P = .030) in the samples pretreated with brimonidine 2× and 5×, respectively. In MIO cells, only the highest dose of brimonidine (5×) was effective in reducing ROS production after exposure to Aβ42 (Figure 1B). Normalized relative fluorescence for the MIO culture exposed to Aβ42 alone was 0.81 ± 0.09, compared to a reduction to 0.45 ± 0.03 (P = .008) in the sample pretreated with high-dose (5×) brimonidine. In both cell cultures, exposure to Aβ42 alone did not significantly increase ROS production. Conversely, the addition of high-dose (5×) brimonidine alone was sufficient to reduce ROS production relative to the control samples in both RPE (1.00 ± 0.17 vs 0.48 ± 0.08, P = .036) and MIO (1.05 ± 0.04 vs 0.47 ± 0.05, P < .001) cells.
Reactive oxygen species (ROS) production. (A) Adult human retinal pigment epithelial (RPE) cells after pretreatment with 1×, 2×, and 5× concentrations of brimonidine and exposure to amyloid-beta 1–42 (Aβ42) 10μM. (B) Müller (MIO) cells after pretreatment with 1×, 2×, and 5× concentrations of brimonidine and exposure to Aβ42 10μM. Data are normalized to set control (untreated) samples at 100% of fluorescence signal for comparison.
Aβ42 was toxic to cell viability in both RPE and MIO cells (Figure 2A). Normalized mean cell viability values were reduced from 1.00 ± 0.07 in untreated RPE cells to 0.59 ± 0.03 (P = .002), and from 1.00 ± 0.04 in untreated MIO cells to 0.84 ± 0.01 (P = .008), when exposed to Aβ42. Brimonidine, at all doses, failed to rescue cell viability in RPE cells after exposure to Aβ42. Although high-dose (5×) brimonidine alone did not reduce RPE cell viability compared to the control sample (1.00 ± 0.07 vs 0.87 ± 0.03, P = .147), high-dose (5×) brimonidine combined with exposure to Aβ42 had a synergistic toxicity reducing RPE cell viability significantly when compared to the sample treated with Aβ42 alone (0.58 ± 0.03 vs 0.51 ± 0.01, P = .043). Brimonidine also failed to rescue cell viability in MIO cells (Figure 2B). MIO cells were even more sensitive to the synergistic toxicity of brimonidine, at both 2× (0.84 ± 0.01 vs 0.67 ± 0.03, P = .001) and 5× (0.84 ± 0.01 vs 0.62 ± 0.01, P < .001) doses, when exposed to Aβ42. Treatment with high-dose (5×) brimonidine alone did not cause toxicity to the MIO cells (1.00 ± 0.04 vs 0.87 ± 0.08, P = .188) when compared to the control.
Cell viability assay (MTT Assay). (A) Adult human retinal pigment epithelial (RPE) cells after pretreatment with 1×, 2×, and 5× concentrations of brimonidine and exposure to amyloid-beta 1–42 (Aβ42) 10μM. (B) Müller (MIO) cells after pretreatment with 1×, 2×, and 5× concentrations of brimonidine and exposure to Aβ42 10μM. Data are normalized to set control (untreated) samples at 100% of signal for comparison.
Mitochondrial Membrane Potential
RPE cells were largely resistant to changes in the mitochondrial membrane potential after addition of Aβ42 (Figure 3A). In RPE cells exposed to Aβ42, mitochondrial membrane potential was reduced only in the sample pretreated with high-dose (5×) brimonidine. The normalized ΔΨm fluorescence ratio in RPE cells exposed to Aβ42 alone was 1.18 ± 0.11 compared to 0.85 ± 0.05 (P = .038) in the sample pretreated with high-dose (5×) brimonidine. In contrast, the mitochondrial membrane potential of MIO cells was exquisitely sensitive to the addition of any compound into cell culture, including the scrambled Aβ peptide and even high-dose (5×) brimonidine alone (Figure 3B). The normalized ΔΨm fluorescence ratio in untreated MIO cells was 1.00 ± 0.05, compared to 0.86 ± 0.02 (P = .018) and 0.66 ± 0.03 (P = .001) samples treated with scrambled Aβ peptide and high-dose (5×) brimonidine, respectively. The synergistic toxicity to the mitochondrial membrane potential with brimonidine and Aβ42 appeared at both the 2× (0.80 ± 0.03 vs 0.67 ± 0.02, P = .011) and 5× (0.80 ± 0.03 vs 0.63 ± 0.02, P = .003) doses of brimonidine in MIO cells.
Mitochondrial membrane potential assay (JC-1 Assay). (A) Adult human retinal pigment epithelial (RPE) cells after pretreatment with 1×, 2×, and 5× concentrations of brimonidine and exposure to amyloid-beta 1–42 (Aβ42) 10μM. (B) Müller (MIO) cells after pretreatment with 1×, 2×, and 5× concentrations of brimonidine and exposure to Aβ42 10μM. Data are normalized to set control (untreated) samples at 100% of signal for comparison.
Brimonidine was capable of significantly reducing ROS production in both RPE and MIO cells after exposure to Aβ42. Brimonidine also significantly reduced ROS production independent of Aβ42 exposure in both cell lines. These results are consistent with previous findings that demonstrate brimonidine as an effective antioxidant in RPE and MIO cells when exposed to hydroquinone, a component of cigarette smoke that is a strong trigger of oxidative stress.13 This protective quality of brimonidine is also observed in experimental models of oxidative stress in rat retinal ganglion cell lines.14 These findings point toward potential utility for brimonidine to protect against ROS-mediated cellular injury in AMD.15
Despite its antioxidant effect, brimonidine was unable to rescue reductions in cell viability in RPE and MIO cell lines after exposure to Aβ42. Instead, high-dose (5×) brimonidine had an additive toxic effect on cell viability in RPE cells, and both higher doses (2× and 5×) of brimonidine had an additive toxic effect on MIO cells. This reduction in cell viability was not seen in cultures treated with high-dose (5×) brimonidine alone, suggesting that this effect was not due to medication toxicity and was indeed a synergistic effect with Aβ42.
A reduction in mitochondrial membrane potential signifies a decline in cellular metabolic function and is considered an early marker for apoptosis. In both RPE and MIO cell lines, brimonidine failed to rescue the reduction in mitochondrial membrane potential after exposure to Aβ42. Again, recapitulating the pattern from cell viability assays, the mitochondrial membrane potential was instead damaged by the addition of high-dose (5×) brimonidine in both cell lines. Interestingly, MIO cells showed a reduction of mitochondrial membrane potential with the addition of any agent in cell culture media, including the scrambled Aβ peptide, as well as high-dose (5×) brimonidine alone, while RPE cells were far more resilient to the addition of Aβ42 and brimonidine. Our laboratory findings have previously echoed this trend that MIO cells are a more vulnerable cell population than RPE cells. (M.C.K., unpublished data)
There are two considerations to why brimonidine was ineffective at rescuing cell viability and mitochondrial membrane potential after exposure to Aβ42. First, Aβ42 did not trigger increased ROS production in both cell lines, suggesting that the toxicity of Aβ42 is mediated through a pathway separate from oxidative stress. Instead, Aβ is thought to trigger the complement pathway in AMD, leading to complement mediated cell lysis and attendant cellular necrosis.10 Aβ appears to be a link between the complement cascade and the pathophysiology of GA, as the complement cascade is an active target of investigation for the treatment of GA.6 Although brimonidine may have utility in preventing cellular damage from oxidative stress, it has no effect in preventing cellular necrosis due to complement activation (Figure 4). Second, brimonidine has recently been shown to alter the processing of amyloid precursor protein (APP) to favor production toward a less cytotoxic end product, αAPP, instead of Aβ peptide (Figure 5).16 Our experimental model bypasses this beneficial effect by directly adding Aβ42 to cell culture, examining the downstream effects of Aβ42-specific cytotoxicity.
Proposed schematic of pathways to cell death in geographic atrophy. The upper arm of the pathway is a consequence of the complement cascade and leads to cellular necrosis. The lower arm of the pathway is a consequence of oxidative stress and leads to both necrosis and apoptosis. Both pathways are rooted from the toxic stimuli contained within drusen, namely, amyloid-beta and oxidized lipoproteins, among many other toxic subcomponents. Treatment of geographic atrophy will likely necessitate a multifactorial approach to prevent oxidative stress as well as complement mediated cellular injury.
Processing of amyloid precursor protein (APP) and brimonidine. APP can be cleaved by α-secretase to produce a more soluble peptide, αAPP, which is considered less toxic than Aβ. Processing of APP by β-secretase leads to the production of Aβ and its attendant cytotoxicity. Brimonidine is thought to preferentially favor production of αAPP, thereby reducing overall cytotoxicity.
This experimental model only represents a one-dimensional model of drusen, examining only the toxic effects of Aβ42. Drusen are composed of a plethora of cytotoxins that do not derive from the amyloid peptide family, such as oxidized lipoproteins and lipofuscin.17 Aβ42 was intentionally chosen in this experiment because of its putative link to GA, and the clinical relevance to the recent findings from a clinical trial using brimonidine to prevent progression of GA.9,10,12 It is possible that brimonidine may be protective against the non-amyloid components of drusen, which cause oxidative stress (Figure 4). Furthermore, although brimonidine has been shown to be neuroprotective for cells of neuronal lineage in experimental models of glaucoma and retinal ischemia, the effect on cells of pigment epithelial and glial lineage is not well characterized.11 Thus, it is not surprising that the neuroprotective effect of brimonidine failed to translate to cells of non-neuronal lineage. However, it was important to use and characterize the cellular responses of RPE cells in this experiment because of their central role in the pathogenesis of AMD.
Taken together, these data show that brimonidine is effective at reducing ROS production even after exposure to Aβ42 in RPE and MIO cell lines, but this protective effect did not translate to cell viability and mitochondrial membrane potential. There is still an opportunity to study brimonidine as a therapy to prevent cellular injury from oxidative stress in AMD, which did not appear to be triggered by the application of Aβ42 to RPE and MIO cells. This in vitro model of AMD may have further applications in elucidating pathways to cell death in GA by examining inhibitors of the apoptosis pathway and complement cascade.
- Friedman DS, O'Colmain BJ, Muñoz B, et al. Prevalence of age-related macular degeneration in the United States. Arch Ophthalmol. 2004;122(4):564–572. doi:10.1001/archopht.122.4.564 [CrossRef]
- Smith W, Assink J, Klein R, et al. Risk factors for age-related macular degeneration: pooled findings from three continents. Ophthalmology. 2001;108(4):697–704. doi:10.1016/S0161-6420(00)00580-7 [CrossRef]
- Klein R, Meuer SM, Knudtson MD, Klein BE. The epidemiology of progression of pure geographic atrophy: the Beaver Dam Eye Study. Am J Ophthalmol. 2008;146(5):692–699. doi:10.1016/j.ajo.2008.05.050 [CrossRef]
- Rofagha S, Bhisitkul RB, Boyer DS, Sadda SR, Zhang KSEVEN-UP Study Group. Seven-year outcomes in ranibizumab-treated patients in ANCHOR, MARINA, and HORIZON: a multicenter cohort study (SEVEN-UP). Ophthalmology. 2013;120(11):2292–2299. doi:10.1016/j.ophtha.2013.03.046 [CrossRef]
- Maguire MG, Martin DF, Comparison of Age-related Macular Degeneration Treatments Trials (CATT) Research Group et al. Five-year outcomes with anti-vascular endothelial growth factor treatment of neovascular age-related macular degeneration: the comparison of age-related macular degeneration treatments trials. Ophthalmology. 2016;123(8):1751–1761. doi:10.1016/j.ophtha.2016.03.045 [CrossRef]
- Boyer DS, Schmidt-Erfurth U, van Lookeren Campagne M, Henry EC, Brittain C. The pathophysiology of geographic atrophy secondary to age-related macular degeneration and the complement pathway as a therapeutic target. Retina. 2017;37(5):819–835. doi:10.1097/IAE.0000000000001392 [CrossRef]
- Murphy MP, LeVine H 3rd, . Alzheimer's disease and the amyloid-beta peptide. J Alzheimers Dis. 2010;19(1):311–323. doi:10.3233/JAD-2010-1221 [CrossRef]
- Ratnayaka JA, Serpell LC, Lotery AJ. Dementia of the eye: the role of amyloid beta in retinal degeneration. Eye. 2015;29(8):1013–1026. doi:10.1038/eye.2015.100 [CrossRef]
- Dentchev T, Milam AH, Lee VM, Trojanowski JQ, Dunaief JL. Amyloid-beta is found in drusen from some age-related macular degeneration retinas, but not in drusen from normal retinas. Mol Vis. 2003;9:184–190.
- Johnson LV, Leitner WP, Rivest AJ, Staples MK, Radeke MJ, Anderson DH. The Alzheimer's A beta-peptide is deposited at sites of complement activation in pathologic deposits associated with aging and age-related macular degeneration. Proc Natl Acad Sci U S A. 2002;99(18):11830–11835. doi:10.1073/pnas.192203399 [CrossRef]
- Saylor M, McLoon LK, Harrison AR, Lee MS. Experimental and clinical evidence for brimonidine as an optic nerve and retinal neuroprotective agent: an evidence-based review. Arch Ophthalmol. 2009;127(4):402–406. doi:10.1001/archophthalmol.2009.9 [CrossRef]
- Kuppermann BD, Patel SS, Boyer DS, et al. Brimonidine Drug Delivery System (DDS) Generation 1 in patients with geographic atrophy: post-hoc analysis of a phase 2 study. Invest Ophthalmol Vis Sci. 2017;58(8):1924.
- Ramírez C, Cáceres-del-Carpio J, Chu J, et al. Brimonidine can prevent in vitro hydroquinone damage on retinal pigment epithelium cells and retinal Müller cells. J Ocul Pharmacol Ther. 2016;32(2):102–108. doi:10.1089/jop.2015.0083 [CrossRef]
- Lee KYC, Nakayama M, Aihara M, Chen YN, Araie M. Brimonidine is neuroprotective against glutamate-induced neurotoxicity, oxidative stress, and hypoxia in purified rat retinal ganglion cells. Mol Vis. 2010;16:246–251.
- Beatty S, Koh H, Phil M, Henson D, Boulton M. The role of oxidative stress in the pathogenesis of age-related macular degeneration. Surv Ophthalmol. 2000;45(2):115–134. doi:10.1016/S0039-6257(00)00140-5 [CrossRef]
- Nizari S, Guo L, Davis BM, et al. Non-amyloidogenic effects of α2 adrenergic agonists: implications for brimonidine-mediated neuroprotection. Cell Death Dis. 2016;7(12):e2514. doi:10.1038/cddis.2016.397 [CrossRef]
- Holz FG, Strauss EC, Schmitz-Valckenberg S, van Lookeren Campagne M. Geographic atrophy: clinical features and potential therapeutic approaches. Ophthalmology. 2014;121(5):1079–1091. doi:10.1016/j.ophtha.2013.11.023 [CrossRef]