Memantine, Simvastatin, and Epicatechin Inhibit 7-Ketocholesterol-induced Apoptosis in Retinal Pigment Epithelial Cells But Not Neurosensory Retinal Cells In Vitro

Abstract Purpose 7-ketocholesterol (7kCh), a natural byproduct of oxidation in lipoprotein deposits is implicated in the pathogenesis of diabetic retinopathy and age-related macular degeneration (AMD). This study was performed to investigate whether several clinical drugs can inhibit 7kCh-induced caspase activation and mitigate its apoptotic effects on retinal cells in vitro. Methods Two populations of retinal cells, human retinal pigment epithelial cells (ARPE-19) and rat neuroretinal cells (R28) were exposed to 7kCh in the presence of the following inhibitors: Z-VAD-FMK (pan-caspase inhibitor), simvastatin, memantine, epicatechin, and Z-IETD-FMK (caspase-8 inhibitor) or Z-ATAD-FMK (caspase-12 inhibitor). Caspase-3/7, -8, and -12 activity levels were measured by fluorochrome caspase assays to quantify cell death. IncuCyte live-cell microscopic images were obtained to quantify cell counts. Results Exposure to 7kCh for 24 hours significantly increased caspase activities for both ARPE-19 and R28 cells (P < 0.05). In ARPE cells, pretreatment with various drugs had significantly lower caspase-3/7, -8, and -12 activities, reported in % change in mean signal intensity (msi): Z-VAD-FMK (48% decrease, P < 0.01), memantine (decreased 47.8% at 1 µM, P = 0.0039 and 81.9% at 1 mM, P < 0.001), simvastatin (decreased 85.3% at 0.01 µM, P < 0.001 and 84.8% at 0.05 µM, P < 0.001) or epicatechin (83.6% decrease, P < 0.05), Z-IETD-FMK (68.1% decrease, P < 0.01), and Z-ATAD-FMK (47.7% decrease, P = 0.0017). In contrast, R28 cells exposed to 7kCh continued to have elevated caspase-3/7, -8, and -12 activities (between 25.7% decrease and 17.5% increase in msi, P > 0.05) regardless of the pretreatment. Conclusion Several current drugs protect ARPE-19 cells but not R28 cells from 7kCh-induced apoptosis, suggesting that a multiple-drug approach is needed to protect both cells types in various retinal diseases.


INTRODUCTION
Apoptosis is a highly regulated process of programmed cell death critical in various disease states and degenerations. These pathways are implicated in various neurodegenerative diseases (e.g., Alzheimer's disease, Parkinson's disease, amyotrophic lateral sclerosis), ocular diseases (e.g., glaucoma, diabetic retinopathy and age-related macular degeneration [AMD]), and immunologic disorders. [1][2][3][4][5][6] While treatments for the wet form of AMD rely on inhibiting choroidal neovascularization via anti-vascular endothelial growth factor (anti-VEGF) injections, [7] there is currently no cure for the dry form of AMD.
A pro-apoptotic oxysterol implicated in AMD is 7-ketocholesterol (7kCh), a toxic metabolite generated from the oxidation of cholesterol-esters in low density lipoprotein (LDL), atheromatous plaques, and drusen. [6,[8][9][10] 7kCh activates three distinct kinase signaling pathways via NFkB, MAPK, and ERK to upregulate proinflammatory cytokines: IL-6, IL-8, and vascular endothelial growth factor (VEGF) which induce neovascularization in the choroid. [7,8,[11][12][13] In previous studies, retinal pigment epithelial (RPE) cells and vascular endothelial cells had an 8-to 10-fold increase in VEGF levels, [14] and a 2-fold increase in endothelial smooth muscle cells after exposure to 7kCh. [15] Our previous studies showed that in both ARPE-19 and rat neuroretinal R28 cell cultures, 7kCh significantly increased the levels of pro-apoptotic caspase-3, -8, and -12 activities. [16][17][18][19] 7kCh is also a chemoattractant that directs microglia to the outer retina to produce metalloproteinases that cause breaks in Bruch's membrane and produce pro-angiogenic substances. [11,14,20] A 2019 study showed that 7kCh intravitreally injected into rat retinas induced apoptosis in photoreceptors and RPE cells and caused microvilli detachment in the outer segment. [10] Thus, the toxic accumulation of 7kCh over decades may activate cellular processes that predispose patients to age-related ocular pathologies such as AMD. [8] Despite the strong evidence of 7kCh's role in age-related diseases, little is known about its detoxifying mechanisms. A 2019 study showed that compounds such as vitamin E, oleic acid, terpenoids, and polyphenols inhibit 7kCh-induced apoptosis, [21] while resveratrol, an antioxidant and anti-inflammatory stilbenoid found in the skin of grapes and red wine inhibits 7kCh-induced VEGF expression. [22] Successful inhibition by these natural compounds advances our knowledge related to functional food therapies. [21] Thus, it is necessary to explore the potential drugs or inhibitors given the limited body of knowledge on 7kCh detoxification.

Cell Culture
R28 rat neurosensory cells have properties that resemble those of various human neurosensory cells. The R28 cells derived from retina of post-natal day 6 rats were cultured in Dulbecco's modified Eagle's media, This is an open access journal, and articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as appropriate credit is given and the new creations are licensed under the identical terms. 10µg/mL gentamicin, 10mM non-essential amino acids, and high glucose (DMEM high glucose; Invitrogen-Gibco, Carlsbad, CA) with 10% fetal bovine serum. Despite their clonal origin, R28 cells are characteristically heterogeneous, display both glial and neuronal cell markers, and have functional neuronal properties. [23] R28's diversity of cell types, ability to respond to a variety of stimuli, and differentiation potential makes it an appropriate model for neuroretinal tissue. [24] A 1:1 mixture (vol/vol) of Dulbecco's modified Eagle's and Ham's nutrient mixture F-12 was used to grow ARPE-19 cells (ATCC, Manassas, VA); (Invitrogen-Gibco, Carlsbad, CA), 0.37% sodium bicarbonate, 0.058% L-glutamine, 10% fetal bovine serum, antibiotics (100U/ml penicillin G, 0.1mg/ml streptomycin sulfate, 10µg/mL gentamicin), 10mM non-essential amino acids, and an anti-fungal agent (amphotericin-B 2.5µg/mL). ARPE-19 cells are a homogeneous, retinal-derived cell line with functional and structural properties similar to human RPE cells. [25] All cells within the ARPE-19 culture express RPE-specific markers such as CRALBP, BEST1, and RPE-65. [26] Cell Plating Cells were plated on 24 well plates (Becton Dickinson Labware, Franklin Lakes, NJ) at 150,000 cells per well and incubated at 37°C in 5% CO 2 until confluent. Cell incubation was performed in a serum-free media for 24 hours to prevent excessive proliferation. Cells in the experimental group were then pretreated with various inhibitors while the cells in the control group were only exposed to DMSO (the vehicle for 7kCh). All experimental cells besides the control cells were then exposed to 30 µg/ml 7kCh for 24 hours before caspase activity was assessed.

Caspase-3/7, -8, and -12 Detection
Carboxyfluorescein FLICA Caspase Apoptosis Detection kits (Immunochemistry Technologies LLC, Bloomington, MN) were used to quantify cells undergoing caspase-mediated apoptosis. The FLICA Detection Reagent contains a caspase inhibitor sequence linked to a fluorescent carboxyfluorescein probe with an optimal excitation range of 488-492 nm and an emission range of 515-535 nm. The probe only fluoresces when the FLICA reagent covalently binds to activated caspase-3/7 and -8 in vitro. The fluorescence intensity quantifies the number of whole, living cells undergoing caspase-mediated apoptosis since cells with premature, inactivated, or no caspase activity will not fluoresce. This was measured using the Fluorescence Image Scanning Unit (FMBIO III) instrument (Hitachi, Yokohama, Japan).
Fresh culture media were used to rinse the wells at 24 hours. Then, 300 µl/well of FLICA solution in culture media were placed in the wells for 1 hour. Phosphate buffered saline (PBS) was used to wash the wells three times. In addition to the experimental groups, we also analyzed the following control groups: (1) untreated R28 cells and ARPE-19 cells without FLICA or CaspGLOW to exclude autofluorescence from cells; (2) untreated R28 and ARPE-19 cells with FLICA or CaspGLOW for comparison of caspase activity of treated cells; (3) tissue culture plate wells (without cells) with buffer alone to represent the background levels; (4) tissue culture plate wells without cells with culture media and DMSO in order to exclude cross-reaction of culture media and/or DMSO with the plastic material of the tissue culture plate or cross-reaction of FLICA or CaspGLOW; (5) R28 cells and ARPE-19 cells with DMSO and FLICA to account for any cross-fluorescence due to the cross-reaction of untreated cells with DMSO; and (6) untreated R28 and ARPE-19 cells with negative control Z-FA-FMK and FLICA as a control for the pan caspase inhibitor.
IncuCyte® Live-Cell Imaging Analysis was used to capture fluorescent images in real time. NucLight Rapid Red and Caspase-3/7 Green reagents were used to stain cells in 96-well plates at 5,000-10,000 cells/well. The NucLight Rapid Red reagent quantifies the amount of cell proliferation in IncuCyte live-cell images by staining the nuclei of living cells. The Caspase-3/7 Green reagent contains an oligopeptide cleavage sequence (DEVD) conjugated to a DNA-binding dye. The green reagent labels apoptotic cells at an emission maximum of 533 nm once the sequence is cleaved by caspase-3/7. Images were captured per well every 3 hours for five days with Brightfield, Red, and Green channels. Apoptosis due to caspase-3/7 activity is discriminated from live and necrotic cells by overlapping the signal counts from NucLight Red with the Caspase-3/7 green.

Statistical Analysis
ANOVA using GraphPad Prism 3.0 version statistics program (GraphPad Software Inc., San Diego, CA) was used for statistical analysis of the data. The data within each experiment were compared using the Newman-Keuls multiple comparison test. Mean ± standard error of mean (SEM) was used to present the data. Experiments were performed in triplicate. P-values < 0.05 were considered statistically significant.
Representative IncuCyte live-cell images of ARPE-19 cells treated with DMSO-Control, 7kCh, and 7kCh with epicatechin, respectively, are shown in Figure 2G. 7kCh increased NucLight Red signal, Caspase-3/7 green signal, and overlap signal counts at 24 hours ( Figure 2G, second row) compared with DMSO-Control ( Figure 2G, first row) in ARPE-19 cells. These counts are reduced when cultures are pretreated with inhibitors such as epicatechin before exposure to 7kCh ( Figure 2G, third row).

The error bars represent the standard error of the mean (SEM) of measurements for the four conditions in three separate runs (n = 12, A-D) and three conditions in three separate runs (n = 9, E-F). Scale bar = 400 μM (G).
pretreatment with Z-ATAD-FMK (28,379 ± 4518 msi, P > 0.05, Figure 3D).

DISCUSSION
This study emphasized 7kCh's role in caspase activation and highlighted several clinical drugs that protected RPE cells but not R28 cells from 7kCh-induced apoptosis. Research in the past two decades has strongly supported 7kCh's role in AMD pathogenesis. 7kCh from LDL is esterified via two enzymes, cPLA2α and SOAT1, to 7kChfatty acid esters (7KFAEs), and effluxed by HDL to the liver for bile acid production. [27] Cholesterol levels may therefore contribute to ocular disease as high serum levels of LDL or 7KFAEs (which revert to 7kCh), or low levels of HDL can result in increased 7kCh levels. [27] 7kCh may also accumulate over time from natural mechanisms such as the rhodopsin cycle and other photo-oxidative processes. [28] Compared to controls, 7kCh was found at 6-to 50-fold higher levels in RPE and photoreceptor inner segments of photodamaged rat retinas, [9] and at 4-fold levels in Bruch's membrane and neuroretinal cells in photodamaged monkey retinas. [14,28] Excess 7kCh can trigger caspase pathways, excess reactive oxygen species (ROS) production, endothelial dysfunction, breaks in Bruch's membrane, induction of cytokines IL-6 and IL-8, and neovascularization in the choroid [16][17][18][19][20][29][30][31] , which are all consistent with AMD. Additional damage includes increased levels of DNA fragmentation and increased chromatin condensation in both ARPE-19 and R28 cells. [32] 7kCh also damages mitochondria by inducing reactive oxygen/nitrogen species, reducing the mitochondrial membrane potential by 2.2-fold (P < 0.001), and decreasing levels of intact 16.2-kb mtDNA. [16] The clinically available inhibitors used in this study all prevented 7kCh induced caspase-3/7 activation in ARPE-19 cells but not R28 cells: (1) Memantine is a moderate-affinity, noncompetitive antagonist of the glutamatergic N-methyl-D-aspartate (NMDA) receptors. It suppresses glutamate-excitotoxicity in animal models associated with neurological diseases such as Alzheimer's, vascular dementia, and retinal ganglion cell death after ischemic strokes. [33][34][35][36][37] By inhibiting downstream p53 and calpain-caspase-3, it reduces Ca 2+ -dependent production of NO and O2(-) and halts neuronal apoptosis in rats with middle cerebral artery occlusion. [36,38,39] A previous study showed that memantine reduces caspase-3/7 activity in staurosporine-treated neuronal cells. [40] The present study showed 1 μM and 1 mM of memantine reduced caspase-3/7 activity by 47.8% and 81.9% in ARPE-19 cells, respectively, but only 5.9% and 4.5% in R28 cells (Figures  2A-2B) despite evidence of immunoreactivity to NMDA and GABAa receptors. [41] This may be because memantine can upregulate brainderived neurotrophic factor through the PI3-K/Akt pathway independent of NMDA receptors. [42] Furthermore, R28 cells derived from six-day old rat retina are not fully differentiated and may not have well-developed NMDA receptors. R28 cultures are also heterogeneous and memantine may preferentially affect neural cells with NMDA receptors over non-neural glial cells. [23] (2) Simvastatin is a hydrophobic drug that inhibits 3-Hydroxy-3-methyl-glytaryl coenzyme A (HMG-CoA) reductase and is frequently utilized to reduce serum LDL in hyperlipidemic patients. [43] Besides its lipid-lowering effects, it is also neuroprotective and upregulates Bcl-2, a major cell survival protein, [44,45] and induces Hsp27, a heat shock protein that promotes retinal ganglion cell survival. [46] In rats with streptozotocininduced diabetes, therapeutic levels of simvastatin slowed progression of diabetic retinopathy by reducing VEGF activation, vascular permeability, and endothelial-leukocyte adherence. [47] The effects of statins are notably biphasic: low concentrations of statins are pro-angiogenic, while high concentrations are angiostatic. [48][49][50] Higher concentrations can also induce caspase-3/7 apoptosis in human monocytes, pericytes, and tumor cells. [50][51][52] Our results show that 0.01 μM and 0.05 μM of simvastatin reduced caspase-3/7 activity by 85.3% and 84.8%, respectively, in ARPE-19 cells (Figure 2C), while 0.05 μM simvastatin increased caspase activity by 17.5% in R28 cells ( Figure 2D). The therapeutic dose in ARPE-19 cells was toxic in R28 cells, suggesting underdeveloped drug import/export systems or excess HMG-CoA reductase concentrations in R28 cells, but further studies are needed.
(3) Epicatechin is a powerful antioxidant flavanol found in many plant-based foods such as grapes, dark chocolate, and green tea. It is commonly used to reduce oxidative damage by inhibiting nicotinamide adenine dinucleotide phosphate (NADPH) oxidase and maintaining nitric oxide (NO) synthase, which prevents oxidized LDL endothelial dysfunction implicated in dementia. [53][54][55] It can also decrease levels ROS in the hippocampus of rat models with hypertension and Alzheimer's. [56] Other studies show that epigallocatechin gallate, a derivative with phenolic hydroxyl, injected into rat retinas attenuated sodium nitroprusside-induced oxidative stress in retinal ganglion cells. [57] Our results show that epicatechin significantly reduced caspase-3/7 activity by 83.6% in ARPE-19 cells, but only 25.7% in R28 cells (P = 0.05. Figures 2E-2F). Unlike the first two inhibitors, epicatechin's borderline p-value suggests it protects R28 cells by decreasing general oxidative burden. Thus, targeting upstream ROS production via NADPH oxidase may protect multiple retinal cell types more effectively than targeting specific downstream enzymes or receptors. These results are consistent with those obtained by IncuCyte® Live-Cell Imaging Analysis which captures caspase-3/7-mediated apoptosis in real-time images ( Figure 2G). The overlapped signal ( Figure 2G, fourth column) shows increased caspase activity with 7kCh ( Figure 2G, second row) that was reduced with epicatechin pretreatment ( Figure 2G, third row). Figure 2G shows the representative images of ARPE-19 cells treated with DMSO-control (first row), 7kCh (second row), and pretreated with epicatechin (third row). Similar image results were obtained with ARPE-19 cells pretreated with simvastatin or memantine.
(6) Z-ATAD-FMK is a caspase-12 inhibitor that disrupts the endoplasmic reticulum stress-induced caspase pathway. In ARPE-19 cells, it reduced caspase-12 activity by 47.7%. In contrast, there was no significant effect on R28 cells, with only 9.1% decrease in caspase activity.
Previous studies demonstrated that a component of cigarette smoke, benzo(e)pyrene [B(e)P], also induces caspase-3/7 activity in ARPE-19 cells and was inhibited by genistein, resveratrol, and memantine but not calpain, BTIC, simvastatin, or epicatechin. [60] When combined with the current study's results, memantine could reverse both B(e)P-and 7kCh-induced caspase-3/7 activities in ARPE-19 cells, but simvastatin and epicatechin only reversed activity induced by 7kCh not B(e)P. This indicates that the same cell type can have distinct pathways to activate and inhibit caspase-3/7, depending on the insult.
Likewise, the current study suggests that different cell types have distinct pathways to activate and inhibit apoptosis induced by 7kCh. None of the drugs significantly reduced caspase activity in R28 cells, which strongly implies that tissue characteristics or processes maintain apoptosis in the presence of inhibitors. As mentioned previously, R28 cells are retinal precursor cells from six-day old rats that express genes and proteins specific to their developmental age, which may not include receptors that respond to the drugs. [41] Furthermore, this heterogeneous population has cell types with distinct morphologies, biochemical characteristics, and neuronal-specific cell markers. A subset of R28 cells have receptors to neurotransmitters such as dopamine, serotonin, glycine, and acetylcholine; other cells have receptors that respond to NMDA and GABA agonists; still others show immunoreactivity to GluR1, GluR2, and GluR3. [24,41] In contrast, ARPE-19 cells are a homogeneous population of RPE cells that grow in a monolayer, express the same receptors and RPE-selective markers like CRALBP, and have the same cobblestone morphology. [61] Thus, receptoror enzyme-specific drugs such as the caspase inhibitors, memantine, or simvastatin will affect all ARPE-19 cells equally and the global response is additive (-47.7% to -85.3% change in caspase activity, all p-values < 0.05). In contrast, only a fraction of R28 cells that express the target enzyme or receptor will be affected, and the global response is represented only by the subset (+17.5% to -25.7% change in caspase activity, all p-values > 0.05).
In summary, the current study quantified caspase-3/7 activity as a marker of apoptosis between human ARPE-19 cells and rat R28 cells after treatment with 7kCh and various inhibitors. 7kCh significantly induced caspase-3/7 for both cultures, yet pretreatment with anti-apoptotic drugs (memantine, simvastatin, epicatechin, Z-VAD-FMK, etc.) consistently reduced caspase activities only in the ARPE-19 cells. This discrepancy in caspase deactivation between ARPE-19 and R28 cells may be related to the differences in the heterogeneity of the cell population or the age and species of the original tissue. Ultimately, the etiology of AMD is multifactorial and future treatments may utilize combination therapy to inhibit common toxins, like 7kCh and B(e)P, that affect multiple retinal cell targets.