4-Hydroxy-7-oxo-5-heptenoic acid (HOHA) lactone induces apoptosis in retinal pigment epithelial cells

Abstract

Retinal pigment epithelial (RPE) cell dysfunction and death play vital roles in age-related macular degeneration (AMD) pathogenesis. Previously we showed that oxidative cleavage of docosahexenoate (DHA) phospholipids generates an α, β-unsaturated aldehyde, 4-hydroxy-7-oxohept-4-enoic acid (HOHA) lactone, that forms ω-carboxyethylpyrrole (CEP) derivatives through adduction to proteins and ethanolamine phospholipids. CEP derivatives and autoantibodies accumulate in the retinas and blood plasma of individuals with AMD and are a biomarker of AMD. They promote the choroidal neovascularization of “wet AMD”. Immunization of mice with CEP-modified mouse serum albumin induces “dry AMD”-like lesions in their retinas as well as interferon-gamma and interleukin-17 production by CEP-specific T cells that promote inflammatory M1 polarization of macrophages. The present study confirms that oxidative stress or inflammatory stimulus produces CEP in both the primary human ARPE-19 cell line and hRPE cells. Exposure of these cells to HOHA lactone fosters production of reactive oxygen species. Thus, HOHA lactone participates in a vicious cycle, promoting intracellular oxidative stress leading to oxidative cleavage of DHA to produce more HOHA lactone. We now show that HOHA lactone is cytotoxic, inducing apoptotic cell death through activation of the intrinsic pathway. This suggests that therapeutic interventions targeting HOHA lactone-induced apoptosis may prevent the loss of RPE cells during the early phase of AMD. We also discovered that ARPE-19 cells are more susceptible than hRPE cells to HOHA lactone cytotoxicity. This is consistent with the view that, compared to normal RPE cells, ARPE-19 cells exhibit a diseased RPE phenotype that also includes elevated expression of the mesenchymal indicator vimentin, elevated integrin a5 promotor strength and deficient secretion of the anti-VEGF molecule pigment-epithelium-derived factor fostering weaker tight junctions.

Graphical Abstract

INTRODUCTION

RPE cell dysfunction plays a vital role in AMD.

Age-related macular degeneration (AMD), a chronic, degenerative disorder in the maculae of the retina, is the leading cause of irreversible blindness among the elderly.[1–4] Although the pathogenesis of AMD is complex and remains poorly understood, the retinal pigment epithelium (RPE) is considered to be a primary site of AMD pathology. [2, 5] The RPE, a monolayer of cells interposed between the photoreceptor cells and the Bruch’s membrane-choroid complex, is critical for the maintenance and survival of the photoreceptor cells. dysfunction and death of RPE cells play a vital role in the pathogenesis of AMD.[2, 5, 6]

Both inflammation and oxidative stress are believed to contribute to RPE dysfunction in AMD.[7, 8] Owing to its structural and functional features, the retina is an ideal environment for the generation of reactive oxygen species (ROS), e.g., H₂O₂, superoxide, or •OH. It experiences high levels of oxygen consumption and cumulative chronic exposure to sunlight and indoor UV/VIS light, and it contains an abundance of photosensitizers and undergoes active phagocytosis of photoreceptor outer segments by the RPE. In addition, the retina contains high levels of unsaturated lipids that are especially susceptible to free radical-induced oxidative damage.

Oxidative stress and lipid peroxidation contribute to RPE dysfunction.

Under physiological conditions, ROS are neutralized by enzymatic and non-enzymatic defense mechanisms. However, excess ROS production results in an imbalance between pro- and anti-oxidant processes, leading to oxidative stress.[3, 9] Under oxidative stress, the lipids in membranes, mainly phospholipids containing PUFAs, are susceptible to oxidative stress-mediated free radical-induced lipid peroxidation (LPO), which has been implicated in the pathogenesis of many degenerative ocular diseases including AMD.[1, 3, 10] LPO products are widely implicated to contribute to RPE dysfunction, leading to AMD. Docosahexaenoate (DHA), an omega-3 fatty acid, comprises 40%, 60% and 80% of the polyunsaturated fatty acids (PUFAs) in the brain, retina and photoreceptor disk membranes, respectively.[11, 12] In the retina it is mainly esterified into phospholipids in photoreceptor disk membranes.[1, 3, 7, 12] DHA is especially prone to non-enzymatic free radical-induced oxidation. Owing to the presence of six double bonds between carbon atoms and five doubly allylic methylene groups (C=C-CH₂-C=C) in its polyene chain, DHA is exquisitely susceptible to lipid peroxidation.[12]

Lipid peroxidation of PUFAs generates a variety of reactive electrophilic molecules. Free radical-induced oxidative cleavage of arachidonyl and linoleyl phospholipids generates the toxic 4-hydroxynonenal (4-HNE).[13, 14] Oxidative cleavage of phospholipids containing DHA generates a variety of reactive aldehydes including the γ-hydroxy-α, β-unsaturated aldehyde 4-hydroxyhexenal (4-HHE), a homologue of 4-HNE, and the 4-hydroxy-7-oxohept-5-enoic acid ester of 2-lyso-phosphatidylcholine (HOHA-PC) (Scheme 1).

Scheme 1.

Generation of CEP and HOHA-lactone from DHA-PC

Oxidative cleavage of DHA phospholipids produces bioactive CEP derivatives.

Previously we discovered that 4-HNE and 4-HHE form 2-pentyl- and 2-ethylpyrrole derivatives of proteins by covalent adduction to primary amino groups of protein lysyl residues (Scheme 1). By analogy, we predicted that HOHA-PC would react with primary amino groups of protein lysyl residues and ethanolamine phospholipids to generate 2-(ω-carboxyethyl)pyrrole (CEP) derivatives, presumably through enzymatic hydrolysis of intermediate CEP phospholipids (Scheme 1).[15–17] We then showed that CEP derivatives and autoantibodies accumulate in the retinas and blood plasma of individuals with AMD and are a biomarker of AMD.[15, 16, 18] Animal model studies demonstrated that CEP derivatives induce choroidal neovascularization and promote wound healing and tumor growth in a toll-like receptor 2-dependent manner.[16, 19, 20] Immunization of mice with CEP-modified mouse serum albumin induces AMD-like lesions in their retinas.[21] It causes interferon-gamma (IFN-γ) and interleukin-17 (IL-17) production by CEP-specific T cells that promote inflammatory M1 polarization of macrophages. Thus, T cells and M1 macrophages activated by oxidative damage to DHA apparently cooperate in AMD pathogenesis.[22] No similar biological activities have been observed for the analogous pentylpyrrole or ethylpyrrole derivatives of 4-HNE or 4-HHE, respectively (Scheme 1).

HOHA-PC produces CEP derivatives nonenzymatically through HOHA lactone.

We discovered that CEP phospholipids, generated through the direct reaction of HOHA-PC with primary amines, are not major intermediates in the production of non-esterified CEPs. Rather, under physiological conditions, HOHA-PC is uniquely unstable owing to the proximity of the hydroxyl group to the carboxylic ester. Intramolecular transesterification results in rapid non-enzymatic deacylation to generate 2-lyso-PC and the five-membered ring lactone of 4-hydroxy-7-oxohept-4-enoic acid (HOHA lactone, Scheme 1).[23] Furthermore, the release of HOHA lactone from HOHA-PC occurs more readily than the reaction of HOHA-PC with primary amines to produce CEP esters of 2-lyso-PC, and those esters do not undergo non-enzymatic hydrolysis to generate non-esterified CEP. Thus, non-esterified CEP derivatives of biomolecules are produced directly through the reaction of HOHA lactone with the primary amino groups.[24]

Oxidative stress contributes to RPE cell death.

Because damage to the RPE is an early and crucial event in the molecular pathways leading to AMD,[3] it is important to determine all of the contributions of HOHA lactone to the degeneration of the RPE. Besides fostering AMD progression by generating bioactive CEP derivatives, we considered the possibility that HOHA lactone contributes to the development of AMD through direct effects on RPE cells. We previously showed that exposure of RPE cells to HOHA lactone causes depletion of cellular GSH owing to Michael adduction of this α,β-unsaturated aldehyde with glutathione (GSH) catalyzed by glutathione S-transferase in ARPE-19 and primary human hRPE cells.[25] Depletion of GSH promotes oxidative stress owing to impaired neutralization of ROS.

In the present study, we first confirmed that the oxidative stress induced by treatment of a human RPE cell line (ARPE-19) or primary human RPE cells (hRPE) with H₂O₂ causes oxidative cleavage of DHA phospholipids and covalent adduction of the resulting HOHA lactone with primary amino groups to generate CEP derivatives. Inflammatory stimulation induced by exposure to lipopolysaccharide (LPS) also generates CEP derivatives in RPE cells. Previously, an α,β-unsaturated aldehyde product of lipid oxidation 4-HNE was shown to induce apoptosis in these RPE cells.[13] Because oxidant-treated RPE cells undergo apoptosis we tested and confirmed the hypothesis that HOHA lactone, the α,β-unsaturated aldehyde precursor of CEP, induces RPE cell death through apoptosis, a possible mechanism by which RPE cells are lost during the early phase of AMD [2]. Because ω−3 fatty acyl precursors of HOHA lactone are more abundant than ra-6 fatty acyl precursors of 4-HNE in the photoreceptor and RPE cells of the retina, their oxidative fragmentation is likely to be the primary source of toxic α,β-unsaturated aldehydes. We now find that the ra-3 fatty acyl-derived HOHA lactone and 4-HHE also exhibit greater pro-apoptotic activity toward RPE cells than the ω−6 fatty acyl-derived 4-HNE.

MATERIALS AND METHODS

Reagents.

Dulbecco’s modified Eagle’s medium (DMEM)/F12, phosphate buffered saline, fetal bovine serum (FBS) and 2’,7’-dichlorofluorescein diacetate (DCFH DA), Pierce BCA protein assay kit, Pierce 660 nm kit and Ionic Detergent Compatibility Reagent for Pierce 660nm Protein Assay Reagent were purchased from Fisher Scientific (Pittsburgh, PA). Mammalian cell lysis buffer was purchased from GoldBio (St. Louis, MO). . Glutathione reductase (250 units ml⁻¹), 5,5’-dithio-bis(2-nitrobenzoic acid) (DTNB), β-NADPH and all other chemicals and reagents were of analytical grade and were obtained from Sigma-Aldrich (St. Louis, MO). The lactone of 4-hydroxy-7-oxohept-5-enoate (HOHA lactone) was synthesized as described elsewhere.[23] 4-Hydroxy-2-nonenal and 4-hydroxy-2-hexenal and 4-oxo-2-nonenal were purchased from Cayman Chemical (Ann Arbor, MI). Retinal Pigment Epithelial Cell Basal Medium (RtEBM) supplemented with an optimized mixture of growth factors and supplements for primary hRPE cells (SingleQuots™ Kit) was obtained from Lonza (Allendale, NJ). Antibodies, monoclonal mouse anti-human caspase-3 (Biolegend #634101), mouse anti-human total p53 (Biolegend #645701), mouse anti-human Bax (Biolegend #633601) and JNK1 (Biolegend #633101) were purchased from Biolegend (San Diego, CA). Mouse anti-human phosphorylated p53 (Ser15) (CST #9286), mouse anti-human p21 (CST #2946), mouse anti-human p-JNK1/JNK2 (Tyr 183/Tyr185; CST #9255) and p53 duplex (SignalSilence p53 siRNA I, CST #6231), scrambled duplex siRNA (SignalSilence Control siRNA, unconjugated, #6568) were obtained from Cell Signaling Technology (Danvers, MA). Mouse monoclonal antibodies against human MDM2 (sc-965) and GAPDH (sc-20358) as well as horseradish peroxidase (HRP)-conjugated secondary antibodies (goat anti-mouse, sc-2005 and goat anti-rabbit, sc-2030) were procured from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal mouse anti-CEP antibodies were raised and authenticated in this laboratory.[15] Restore Plus Western blot stripping buffer as well as the enhanced chemiluminescence (ECL) Western blot detection system were from Pierce Biotechnology (Rockford, IL) via Fisher Scientific (Pittsburgh, PA). Reagents and pre-cast gels (4–20% gradient and 16%) for SDS-PAGE were purchased from Invitrogen Life Technologies (Carlsbad, CA) via Fisher Scientific (Pittsburgh, PA). In situ apoptosis detection kits, NucView 488 Caspase 3 assay (DEVD-NucView 488-substrate) for live cells and CF488A-Annexin V - propidium iodide apoptosis assay kits were purchased from Biotium (Hayward, CA).

Cell Culture.

The cell line ARPE-19 (ATCC; CRL-2302) derived from spontaneously arising retinal pigment epithelia of a healthy person, as described by Dunn et al.[26] was obtained from American Type Culture Collection (Manassas, VA). The stock cells were grown on 100-mm dishes in a humidified CO₂ incubator at 37°C and 5% CO₂ in Ham’s F12 medium and DMEM (50:50 ratio), containing L-glutamine and 10% heat-inactivated FBS. Cells were trypsinized and passaged every 2–3 days. Clonetics™ human primary retinal pigmented epithelial cells (hRPE) (passage 2) were obtained from Lonza (Allendale, NJ) and maintained and sub-cultured in RtEBM supplemented with an optimized mixture of growth factors and supplements (SingleQuots™ Kit) and 10% heat-inactivated FBS. hRPE cells were passaged every 4–5 days.

Cell Viability.

The sensitivity of hRPE and ARPE19 cells to HOHA lactone was evaluated using MTT and LDH assays as described by van Meerloo et al.[27] and in accordance with the Pierce LDH assay manual, respectively.

MTT Assay.

Briefly, either hRPE (45,000 cells/ per well) or ARPE19 cells (45,000 cells/ per well) were seeded in a 96-well flat bottom clear plate in 200 μl of the respective cell culture medium supplemented with 10% FBS and allowed to attach to culture plates in a humidified CO₂ incubator at 37 °C and 5% CO₂. The following day, cells were washed with the respective basal medium and starved in 180 μl of the respective basal medium for an additional 4–5 hours. To these cells (8 replicate wells were used for each concentration), 20 μl of stock solutions of HOHA lactone in the respective basal medium were added to create 0 to 100 μM final concentrations of HOHA lactone followed by 24-hour incubation in a humidified CO₂ incubator at 37 °C and 5% CO₂. At the end of the incubation, 20 μl of sterile 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT; 5 mg/ml in sterile PBS) solution was added to each well, and the plates were incubated for an additional 4 h in a CO₂ incubator at 37 °C and 5% CO₂. The plates were then centrifuged at 1,000 g for 5 min and the medium was aspirated from each well. Dimethylsulfoxide was added to each well (200 μl) and the water-insoluble intracellular formazan crystals were dissolved by carefully pipetting the content of the wells. The optical density (OD) of the formazan solutions was measured by using a plate reader (Model M2, Molecular Device) set at λ = 540 nm and λ = 620 nm. The concentration of HOHA lactone resulting in a 50% decrease in formazan formation was calculated as the IC₅₀ value of HOHA lactone.

LDH Assay.

Briefly, ARPE19 cells (45,000 cells/ per well) were plated on 96-well plates in 200 μl of DMEM/ F12 medium supplemented with 10% of inactivated FBS. The following day, the cells were starved in basal DMEM/ F12 medium for 4–5 hours and washed by DMEM/ F12 medium. Solutions of 0–100 μM HOHA lactone were prepared in basal DMEM/ F12 cell culture medium and added to the respective wells (8 replicate wells were used for each concentration) followed by 24 h incubation in a CO₂ incubator at 37 °C and 5% CO₂. ARPE19 cell viability was estimated by the LDH assay according to the manufacturer’s instructions. The optical density (OD) was measured by using a plate reader (Model M2, Molecular Device).

Estimation of Intracellular GSH and GSSG in hRPE and ARPE-19 Cells.

Aliquots (50 μl) of hRPE and ARPE cell lysates from time-course and dose-dependent studies were taken to estimate intracellular GSH and GSSG levels using a spectrofluoremetric microplate method described earlier.[28]

Post-translational Modification of hRPE and ARPE-19 Cells by HOHA Lactone.

Cells (30,000 cells/ per well) were plated on an 8-chamber well (Lab-Tek II Chamber Slide System, Nunc, Rochester, NY) in the respective complete medium (with 10% FBS). The following day, the cells were starved in the respective basal medium for an additional 4–5 h, washed three times with basal cell culture medium and the wells were aspirated. Solutions of HOHA lactone (0–30 μM) in the respective basal medium were added to the respective wells. The hRPE and ARPE19 cells were incubated with these cell culture media overnight at 37 °C and 5% CO₂. After incubation, the cells were aspirated, fixed and permeabilized with acetone (−20 °C). The slides were blocked with 1:100 diluted normal goat serum (Fisher Scientific, Pittsburgh, PA). The cells were probed with mouse anti-CEP monoclonal antibody (1:100) followed by incubation with goat anti mouse FITC antibody (1:200; Invitrogen, Carlsbad, CA). The slides were mounted in VectaShield containing DAPI mounting medium (Vector Laboratories, Burlingame, CA). Cell images were taken at 10x magnification using a Leica DMI 6000 B fluorescence inverted microscope.

Visualization of Intracellular Reactive Oxygen Species.

Intracellular generation of ROS in hRPE and ARPE19 cells as the result of HOHA lactone treatment was evaluated using DCFHDA according to the method of Wang and Joseph[29] with minor modifications. Briefly, either hRPE or ARPE19 cells (45,000 cells/ per well) were plated on 8-chamber well (Lab-Tek II Chamber Slide System, Nunc, Rochester, NY) in the respective complete medium (with 10% FBS). The following day, the cells were starved in a basal medium for 4–5 hours. Cells were pre-treated with DCFH-DA for 45 minutes with 13.3 μM DCFH-DA in the respective basal medium at 5% CO₂/95% air at 37 °C. The cells were washed, and then further treated with 0, 15, or 30 μM of HOHA lactone solution for 30 minutes at 5% CO₂/95% air at 37 °C. The images were taken at 10X magnification with a Leica DMI 6000 B fluorescence inverted microscope.

Detection of CEP in RPE Cells Incubated with H₂O₂ or LPS.

Either ARPE-19 cells or hRPE cells (2.5 × 10⁴ cells/ per well) were plated on an 8-chamber slide (Lab-Tek II Chamber Slide System, Nunc, Rochester, NY) in the corresponding complete medium (with 10% FBS) at 37 °C and 5% CO₂ overnight. After starving the cells in the corresponding basal medium overnight, the cells were washed three times with the corresponding basal cell culture medium. Then the cells were exposed to 0 or 25 μM H₂O₂ in PBS buffer for 2 h at 37 °C and 5% CO₂ The cells were aspirated and washed three times with the corresponding basal cell culture medium followed by overnight incubation at 37 °C and 5% CO₂. Alternatively, the cells were exposed to 0 or 12 μg/mL LPS in PBS buffer overnight at 37 °C and 5% CO₂. After incubation, the slides were centrifuged at 500g for 5 min and the medium was aspirated from each chamber. Cells were fixed with dry-ice cold acetone and immunostaining was performed as described above for HOHA-lactone treated RPE cells. Images were taken at 10X magnification with a Leica DMI 6000 B fluorescence inverted microscope. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA).

In Situ Detection of Apoptosis.

Activation of caspase 3/7 activity in ARPE19 cells after treatment with HOHA lactone was visualized using an in situ apoptosis detection kit, NucView 488 Caspase 3 assay (DEVD-NucView 488-substrate) for live cells. Briefly, hRPE or ARPE19 cells (20,000 cells/ per well) were plated on an 8-chamber slide (Lab-Tek II Chamber Slide System, Nunc, Rochester, NY) in the respective complete medium (with 10% FBS) and maintained in 5% CO₂/95% air at 37 °C. The following day, the cells were starved in the basal medium for additional 4–5 hours. The cells were then incubated overnight with 0 and 15 μM concentrations of HOHA lactone at 5% CO₂/95% air at 37 °C. Following incubation with HOHA lactone, the chamber slides with cells were centrifuged at room temperature at 200 g and stained with 5 μM DEVD-NucView 488-substrate following the manufacturer’s instructions. Images were taken at 10X magnification with a Leica DMI 6000 B fluorescence inverted microscope. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA).

Annexin V Binding.

Detection of enhanced annexin V binding to ARPE19 cells as the result of HOHA lactone treatment was achieved using the CF488A-annexin V - propidium iodide apoptosis assay kit (Biotium, Hayward, CA). ARPE 19 cells (20,000 cells/ per well) were plated in 8-chamber wells (Lab-Tek II Chamber Slide System, Nunc, Rochester, NY) in complete DMEM/ F12 medium (10%FBS) and allowed to attach to the slide’s surface overnight in 5% CO₂/95% air at 37 °C. The following day, the cells were starved in basal DMEM /F12 medium for 4–5 h. The cells were then incubated overnight (16 h) with 0 and 15 μM HOHA lactone in 5% CO₂/95% air at 37 °C. Following incubation with HOHA lactone, the cells were stained with a mixture of annexin V/propidium iodide following the protocol described in manufacturer’s manual. Images were taken at 10X magnification with a Leica DMI 6000 B fluorescence inverted microscope. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA).

Dose-dependent Exposure of ARPE19 Cells to HOHA Lactone.

Serum starved monolayers of hRPE or ARPE19 cells (80–90% confluence) on 60 mm plates in 5% CO₂/95% air at 37 °C were challenged for 2 h with 0–40 μM HOHA lactone solutions (2 ml) in the respective basal cell culture medium. Cells were then scraped with a rubber policeman, transferred to 5-ml conical tubes and centrifuged at 480 g at 4 °C for 10 min. The medium was carefully aspirated and the cells were washed three times with ice-cold PBS followed by centrifugation at 480 g at 4 °C for 10 min. Cells were disrupted upon incubation with an ice-cold mammalian cell lysis buffer (GoldBio, St. Louis, MO) supplemented with 1x HALT protease inhibitor cocktail (Fisher Scientific, Pittsburgh, PA) for 15 min at 4 °C. Cell lysate was further placed on ice and sonicated at 40% power (5 cycles of 5 secs on and 5 secs off). The cell lysate was spun at 14,000 g at 4 °C for 20 min and the supernatant was collected and snap-frozen in liquid nitrogen. For PAGE analysis the supernatants were thawed out on ice and protein concentrations were determined using the 660 nm Pierce protein quantitation assay supplemented with ionic detergent compatibility reagent according to the manufacturer’s manual.

Time-course Exposure of ARPE19 Cells to HOHA Lactone.

After 16 h of starvation in the respective basal cell culture medium, hRPE or ARPE19 cells (80–90% confluence) in 60-mm plates were treated with 20 μM HOHA lactone solution made with a basal medium for 0, 15, 30, 60, 90, or 120 min in 5% CO₂/95% air at 37 °C. Handling cells and preparation of the cell lysates was carried out in a manner similar to that described in the previous paragraph.

Western Blot Analysis.

ARPE19 cell extracts (25–30 μg protein/lane) were separated by SDS-PAGE (4–20% gels) and the proteins were electroblotted onto a nitrocellulose membrane (Bio-Rad, Hercules, CA). The blot was blocked with 5% bovine serum albumin (fatty acid-free BSA; Equitech-Bio, Kerrville, TX) in Tris-buffered saline containing 0.1% Tween-20 (TBST) for an hour at room temperature. The blots were subsequently probed overnight with the indicated antibody at 4 °C in a 5% BSA TBST buffer. After washing with TBST, the membrane was incubated with the appropriate secondary antibody-HRP conjugate at room temperature for 1 h. The membrane was washed with TBST and the immunoblot was developed with the SuperSignal West Pico Chemiluminescent Substrate (#34580) or SuperSignal West Femto Maximum Sensitivity Substrate (#34095) from Pierce Biotechnology (Rockford, IL) according to the manufacturer’s instructions. For the detection of caspase-3, either 4–20% or 16% gels were used throughout this project. Representative Western blots from three independent experiments which show very similar results are shown. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA).

Detection of Nuclear Accumulation of p53 in ARPE-19 cells challenged with HOHA lactone.

ARPE-19 cells (4 × 10⁶ cells in 100 mm petri dishes) were treated with HOHA lactone (20 μM) made with a basal medium for 0, 15, 30, 60, 90, or 120 min in 5% CO₂/95% air at 37 °C for two hours. Cells were washed with room temperature PBS, harvested from the petri dishes and their cytoplasmic and nuclear extracts were prepared by using the method of Baghirova et al. [30] Briefly, after centrifugation of harvested cells at 500 × g for 10min at 4° C the supernatant was discarded and ice cold lysis buffer A (containing 25 μg/ml digitonin, 1 M hexylene glycol and 1% protease inhibitor cocktail) to the cell pellet the cells, vigorously vortexed followed by an incubation for10 min at 4 °C. After centrifugation at 2000g for 10 min at 4 °C, the supernatant was collected (designated as “cytosolic protein fraction”). The pellet was further incubated in an ice cold lysis buffer B (50 mM Igepal, 1 M hexylene glycol and 1% protease inhibitor cocktail) for 30 min at 4 °C and centrifuged again at 7000xg for 10 min at 4 °C. The supernatant was collected (designated as “proteins from membrane-bound organelles”). The remaining pellet was treated with ice cold lysis buffer C (containing 0.5% sodium deoxycholate, 0.1% SDS, 1% protease inhibitor cocktail and benzonase), vigorously vortexed, and let to stand for 30 min at 4°C to allow complete solubilization of nuclei and digestion of genomic DNA. Centrifugation at 7800xg for 10 min at 4 °C and collection of the supernatant yielded a fraction designated as “nuclear proteins”. The final pellet that contained insoluble nuclear proteins was discarded. Both cytosolic and nuclear lysates were further placed on ice and sonicated at 40% power (5 cycles of 5 sec on and 5 sec off). Protein concentrations in both fractions were determined by the 660 nm Pierce protein quantitation assay supplemented with ionic detergent compatibility reagent according to the manufacturer’s manual. Cell extracts (50 μg protein/lane) were separated on 4–20% SDS-PAGE and immunoblotted using mouse anti-p53 and rabbit anti-MDM2 antibodies. GAPDH was used as the loading control and to demonstrate, by its absence, that the p53 in the extracted proteins is of nuclear and not cytosolic origin. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA).

Transfection of ARPE19 Cells with p53 siRNA.

ARPE-19 cells were seeded in six-well plates at a density of 0.5 × 10⁶ cells/well in antibiotic-free normal growth cell culture medium to achieve 70% to 80% confluence on day 2. For transfection of each well 12 μl of transfection reagent (Lipofectamine RNAiMax, Invitrogen) was diluted in 200 μl of in serum-free basal medium, vortexed and incubated for 45 min. Then of 12 μL siRNA duplex (from 10 μM stock either scrambled or p53) were added and incubation continued for another 30 to 45 min at room temperature. The cell monolayer was rinsed in serum free-basal cell culture medium, and the siRNA-transfection mixture was added dropwise onto the monolayer and incubated for 24 hours at 37 °C. Fresh complete cell culture medium containing 10% FBS heat-inactivated serum was added onto the monolayer without removing the transfection mixture followed by incubation for an additional 24 h. Total cell lysates for Western blot analysis were prepared using GoldBio Mammalian Lysis Buffer (St. Louis, MO) instead of RIPA lysis buffer. Protein concentrations were determined by the BCA protein quantitation assay.

RESULTS

Both inflammatory stimulus or oxidative stress generates CEP in RPE cells.

ARPE-19 cells, a cell line developed from the RPE of a 19 year-old adult male donor, retain many of the characteristics of RPE cells and have been widely used as an alternative to primary RPE cells to study the impact of oxidative stress due to its ready availability and feature stability even after prolonged cultivation.[26],[31] We previously established that HOHA lactone is an important precursor for the generation of CEP derivatives, and observed the development of CEP-positive immunostaining upon exposure of ARPE-19 cells to HOHA lactone.[24] However, ARPE-19 cells are transformed and therefore may respond differently to oxidative stress compared to primary RPE cells in vitro and in vivo. Therefore, it was desirable confirm the conclusions drawn with ARPE-19 cells by parallel experiments with primary RPE cells. Because both local inflammation and oxidative stress are associated with RPE damage in AMD, and because CEP is a biomarker of AMD,[2, 15] we examined the effect of an inflammatory stimulus and of oxidative stress on the generation of CEP derivatives in both ARPE-19 and hRPE cells. Inflammatory stimulation was accomplished by treatment with LPS and oxidative stress was induced by treatment with H₂O₂. As shown in Figure 1, oxidative stress induced in RPE cells by treatment with 25 μM H₂O₂ generated CEP derivatives. Treatment with LPS also caused CEP generation (Figure 2).

Figure 1.

Generation of CEP epitopes in RPE cells under oxidative stress. (A) Images of the generation of CEP in RPE cells. (B) Quantification of the generation of CEP in ARPE-19 cells. (C) Quantification of the generation of CEP in hRPE cells. RPE cells were treated with 25 βM H₂O₂ for 2 hours followed by overnight recovery in a basal cell culture medium and then fixed in dry-ice cold acetone and immunostained with mouse monoclonal anti-CEP antibody (1°)/goat anti-mouse Alexa Fluor 488 antibodies (2°) and DAPI. Images were taken at 10X magnification with a Leica DMI 6000 B inverted fluorescence microscope. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA).The figure is representative of three independent experiments that showed very similar results. ***P<0.0001

Figure 2.

Generation of CEP adducts in RPE cells under inflammatory stimulus. (A) Images of the generation of CEP in RPE cells. (B) Quantification of the generation of CEP in ARPE-19 cells. (C) Quantification of the generation of CEP in hRPE cells. RPE cells were treated with 12 μg/ml LPS for overnight and then immunostained with mouse monoclonal anti-CEP antibody/goat anti-mouse Alexa Fluor 488 antibodies and DAPI. Images were taken at 10X magnification with a Leica DMI 6000 B inverted fluorescence microscope. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA). The figure shows a representative of three independent experiments, which showed very similar results. ***P<0.0001

HOHA lactone fosters the accumulation of ROS in RPE cells.

The accumulation of CEP in RPE cells exposed to an inflammatory stimulus or oxidative stress results primarily from the production of HOHA lactone and its reaction with the primary amino groups of protein lysyl residues and ethanolamine phospholipids.[15, 16, 24] We previously showed that ARPE-19 and hRPE cells metabolize HOHA lactone by forming Michael adducts with GSH resulting in depletion of intracellular GSH.[25] Depletion of GSH is expected to result in oxidative stress owing to impaired neutralization of ROS. Oxidative stress engendered by the release of reactive oxygen species ROS is a crucial trigger for AMD pathogenesis.[2] Therefore, we investigated the induction of ROS in ARPE-19 and hRPE cells upon exposure to HOHA lactone using DCFH-DA to detect oxidative stress. As shown in Figure 3, fluorescence levels increased in both ARPE-19 and hRPE cells exposed to HOHA lactone compared to control RPE cells, indicating that HOHA lactone treatment promoted intracellular ROS accumulation in a dose-dependent manner. The cell morphology of ARPE-19 cells changed significantly as the concentration of HOHA lactone increased, and ARPE-19 cells treated with 30 μM HOHA lactone exhibited characteristics typical of apoptosis including cell shrinkage, nuclear chromatin condensation, and segmentation of the nucleus (see phase-contrast panel corresponding to 30 μM HOHA lactone in Figure 3). However, 30 μM of HOHA lactone was less detrimental to primary hRPE cells. This is consistent with the results of the cell viability MTT assays, and suggests that hRPE cells are more effective at detoxifying HOHA lactone, or protecting against HOHA lactone induced oxidative injury, or both.

Figure 3.

Detection of ROS in RPE cells challenged with various concentrations of HOHA lactone. (A) Images of the generation of ROS in RPE cells. (B) Quantification of the generation of ROS in ARPE-19 cells. (C) Quantification of the generation of ROS in hRPE cells. Cells were treated with DCFHDA (15 μg/ml) in the basal medium under 5% CO₂/95% air at 37 °C for 30 min, washed with a basal cell culture medium and then were exposed to HOHA lactone for two hours. Images were taken at 10X magnification with a Leica DMI 6000 B inverted fluorescence microscope. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA). Data shown are representative of three independent experiments that showed very similar results. *P<0.05,***P<0.0001

ARPE-19 cells are more sensitive to HOHA lactone cytotoxicity than hRPE cells.

To assess the cytotoxicity of HOHA lactone to RPE cells, we determined the cell viability after overnight exposure to various concentrations of HOHA lactone. Results of an MTT viability assay showed a progressive decrease in cell viability of ARPE-19 cells with increasing concentrations of HOHA lactone (Figure 4A), and the LC₅₀ ≈20 μM is lower than for hRPE cells. Furthermore, compared to ARPE-19 cells, HOHA lactone was even less cytotoxic to primary hRPE cells (Figure 4B) and showed little or no cytotoxic effect (possibly even weak growth stimulation) within the range of 0–50 μM. Cell viability only declined precipitously at concentrations in excess of 50 μM for primary hRPE cells versus ≈10 μM for ARPE-19 cells. This contrasts with the similar cytotoxicity of HNE toward transformed hRPE versus ARPE-19 cells, 52 versus 46 μM reported previously.[13] Our study demonstrates that results obtained with transformed cells may not be reliable indicators of the response of primary cells toward lipid oxidation derived α,β-unsaturated aldehyde electrophiles.

Figure 4.

Cytotoxicity of HOHA lactone to (A) ARPE-19 cells and (B) primary hRPE cells (both cell types were seeded at 4.5×10⁴ cells/well density) as measured by MTT assay. Eight replicate wells were used for each concentration of HOHA lactone in these studies. Results presented are percent cell survival in HOHA lactone treated groups with respect to vehicle (PBS)-treated cells (mean ± SD; n = 8). Data are representative of three independent experiments that show very similar results. *P<0.05, P<0.008, ***P<0.00001

HOHA lactone induces RPE apoptosis.

Because HOHA lactone can induce ROS accumulation, and given that one potential important consequence of ROS signaling is the induction of apoptosis, a key mechanism of RPE loss during AMD, we next examined whether HOHA lactone could induce RPE apoptosis. We used fluorescence microscopy to characterize HOHA lactone-induced RPE cell death by staining with CF488A-annexin V and propidium iodide (PI). Most of the ARPE-19 cells negative to both annexin V and PI after treatment with vehicle (PBS, upper panels of Figure 5A), indicating that they are viable with only a few early apoptotic cells. In contrast, most of the ARPE-19 cells were annexin V positive and PI negative after 16 h treatment with 15 μM HOHA lactone, indicating that most of these cells were in the early stages of apoptosis with few late apoptotic or necrotic cells.

Figure 5.

HOHA lactone induces apoptosis in ARPE-19 cells. The fluorescence images of ARPE-19 cells stained with CF488A-annexin V and propidium iodide (PI). Panel A shows ARPE-19 cells stained with annexin V to detect the presence of phosphatidylserine on the outer leaflet of the plasma membrane and PI was used to establish the integrity of the plasma membrane. Panel B Densitometric analysis of the results shown in panel A. Images were taken at 10X magnification with a Leica DMI 6000 B inverted fluorescence microscope. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA). Data are representative of three independent experiments that show very similar results. *−p <0.04, **−p<0.0025, ***−p<0.001 significantly different compared with control cells (PBS-treated cells).

HOHA lactone promotes ARPE-19 cell apoptosis through the intrinsic pathway.

In a damaged cell, posttranslational modifications of p53 block the sequestration of p53 by its negative regulator MDM2, thus resulting in an increase in free p53 and of p53 stimulation of apoptosis through increased expression of PUMA (p53 upregulated modulator of apoptosis) and activation of caspase-3.[32] Dissociation of p53 protein from MDM2 results in degradation of MDM2, phosphorylation and translocation of p53 to the nucleus where it promotes apoptosis. We found that treatment of ARPE-19 cells with HOHA lactone causes translocation of p53 to the nucleus (Figure 6A) and the disappearance of MDM2 (Figure 6B). In contrast, silencing p53 protected the cells against HOHA lactone-induced death (Figure 7).

Figure 6.

HOHA lactone induces nuclear accumulation of p53 and degradation of MDM2 in ARPE19 cells. Western blot analyses of HOHA lactone treated ARPE-19 cells showing time-dependent nuclear accumulation of p53 (Panel A) and the time-dependent degradation of MDM2 (Panel B). Cells (4 × 10⁶) were treated with HOHA lactone (20 μM) for one hour. Cell extracts (50 μg protein/lane) were separated on 4–20% SDS-PAGE and immunoblotted using mouse anti-p53 (A) and rabbit anti-MDM2 antibodies. GAPDH was used as the loading control (Panel A) and the nuclear fraction purity control (Panel B). The optical density of the protein bands on the X-ray films after their exposure to the respective Western blots were measured using MetaMorph software.

Figure 7.

Silencing p53 increases cell survival of ARPE-19 cells treated with HOHA lactone. (A) ARPE-19 cells (5×10⁵ cells/well) were transfected with scrambled (scr) or p53 siRNA (150 pM, respectively 2 times for 24 hours and the knockdown of p53 was confirmed by Western blot analysis. (B) Both p53-silenced and p53 expressing ARPE-19 cells were treated with 0–30 μM HOHA lactone for two hours followed by a MTT survivability assay. The optical density of the protein bands on the X-ray films after their exposure to the respective Western blots were measured using MetaMorph software. The graph shows cell survival (mean ± SD; n = 4) in the presence of different concentrations of HOHA lactone. *−p <0.03, **−p<0.002 significantly different compared with ARPE-19 cells treated with scrambled siRNA and p53 siRNA in the presence of the same HOHA-lactone concentration.

HOHA lactone induces expression and phosphorylation of apoptotic proteins.

The levels of several apoptosis-related proteins were monitored after exposure of ARPE-19 cells to 15 μM HOHA lactone (Figure 8). While the level of p53 remained fairly constant, that of phosphorylated p53 rose dramatically after 60 min and continued to rise at 90 and 120 min. Small increases in p21 and JNK occurred gradually, while the level of phosphorylated JNK jumped dramatically within 15 min.

Figure 8.

Effect of HOHA lactone on upregulation of pro-apoptotic proteins in ARPE19 cells. Western blot analyses of HOHA-lactone-induced dose- (panel A) and time-dependent (panel B) phosphorylation of Ser15-p53 and activation of JNK, and p-JNK, Bax, p21 in ARPE-19 cells: 1 × 106 cell were treated with 0–40 μM HOHA-lactone for two hours in panel A or with 15 μM HOHA-lactone and incubated for 0, 15, 30, 60, 90 and 120 min in panel B and cell extracts (15 mg protein/lane) were separated on 4–20% gradient SDS-PAGE gels and electroblotted on nitrocellulose. The blots were probed with anti-Ser 15, anti-Bax, anti-p21, anti-JNK, and anti-p-JNK antibodies, respectively. GAPDH was used as the loading control. Blots were developed with West Pico-chemiluminescence reagent (Pierce). The optical density of the protein bands on the X-ray films after their exposure to the respective Western blots were measured using MetaMorph software. The blots shown (from one set) are representative of 4 independent experiments that show very similar results.

HOHA lactone-induced expression of Bax leads to upregulation of caspase 3/7.

Upregulation of Bax results in mitochondrial damage and release of cytochrome c, assembly of the apoptosome and, ultimately, the production of active caspases 3/7. These caspases are central to the execution of programmed cell death and their activation constitutes the biochemical hallmark of apoptosis.[33] We assessed caspase in situ specific activity in HOHA lactone-treated versus untreated ARPE-19 cells using a NucView 488 caspase-3 assay kit. The top two panels of Figure 9 show fluorescence images of ARPE-19 cells challenged overnight (16 h) with or without 15 μM HOHA lactone followed by staining with the green fluorogenic DEVDNucView488™ caspase-3 substrate. Exposure of ARPE-19 cells to 15 μM HOHA lactone in PBS induced a marked increase in fluorescence compared to control cells treated with PBS alone. In addition, the specificity of the cytosolic enzyme capable of proteolysis of DEVDNucView-488 in ARPE-19 cells was verified by pre-incubating the cells with the plasma membrane permeable caspase 3/7 inhibitor Ac-DEVD-CHO before incubation with DEVDNucView 488 substrate. As shown in the bottom two panels of Figure 9, the NucView488™ fluorescence signal generated in the presence of the inhibitor was suppressed and resembled that found for cells treated with PBS alone. Thus, pre-treatment with the inhibitor essentially abolished the proteolysis of the caspase-3/7 specific substrate in the cells challenged with HOHA lactone.

Figure 9.

HOHA lactone upregulates caspase 3/7 in ARPE-19 cells. The microscopy imaging of caspase activation in ARPE-19 cells using the NucView 488 caspase-3 assay kit. ARPE-19 cells were challenged with 15 μM HOHA lactone in DMEM/F12 medium for 18 hours. Cells were then stained using 5 μM DEVD-NucView 488-substrate to detect the activation of caspase 3/7. The inhibitor DEVD-CHO was used to confirm that the fluorescence generated in the HOHA lactone-treated cells is dependent upon active caspase 3/7. Images were taken at 10X magnification with a Leica DMI 6000 B inverted fluorescence microscope.

A previous study showed that 4-HNE causes a dose-dependent cleavage of a 17-kDa fragment from procaspase-3.[13] To confirm the increased truncated caspase 3 up-regulation indicated by DEVD-NucView 488 staining, we evaluated intracellular levels of pro-caspase and activated caspase-3 in HOHA lactone-treated ARPE-19 cells. Activation of caspase 3 was detected by the appearance of a band at 17 kDa. Immunoblot analyses presented in Figure 10 show that HOHA lactone caused a dose-dependent cleavage of the caspase-3 zymogen to a 17-kDa fragment with the highest level of the active enzyme being generated upon treatment with 10–20 μM HOHA lactone, and early indication of apoptosis is detectable even at 5 μM HOHA lactone. These immunoblotting results correlate well with the data obtained from in situ immunofluorescence studies described above (Figure 5) which also show enhanced activity of caspase-3 in HOHA lactone-treated ARPE-19 cells compared to the untreated cells. Taken together, the present observations clearly demonstrate that HOHA lactone can induce cell death that coincides with activation of caspase-3 in ARPE-19 cells.

Figure 10.

The effect of HOHA lactone on caspase-3 activation in ARPE-19 cells. (A)Western blot analysis of caspase-3 in ARPE-19 cells incubated overnight with various concentrations of HOHA lactone (B) Densitometric analysis of the Western blot shown in (A). Immunoblotting was conducted using a rabbit polyclonal anti-caspase-3 antibody. GAPDH was used as a loading control. The optical density of the protein bands on the X-ray films after their exposure to the respective Western blots were measured using MetaMorph software. Results are representative of three independent experiments that show similar results. *P<0.05, **P<0.02.

HOHA lactone and HHE are more potent activators of caspase-3 than their longer chain homologues HOOA lactone and HNE.

Docosahexenoic acid, an ω−3 fatty acid, is one of the most abundant PUFA esterified in membrane phospholipids of photoreceptor and RPE cells, produces HOHA lactone and 4-HHE through oxidative fragmentation (Scheme 1). The homologous 4-hydroxy-8-oxohept-5-enoic acid (HOOA) lactone and 4-HNE are derived from ω−6 PUFAs that are less abundant. We now find that DHA-derived HOHA lactone, HOOA lactone and 4-HHE are potent inducers of caspase-3 upregulation and apoptosis in RPE cells while 4-HNE evokes a relatively minor response (Figure 11).

Figure 11.

Change of active caspase-3 levels in ARPE-19 cells (cleaved17 kDa fragment) induced 2 h after treatment with 15 or 30 μM of different electrophiles, expressed as mean ± SD. The optical density of the protein bands on the X-ray films after their exposure to the respective Western blots were measured using MetaMorph software. Results are representative of three independent experiments that show similar results (n = 3). *P<0.05, **P<0.002, ***P<0.0001,

DISCUSSION

Oxidative stress or inflammatory stimulus cause the accumulation of CEP in both ARPE-19 and primary human RPE cells.

Chronic low levels of oxidative stress are thought to play an important role in AMD pathogenesis. HOHA lactone, generated through oxidative fragmentation of DHA phospholipids, produces CEP derivatives of primary amino groups of protein lysyl residues and ethanolamine phospholipids.[15, 16, 24] Production of Michael adducts of GSH with HOHA lactone competes with CEP generation in ARPE-19 and hRPE cells.[25] We previously showed that irradiation of bovine retina extracts induces an exponential irradiation time-dependent increase in the formation of HOHA lactone-GSH derivatives and decrease in GSH levels.[34] The resulting depletion of cellular GSH fosters more oxidative stress owing to impaired neutralization of ROS. In other words, HOHA lactone participates in a vicious cycle by promoting intracellular oxidative stress leading to further oxidative cleavage of DHA to produce more HOHA lactone.

The present study confirmed that the oxidative stress or inflammatory stimulus promotes oxidative fragmentation of DHA phospholipids in ARPE-19 cells that produces CEP derivatives. Primary human RPE cells responded similarly. The present study confirms that exposure to HOHA lactone induces the production of ROS in ARPE-19 and hRPE cells, concomitant with the induction of apoptosis. ARPE-19 cells displayed a higher level of apoptotic changes than primary hRPE cells. This proclivity toward oxidative injury-induced pathology suggests that ARPE-19 cells are an in vitro model for diseased RPE.

ARPE-19 cells exhibit pathological hypersensitivity toward HOHA lactone toxicity.

In vitro models of the retinal pigment epithelium (RPE) include primary human hRPE, fetal human fhRPE and ARPE-19 cells as well as immortalized hRPE cell lines. ARPE-19 cells are a nontransformed human RPE cell line which arose spontaneously from a primary culture of RPE cells.[26] While hRPE and fhRPE cells recapitulate the phenotype of normal RPE in vivo, ARPE-cells can adopt two different phenotypes. They can exhibit epithelial cell morphology and express several genes specific for the RPE and perform many of the known functions of the human RPE. However, ARPE-19 cells can lose these epithelial characteristics and exhibit migratory mesenchymal cell-like properties.[35] When cultured for only a few days, ARPE-19 cells are distinguished from other primary RPE cultures by a rapid rate of proliferation characteristic of a mesenchymal phenotype.[26] Mesenchymal cells can transition into epithelial cells via mesenchymal to epithelial transition (MET), and revert to mesenchymal cells via epithelial to mesenchymal transition (EMT). Compared to hRPE or fhRPE cells, ARPE-19 cells when cultured for only a few days have a relatively low capacity to express critical proteins that regulate and maintain a barrier with strong tight junctions. The differences (occasionally >1000 fold) in gene expression and phenotype between these ARPE-19 cells cultured for only a few days, which are referred to as undifferentiated, and hRPE cells or RPE in vivo can be largely eliminated by prolonged incubation (4 months) during which they undergo MET-like differentiation acquiring an epithelial phenotype closely resembling that of hRPE cells and the RPE in vivo.[36] As expected for MET-mediated epithelial differentiation, expression of the canonical mesenchymal indicator vimentin is markedly decreased in differentiated compared to undifferentiated ARPE-19 cells.[36]

Our cell viability MTT assays using undifferentiated ARPE-19 cells revealed that the LC₅₀ of HOHA lactone toward ARPE-19 cells is at least 5-fold lower than toward hRPE cells. The proclivity of undifferentiated ARPE-19 cells toward a pathological response to HOHA lactone is consistent with the hypothesis that they are a model for diseased RPE. The hypersensitivity of undifferentiated ARPE-19 cells to VEGF action and their weaker tight junctions is reminiscent of the RPE in aged eye or in retinal pathology. Strong tight junction complexes between RPE cells are essential for the outer blood-retina barrier function of the RPE. Disrupting this barrier alters the metabolic circuits and function of the RPE. Deficient secretion of the anti-VEGF molecule pigment-epithelium-derived factor (PEDF) by undifferentiated ARPE-19 cells presumably contributes to their inferior barrier function because PEDF is an important antagonist that limits the mitogenic activity of VEGF. PEDF is secreted in large quantities by the RPE in the normal eye, but undifferentiated ARPE-19 cells exhibit secretion rates a thousand-fold lower than in fhRPE cells.[37] The angiogenic activity and elevated expression of VEGF by RPE is an important disease mechanism in advanced wet AMD. While hRPE cells serve as a model to study disruption of normal RPE function, undifferentiated ARPE-19 cells have been used to study how a pathologic phenotype can be ameliorated, e.g., by administering exogenous agents such as PEDF.[37]

Complement activation and cellular senescence in AMD pathology.

Previously, we discovered that HOHA lactone is a potent inducer of the alternative complement pathway in ARPE-19 cells resulting in sublethal assembly of membrane attack complexes.[38] This is especially problematic for AMD pathogenesis because immunoreactivity for complement factor H (CFH), an inhibitor of complement activation, is reduced in human AMD donor eyes compared to age-matched controls. We also discovered that HOHA lactone induces senescence in undifferentiated ARPE-19 cells. Incubation of ARPE-19 cells with 5 μM HOHA lactone for 12 h, induces a 3.5-fold increase in senescence-associated β-galactosidase staining. This indicates that a sub-lethal dose of HOHA lactone produces a prematurely senescent phenotype in ARPE-19 cells which results in a permanent loss of the ability to proliferate and a diminished resistance to apoptosis [39]. HOHA lactone presumably contributes to oxidative stress-induced senescence caused by exposure of undifferentiated ARPE-19 cells to hydrogen peroxide (H₂O₂), an endogenous stress source.[40] Senescent cells upregulate the proinflammatory cytokines IL-6 and IL-8, the main markers of the senescence-associated secretory phenotype. Senescent ARPE-19 cells also upregulate VEGF and simultaneously downregulate CFH expression. The finding of reduced CFH expression in senescent ARPE-19 cells implies that oxidative stress-induced premature senescence may foster CFH deficiency, enhancing complement pathway activation, and contributing to retinal damage involved in the induction and progression of AMD.[41]

ARPE-19 cells as a model for diseased RPE.

Other properties of undifferentiated ARPE-19 cells support the view that they are a model for diseased RPE. They show a 2 to 10 fold higher integrin a5 promoter strength compared to primary hRPE cells.[31] Signals transduced by integrins can influence cell adhesion, migration, differentiation and proliferation. Elevated transcriptional activity driven by the a5 promoter could contribute to a breakdown in adhesion between the photoreceptor cell layer of the neural retina and the RPE eventually leading to photoreceptor degeneration and blindness. A lack of adhesion between the RPE and the choroid results in proliferative vitreoretinopathy involving detachment from the underlying Bruch’s membrane, migration, proliferation, and secretion of extracellular matrix molecules and formation of an epiretinal membrane leading to retinal detachment.

HOHA lactone activates the intrinsic apoptotic pathway in RPE cells.

The basal rate of p53-dependent apoptosis increases in an age-dependent manner in human RPE owing to alterations in several aspects of the p53 pathway and activation of caspase-3.[32] We found that HOHA lactone induces a rapid increase in the expression and phosphorylation of JNK followed by initiation of the intrinsic apoptotic pathway through phosphorylation of p53[42–45], nuclear translocation of phosphorylated p53, expression of Bax and ultimately caspase-3. Phosphorylation that activates JNK is the initial pro-apoptotic event that we detected upon exposure of RPE cells to HOHA lactone. Prolonged JNK activation can promote apoptosis by inactivating suppressors of the mitochondrial-dependent death pathway.[46] This pathway can be initiated by phosphorylation of p53 a transcription factor that regulates cellular processes including apoptosis. Dissociation of p53 from its negative regulator MDM2 results in degradation of MDM2, phosphorylation and translocation of p53 to the nucleus where it can promote apoptosis. We found that HOHA lactone caused the translocation of p53 to the nucleus of ARPE-19 cells and the disappearance of MDM2. We confirmed the role of p53 in the induction of apoptosis by demonstrating that silencing p53 protected the cells against HOHA lactone-induced death.

Phosphorylated p53 accumulates in the nucleus where it triggers the intrinsic apoptotic pathways. It promotes expression of Bax that moves to the mitochondrial membrane, after a conformational change, and causes release of mitochondrial cytochrome C into the cytoso1,[47, 48] assembly of the apoptosome and, ultimately, the production of active caspases 3/7 that are central to the execution of programmed cell death.[33] We showed that exposure of ARPE-19 cells to 15 μM HOHA lactone markedly upregulated caspase-3 activity and intracellular levels of pro-caspase and activated caspase-3 protein.

CONCLUSIONS AND FUTURE PROSPECTS

Our previous studies of the consequences of DHA oxidation for the pathogenesis of AMD revealed that oxidative cleavage DHA-PC generates the HOHA-PC that spontaneously undergoes nonenzymatic deacylation generating HOHA lactone. Adduction of this α,β-unsaturated aldehyde to primary amines produces CEP derivatives that promote the geographic retinal atrophy of “dry AMD” and the choroidal neovascularization of “wet AMD”. The present study confirmed that oxidative stress or inflammatory stimulus produces CEP in RPE cells, and showed that HOHA lactone also participates in a vicious cycle, promoting intracellular oxidative stress leading to oxidative cleavage of DHA to produce more HOHA lactone.

The present study revealed another consequence of DHA oxidation for the pathogenesis of AMD. HOHA lactone and 4-HHEare cytotoxic, inducing apoptotic cell death through activation of the intrinsic pathway. Previously, the α,β-unsaturated aldehyde 4-HNE, a product of arachidonate oxidative fragmentation, was shown to induce nuclear accumulation of p53 and degradation of MDM2 in RPE cells.[13] Now we find that 4-HHE and HOHA lactone, products of docosahexaenoate oxidative fragmentation, are significantly more pro-apoptotic. Furthermore, the pro-apoptotic activity of HOHA lactone shows that a γ-hydroxyl group is not essential for toxicity of α,β-unsaturated aldehyde electrophiles toward RPE cells. These observations are a starting point for understanding the mechanistic basis of lipid oxidation-induced apoptotic RPE cell loss during the early phase of AMD and for developing therapeutic counter measures.

Induction of apoptosis versus enzymatic detoxification of α,β-unsaturated aldehydes.

The pro-apoptotic toxicity of 4-HNE is eliminated in RPE cells that overexpress glutathione S-transfrase (GST).[13] This suggests that the pro-apoptotic activity of α,β-unsaturated aldehydes can be neutralized by glutathionylation. Cell mediated conjugation of 4-HNE and HOHA lactone with GSH followed by reduction of the aldehyde and excretion of the reduced GSH adducts and reduction of 4-HNE-GSH adducts have been demonstrated.[25, 49, 50] It is tempting to speculate that the lower toxicity of 4-HNE toward RPE cells results from more efficient detoxification of HNE versus 4-HHE or HOHA lactone. More efficient GST-catalyzed detoxification of 4-HNE compared to HOHA lactone and 4-HHE might be a consequence a higher rate of diffusion of 4-HNE versus 4-HHE or HOHA lactone into the cell through the lipophilic cell membrane bilayer that is required for glutathionylation. The lipophilicity of 4-HNE (ClogP = 1.47) is higher than that of HOHA lactone (ClogP = −1.02) and 4-HHE (ClogP = −0.12).

The present study shows that one mechanism for the induction of apoptosis by HOHA lactone involves dissociation of the negative regulator MDM2 protein from the p53 protein, perhaps owing to covalent modification of p53 by the α,β-unsaturated aldehyde. The γ-hydroxyl present in HHE and HNE is not present in HOHA lactone, and therefore must not be required for induction of apoptosis by α,β-unsaturated aldehydes. Subsequent phosphorylation of p53 and translocation to the nucleus induces expression of caspase-3 and consequent apoptosis (Figure 12). However, this is not necessarily the only mechanism by which lipid oxidation-derived α,β-unsaturated aldehydes induce apoptosis in RPE cells. The present study does not rule out additional mechanisms. Diffusion into the cell would not be required if apoptosis were triggered through extracellular binding, e.g., to a putative receptor such as GPR109A.[51] Michael addition of the α,β-unsaturated aldehyde electrophiles HOHA lactone, 4-HHE or 4-HNE, e.g., to a histidyl residue of a receptor, might also trigger apoptosis (Figure 12). Reversible Michael addition of to a histidyl residue of the C3 convertase was suggested previously for complement activation by HOHA lactone.[38]

Figure 12.

Competition between pro-apoptotic activities of α,β-unsaturated aldehyde electrophiles and their detoxification through conjugation with GSH or carcinine.

Protection of RPE cells by chemical detoxification of α,β-unsaturated electrophiles.

The inferior resistance of ARPE-19 compared with hRPE cells to HOHA lactone-induced death may be a consequence of less effective detoxification of HOHA lactone, e.g., through conjugation with GSH, less effective inhibition of HOHA lactone-induced production of ROS, or both. Because they are less resilient than hRPE cells, undifferentiated ARPE-19 cells may better reveal pathways and mechanisms through which inflammation and oxidative injury can compromise RPE viability.

An alternative to cell-mediated detoxification of α,β-unsaturated aldehydes is scavenging with a therapeutic nucleophile. Drugs that intercept α,β-unsaturated aldehyde products of free radical-induced lipid oxidation might have prophylactic utility for preventing RPE apoptosis. Oral administration of the imidazole-containing peptide carcinine (β-alanyl-histamine) protects mouse retina against light-induced photoreceptor degeneration.[52] This protection may be the result of adduction involving Michael addition of the imidazole residue and Schiff base formation with the β-alanyl primary amino group (Figure 12).

Mitochondrial permeability transition (PT) initiated apoptosis.

Despite their close structural analogy, 4-HHE and 4-HNE can exhibit strikingly different biological activities. Our observation that 4-HHE is a potent inducer of RPE cell apoptosis, while 4-HNE is not, suggests an alternative mechanism for the induction of apoptosis by 4-HHE and HOHA lactone. 4-HHE enhances calcium-mediated induction of the mitochondrial permeability transition (PT) at femtomolar concentrations while 4-HNE is inactive at concentrations <1 μM.[53] Mitochondrial PT can initiate apoptosis.[54] This suggests that induction of the mitochondrial PT is a potential additional mechanism for the pro-apoptotic activity of HOHA lactone and 4-HHE. We recently reported that treatment with 5 to 20 μM HOHA lactone for 24 h induced a dose-dependent mitochondrial depolarization [34]. This demonstrated severe damage of mitochondria by HOHA lactone that caused an irreversible loss of mitochondrial membrane potential characteristic of late apoptotic cells [34].

Highlights.

• HOHA lactone and 4-HHE are pro-apoptotic toward retinal pigmented epithelial cells

• HOHA lactone causes apoptosis through the p53-MDM2 intrinsic pathway

• Primary hRPE cells are less susceptible to HOHA lactone cytotoxicity than ARPE-19 cells

• ARPE-19 cells are more sensitive than primary hRPE cells to HOHA lactone cytotoxicity

ABBREVIATIONS

4-HHE

4-hydroxyhex-2-enal

4-HNE

4-hydroxynon-2-enal

AMD

age-related macular degeneration

BSA

bovine serum albumin

CEP

2-(ω-carboxyethyl)pyrrole

DCFH DA

2’,7’-dichlorofluorescein diacetate

DHA

docosahexaenoic acid

DHA-PC

1-palmityl-2-docosahexaenoyl-sn-glycero-3-phosphocholine

DMEM

Dulbecco’s modified Eagle’s medium

DTNB

5,5’-dithio-bis(2-nitrobenzoic acid)

FBS

fetal bovine serum

GSH

reduced glutathione

GST

glutathione S-transferase

HOHA

4-hydroxy-7-oxo-hept-5-eonic acid

HOHA-PC

1-palmityl-2-(4-hydroxy-7-oxo-5-heptenoyl)-sn-glycero-3-phosphatidylcholine

HOOA

4-hydroxy-8-oxohept-5-enoic acid

HRP

horseradish peroxidase

LPO

lipid peroxidation

LPS

lipopolysaccharide

MTT

3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide

OD

optical density

PBS

phosphate buffered saline

PMSF

phenylmethanesulfonyl fluoride

PT

permeability transition

PUFA

polyunsaturated fatty acid

ROS

reactive oxygen species

RPE

retinal pigmented endothelium

RtEBM

Retinal Pigment Epithelial Cell Basal Medium

TBST

Tris buffered saline containing 0.1% Tween