4-Hydroxy-7-oxo-5-heptenoic acid lactone can induce mitochondrial dysfunction in retinal pigmented epithelial cells

Abstract

Oxidation of docosahexaenoate (DHA)-containing phospholipids in the cell plasma membrane leads to release of the α,β-unsaturated aldehyde 4-hydroxy-7-oxo-5-heptenoic acid (HOHA) lactone which is capable of inducing retinal pigmented epithelial (RPE) cell dysfunction. Previously, HOHA lactone was shown to induce apoptosis and angiogenesis, and to activate the alternative complement pathway. RPE cells metabolize HOHA lactone through enzymatic conjugation with glutathione (GSH). Competing with this process is the adduction of HOHA lactone to protein lysyl residues generating 2-(ω-carboxyethyl)pyrrole (CEP) derivatives that have pathological relevance to age-related macular degeneration (AMD). We now find that HOHA lactone induces mitochondrial dysfunction. It decreases ATP levels, mitochondrial membrane potentials, enzymatic activities of mitochondrial complexes, depletes GSH and induces oxidative stress in RPE cells. The present study confirmed that pyridoxamine and other primary amines, which have been shown to scavenge γ-ketoaldehydes formed by carbohydrate or lipid peroxidation, are ineffective for scavenging the α,β-unsaturated aldehydes. Histidyl hydrazide (HH), that has both hydrazide and imidazole nucleophile functionalities, is an effective scavenger of HOHA lactone and it protects ARPE-19 cells against HOHA lactone-induced cytotoxicity. The HH α-amino group is not essential for this electrophile trapping activity. The N^α-acyl L-histidyl hydrazide derivatives with 2- to 7-carbon acyl groups with increasing lipophilicities are capable of maintaining the effectiveness of HH in protecting ARPE-19 cells against HOHA lactone toxicity, which potentially has therapeutic utility for treatment of age related eye diseases.

Graphical Abstract

INTRODUCTION

Age-related macular degeneration (AMD) is a progressive loss of central vision resulting from damage to the retinal pigmented epithelium (RPE) and neural retina that affects approximately 30–50 million people worldwide (1), and its prevalence increases with age. The RPE, a monolayer of cells interposed between the photoreceptors and the Bruch’s membrane-choroid complex, is critical for the maintenance of retinal homeostasis (2). RPE cells with impaired functions can cause the formation of extracellular deposits called drusen which accumulate between the RPE and Bruch’s membrane (dry AMD). Choroidal neovascularization leads to leakage of blood into the macula and detachment of the neural retina (wet AMD). AMD-associated RPE cell dysfunction and death results in the loss of photoreceptor function and irreversible blindness.

Oxidative cleavage of polyunsaturated lipids generates α,β-unsaturated aldehydes.

ROS induce peroxidation of polyunsaturated fatty acids within the cellular plasma and organelle membranes (3). Oxidative fragmentation of polyunsaturated fatty acids that follows, generates a plethora of aldehydes, which include the α,β-unsaturated aldehydes, acrolein, 4-hydroxynon-2-enal (HNE), 4-hydroxyhex-2-enal (HHE), etc. Previously, we showed that phospholipid esters of docosahexaenoate (DHA), the most abundant fatty acid in the retina (in photoreceptor and in RPE cells) and brain, are oxidatively cleaved to generate and 4-hydroxy-7-oxohept-5-enoic acid (HOHA) phospholipids. HOHA ester of 2-lysophosphatidylcholine (HOHA-PC), readily undergoes intramolecular transesterification to release HOHA lactone (4). HOHA lactone can induce angiogenesis (5), activate the alternative complement pathway (6) and cause RPE cell dysfunction including cellular senescence, lysosomal membrane permeabilization and dissipation of mitochondrial membrane potential (7) and apoptosis (8).

Covalent adduction of α,β-unsaturated aldehydes with biological nucleophiles.

RPE cells metabolize HOHA lactone by adduction with glutathione (GSH) followed by the reduction of the adduct, and excretion of aldehyde and alcoholic forms of HOHA lactone-GSH (9). These metabolites are biologically active. HOHA lactone-GSH significantly accelerates cell migration and increases tube formation of human umbilical vein endothelial cells at sub-micromolar concentrations (5).

One of the major inducers of HOHA lactone formation in RPE cells is thought to be short wavelength blue light in the visible light spectrum, which promotes both the generation of HOHA lactone glutathione (GSH) adducts and production of CEP derivatives of proteins and ethanolamine phospholipids through adduction of HOHA to primary amino groups (7). Levels of CEP derivatives are elevated in plasma from individuals with AMD and in drusen from AMD donor eye tissues. CEP promotes vascular angiogenesis (10) and platelet activation via Toll-like receptor-2 and Toll-like receptor-9 dependent signaling pathways (11).

Mitochondrial dysfunction and AMD.

The pathogenesis of AMD is a complex multifactorial process and it remains poorly understood. Converging evidence from multiple studies implicate the role of mitochondrial dysfunction in AMD pathogenesis (12). Decline in the number and structural integrity of mitochondria (13) and decreased levels of electron transport chain proteins were observed in RPE cells from human donors with AMD (14).Mitochondria produce energy for the cell in the form of ATP, through the citric acid cycle and oxidative phosphorylation (OXPHOS), and regulate cellular metabolism and homeostasis, e.g., through maintenance of membrane potential (ΔΨ_m), removal of reactive oxygen species (ROS) and mitophagy. Failure in the balance of these processes triggers mitochondrial dysfunction. The collapse of ΔΨ_m and excessive production of mitochondrial ROS are canonical precursors of cellular death. RPE cells from AMD donors produce significantly higher ROS levels than normal RPE cells under oxidative stress (15). Also, the levels of ATP produced by mitochondria are significantly lower in AMD RPE cells. This indicates lower mitochondrial activity and is in a good agreement with the discovery that the levels of ATP synthase α, β, and δ subunits are lower in RPE mitochondria isolated from AMD than from normal donor eye retina (16).

The present study shows that HOHA lactone has the ability to induce mitochondrial dysfunction in RPE cells, determines the mechanistic basis of this toxicity, and explores theraputic countermeasures. HOHA lactone inactivates mitochondrial OXPHOS complexes, especially I and II, causes oxidative stress in ARPE-19 cells and is toxic to isolated mitochondria. The study compares the efficacy of HH derivatives for prophylactic trapping HOHA lactone. We now report that HH captures HOHA lactone 20x faster than carnosine and that certain Nα-acyl HH derivatives retain full theraputic α,β-unsaturated aldehyde scavenging activity that protects RPE cells against HOHA lactone toxicity.

MATERIALS AND METHODS

Reagents

Dulbecco’s modified Eagle’s cell culture medium and Ham’s F12 cell culture medium F-12 (1:1 mixture, DMEM/F12), Dulbecco’s phosphate-buffered saline (DPBS), Hank’s balanced salt solution (HBSS), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) were purchased from Fisher Scientific (Pittsburgh, PA). Heat-inactivated fetal bovine serum (FBS) was from Equitech-Bio, Inc. (Kerrville, TX). Texas Red-X Goat anti-Rabbit IgG (H+L) cross-adsorbed secondary antibody (T-6391) was from ThermoFisher Scientific (Waltham, MA). Flash Phalloidin™ Green 488 was from Biolegend (San Diego, CA). β-NADPH was obtained from Cayman Chemical (Ann Arbor, MI). All other chemicals and reagents, including L-glutathione (reduced), glutathione reductase (250 units/mL) and 5,5’-dithiobis(2-nitrobenzoic acid) (DTNB) were purchased from Sigma-Aldrich (St. Louis, MO). 4-Hydroxy-7-oxohept-5-enoic acid (HOHA) lactone was synthesized as described previously (9). A polyclonal rabbit anti-CEP antibody was raised and characterized as described previously (17). The Pierce 660 nm protein assay kit and the Pierce BCA Protein Assay kit were obtained from ThermoFisher Scientific (Waltham, MA) and used to determine a protein concentration in the lysates in accordance with the manufacturer’s instructions. The MitoSOX Red reagent was obtained from ThermoFisher Scientific (Waltham, MA) and used to determine mitochondrial superoxide levels in accordance with the manufacturer’s instructions.

General methods

NMR spectra were acquired on a 500 MHz Bruker Ascend Avance III HDTM (Bruker, Billerica, MA) equipped with a Prodigy ultra-high sensitivity multinuclear broadband CryoProbe operating at 500 and 125 MHz for 1H and 13C, respectively. They were referenced internally according to residual solvent signals. All ESI mass spectra were obtained from a Thermo Finnigan LCQ Deca XP (ThermoFisher Scientific, Waltham, MA). High-performance liquid chromatography (HPLC) was performed on a Shimadzu UFLC system equipped with a 5 μm Phenomenex Luna C-18 column (Torrance, CA). Flash column chromatography was performed on 230−400 mesh silica gel supplied by Sigma-Aldrich (St. Louis, MO) with ACS grade solvents. R_f values are quoted for glass plates coated with SiO₂ with silica gel layer thickness of 0.25 mm. The plates were visualized with iodine, UV, and phosphomolybdic acid reagents. All reactions were carried out under an argon atmosphere. All reagents were obtained commercially unless otherwise noted. Reactions were performed using glassware that was oven-dried at 120 °C. Air and moisture sensitive liquids and solutions were transferred via syringe or stainless-steel cannula.

Syntheses.

The following HOHA lactone chemical traps (see Fig. 1) were synthesized, characterized and tested for their scavenging activity toward HOHA-lactone in ARPE-19 cells: histidyl hydrazide (HH), Nα-acetyl HH (Ac-HH), Nα-trimethylacetyl HH (tBu-HH), Nα-t-Boc-Histidyl hydrazide (boc-HH), Nα-cyclohexanecarbonyl HH (cH-HH), Nα-cyclopentanecarbonyl HH (cP-HH), Nα-nonanoyl HH (n-HH),Nα-decanoyl HH (d-HH), Nα-undecanoyl HH (u-HH), Nα-dodecanoyl HH (Dd-HH) (see the description of the respective ¹H NMR spectra of these compounds along with their ¹H and ¹³C spectra in the Supporting Information section of this manuscript).

Figure 1.

Synthesis of histidyl hydrazide and its Nα-acyl HH derivatives. (A) Synthetic approach used for the preparation of HH and its analogues. (B) Chemical structures of HH derivatives.

Cell culture.

The cell line ARPE-19 (ATCC; CRL-2302) derived from spontaneously arising retinal pigment epithelia of a healthy person (18) was obtained from the American Type Culture Collection (Manassas, VA). The cells were grown on 100-mm dishes in a humidified CO₂ incubator at 37 °C and 5% CO₂ in Ham’s F12 /Dulbecco’s modified Eagle’s medium (DMEM) (50:50 ratio), containing L-glutamine and 10% heat-inactivated FBS. Cells were trypsinized and passaged every 2–3 days. Cells with passages 21−30 were used.

Microscopy.

Images were collected on a Leica DMI 6000B inverted fluorescence microscope (Leica Microsystems, Wetzlar, Germany) using a Retiga EXI camera (QImaging, Vancouver, British Columbia). Image analysis was performed using MetaMorph Imaging Software (Molecular Devices, Downington, PA).

MTT cell viability assay.

We performed this assay as described previously (19). Briefly, ARPE-19 cells (10,000 cells/ per well) were seeded into a sterile 96-well flat bottom microplate in 200 μL of complete DMEM/F12 medium supplemented with 10% FBS and allowed to attach to the surface of culture plates in a humidified CO₂ incubator at 37 °C and 5% CO₂. After the cells reached ≈ 80% confluence, they were starved in 200 μL of basal DMEM/F12 cell culture medium for 5 h to overnight. The cells were then pre-incubated with HH or its derivative for 1 h, 6 replicate wells for each concentration, followed by 24 h incubation with 20 μM of HOHA lactone in under the conditions described above. At the end of the incubation period, the cell culture medium was aspirated from each well and the cells were incubated for 2 h at 37 °C and 5% CO₂ in 180 μL of DMEM/F12 basal medium supplemented with 20 μL of 0.2 μ PES filtersterilized MTT solution (5 mg/mL in DPBS). The cell culture medium was aspirated from each well, dimethyl sulfoxide (DMSO) was added to each well (200 μL) to dissolve water-insoluble intracellular formazan crystals by carefully pipetting the contents of the wells. The optical density (OD) of formazan solutions at λ = 540 nm were measured with a plate reader (SpectraMax M2, Molecular Devices, San Jose, CA) with a reference wavelength set at λ = 670 nm to account well content light scattering.

Assay of RPE cell mitochondrial membrane potential.

The JC-10 assay was performed by the method described previously (20) with slight modifications. Briefly, after HOHA lactone treatment described above, to ARPE-19 cells in a 96-well flat bottom plate in 100 μL of basal DMEM/F12 cell culture medium was added 100 μL of 5 μg/mL stock solutions of JC-10 in DPBS to establish 2.5 μg/mL final concentrations of JC-10 in a well. After subsequent incubation in a humidified CO₂ incubator for 30 min at 37 °C and 5% CO₂, the fluorescence was immediately measured with a fluorescence microplate reader (SpectraMax M2, Molecular Device, San Jose, CA) set at λ_Ex/Em = 490/525 nm (cut off at λ = 515 nm) for the green monomer and λ_Ex/Em = 540/590 nm (cut off at λ = 570 nm) for the red aggregate. Changes in mitochondrial membrane potential were derived from the green/red fluorescence ratio.

Assay of RPE cell mitochondrial reactive oxygen species.

The production of superoxide in mitochondria in ARPE-19 cells was visualized with MitoSOX Red (Life Technologies, Waltham, MA). ARPE-19 cells (10,000 cells/ per well) were plated on an 8-chamber well (Lab-Tek II Chamber Slide System, Nunc, Rochester, NY) in the DMEM/F12 with 10% FBS and incubated at 37 °C in 5% CO₂ for three days. After being starved overnight, the cells were treated with 0 – 10 μM HOHA lactone and incubated for an additional 1 h. Following cell culture medium aspiration, the cells were incubated with 5 μM MitoSOX Red reagent in HBSS for 15 minutes at 37 °C. Cells were washed in warm DMEM/F12, fixed with 3.7% PFA for 10 minutes, washed with DPBS, and mounted with DAPI Fluoromount-G (Southern Biotech, Birmingham, AL). All images were acquired with a Leica DMI 6000 B inverted fluorescent microscope using a Retiga EXI camera. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA). The images were acquired at 20x magnification.

For quantitation, the assay was performed in 96-well microplate (5,000 cells/well) format by the method described above. After MitoSOX Red staining, the cells were washed with HBSS and the fluorescence was immediately measured without medium using a fluorescence microplate reader (SpectraMax M2, Molecular Device) set at λ_Ex/Em = 510/580 nm (cut off at λ = 570 nm) and normalized by the total protein concentration in each well measured by BCA assay (21).

Assay of total RPE cellular reactive oxygen species.

ARPE-19 cells (5,000 cells/ per well) were seeded into a 96-well flat bottom for three days. After they have been starved overnight, the cells were incubated with 2 μM CM-H₂DCFDA in HBSS for 45 minutes at 37 °C. After washing with HBSS, cells were treated with 0 – 20 μM HOHA lactone and well fluorescence was measured at indicated time points using a fluorescence microplate reader (SpectraMax M2, Molecular Device) set at λ_Ex/Em = 485/535 nm (cut off at λ = 530 nm) and normalized by the total protein concentration in each well as measured by BCA assay.

Quantification of total RPE intracellular GSH.

ARPE-19 cells (75,000 cells/dish) were seeded into 35-mm dishes in DMEM/F12 for three days. After cells became confluent, they were starved for 3 h. The cells were then pre-incubated in HH for 1 h followed by 24 h incubation in 20 μM of HOHA lactone. Aliquots (20 μL) of ARPE-19 cell lysates were assayed to determine total intracellular GSH levels using a 96-well microplate format described earlier (22). In these experiments, all of the reagents were prepared in 0.1 M potassium phosphate buffer with 5 mM EDTA disodium salt, pH 7.5 (KPE buffer). Briefly, 20 μL of KPE buffer, GSH standards, or samples were added to the respective microplate wells, followed by the addition of 120 μL of a freshly prepared 1:1 mixture of DTNB (2 mg/3 mL) and glutathione reductase (10 U/3 mL). Then, 60 μL of NADPH (2 mg/3 mL) was quickly introduced into the wells and the plate was mixed well. The absorbance was read immediately at λ = 412 nm in a microplate reader (SpectraMax M2, Molecular Devices). Measurements were taken every 20 s for 5 min (15 readings in total from 0–300 s). The total GSH concentration in the samples was determined by linear regression to calculate the values obtained from a standard curve.

Determination of RPE cellular ATP.

To measure cellular ATP level, CellTiter-Glo (Promega, Madison, WI, USA) was used according to manufacturer’s directions. In brief, after equilibrating to room temperature, equal volume of the CellTiter-Glo reagent was added to the cell medium of HOHA lactone treated cells in a 96 well white polystyrene cell culture microplate (Greiner BioOne CELLSTAR white plate with solid bottom, Cat. # 655083) and mixed for 2 min on an orbital shaker. The microplate was incubated for additional 10 minutes at room temperature to stabilize the luminescence and was immediately measured with a luminescence microplate reader (SpectraMax M2, Molecular Device).

Detection of CEP in ARPE-19 cells.

ARPE-19 cells (10,000 cells/chamber) were plated on an 8-chamber slide (Lab-Tek II Chamber Slide System, Nunc, Rochester, NY) in DMEM/F12 supplemented with 10% FBS and then incubated at 37 °C under 5% CO₂ for three days. After 5 h of starvation in basal DMEM/F12 cell culture medium, the cells were incubated with HH for 1 h, and were subsequently added HOHA lactone to a final concentration of 20 μM. After 24 h incubation in a humidified CO₂ incubator at 37 °C under 5% CO₂, the cell culture medium was aspirated from each chamber and the chambers were washed twice with DPBS. The cells were fixed with cold acetone (−20 °C) for 12 min at −25 °C. After washing with ice-cold PBST three times, the slides were blocked with 3% BSA in PBST for 1 h at 23 °C. The cells were probed with rabbit anti-CEP polyclonal antibody (in 3% BSA in PBST, 18 μg/mL) overnight at 4 °C, and washed exhaustively with PBST the next day. The slides were treated with Texas Red-X goat anti-rabbit antibody (1:100 dilution in 3% BSA in PBST; T-6391, ThermoFisher Scientific, Waltham, MA) overnight at 4 °C, washed exhaustively with PBST. The slides were further incubated with 1:20 diluted in PBS stock solution of Flash Phalloidin™ Green 488 (0.2 U/μL, Biolegend, San Diego, CA) in the dark for 30 min at 23 °C. After washing with PBST, slides were mounted with DAPI containing Fluoromount-G (Southern Biotech, Birmingham, AL). All images were acquired with a Leica DMI 6000 B inverted fluorescent microscope using a Retiga EXI camera. Image analysis was performed using Metamorph imaging software (Molecular Devices, Downington, PA). The images were acquired at 20x magnification.

Isolation of mitochondria from ARPE-19 cells.

ARPE-19 cells were sub-cultured in T-150 flasks and trypsinized once they reached confluency (10 – 15 ×10⁶ cells). After being washed twice with DPBS and centrifuging at 1,000g for 5 min at 4 °C, the cell pellet was flash-frozen in liquid nitrogen, then thawed and suspended in 1 mL of 10 mM ice-cold hypotonic Tris buffer (pH 7.6). After the suspension was homogenized carefully with a Teflon tissue homogenizer on ice, 200 μL of 1.5 M sucrose solution was added, vortexed thoroughly and centrifuged at 600g for 10 min at 4 °C. The supernatant was collected and centrifuged additionally at 14,000g for 10 min at 4 °C. The resulting mitochondrial pellet was resuspended in 0.5 ml of 10 mM ice-cold hypotonic Tris buffer (pH 7.6), divided into aliquots and stored at −80 °C. The cell mitochondrial solution was thawed and subjected to three cycles of freeze-thawing in liquid nitrogen/37 °C water bath to disrupt the mitochondrial membranes just before use. The mitochondrial protein concentration was determined using the Pierce 660 nm Protein Assay kit ThermoFisher Scientific (Waltham, MA).

Assays for mitochondrial enzyme activities.

Mitochondrial respiratory chain enzymatic activities (complexes I-IV) were assessed as previously described (23) and accommodated to 96-well microplate format.

Complex I activity:

To isolated mitochondrial fraction (2 μg of protein) was added the assay medium containing potassium phosphate buffer (50 mM, pH 7.5), fatty acid-free BSA (3 mg/mL), KCN (300 μM), NADH (100 μM) in distilled and deionized water (200 μL). After reading the baseline at 340 nm for 2 minutes, ubiquinone1 (60 μM final concentration) was quickly introduced. The decrease in reaction mixture absorbance was recorded at λ = 340 nm for 60 minutes. In parallel, the same quantity of reagents and samples but with the addition of rotenone solution (10 μM final concentration) was used.

Complex II activity:

To isolated mitochondrial fraction (1 μg of protein) was added the assay medium containing potassium phosphate buffer (25 mM, pH 7.5), fatty acid-free BSA (1 mg/mL), KCN (300 μM), succinate (20 mM), 2,6-dichlorophenolindophenol (DCPIP) sodium salt (80 μM) and distilled water. The mixture was incubated at 37 °C for 10 minutes, and then the baseline was recorded at 600 nm for 2 minutes. Then decylubiquinone (DUB, 50 μM final concentration) was applied and the decrease in absorbance at 600 nm was recorded for 60 minutes.

Complex III activity:

To isolated mitochondrial fraction (1 μg of protein) was added the assay medium containing potassium phosphate buffer (25 mM, pH 7.5), KCN (500 μM), EDTA (100 μM), oxidized cytochrome c (75 μM), Tween-20 (0.025% (vol/vol)) and distilled water. After reading the baseline at λ = 550 nm for 2 minutes, decylubiquinol (100 μM final concentration) was introduced. The increase in absorbance at 550 nm was recorded for 15 minutes.

Complex IV activity:

To isolated mitochondrial fraction (1 μg of protein) was added the assay medium containing potassium phosphate buffer (25 mM, pH 7.0), KCN (300 μM), reduced cytochrome c (50 μM) in distilled and deionized water (200 μL). The decrease in absorbance at λ = 550 nm was recorded for 15 minutes.

HPLC analysis of HOHA lactone scavenging activity.

Both 10 mM or 100 mM solutions of the compounds and 0.5 mM solution of HOHA lactone in DPBS were prepared. In a screw cap-equipped vial were mixed 10 μL of the compound solution and 100 μL of the HOHA lactone solution (final molar ratio between compound and aldehyde = 1:2 or 1:20). The resulting mixture was sealed and incubated at 37°C. At time (T) = 0, 15, 30, 45 and 60 min a 20 μL sample was pooled from the vial and injected into the HPLC system. Reverse-phase HPLC was conducted using a Shimadzu UFLC system equipped with a 5 μm 4.6 × 250 mm Phenomenex Luna C18 column (isocratic mode; mobile phase, H₂O/methanol/formic acid 70:30:0.1; flow rate 1 ml/min). Peak areas corresponding to the unreacted HOHA lactone were integrated and the residual concentration was calculated according to the formula: HOHA lactone residual concentration (%) = (A₀ / A_t) × 100 where A₀ is the peak area at time 0 and A_t is the peak area at each sampling time. The data were plotted into a graph reporting the residual % or consuming % concentration of compound vs. time interval (for each compound the analysis was run in duplicate; values are the mean ± SD).

Quantitation of HH uptake by LC-MS/MS.

ARPE-19 cells (200,000 cells/dish) were seeded in a 60-mm dish and were grown to confluency. After being starved in basal DMEM/F12 cell culture medium overnight, the cells were treated with 0–800 μM HH for 24h. Before harvest, the cells were washed extensively with HBSS to remove any residual HH. The cell pellets were frozen in dry ice and resuspended in methanol. The suspensions were then vortexed and sonicated to extract the cellular HH.

The cells were then scraped and pelleted by centrifugation (1,000g, 5 min, 4 °C). The supernatant was discarded and the pellet was snap-frozen on dry ice. For analysis, the pellet was resuspended in 500 μL of neat methanol (MeOH) was bath-sonicated for 5 min. After the sample was carefully vortexed for 1 min and centrifuged (14,000g, 15 min, 4 °C), 400 μL of supernatant was transferred to another fresh tube and dried by nitrogen stream. The sample was then resuspended in 200 μL of 0.1% formic acid in water, sonicated for 5 min, vortexed for 1 min and the resulting solution was filtered into a fresh tube before LC-MS/MS analysis. The LC-MS/MS system for the analysis consisted of a Thermo Finnigan LCQ Deca XP with a Surveyor LC system. Liquid chromatography was performed using a Gemini-NX C18 column (2.1 × 150 mm, 5 μm) with a Gemini-NX C18 guard column (2 × 4 mm) (both from Phenomenex). An isocratic HPLC method with water/methanol/formic acid 95:5:0.1 mobile phase was used and the total run time was 6 min with flow rate 200 μL/min and a 10 μL sample volume was injected. ESI mass spectrometry in the positive ion mode was employed. The instrument parameters were as follows: the heated capillary temperature was 200 °C, the source voltage 4.0 kV, and the capillary voltage 11.00 V. Nitrogen was used as sheath and auxiliary gas. All data were processed with the Qual browser in Xcalibur software. MS/MS experiments were performed by selecting an ion with an isolation width of 2 m/z units and collision energy at 26%. Selected ion recording at m/z 153.0 from CID fragmentation of m/z 170.1 in the positive ion mode was used to identify HH. The cellular amount was calculated using a calibration curve built from the authentic standard and normalized by the total protein concentration of the precipitated protein obtained after methanol extraction.

HH inhibition of HOHA lactone-protein adduction in ARPE-19 cells.

ARPE-19 cells (200,000 cells/dish) were seeded in a 60-mm cell culture dish and were grown to confluency. After being starved in basal DMEM/F12 cell culture medium overnight, the cells were treated with 0–800 μM HH for 24h and then washed extensively with HBSS to remove any residual HH. To evaluate the efficacy of HH pretreatment to prevent CEP production in ARPE-19 cells, the cells were then exposed to 10 μM HOHA lactone for 24 h with or without HH pretreatment and then immunostained with primary rabbit anti-CEP polyclonal antibody/secondary goat anti-rabbit Texas Red-X antibody.

Statistical Analysis.

Unless specified in the text or the figure legends, comparisons were made using one-way ANOVA followed by Holm-Sidak’s post-hoc multiple comparison test. Statistical significance is shown as follows: *, p < 0.05; **, p < 0.01; ***, p < 0.001. Data are presented as mean ± standard error of mean (SEM).

RESULTS

HOHA lactone damages OXPHOS complexes in RPE cell mitochondria.

Exposure of mitochondria (0.2 mg/mL protein), isolated from cultured ARPE-19 cells, to HOHA lactone impairs the enzymatic activities of OXPHOS complexes I and II (see Fig. 2). Mitochondrial complex I total activity decreases 28% after treatment with 10 μM (50 nmol/mg protein), 49% after treatment with 50 μM HOHA lactone but remains flat at 47% even after treatment with 100 μM HOHA lactone (Figure 2A). In contrast, rotenone-sensitive activity of the mitochondrial complex I – calculated by subtracting total complex I specific activity (mitochondria without rotenone) and rotenone-resistant activity (mitochondria with rotenone) – decreases by 53% after exposure to 10 μM HOHA lactone, by 90% after exposure to 50 μM HOHA lactone, and all of this activity is lost after exposure to 100 μM HOHA lactone (Figure 2A). There is also a strong HOHA lactone dose-dependent decrease of mitochondrial complex II activity. After treatment with 20 μM HOHA lactone, complex II activity decreases by 19% and by 78% after 100 μM HOHA lactone (Figure 2B). The activity of mitochondrial complex III is not significantly affected by HOHA lactone treatment even at 100 μM (Figure 2C). Mitochondrial complex IV is less susceptible to HOHA lactone than complexes I or II. Only after treatment with 100 μM HOHA lactone, the complex IV activity only decreases by 20% (Figure 2D).

Figure 2.

Decline in the specific enzymatic activities of mitochondrial OXPHOS complexes induced by exposure of ARPE-19 cells to HOHA lactone. Isolated mitochondria were incubated with HOHA lactone at the indicated concentrations for 0.5 h, and then the specific enzymatic activities of complexes (A) I, (B) II, (C) III and (D) IV were assayed spectrophotometrically (E; blue arrows in panel E show the wavelengths used to measure the enzymatic activity of respiratory chain complexes I–IV) by monitoring NADH, 2,6-dichlorophenolindophenol (DCPIP) or cytochrome c (Cyt c; see Materials and Methods for details). The data are presented as mean ± SEM of % control, n = 3; **P<0.01, ***P<0.001 compared with the respective non-treated controls.

HOHA lactone promotes mitochondrial ROS generation and cellular dysfunction.

The main sources of mitochondrial ROS are complex I and complex III of the respiratory chain (24). Under pathological conditions, complex I becomes a major ROS source (25). Because HOHA lactone inhibits mitochondrial complex I, we expected that the resulting electron leakage would lead to superoxide generation and its subsequent release in the mitochondria. Monitoring the generation of total cellular ROS in RPE cells with the CM-H₂DCFDA probe, we observed a time- and concentration-dependent elevation of ROS in ARPE-19 cells (Figure 3A). To specifically measure the level of mitochondrial ROS, which is the main source of cellular ROS, we used MitoSOX Red fluorescent probe. Even at low micromolar HOHA lactone concentrations we detected an increase of mitochondrial ROS levels in ARPE-19 cells which peaked at 20 μM (3-fold) after 1h incubation with HOHA lactone followed by a precipitous decline at 40 μM HOHA lactone, presumably owing to cytotoxicity at this level (Figures 3B, 3C). We estimate that the endogenous level of HOHA under stressed conditions, e.g., UVA irradiation, is approximately 20 μM (see Supporting Information).

Figure 3.

HOHA lactone induces the generation of mitochondrial ROS in ARPE-19 cells. (A) Time-course for the generation of total cellular ROS measured with the CM-H₂DCFDA assay. Cells were preloaded with CM-H₂DCFDA dye for 1 h were washed with PBS and then exposed to various concentrations of HOHA lactone at the time points indicated. (B) The mitochondrial ROS level was quantified with MitoSOX Red reagent. Cells were exposed to HOHA lactone for 1 h and then incubated with MitoSOX Red reagent before measurement (see Materials and Methods section for details). The data are presented as mean ± SEM of % control, n = 3 – 6; *P<0.05, ***P<0.001. (C) Representative images showing MitoSOX Red fluorescence were acquired using Leica DMI 6000 B inverted fluorescent microscope. Cells were exposed to HOHA lactone and then incubated with MitoSOX Red reagent before fixation, mounting and staining with DAPI.

In mitochondria with impaired respiration or a leaky inner membrane, the F₀F₁-ATPase reverses and consumes ATP to maintain the ΔΨm at a suboptimal level by pumping protons out of the matrix (26). However, continuous operation of F₀F₁-ATPase in reverse consumes cellular ATP reserves driving the cell into an energy crisis and eventually leading to cell demise (26). The lowered driving force (proton motive force) across the inner mitochondrial membrane may also inhibit ATP synthesis. Our data show that HOHA lactone is a potent inhibitor of ATP synthesis in ARPE cells. Incubation of ARPE-19 cells with 40 μM HOHA lactone for 24 h causes a 75% decrease in the cellular ATP content compared with that of control cells (Figure 4A). Under the same culture conditions, incubation of ARPE-19 cells with oligomycin A caused a 45% decrease of the ATP level (Figure 4A). Oligomycin A is a well-known ATP synthase inhibitor, which blocks the proton channel of ATP synthase (F₀ subunit) and prevents oxidative phosphorylation of ADP to ATP (27). It remains to be determined whether blocking of mitochondrial energy metabolism and depletion of ATP in ARPE-19 cells by HOHA lactone results from damaging the mitochondrial respiratory chain, ATP synthase, or the citric acid cycle.

Figure 4.

HOHA lactone induces depletion of ATP, and mitochondrial membrane potential in ARPE-19 cells. Cells were exposed to various concentrations of HOHA lactone or 0.1 μM oligomycin A for 24 h (A) ATP level was quantified by a luciferin-luciferase based assay, and the ATP level was measured; (B) mitochondrial membrane potential (Δψ_m) was quantified with the JC-10 probe. The data are presented as mean ± SEM of % control, n = 3 – 6; *P<0.05, ***P<0.001.

The decrease of mitochondrial membrane potential (Δψ_m) is a point of no return in the classical apoptosis pathway (28). Exposure to HOHA lactone led to a decrease of mitochondrial membrane potential in ARPE-19 cells since ATP synthase was not able to maintain Δψ_m under our experimental conditions. We detected a 60% decrease of Δψ_m by JC-10 assay after incubation with 20 μM HOHA lactone (Figure 4B). Exposure to HOHA lactone caused a drastic decrease of total GSH levels (the levels of GSH and GSSG) in ARPE-19 cell. At 1 μM HOHA lactone, the GSH level dropped from 31 nmol/mg protein to 19 nmol/mg protein (39% decrease) and at 10 μM, only 10 nmol/mg protein was retained (68% decrease), respectively (Figure 5A). Glutathione (GSH) is an important endogenous antioxidant, and excessive generation of ROS due to mitochondrial dysfunction leads to the depletion of GSH levels in the cell, which, in turn, leads to the cell’s inability to protect itself against oxidative stress. Mitochondrial and cellular dysfunction can eventually lead to cell death. We showed that incubation ARPE-19 cells with even as little as 5 μM HOHA lactone for 24 h caused a significant decrease in cell viability down to 85%. The cell viability dropped to 27% after 24 h exposure 20 μM HOHA lactone and dropped to 19% after exposure to 30 μM HOHA lactone (Figure 5B).

Figure 5.

HOHA lactone induces depletion of GSH and a decrease in cell viability in ARPE-19 cells. Cells were exposed to various concentrations of HOHA lactone and after 24 h (A) intracellular GSH levels in the ARPE-19 cell lysates were quantified by the DTNB method. (B) cell viability was quantified by an MTT assay. The data are presented as mean ± SEM of % control, n = 3 – 6; *P<0.05, ***P<0.001.

Scavenging α,β-unsaturated aldehydes by electrophile traps.

DHA is especially abundant in photoreceptor disk membranes of the human retina. The light-induced oxidative fragmentation of DHA phospholipids in these membranes releases HOHA lactone that diffuses into RPE cells where GSH adducts are formed and secreted (Figure 6). This depletes GSH in RPE cells and favors further lipid oxidation and post-translational modification of biomolecules, e.g., CEP formation. Mitochondrial dysfunction results from adduction of HOHA lactone, e.g., to mitochondrial proteins in RPE cells, especially under pathological conditions where GSH levels are depressed.

Figure 6.

HOHA lactone produced in photoreceptor disks dissociates into RPE cells (29) where formation of covalent adducts with GSH, proteins and ethanolamine phospholipids may cause mitochondrial dysfunction. A competing covalent adduction with HH may protect RPE cells against HOHA lactone-induced GSH depletion and cytotoxic covalent modification.

To block the depletion of GSH by α,β-unsaturated aldehydes and the cytotoxic consequences of HOHA lactone on RPE cells, we evaluated the efficacy of various electrophile scavengers for preventing cell death. First, we examined the ability of some known aldehyde traps to scavenge the HOHA lactone using RP-HPLC. At high concentration (20 to 1 ratio of scavenger to α,β-unsaturated aldehyde, 10 mM scavenger to 0.5 mM HOHA lactone), HH (100 %), L-carnosine (95 %), methoxyamine (100 %) and glutathione (99 %) effectively scavenge HOHA lactone within an hour (Figure 7A).

Figure 7.

Scavengers of HOHA lactone. (A) Levels of HOHA lactone (initial concentration 0.5 mM) after incubation scavengers (initial concentration 1 mM) for 1 h at 37 °C. (B) HOHA lactone consumption during 1h incubation with scavengers (1 mM, scavenger/aldehyde ratio = 2:1; 10 mM, scavenger/aldehyde ratio = 20:1), and reaction rate constants (k_obs) of scavengers with HOHA lactone determined by sigmoidal curve fitting (%A_t = A₀ × e^Kobs).

However, at lower concentrations (a 2:1 ratio of scavenger to α,β-unsaturated aldehyde, 1 mM scavengers to 0.5 mM HOHA lactone), L-carnosine (20 %) did not scavenge HOHA lactone efficiently; instead, only HH (99 %) showed an appreciable ability to scavenge HOHA lactone as effectively as glutathione (100 %) in one-hour incubation experiments (Figure 7B).

HH forms covalent adducts with HOHA lactone and HOHA lactone-GSH.

L-Carnosine forms a cyclic adduct with 4-HNE. Since HH, like L-carnosine, has a histidyl functional group and a terminal NH₂ group, we expected that HOHA lactone would form a similar adduct. After 24 hours incubation of an equimolar mixture of HOHA lactone and HH at 37 °C we were able to detect two ions in the positive mode of the mass spectrum, [M+ H]⁺ = 292 m/z and [M+H₂O+H]⁺= 310 m/z (Figure 8A) that correspond to the anticipated adducts. We previously reported that HOHA lactone could be metabolized by RPE cells to form HOHA lactone GSH adducts (9). HOHA lactone-GSH is generated by Michael addition of glutathione to HOHA lactone, therefore, it still possesses an aldehyde group that can react with primary amino groups (9). We found that HH reacts with HOHA lactone-GSH to form a Shiff-base adduct after 24 h incubation at 37 °C. In the positive mode of the mass spectrum, [M+ H]⁺ = 599 m/z and [1/2M+ H]⁺ = 300 m/z were found (Figure 8B).

Figure 8.

HH is a scavenger of HOHA lactone in RPE cells in culture. (A) Mass spectra of HH covalent adducts with HOHA lactone and (B) HOHA lactone-GSH adduct.

Cellular uptake and cytotoxicity of HH.

To examine if HH is taken up by ARPE-19 cells, we used LC-MS/MS to quantify its presence in ARPE-19 cells. After incubation with 50 to 800 μM HH for 24 h, a dose-dependent increase of the cellular levels of HH (88 pmol per mg protein at 50 μM to 853 pmol per mg protein at 800 μM) was detected. This indicates that RPE cells are able to acquire HH from exogenous sources (Figure 9A). We evaluated cell viability by the MTT assay to examine the toxicity of HH toward ARPE-19 cells. After exposure to HH concentrations up to 1 mM, ARPE-19 cells did not show any significant decrease in cell viability. Thus, within this concentration range, HH is not cytotoxic toward RPE cells (Figure 9B).

Figure 9.

HH is an effective scavenger of HOHA lactone in RPE cell culture. (A) Uptake of HH by ARPE-19 cells. HH in the cells was quantified by LC-MS/MS using a standard curve after 24 h incubation in a cell culture medium containing 50 – 800 μM HH. The data are presented as mean ± SEM, n = 3. (B) HH is not cytoxic toward ARPE-19 cells. ARPE-19 cells were exposed for 24 h to the indicated concentrations of HH and cell viability was measured by MTT assay. The data are presented as a box plot, n = 6. (C) Cell viability was measured using an MTT assay. (D) Mitochondrial membrane potential (Δψ_m) was measured using a JC-10 assay. Cells were exposed to HOHA lactone after HH pretreatment (1h), and Δψ_m was measured after 24 h incubation. The data are presented as mean ± SEM of % control, n = 6. (E) Cells were exposed to HOHA lactone after HH pretreatment (1 h) and harvested after 24 h. The GSH level in the cell lysate was quantified by a DTNB assay. The data are presented as mean ± SEM, n = 3; **P<0.01, ***P<0.001 compared with the value for HOHA lactone treatment in the absence of HH. (F) Mitochondrial ROS levels were quantified by MitoSOX Red fluorescence. Cells were exposed to HOHA lactone for 1 h after pretreatment with HH, and then incubated with MitoSOX Red before measuring the red fluorescence of the oxidized reagent in a microplate reader. (G) ATP level was quantified by a luciferin-luciferase based assay. Cells were exposed to HOHA lactone after 1 h HH pretreatment, and the ATP level was then measured after 24 h incubation with HOHA lactone. The data are presented as mean ± SEM of % control, n = 3 for A and n = 5 for B; *P<0.05, **P<0.01 compared with HOHA lactone treatment in the absence of HH.

Protective effect of HH on GSH levels, cell viability and mitochondrial health.

HOHA lactone is cytotoxic to ARPE-19 cells, as are other α,β-unsaturated aldehydes such as 4-HNE (30–32) and acrolein (33–35). Incubation of ARPE-19 cells with 20 μM HOHA lactone for 24 h results in 80% loss of metabolic viability (Figure 9C). The efficacy of HH for preventing RPE cell death induced by HOHA lactone was assessed with an MTT assay. After pretreatment for 1 h with 50 μM HH, incubation for 24 h with 20 μM HOHA lactone caused ≈ 50% decrease in viable cells compared to ≈ 80% decrease in the absence of HH. With 600 μM HH pretreatment, most of the cells were completely protected against HOHA lactone toxicity. Pretreatment with HH also diminished the decrease of mitochondrial membrane potential induced by HH (Figure 9D). In addition, while the levels of GSH after treatment with 20 μM HOHA lactone for 2 h, decreased from 25 nmol/mg protein to 9 nmol/mg protein (70% decrease), 1 h pretreatment with HH attenuated the decrease of GSH (Figure 9E). In the presence of 25 μM HH, the GSH levels in RPE cells only decreased to 12 nmol/mg protein (≈ 50% of control) and in the presence of 500 μM HH, the cells retained >90% of their GSH (Figure 9E).

The efficacy of HH for preventing HOHA lactone-induced mitochondrial damage in RPE cells was assessed. Pretreatment with 200 μM HH prior to incubation for 24 h with 20 μM HOHA lactone, allowed 80% retention of of Δψ_m (Figure 9D). Pretreatment with HH, also caused a dose-dependent decrease of mitochondrial ROS production. The levels dropped to basal in the presence of 400 μM HOHA lactone (Figure 9F). Incubation of ARPE-19 cells with 40 μM HOHA lactone for 24 h also reduced the cellular ATP content to 25% that of control cells, while pretreatment with HH effectively attenuated the decrease of ATP level (Figure 9G). These observations demonstrate that HOHA lactone can impair mitochondrial energy metabolism, and HH is able to ameliorate the metabolic damage engendered by this α,β-unsaturated aldehyde electrophile.

HH reduces CEP generation in RPE cells.

HOHA lactone reacts with primary amino groups in biomolecules to generate CEPs. These DHA-derived modifications have significant pathological and physiological consequence to AMD, cancer and wound healing. Treatment of ARPE-19 cells with 10 μM HOHA lactone caused a significant increase in levels of CEP indicated by a 100% increase in red fluorescence intensity owing to immunostaining with anti-CEP antibody (Figure 10A). The presence of HH at concentrations higher than 100 μM caused a decrease in red fluorescence to almost the basal level (Figure 10B). Thus, HH in the cell culture medium at concentrations above 100 μM efficiently prevents the generation of CEP derivatives consequent to adduction of HOHA lactone with proteins and/or ethanolamine phospholipids.

Figure 10.

HH prevents adduction of HOHA lactone to the primary amino groups in protein lysyl residues and ethanolamine phospholipids to form CEP derivatives in ARPE-19 cells. Cells were exposed to 10 μM HOHA lactone for 24 h in the presence or absence of HH and then immunostained with primary rabbit anti-CEP polyclonal antibody/secondary goat anti-rabbit Texas Red-X antibody. (A) Images were acquired using a Leica DMI 6000 B inverted fluorescent microscope. (B) Red fluorescence intensity was quantified as the CEP level. The figure is representative of four independent experiments that showed very similar results. The data are presented as mean ± SEM; ***P<0.001 compared with the value for HOHA lactone treatment in the absence of HH.

Efficacy of Nα-acyl HH derivatives as HOHA lactone scavengers to prevent HOHA lactone cytotoxicity to ARPE-19 cells.

To assess the influence of lipophilicity on the scavenging of HOHA lactone by HH derivatives, we prepared a series of Nα-acyl HH derivatives, including straight chains (n = 0, 8 – 11) and branched chains or rings, by reacting L-histidine methyl ester with either acyl chlorides or anhydrides followed by nucleophilic substitution of the ester alkoxyl group by hydrazine (Figure 1).

AcHH (C2), tBocHH (C5) and tBuHH (C5), like HH are water soluble and were dissolved in phosphate buffered saline. However, cPHH (C6), cHHH (C7), nHH (C9), dHH (C10), uHH (C11) and DdHH (C12) are water-insoluble and were added to the PBS as solutions in DMSO. Derivatives with long aliphatic chains were only slightly soluble even in DMSO. Thus, to avoid exceeding the tolerable concentration of DMSO in the culture medium for the RPE cells, for the compounds with aliphatic chains longer than 9 carbons, 50 μM in PBS contained a ≈ 1% final concentration of DMSO.

To compare their abilities to prevent HOHA lactone induced cell death, RPE cells were incubated with 20 μM HOHA lactone and 50 μM of either HH or the Nα-acyl derivatives. Derivatives with carbon chains shorter than 7 demonstrated protection comparable to that of unmodified HH (Figure 11). This confirms the conclusion implicit in the structure of the HH-HOHA lactone adduct (Figure 8A) that the α-amino group in HH is not important for the protective efficacy of HH or its Nα-acyl derivatives. While longer and more hydrophobic aliphatic side chains at the N-terminus of the HH increased its lipophilicity, they also caused these derivatives to become cytotoxic, perhaps owing to the damage they can cause to the cell plasma and intracellular membranes.

Figure 11.

RPE cell viability protection by HH analogues. ARPE-19 cell viability (measured using MTT assay). Cells were exposed to HOHA lactone (20 μM) after pretreatment with 50 μM of HH or its Nα-acyl derivatives (cell viability was measured after 24 h). The data are presented as mean ± SEM of % control, n = 6; *P<0.05, ***P<0.001 compared with the HOHA lactone alone value.

To examine the ability of HH and its derivatives to protect the mitochondrial respiratory chain complexes against the toxicity of HOHA lactone, isolated ARPE-19 cell mitochondria were incubated with HOHA lactone both in the presence or absence of HH, AcHH, tBuHH or cHHH. HOHA lactone at 20 μM was chosen as an optimal concentration to reduce, but not completely inhibit the enzyme activity for complex I, while 50 μM was chosen for complex II and 100 μM was selected for complex IV.

Complex I specific activity of mitochondria exposed to 20 μM HOHA lactone was only protected by pretreatment with 400 μM (a 20x excess) of histidyl hydrazide derivatives except for cHHH that did not show significant protection at any concentration tested. At 400 μM, HH showed a 2.5-fold increase of complex I activity, AcHH showed a 1.5-fold and tBuHH showed a 2-fold increase, but cHHH did not provide any protection (Figure 12A). Nα-acylation did not improve the therapeutic efficacy of any of the derivatives relative to HH itself.

Figure 12.

Protective efficacy of HH and its Nα-acyl derivatives toward the activities of mitochondrial complexes I, II and IV against HOHA lactone toxicity. Isolated mitochondria were incubated with HH or its Nα-acyl derivatives for 10 min followed by HOHA lactone for 30 min and the specific enzymatic activities of complexes (A), complex II (B) and complex IV (C) were then measured spectrophotometrically (see Materials and methods for details). The data are presented as mean ± SEM of % control, n = 3; *P<0.05, **P<0.01, ***P<0.001 compared with the HOHA lactone alone value.

For mitochondria exposed to 50 μM HOHA lactone, significant protection of complex II activity was observed even by pretreatment with relatively low concentrations of the histidyl hydrazide derivatives except for cHHH. At 80 μM, HH, AcHH and tBuHH all demonstrated 90 – 100% protection of enzyme activity while cHHH did not provide any protection (Figure 12B). Presumably because HOHA lactone did not exert much inhibition of complex IV activity in mitochondria exposed to 100 μM HOHA lactone, all of the derivatives provided full or almost full protection against the toxicity of HOHA lactone against complex IV (Figure 12C). Even 1 mole % of HH, AcHH or cHHH relative to HOHA lactone was effective.

DISCUSSION

Mechanisms of HOHA lactone-induced mitochondrial dysfunction and oxidative stress.

The pathogenesis of AMD involves apoptosis of RPE cells followed by death of the underlying photoreceptors (36). Mitochondrial dysfunction is a factor in the development and progression of AMD. Evidence supporting this view includes loss in mitochondrial mass, disrupted cristae, ruptured membranes, decreased content of proteins in the electron transport chain and fragmented mitochondrial (mt)DNA in RPE cells from human donors with AMD. Mitochondria are responsible for aerobic respiration and ATP synthesis by oxidative phosphorylation. In addition to being crucial for energy production and metabolic pathways, they also play key roles in integrating cell death stimuli and executing the apoptotic program. Impaired function of the electron transport chain in the oxidative phosphorylation (OXPHOS) system can result in increased formation of reactive oxygen species and cause disturbances of energy metabolism (24, 25). Decreased ATP production leads to impairment of ATP-dependent processes, where all cellular functions are involved (37–39). Decrease of mitochondrial membrane potential is followed by opening of mitochondrial permeability transition pores (MPTPs) (40–42). Subsequent release of cytochrome c and other proapoptotic factors from the intermembrane space of mitochondria induces the formation of apoptosomes and consequently triggers activation of caspases and ultimately leads to apoptosis (39). The present study found that the cytotoxicity of HOHA lactone toward RPE cells includes the induction of mitochondrial dysfunction.

Previously, by the MTT, Alamar blue and LDH assays, we showed that HOHA lactone reduces ARPE-19 cell viability. The MTT assay measures mitochondrial succinate dehydrogenase (Complex II) activity (43). It shows the ability of HOHA lactone to affect mitochondrial metabolic activity. To further understand the effect of HOHA lactone on mitochondria, we studied the individual activities of OXPHOS complexes I to IV, mitochondrial ROS, cellular ATP level and mitochondrial membrane potential. The results of those studies provide mechanistic insights into HOHA lactone-induced damage of mitochondria that may lead to apoptosis or necrosis of RPE cells (Figure 13). By inducing an inhibition of complex I, HOHA lactone causes a decrease of electron flow through upstream sites that are prone to electron leakage. With electrons remaining at the site longer than normal, molecular oxygen can be reduced via a single electron transfer to produce superoxide that is released from the mitochondria. HOHA lactone-induced increased mitochondrial ROS generation was confirmed by MitoSOX Red reagent fluorescence. HOHA lactone also induced inhibition of complexes II and IV, contributing further to HOHA lactone-induced OXPHOS dysfunction and aberrant mitochondrial respiration. To prevent this, the F₀F₁-ATPase starts to function as a proton pump and hydrolyses ATP. The anticipated decrease of cellular ATP level induced by HOHA lactone was observed using the Celltiter-Glo assay. When severe damage of OXPHOS by HOHA lactone becomes irreversible, the ATPase can no longer maintain the mitochondrial membrane potential causing collapse of the ΔΨ_m that was detected by the JC-10 probe (Figure 4B). The increase of mitochondrial ROS levels gradually spreads to other cellular compartments and leads to a global increase of oxidative stress as was detected by CM-H₂DCFDA staining (Figure 3A).

Figure 13.

HOHA lactone depletes endogenous antioxidant GSH and impairs mitochondrial respiratory chain complexes and induces consequent ROS production, ATP depletion and mitochondrial membrane potential dissipation, leading to mitochondrial damage and cell death.

Dissipation of the mitochondrial membrane potential is believed to be followed by opening of mitochondrial permeability transition pores (40–42) and release of pro-apoptotic factors which then further trigger apoptosis and induce cell death, which could be intensified by the decrease of the endogenous cellular antioxidant GSH (39).

Prophylactic scavenging of lipid oxidation products.

We previously found that astrocytes derived from the glaucomatous optic nerve head are protected by pyridoxamine against pressure-induced oxidative protein modification by isoLGs, γ-ketoaldehyde electrophiles that are generated through free radical-induced oxidation of arachidonate (44). Pretreatment of mice with pyridoxamine prior to exposure to bright light reduces the levels of retinal adducts of isoLGs and morphological changes in photoreceptor mitochondria compared to untreated animals (45). In sharp contrast, we now find that pyridoxamine is ineffective in protecting ARPE-19 cells against the cytotoxicity of HOHA lactone (Figure 7 above and Figure 14). This is consistent with the finding that pyridoxamine is uniquely effective in scavenging γ-ketoaldehydes but not α,β-unsaturated aldehydes (46). Taurine and carnosine, primary amines that are present in μM concentrations in vivo, were similarly ineffective. In contrast, methoxyamine showed superior scavenging efficacy. This is presumably because, owing to the alpha effect (47, 48), methoxyamine is a “supernucleophile” that forms Schiff base adducts with aldehydes. Similarly, hydrazides are also supernucleophiles. FDA approved electrophile-trapping hydrazine derivatives, hydralazine and phenelzine protect brain mitochondrial function in vitro and in vivo following stroke or traumatic brain injury by scavenging the 4-HNE and acrolein (3). For brain mitochondria exposed to 4-HNE or acrolein, phenelzine prevents adduction of 4-HNE and acrolein with proteins, as detected by immunoblotting with the corresponding antibodies (3). Phenelzine dose-dependently improves cortical tissue sparing at 14 days after traumatic brain injury (TBI) presumably owing to mitochondrial protection by scavenging of aldehydic lipid peroxidation products (3, 49).

Figure 14.

Relative reaction rates of various nucleophilic scavengers with HOHA lactone.

Carnosine is a natural dipeptide consisting of β-alanine and histidine. Endogenously produced HNE-carnosine Michael adducts were detected in biological matrices, including the urine of Zucker obese rats (50) and oxidized skeletal muscle (51). Carnosine apparently competitively inhibits HNE-induced protein cross-linking (52) by quenching HNE in vivo. Its inhibition of α,β-unsaturated aldehyde-induced mitochondrial damage might involve a variety of mechanisms (53). It inhibits α,β-unsaturated aldehyde production by lipid oxidation and oxidative cleavage by scavenging ROS, such as hydroxyl radical (54–56), peroxyl radical (55, 57, 58), singlet oxygen (59) and nitric oxide (60). The imidazole ring provides antioxidant, metal ion chelating and buffering activity. Carnosine also scavenges α,β-unsaturated aldehydes (61–63) and reacts with and detoxifies carbonylated proteins (64–66). It is a bis-nucleophile that can trap the two electrophilic sites in α,β-unsaturated aldehydes. Because the histidyl nucleophile selectively undergoes Michael addition to the electrophilic C=C bond while the amino group of the β-alanyl moiety forms a Schiff base with the electrophilic carbonyl group, the imidazole ring of the histidyl moiety and primary amino group of the β-alanyl moiety act synergistically in trapping and inhibiting pathological consequences of α,β-unsaturated aldehyde formation. Protection of cells against the toxicity of HNE by L-carnosine was shown to benefit from the formation of a cyclic adduct between the imidazole and C=C bond and between the primary amino and carbonyl groups (61).

Cell viability assays previously suggested that histidyl hydrazide (HH) is more effective than carnosine and other histidine analogues in protecting neurons against HNE toxicity (67), and may have therapeutic potential for the treatment of stroke and some neurodegenerative conditions (68). However, its utility for ameliorating damage of RPE induced by reactive carbonyl species leading to AMD was unknown. We now find that HH is especially effective in alleviating the toxicity of HOHA lactone that causes mitochondrial dysfunction in RPE cells. It also prevents the depletion of glutathione that contributes to oxidative stress and the production of CEP derivatives that contribute to the pathogenesis of both “dry” and “wet” AMD. Two factors may contribute to the 20x faster reaction with HOHA lactone of histidyl hydrazide versus carnosine: (1) because hydrazides are supernucleophiles, they are more reactive than simple primary amines, and (2) the formation of an 11-member cyclic adduct is entropically favored over the formation of a 13-member cyclic adduct.

Prevention of RPE cell glutathione depletion.

To defend against toxic oxidative metabolites, especially reactive α,β-unsaturated aldehydes, such as HOHA lactone, 4-HNE, 4-HHE or acrolein, RPE cells detoxify them by Michael addition of endogenous GSH. However, the cellular level of GSH may be insufficient if large quantities of these aldehydes are generated of it there is an abnormally low level of GSH in the cells. GSH levels are depleted by the formation of HOHA lactone adducts. HH inhibits the depletion of GSH in ARPE-19 cells (Figure 9E). The supernucleophilicity of the hydrazide moiety favors capture of HOHA lactone aldehyde carbonyl by HH versus Michael addition of GSH with the HOHA lactone C=C bond. However, GSH could still be depleted by the formation of HOHA lactone-HH-GSH-bis-adducts (Figure 8 above and Figure 15). Preferential cyclization of the HH-HOHA lactone Schiff base adduct contributes to the efficacy of HH for preventing the depletion of GSH in ARPE-19 cells by bimolecular addition of GSH to the Schiff base adduct of HH with HOHA lactone. Rapid extracellular capture of HOHA lactone by HH is also expected to impede its diffusion into ARPE cells where it can form GSH adducts.

Figure 15.

The supernucleophilicity of HH favors rapid Schiff base formation. Subsequent cyclization is favored over bi-molecular addition of glutathione (GSH). Protonation of the resulting cyclic enamine delivers the final cyclic Schiff base adduct.

Incubation of HH with ARPE-19 cells resulted in a dose dependent uptake of this inhibitor (Figure 9). We postulated that increasing the lipophilicity of HH derivatives by appendage of an N-acyl group might enhance association with or permeation of ARPE-19 cell membranes, and result in more efficient interception of HOHA lactone before it could enter the cells, deplete GSH and damage proteins. The present study established that the α-amino group in HH is not essential for its efficacy for preventing GSH depletion and the cytotoxicity of HOHA lactone. Various Nα-acyl derivatives of HH (Ac-HH, Boc-HH, tBu-HH and cH-HH) retain full prophylactic α,β-unsaturated aldehyde scavenging activity that prevents mitochondrial and ARPE-19 cell damage by HOHA lactone (Figure 11). But highly lipophilic derivatives are toxic to ARPE-19 cells possibly indicating that these highly lipophilic HH derivatives disrupt cellular membranes.

The toxicity of α,β-unsaturated aldehydes such as HOHA lactone, HNE and acrolein might include protein-protein cross linking. We previously reported that HOHA lactone is metabolized by RPE cells forming HOHA lactone-GSH that is then reduced to an alcohol, 5,7-dihydroxyheptanoic acid lactone-GSH by aldose reductase (7). However, deficiency of aldose reductase or the NADPH cofactor might lead to incomplete metabolism and stop at HOHA lactone-GSH, which still possesses an aldehyde functional group and retains the potential to form Schiff base cross links with amino groups. By scavenging carbonylated proteins, e.g. Michael adducts of protein histidyl or cysteinyl groups, HH can prevent their involvement in the formation of protein-protein crosslinks. Now we discovered that HH can also scavenge HOHA lactone-GSH and interfere with its ability to form Schiff-base cross links.

HH protects RPE cells against HOHA lactone-induced mitochondrial impairment.

The retina contains an abundance of polyunsaturated fatty acyls, especially DHA phospholipids that are extremely susceptible to light-induced peroxidation. Products of lipid oxidation include α,β-unsaturated aldehydes HOHA lactone, 4-HHE and HNE. In the present study, we showed that HH can protect RPE cells against HOHA lactone-induced cell death. By scavenging HOHA lactone, it prevents depletion of cytosolic GSH and the formation of CEP derivatives of proteins or ethanolamine phospholipids in RPE cells. HH can prevent HOHA lactone-induced: (1) impairment of the activities of the mitochondrial respiratory chain complexes I, II and IV, (2) decrease of ATP levels and loss of mitochondrial membrane potential and (3) increase of mitochondrial ROS generation. Therefore, HH has potential therapeutic utility for preventing oxidative stress-induced retinal damage involving the generation and toxicity of α,β-unsaturated aldehydes, such as HOHA lactone, that damage RPE cells contributing to the pathogenesis of AMD and other retinal disorders.

However, HH is a highly hydrophilic compound. To protect RPE cells against exogenous HOHA lactone that diffuses into or is generated endogenously in RPE cells, it is necessary for HH to diffuse efficiently into the cells. To be effective as an intracellular inhibitor it must penetrate the RPE cell membrane. Therefore, the present study tested the hypothesis that more lipophilic N-acylated derivatives of the histidyl amino group might possess superior protective activity against α,β-unsaturated aldehydes such as HOHA lactone. We found that certain Nα-acyl HH derivatives retain full prophylactic α,β-unsaturated aldehyde scavenging activity that protects RPE cells against HOHA lactone toxicity. However, derivatives with acyl chains longer than 7 carbons are cytotoxic to the cells, while compounds with side chain shorter than 7 showed protection comparable to HH against cell death induced by HOHA lactone, and only Nα-acyl derivatives of HH with 5 carbons or less protected mitochondrial OXPHOS complexes I, II and IV as effectively as HH.

CONCLUSIONS AND FUTURE PROSPECTS

The present study tested the hypothesis that mitochondrial dysfunction may contribute to HOHA lactone-induced damage of the RPE found in AMD. It showed that HOHA lactone inhibits mitochondrial respiration and leads to (1) electron leakage resulting in increasing oxidative stress and (2) imbalance of the charge distribution resulting in collapse of the ΔΨ_m and accelerating the ATP consumption. The dissipation of ΔΨ_m triggers apoptosis and induces cell death, which is intensified by depletion of the cellular intrinsic antioxidant GSH. Scavenging products of lipid oxidation is a strategy, which might complement or supplant antioxidant therapy, to protect RPE cells against lipid peroxidation product-induced damage. We previously found that scavenging γ-ketoaldehyde products of lipid oxidation with pyridoxamine protects mitochondria against light-induced oxidative injury in vivo. However, pyridoxamine is ineffective against the toxicity of HOHA lactone. The present study supports the possibility that HH in the interphotoreceptor matrix can intercept exogenous HOHA lactone, e.g., the one that is shed from oxidatively damaged photoreceptor disk membranes. To efficiently intercept HOHA lactone generated endogenously in RPE cells and prevent GSH depletion and mitochondrial dysfunction, HH must readily diffuse into the RPE cell cytosol. The present study established that the α-amino group in HH is not essential for its protective activity against the HOHA lactone. Nα-acyl derivatives of HH, e.g., Ac-HH, boc-HH, tBu-HH and cH-HH are equally, but not more, effective as HH in protecting ARPE-19 cells against HOHA lactone toxicity (Figure 12). These Nα-acyl HH derivatives retain full prophylactic α,β-unsaturated aldehyde scavenging activity that protects mitochondria against HOHA lactone toxicity, but more lipophilic Nα-acyl HH derivatives are cytotoxic. Perhaps most important for improving the scavenging efficacy of HH, concentrations as high as 1 mM did not show any intrinsic cytotoxicity. This may allow the use of higher concentrations for prophylactic applications of HH in vivo.

Highlights.

• HOHA lactone is toxic to mitochondrial complexes I, II and IV in RPE cells

• HOHA lactone causes loss of mitochondrial membrane potential in RPE cells

• Histidyl hydrazide traps HOHA lactone preventing mitochondrial dysfunction

• Histidyl hydrazide traps HOHA lactone preventing GSH depletion and oxidative stress

ABBREVIATIONS

AMD

age-related macular degeneration

BCA

bicinchoninic acid

BSA

bovine serum albumin

CEP

carboxyethylpyrrole

DAPI

4,6-diamidino-2-phenylindole

DHA

docosahexaenoate

DMEM

Dulbecco’s modified Eagle’s medium

FBS

fetal bovine serum

GSH

glutathione

GST

glutathione S-transferase

HOHA

4-hydroxy-7-oxo-5-heptenoate

LDH

lactate dehydrogenase

NAD(P)H

dehydrogenase quinone 1

PBS

phosphate-buffered saline

RPE

retinal pigmented endothelium

JC-10

a 5,5’,6,6’-tetrachloro-1,1’,3,3’-tetraethylbenzimidazoylcarbocyanine iodide derivative

CM-H₂DCFDA

carboxymethyl-H₂-dichlorofluorescein diacetate