4-Hydroxy-7-oxo-5-heptenoic Acid (HOHA) Lactone is a Biologically Active Precursor for the Generation of 2-(ω-Carboxyethyl)pyrrole (CEP) Derivatives of Proteins and Ethanolamine Phospholipids

Abstract

2-(ω-Carboxyethyl)pyrrole (CEP) derivatives of proteins were previously shown to have significant pathological and physiological relevance to age-related macular degeneration, cancer and wound healing. Previously, we showed that CEPs are generated in the reaction of ε-amino groups of protein lysyl residues with 1-palmityl-2-(4-hydroxy-7-oxo-5-heptenoyl)-sn-glycero-3-phosphatidylcholine (HOHA-PC), a lipid oxidation product uniquely generated by oxidative truncation of docosahexanenate-containing phosphatidylcholine. More recently, we found that HOHA-PC rapidly releases HOHA-lactone and 2-lyso-PC (t_1/2 = 30 min at 37 °C) by non-enzymatic transesterification/deacylation. Now we report that HOHA-lactone reacts with Ac-Gly-Lys-OMe or human serum albumin to form CEP derivatives in vitro. Incubation of human red blood cell ghosts with HOHA-lactone generates CEP derivatives of membrane proteins and ethanolamine phospholipids. Quantitative analysis of the products generated in the reaction HOHA-PC with Ac-Gly-Lys-OMe showed that HOHA-PC mainly forms CEP-dipeptide that is not esterified to 2-lysophosphatidycholine. Thus, the HOHA-lactone pathway predominates over the direct reaction of HOHA-PC to produce the CEP-PC-dipeptide derivative. Myleoperoxidase/H₂O₂/NO₂⁻ promoted in vitro oxidation of either 1-palmityl-2-docosahexaneoyl-sn-glycero-3-phosphatidylcholine (DHA-PC) or docosahexaenoic acid (DHA) generates HOHA-lactone in yields of 0.45% and 0.78%, respectively. Lipid oxidation in human red blood cell ghosts also releases HOHA-lactone. Oxidative injury of ARPE-19 human retinal pigmented epithelial cells by exposure to H₂O₂ generated CEP derivatives. Treatment of ARPE-19 cells with HOHA-lactone generated CEP-modified proteins. Low (submicromolar), but not high, concentrations of HOHA-lactone promote increased vascular endothelial growth factor (VEGF) secretion by ARPE-19 cells. Therefore, HOHA-lactone not only serves as an intermediate for the generation of CEPs but also is a biologically active oxidative truncation product from docosahexaenoate lipids.

INTRODUCTION

Our basic research on the chemical biology of lipid oxidation led to the discovery of 2-(ω-carboxyethyl)pyrrole (CEP)-modifications of proteins and ethanolamine phospholipids.¹ Using anti-CEP antibodies raised against CEP-protein,² CEP immunoreactivity was first detected in vivo in photoreceptor rod outer segments and in retinal pigmented epithelium. CEP derivatives were found in plasma² from individuals with age-related macular degeneration (AMD) and in drusen from AMD donor eye tissues,³ in brain tissue from autistic children,⁴ in human melanoma skin tissue and healing wounds.⁵ CEP-protein, CEP-peptide and CEP-ethanolamine phospholipids promote vascular angiogenesis and platelet activation through Toll-like receptor-2⁵ and Toll-like receptor-9⁶ dependent signaling pathways, respectively. In addition, immunization of mice with CEP-modified mouse serum albumin induces AMD-like lesions in their retinas.⁷ Such important physiological and pathological activities tightly link CEPs to human diseases and oxidative stress.

The generation of CEP-protein derivatives 4 (Scheme 1) results from an oxidative cleavage of docosahexaenoic acid (DHA)-containing phospholipids, such as 1-palmityl-2-docosahexaenoyl-sn-glycero-3-phosphocholine (DHA-PC, 1). Oxidative cleavage might occur after PLA2-catalyzed hydrolysis to DHA (2) to generate 4-hydroxy-7-oxo-hept-5-eonate (HOHA, 3) that produces CEP-protein derivatives 4 through condensation with the primary amino groups of protein lysyl residues. Release of free fatty acid from membrane phospholipids under catalysis of phospholipase (PL) A2 occurs during oxidative stress.⁸ Alternatively, oxidative cleavage of DHA-PC (1) might occur first to deliver 1-palmityl-2-(4-hydroxy-7-oxo-5-heptenoyl)-sn-glycero-3-phosphatidylcholine (HOHA-PC, 5). Condensation of this oxidatively-truncated phospholipid with the primary amino groups of protein lysyl residues could be followed by PLA2-catalyzed hydrolysis of an intermediate CEP-PC-protein derivative 6 to deliver 4.^{9, 10} In vitro experiments demonstrated that CEP-PC derivatives could be generated during autoxidation of DHA-PC in the presence of protein.

Scheme 1.

Another plausible pathway for the generation of CEPs was suggested by our finding that HOHA-PC (5) undergoes spontaneous deacylation with a half-life of 30 minutes under physiological conditions, i.e., 37 °C and pH 7.4.¹¹ Intramolecular transesterification of HOHA-PC (5) generates 2-lyso-PC and a five-member ring lactone aldehyde, HOHA-lactone (7, Scheme 1). It seemed reasonable to expect that HOHA-lactone would react with primary amino groups in biomolecules to generate CEPs. We now report that HOHA-lactone readily forms CEP derivatives by reaction with Ac-Gly-Lys-OMe, proteins and ethanolamine phospholipids. We also find that HOHA-lactone is generated through autoxidation not only of DHA-PC but also of unesterified DHA. Thus, the equilibrium between lactone 7 and γ-hydroxy acid 3 strongly favors the lactone even in an aqueous environment. Evidence is also presented showing that deacylation of HOHA-PC (5) can predominate over its reaction with primary amines, and therefore the production of CEPs from HOHA-PC in vivo can occur entirely nonenzymatically, e.g., without the need for phospholipase activity that is required to convert CEP-PC-protein (6) into CEP-protein (4).

EXPERIMENTAL PROCEDURES

Materials

Human serum albumin (HSA, fraction V), DHA and Phospholipase D Streptomyces Chromofuscus (PLD) were purchased from Sigma (St. Louis, MO). The Ac-Gly-Lys-OMe dipeptide was purchased from Bachem (Torrance, CA). DHA-PC and LysoPC (13:0) (1-tridecanoyl-2-hydroxy-sn-glycero-3-phosphocholine) were purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). CEP-dipeptide (7) and CEP-lysine,¹² HOHA-PC and HOHA-lactone,¹³ mouse monoclonal anti-CEP antibody,¹⁴ CEP-ethanolamine (14) and d₄-CEP-ethanolamine,¹ were prepared as described previously. CEP-PC-dipeptide (11) was prepared through a Paar-Knorr pyrrole synthesis as described in the Supporting Information. 2,2'-Azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt (ABTS) was from Thermo Scientific (Rockford, IL). Dulbecco’s modified Eagle’s medium, high glucose (DMEM/HG), fetal bovine serum, penicillin–streptomycin and L-glutamine were obtained from Gibco (Life Technologies, Grand Island, NY). 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) was obtained from Invitrogen. ARPE-19 cells were obtained from American Type Culture Collection (CRL-2302; ATCC, Rockville, MD). A human VEGF-A ELISA kit was purchased from Piercenet (Thermo Fisher Scientific, Rockford, IL).

Reaction of Ac-Gly-Lys-OMe with HOHA-PC or HOHA-lactone

Ac-Gly-Lys-OMe (2 mM) was incubated with either HOHA-PC (0.2 mM) or HOHA-lactone (1 mM) in a binary solution mixture of phosphate buffered saline (PBS) and methanol (9:1, v/v). Reaction vials were blanketed with argon, and were incubated at 37 °C. For the reaction with HOHA-PC, 50 µL aliquots of the reaction product mixture were withdrawn after various reaction times. Methanol (200 µL), internal standards CEP-lysine (50 ng) and lysoPC (13:0) (50 ng) were added. The above solution (10 µL) was injected into the LC-MS/MS for analysis. CEP-dipeptide and CEP-PC-dipeptide concentrations were determined using calibration curves that were obtained by using pure CEP-dipeptide and CEP-PC-dipeptide standards at various known concentrations spiked with 50 ng of CEP-lysine and lysoPC (13:0). For the reaction of Ac-Gly-Lys-OMe with HOHA-lactone, 20 µL aliquots of the reaction mixture were withdrawn after various reaction times. Methanol (200 µL) and 50 ng of CEP-lysine internal standard were added and the resulting solution (10 µL) was injected into HPLC-ESI/MS/MS for analysis.

Reaction of human serum albumin with HOHA-lactone

A sterile solution of human serum albumin (HSA, 1 mg/mL) was incubated with 0.5 mM HOHA-lactone in 1 mL PBS for up to 3 days at 37 °C. Aliquots of the reaction mixture were withdrawn at various times and then quenched by four dialyses against 1 L of 1× PBS. The final protein concentration after dialysis was determined by BCA protein assay (Piercenet, Rockford, IL). An aliquot (10 µL) of each modified HSA-protein was loaded onto an SDS PAGE gel and then analyzed by either western blot with monoclonal mouse anti-CEP antibody or by Coomassie blue staining. For western blot analysis, proteins were transferred to a nitrocellulose membrane (0.45 µ). The membrane was subsequently blocked with 5% BSA-TBST. The CEP modifications on proteins were detected with 0.2 µg /mL monoclonal anti-CEP antibody and 1: 10,000 diluted HRP conjugated goat anti-mouse antibody. The membrane was developed with SuperSignal West Pico Chemiluminescent substrate solution (Thermo Scientific, Rockford, IL) and was then exposed to X-ray film.

Incubation of human erythrocyte ghosts with HOHA-lactone

Human erythrocyte ghosts in PBS (1.15 mg/mL protein) were prepared by hemolysis of packed human red blood cells in isotonic solution as described previously.¹⁵ Then, 1 mL of erythrocyte ghosts were incubated with 1 mM HOHA-lactone at 37 °C. Untreated erythrocyte ghosts that were incubated at 37°C for three days were used as control. Aliquots (200 µL) withdrawn after 1 day and 3 days as well as a control sample (200 µL) were extracted with chloroform/methanol by the Bligh & Dyer method to separate proteins and phospholipids.¹⁶ The lower layer fractions containing phospholipids were dried under a nitrogen gas stream, and then re-dissolved in diethyl ether (1 mL). Then 500 Units of PLD in MES buffer (250 µL, pH 5.5) supplemented with 15 mM of CaCl₂ were added. The biphasic solution was shaken vigorously overnight at 37 °C. Then diethyl ether was evaporated under a stream of nitrogen. The residue was mixed with 1 mL of methanol and then centrifuged at 10,000 × g for 10 min. The supernatant was dried under nitrogen, and was redissolved in 100 µL of methanol. Deuterium labeled CEP-ethanolamine (d₄-CEP-ETN, 50 ng) was spiked into the solution as internal standard. 20 µL of solution was injected into the LC-MS for analysis.

Protein disks obtained during the Bligh & Dyer extraction were resuspended in 200 µL of PBS, mixed with 200 µL of 2× Laemelli SDS sample buffer, and then heated at 95 °C for 5 min. An aliquot (30 µL) of the protein solution (approximately 15 µg of protein) was separated by SDS PAGE gel electrophoresis followed by transfer of the proteins to a PVDF membrane. CEP modified proteins were then analyzed by western blot as described above.

Myeloperoxdase/H₂O₂/NO₂⁻ induced autoxidation of DHA-PC vesicles, DHA and human erythrocyte ghosts

DHA-PC vesicles in PBS (1 mL of 1.24 mM solution) were prepared by passing hydrated DHA-PC through a 0.1 µm polycarbonate filter 20 times using an Avanti Mini-Extruder Set (Avanti Polar Lipids, Inc., Alabaster, AL). DHA (2 mM) was mixed with 1 mL of PBS, and then the mixture was sonicated for 5 min to obtain a homogeneous solution. Human erythrocyte ghosts in PBS (1.15 mg/mL in protein concentration) were used without further treatment for MPO/H₂O₂/NO₂⁻ system-induced oxidation. DHA-PC vesicles, DHA or human erythrocyte ghosts in PBS solution (1.0 ml) were incubated in air at 37 °C in the presence of 30 nM myeloperoxidase (MPO), 100 ng/ml glucose oxidase, 100 µg/ml glucose and 500 µM NaNO₂. Aliquots (50 µL) of the reaction mixture were withdrawn after various reaction times, and were extracted three times with 200 µL of ethyl acetate. Organic layers were pooled and evaporated to dryness under a stream of nitrogen. The residues were stored under −80 °C until analysis.

Pentafluorobenzyl oxime derivatization of HOHA-lactone (HOHA-lactone-PFBHA)

Reaction mixtures from MPO/H₂O₂/NO₂⁻ promoted lipid oxidation were derivatized as described previously¹⁷ with pentafluorobenzyl hydroxylamine (PFBHA) for LC-MS analysis. In brief, the residue obtained from 50 µL of oxidation reaction mixture was treated for 1 h with 0.05 M pentaflurobenzyl hydroxylamine hydrochloride aqueous solution (pH was adjusted to 7.0 by 0.1 N NaOH). Then 500 µL methanol and 2.5 mL hexane were added. After adding 6 drops of concentrated sulfuric acid, the mixture was immediately vortex vigorously and centrifuged for 5 min at 3,000 × g. The upper organic layer was collected and evaporated to dryness under a stream of nitrogen. The residue was re-dissolved in methanol (100 µL). An aliquot of this solution (10 µL) was injected into an LC-MS for analysis. Authentic HOHA-lactone-PFB oxime was prepared in the same way by incubating a sample of pure HOHA-lactone with PFB-HA hydrochloride in methanol.

High Performance liquid chromatography/mass spectrometry

Chromatographic separation was achieved with a Waters Alliance 2690 HPLC system (Waters, Milford, MA) equipped with a Luna C18(2) column (2.0 mm i.d. × 150 mm length, 5 µm, Phenomenex). Mobile phase A consisted of HPLC grade water containing 0.1% formic acid. Mobile phase B was HPLC grade methanol containing 0.1% formic acid. For analysis of Ac-Gly-Lys-OMe after treatment with HOHA-PC or HOHA-lactone, the total run time was 36 min. HPLC gradient steps were as follows: 0–15 min, linear gradient from 2 to 100% solvent B; 15–25 min, isocratic at 100% solvent B; 25–26 min, linear gradient from 100 to 2% solvent B; 26–36 min, isocratic at 2% solvent B. The flow rate was 200 µL/min. For analysis of PLD-digested RBC ghost phospholipids, the total run time was 40 min. The HPLC gradient steps were set as follows: 0–5 min, isocratic at 5% solvent B; 5–22 min, linear gradient from 5 to 100% solvent B; 22–30 min, isocratic at 100% solvent B; 30–31 min, linear gradient from 100 to 5% solvent B; 31–40 min, isocratic at 5% solvent B. The flow rate was 200 µL/min. For analysis of HOHA-lactone-PFB oxime, the total run was 25 min. HPLC gradient steps were set as follows: 0–7 min, linear gradient from 25 to 100% solvent B; 7–15 min, isocratic at 100% solvent B; 15–16 min, linear gradient from 100 to 25 % solvent B; 16–25 min, isocratic at 25% solvent B. The flow rate was 200 µL/min.

The analytes were detected with a Quattro Ultima triple-quadrupole mass spectrometer (Micromass, Wythenshawe, UK) operated in the positive ion mode. The source temperature was maintained at 120, the desolvation temperature was kept at 250 °C, the drying gas (N₂) was maintained at ca. 450 L/h, the cone gas flow was kept at ca. 70 L/h, and the multiplier was set at an absolute value of 600. Optimized parameters for detecting each analyte were determined using authentic samples. MS scans from 50–1000 were obtained for standard compounds. For multiple reaction monitoring (MRM) experiments, argon was used as collision gas at a pressure of 5 psi. The optimum collision energy and other parameters were determined for each individual analyte. (Supporting information Table S1)

¹H-NMR analysis of HOHA-lactone incubated in aqueous buffer under physiomimetic conditions

To evaluate the stability of HOHA-lactone under physiological conditions, 2.5 mg of pure HOHA-lactone was dissolved in 725 µL of D₂O-reconstituted phosphate buffered saline solution (pH 7.4), HOHA-lactone solution was then transferred into a NMR tube and incubated in an oven set at 37 °C. ¹H NMR data was collected on Varian Inova AS400 spectrometer operating 400 MHz after 0, 12 and 24 h of incubation.

Cell culture

Human retinal pigmented epithelial cells (ARPE-19) were cultured in complete DMEM/F12 medium supplemented with 10% FBS, and allowed to attach to culture plates in a humidified CO₂ incubator at 37 °C and 5% CO₂overnight. All studies were conducted by using ARPE-19 cells of passages 20–40 following 12 h quiescence in DMEM/ F12 serum-free medium. All the experiments were repeated two to three times with n = 8 in each experiment for cell viability studies and VEGF secretion, and n = 3 for measurements of metabolites in cell lysates and extracellular medium by either LC-MS or GC-MS. For metabolism studies, 1×10⁶ of ARPE-19 cells/plate were grown in a 60 mm culture dish and washed with PBS, then 2 mL of PBS containing HOHA-lactone were added and incubated at 37 °C for up to 2 h. For cell viability assay and VEGF assay, 25,000 cells/well were grown in 96 well plates with 200 µL culture medium per well before analysis.

Cell viability assay (MTT assay)

ARPE19 cells (25,000 cells/ per well) were seeded in a 96-well flat bottom plate in 200 µl of complete DMEM/ HG 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 basal DMEM/ HG medium and starved in 180 µl of basal DMEM/ HG medium for 4–5 hours. To these cells (8 replicate wells were used for each concentration), 20 µl of stock solutions of HOHA-lactone in PBS were added to create 0 to 50 µM final concentrations of HOHA-lactone followed by 2-hour incubation in a humidified CO₂ incubator at 37 °C and 5% CO₂. The plates were then centrifuged at 1,000g for 5 min and the medium was aspirated from each well and the cells were washed with DMEM/HG medium three times and then incubated for 4 h at 37 °C with 20 µL of filter-sterilized 3-[4,5–dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) solution (5 mg/ml in PBS) plus 180 µL of DMEM/HG medium. The plates were then centrifuged at 1,000g for 5 min again and the medium was aspirated from each well. Dimethylsulfoxide (DMSO) 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 formazan solutions were measured by using a plate reader (Model M3, Molecular Device) set at λ = 540 nm and a reference wavelength set λ = 620 nm.

Measurement of secreted VEGF-A from ARPE-19 cells

After starving for 12 hours, ARPE-19 cells grown in 96-well plates (25,000 cells/well) were incubated with HOHA-lactone (0–25 µM) dissolved in 200 µL of serum-free DMEM /HG medium. After 12 h, supernatants were collected to estimate secreted VEGF using an enzyme-linked immunosorbent assay (ELISA) according to the manufacturers protocol (Human VEGF-A ELISA kit, Thermo Scientific, Rockford, IL). Briefly, 100 µL of supernatant samples, recombinant VEGF standards, and blank samples were added to an ELISA plate, which was pre-coated overnight with the coating antibody (anti-human VEGF-A antibody) and blocked 1 h with 4% BSA in PBS. After incubation for 1 h at room temperature, the plates were aspirated and washed three times with 300 µL per well of PBS with 0.05% Tween-20 (wash buffer). Then 1:100 diluted anti-human VEGF-A detection antibody in 4% BSA in PBS was added to each well and incubated for 1 h at room temperature. Then the plate was washed three times with wash buffer described above, and 100 µL of 1:400 diluted Streptavidin-HRP reagent was added to each well. After incubation at room temperature for 30 min, the plate was washed three times with wash buffer. Substrate solution (100 µL) was added to each well and incubated for 20 min until the reaction was stopped with 0.16 M sulfuric acid. All the absorbances were measured using a microplate reader with a test wavelength of 450 nm and a reference wavelength set at 550 nm.

3-(Furan-2-yl)propanoyl amide 10a

N-hydroxysulfosuccinimide (80mg, 0.37mmol) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (88mg, 0.46mmol) were added into 6.0 ml of anhydrous dimethylformamide solution containing 2-furanpropanoic acid (60.0 mg, 0.43 mmol). After stiring for 3.0 hour at room temperature under Ar protection, the reaction was filtered and 2.0 ml DMF was used to wash the filter three times. The filtrate was added to a solution of methyl 6-amino-2-((2-acetylamino)acetyl)amino)hexanoate (Ac-Gly-Lys-OMe, 78 mg, 0.30 mmol) with N,N-diisopropylethylamine (100.0 µl, 0.76 mmol) in the 3.0 ml anhydrous DMF at room temperature. The resulting mixture was stirred for another 2.0 h. Then ddH₂O was added to the reaction mixture and 5N HCl was used to adjust the pH to 2–3. The mixture was extracted with ethyl acetate three times. The organic phase was collected and dried with anhydrous sodium sulfate. After the removal of solvent by rotarary evaporation, the crude product was purified by silica gel chromatography to give the target compound (80.0 mg, 70 %, R_f = 0.30, 6.3 % methanol in chloroform). ¹H NMR (400M Hz, CD₃OD): 7.99 (broad, 1 H), 7.35 (dd, J = 0.8, 1.6 Hz, 1 H), 6.28 (dd, J = 2.0, 2.8 Hz, 1H), 6.04 (dd, J = 0.8, 3.2 Hz, 1 H), 4.41 (dd, J = 5.2, 9.2 Hz, 1 H), 3.90 (dd, J = 16.4, 28.8 Hz, 2 H), 3.71 (s, 3 H), 3.16 (dd, J = 6.4, 12.4 Hz, 2 H), 2.92 (t, J = 7.6 Hz, 2 H), 2.49 (t, J = 7.6 Hz, 2 H), 2.00 (s, 3 H), 1.82 (m, 1 H), 1.69 (m, 1 H), 1.47 (m, 2 H), 1.34 (m, 2 H). ¹³C NMR (101 MHz, CD₃OD): δ 173.55, 172.79, 172.53, 170.45, 154.53, 141.19, 109.98, 105.09, 52.38, 51.53, 42.10, 38.90, 34.36, 30.93, 28.62, 23.95, 22.84, 21.24. ESI-MS: m/z calcd for C₁₈H₂₆N₃O₆ [M-H]⁻, 380.18; found 380.73; calcd for C₁₈H₂₈N₃O₆ [M+H]⁺382.20, found 382.60.

LC-MS comparison of furan-dipeptide 10a and CEP-dipeptide 10

ESI mass spectrometry was performed on a Thermo Finnigan LCQ Deca instrument in the positive ion mode using nitrogen as the sheath and auxiliary gas. The Surveyor LC system was equipped with a Luna C18 (2) column (2.0 mm i.d. × 150 mm length, 5 µm, Phenomenex). Mobile phase A consisted of HPLC grade water containing 0.1% formic acid. Mobile phase B was HPLC grade methanol containing 0.1% formic acid. The total run time was 36 min. HPLC gradient steps were as follows: 0–15 min, linear gradient from 2 to 100% solvent B; 15–25 min, isocratic at 100% solvent B; 25–26 min, linear gradient from 100 to 2% solvent B; 26–36 min, isocratic at 2% solvent B. The flow rate was 200 µL/min. The heated capillary temperature was 300 °C, the source voltage was 4.5 kV, and the capillary voltage was 31.00 V. The data were processed with Qual browser in Xcalibur software.

RESULTS

HOHA-lactone (7) reacts with Ac-Gly-Lys-OMe (8) to form CEP-dipeptide 9

Incubation of HOHA-lactone (7) with the dipeptide Ac-Gly-Lys-OMe (8) in pH 7.4 phosphate buffered saline (PBS) at 37 °C generated CEP-dipeptide 10 (Scheme 2) that was analyzed by reverse phase HPLC/ESI/MS/MS using multiple reaction monitoring (MRM) in the positive ion mode. A MRM method was developed based on the collision induced dissociation spectrum for authentic CEP-dipeptide 10 (Supporting Information Figure S1). Four representative daughter ions, m/z 206.3, 265.3, 283.0 and 322.2, were chosen as MRM transitions for fragmentations of the parent ion m/z 382.3 for CEP-dipeptide analysis. HOHA-lactone treated dipeptide after 24 h has a major peak in LC-MS/MS spectrum at 12.24 min in all four MRM channels, as does the authentic CEP-dipeptide (Figure 1A, B). The peak that elutes at 9.37 min in the transitions 382.3 → 283.0 and 382.3 → 206.3 might correspond to the isobaric Schiff base adduct (9) or 3-(furan-2-yl)propanoyl amide (10a). An authentic sample of 10a was readily distinguished from CEP-dipeptide (10) by LC retention time (Figure S2). The concentration of CEP-dipeptide generated from the reaction of HOHA-lactone (7) with dipeptide (8) was calculated from the peak area of the transition 382.3→283.0 by comparison with a calibration curve generated from authentic CEP-dipeptide standard (Figure 2A). The profile for the evolution of CEP-dipeptide shows a time dependent increase (Figure 2B).

Scheme 2.

Figure 1.

HPLC/ESI/MS/MS analysis of CEP-dipeptide production in the reaction of HOHA-lactone with Ac-Gly-Lys-OMe dipeptide. (A) Authentic CEP-dipeptide standard. (B) Reaction product mixture after 24 hours incubation at 37 °C.

Figure 2.

(A) Calibration curve for CEP-dipeptide MRM transition 382.3→283.0 and (B) evolution profile of CEP-dipeptide generation in the reaction of HOHA-lactone with Ac-Gly-Lys-OMe.

CEP-dipeptide (10), instead of CEP-PC-dipeptide (11), is the major CEP derivative generated in the reaction between HOHA-PC (2) and Ac-Gly-Lys-OMe (8)

Previously, we found that reaction of HOHA-PC (5) with Ac-Gly-Lys-OMe (8) produced CEP-PC-dipeptide (11, Scheme 3) in 0.7% yield after 30 h incubation.⁹ Since we now know that HOHA-PC (5) produces to HOHA-lactone (7) under the reaction conditions, it seemed likely that CEP-dipeptide (10) would be generated through the intermediacy of 7 in addition to CEP-PC-dipeptide (11). Using HPLC/ESI/MS/MS analysis, we now find that both 10 and 11 are indeed cogenerated in the reaction of 5 with 8 after 24 h incubation at 37 °C (Figure 3). CEP-dipeptide (10) and CEP-PC-dipeptide (11) in reaction mixture were simultaneously measured by LC-MS/MS analysis using MRM. Quantification of concentrations was accomplished using CEP-lysine and 2-lyso-PC (13:0) as internal standards, the peak area ratio of the CEP-dipeptide MRM transition 382.3 → 206.3 to the CEP-lysine transition 269.2 → 84, and the CEP-PC-dipeptide MRM transition 859.6 → 184 to the 2-lyso-PC (13:0) transition 454 → 184. Calibration curves of each analyte were generated using authentic standards (Figure 4A). CEP-dipeptide was initially generated at the same rate as CEP-PC-dipeptide. But after the first two days, CEP-dipeptide generation was faster than CEP-PC-dipeptide (Figure 4B). The concentration of CEP-dipeptide continued to increase and it became the major CEP product after 3 days incubation. After five days incubation, the concentration of CEP-dipeptide was 2.36 ± 0.23 µM, nearly 3 times that of CEP-PC-dipeptide, 0.71 ± 0.10 µM.

Scheme 3.

Figure 3.

HPLC/ESI/MS/MS analysis of CEP-dipeptide and CEP-PC-dipeptide adducts. All chromatograms were monitored by LC-MS in the positive ion mode with MRM of appropriate mass transitions as noted. (A) Authentic CEP-dipeptide. (B) Authentic CEP-PC-dipeptide. (C) CEP-dipeptide produced from the reaction mixture of HOHA-PC and Ac-Gly-Lys-OMe dipeptide for 24 h incubation. (D) CEP-PC-dipeptide produced from the reaction mixture of HOHA-PC and Ac-Gly-Lys-OMe dipeptide for 24 h incubation. (E) Internal standard, CEP-lysine. (F) Internal standard, lysoPC (13:0).

Figure 4.

(A) Calibration curve for CEP-dipeptide and CEP-PC-dipeptide generated by peak area ratio of each analyte to internal standard. (Peak area ratio of MRM transition 382.3 → 206.3 to 269.2 →84.0 for CEP-dipeptide, 859.6 → 184.0 to 454.0 → 184.0 for CEP-PC-dipeptide) (B) Evolution profile of CEP adducts generation in the reaction mixture of HOHA-PC and Ac-Gly-Lys-OMe dipeptide and the yield of CEP-dipeptide generated from spontaneous hydrolysis of 20 µM CEP-PC-dipeptide in aqueous solution.

After four days, the CEP-PC-dipeptide concentration reached a maximum of 0.94 ± 0.11 µM and then its concentration started to decrease owing presumably to non-enzymatic hydrolysis. Nevertheless, the majority of non-esterified CEP-dipeptide was apparently generated through the HOHA-lactone pathway. To explore the possibility that non-enzymatic hydrolysis of CEP-PC-dipeptide accounts for the majority of CEP-dipeptide generation, we determined the yield of CEP-dipeptide generated upon incubation of 20 µM CEP-PC-dipeptide (Figure 4B). CEP-PC-dipeptide only generated CEP-dipeptide in 17.6% yield after incubation for 5 days. Thus, the concentration of CEP-dipeptide produced by spontaneous hydrolysis of CEP-PC-dipeptide observed at 5 days is expected to be at most 0.35 µM based on the maximum concentration of CEP-PC-dipeptide produced in the reaction of the HOHA-PC with dipeptide 8. This only accounts for 15% of the total CEP-dipeptide detected in latter reaction mixture. This confirms that the majority of CEP-dipeptide generated in the reaction of HOHA-PC with Ac-Gly-Lys-OMe is produced through the HOHA-lactone pathway. This suggests that HOHA-lactone is an important precursor of CEPs during oxidative damage in vivo.

HOHA-lactone forms CEP derivatives by reaction with proteins and amino phospholipids

The reactivity of HOHA-lactone towards biomolecules such as protein and ethanolamine phospholipids was tested by treating human serum albumin (HSA) and human red blood cell ghosts (RBC ghosts) with HOHA-lactone. Reaction of HSA in PBS with HOHA-lactone at 37 °C followed by removal of excess HOHA-lactone by dialysis generated CEP-protein. Thus, enzyme-linked immunosorbent assay (ELISA) of HSA protein after incubation with HOHA-lactone for 3 days exhibited significant CEP immunoreactivity to mouse monoclonal anti-CEP antibody, while control HSA showed no CEP immunoreactivity (Figure 5A).

Figure 5.

HOHA-lactone modifies biomolecules to form CEP adducts. (A) Ten µg per well of HOHA-lactone (0.5 mM) treated HSA protein (1 mg/mL) from 0 to 72 h incubation was either analyzed by western blot (Left) using monoclonal mouse anti-CEP IgG antibody or Coomassie blue staining (Right). (B) Western blot analysis of the protein fraction isolated from HOHA-lactone treated human RBC ghosts (incubation for 0, 24 and 72 h) by monoclonal anti-CEP mouse IgG antibody. (C) LC-MS/MS analysis of CEP-ethanolamine adducts in PLD digested phospholipids isolated from HOHA-lactone treated human RBC ghosts. A fixed amount of deuterium labeled CEP-ethanolamine (d₄-CEP-ETN) was added as internal standard.

Human erythrocyte ghosts were incubated with 1 mM HOHA-lactone in PBS for 0, 24 and 72 h and then were extracted with chloroform and methanol to isolate proteins and phospholipids. Immunoblot analysis of the protein fraction showed that HOHA-lactone formed CEP protein adducts in human erythrocyte ghosts in a time-dependent manner (Figure 5B). The phospholipid fraction was dried, and then hydrolyzed by phospholipase D (Streptomyces Chromofuscus). The presence of CEP modified ethanolamine phospholipids (CEP-EA-PL, 13, Scheme 4) was then determined by isotope dilution analysis of the resulting CEP-ethanolamine (CEP-EA, 14) by HPLC/ESI/MS/MS. MRM was performed in the positive ion mode monitoring the transition 184 → 124 for CEP-EA and 188 → 128 for d₄-CEP-EA standard, according to daughter ion scan of CEP-EA and d₄-CEP-EA, respectively (Supporting information Figure S2). LC-MS/MS showed that treatment of the RBC ghosts with HOHA-lactone for 24 and 72 h resulted in a significant peak that elutes at 15.47 min in the CEP-EA channel, at the same retention time as the internal standard d₄-CEP-EA, while the untreated RBC ghosts had no peak at this retention time (Figure 5C).

Scheme 4.

Oxidation DHA-containing phospholipidipids or DHA generates HOHA-lactone

DHA-containing phospholipids (DHA-PC vesicle), free docosahexaenoic acid (DHA), and human red blood cell ghosts were subjected to myleoperoxidase/H₂O₂/NO₂system-induced lipid oxidation. Measurement of HOHA-lactone in those lipid peroxidation products was achieved by detection of the derived pentafluorobenzyl oximes by LC-ESI-MS/MS (Scheme 5). MS scan spectrum of authentic HOHA-lactone-PFB oxime (11) in positive ion mode showed a strong peak at m/z 336, which corresponds to protonated HOHA-lactone-PFB oxime. Collision induced dissociation spectrum of m/z 336 with 25 eV collision energy showed three major daughter ions, m/z 181, 139 and 95 (Supporting Information Figures S3 and S4).

Scheme 5.

A selected reaction monitoring (SRM) of the transition 336→139 was used to analyze HOHA-lactone-PFB oxime. A typical LC-MS/MS chromatogram of authentic standard shows that HOHA-lactone-PFB oxime eluated at 12.34 min. DHA-PC vesicle, free DHA acid and RBC ghosts were oxidized by the MPO/H₂O₂/NO₂⁻ system, and then the lipid oxidation products were isolated by ethyl acetate extraction and derivatized with pentafluorobenzyl hydroxyamine. LC-MS/MS analysis showed that peaks that eluted at the same retention time as authentic HOHA-lactone-PFB oxime were present in the products from oxidation of DHA-PC vesicles, DHA or RBC ghosts, but not in those non-oxidized samples (Figure 6). The yield of HOHA-lactone from DHA-PC oxidation and DHA were determined as 0.45% and 0.78% after 24 h incubation, by quantifying peak area of HOHA-lactone-PFB oxime at 12.34 min in the 336→139 transition and comparing to calibration curve generated with authentic HOHA-lactone-PFB oxime (Figure 7).

Figure 6.

HOHA-lactone is generated during MPO//H₂O₂/NO₂⁻ system induced lipid oxidation. HOHA-lactone was measured by LC-MS/MS as pentafluorobenzyl oxime derivative. Left: LC chromatograms of (A) authentic HOHA-lactone-PFB oxime, (B) oxidized DHA-PC vesicle, (C) oxidized free DHA acid, (D) oxidized human RBC ghosts. Right: LC chromatograms of (E) non-oxidized DHA-PC vesicle, (F) non-oxidized free DHA acid, (G) non-oxidized human RBC ghosts.

Figure 7.

(A) Calibration curve of HOHA-lactone-PFB oxime determined by LC-MS/MS with SRM method 336.3 → 139.1. (B) Evolution profile of HOHA-lactone generation in MPO//H₂O₂/NO₂⁻ system induced oxidized DHA-PC vesicle (open circle) and free DHA acid (open square).

HOHA-lactone is stable under physiological conditions of pH and temperature

To confirm the stability of HOHA-lactone relative to the free acid HOHA under physiological conditions, pure HOHA-lactone (2.5 mg) was dissolved in 725 µL of phosphate buffered saline solution (pH 7.4) in D₂O and transferred into a NMR tube. ¹H NMR data was collected immediately and then the tube was placed in an incubator set at 37 °C. Another two ¹H NMR analyses were recorded after incubation for 12 and 24 h at 37 °C. Comparison of the NMR data at 0, 12 and 24 h showed no significant differences (Figure 8). Moreover, peaks with chemical shift range from 5 to 10 ppm correspond to four hydrogen atoms: an aldehydic hydrogen H_aat 9.33 ppm; two vinyl hydrogens, H_b and H_c, at 6.18 and 6.94 ppm and H_d, the allylic ring hydrogen α to the ester oxygen that exhibits a unique chemical shift at 5.25 ppm. The peaks of these four hydrogens were integrated, and the integration values were compared at different time points. The ratio of aldehyde proton to other three protons suggests that there is no significant structure change of HOHA-lactone.

Figure 8.

¹H-NMR spectrum analysis of 2.5 mg of HOHA-lactone in 725 µl of PBS-D₂O solution, pH 7.4, after incubated in 37 °C for 0, 12 and 24 h. The proton peaks that correspond to hydrogen atoms H_a, H_b, H_c and H_d on HOHA-lactone was integrated and then the ratio of these four H atoms were determined.

HOHA-lactone induces ARPE-19 cell growth and vascular endothelial growth factor (VEGF) secretion

Oxidative injury of ARPE-19 human retinal pigmented epithelial cells by exposure to H₂O₂ generated CEP derivatives (Supporting Information Figure S5). Treatment of ARPE-19 cells with HOHA-lactone generated CEP-modified proteins (Figure S6). A recent study showed that low levels of 4-hydroxynonenal (4-HNE), a structural analog of HOHA-lactone, cause secretion of VEGF from RPE cells.^{18, 19} To test whether HOHA-lactone could induce cell signaling and stimulate the secretion of VEGF from RPE cells in a similar pattern as 4-HNE, we treated ARPE-19 cells with various concentrations of HOHA-lactone. The cytotoxicity of HOHA-lactone to ARPE-19 cells was assessed by MTT assay. And VEGF-A levels in the extracellular medium were examined using a human VEGF-A ELISA kit (Thermo Scientific, Rockford, IL). Cell viability results showed that HOHA-lactone significantly stimulated cell proliferation at relatively low concentrations (< 0.5 µM). However, it significantly decreased cell viability at higher concentration levels (>20 µM, Figure 9A).

Figure 9.

(A) Cytotoxicity of HOHA-lactone to ARPE-19 cells (25,000 cells/well). Cells were treated with 0–50 µM HOHA-lactone for 2 h and then assayed for cytotoxicity by MTT assay. Cell survival values (%) were presented as means ± S.D. of 8 independent experiments. (B) Effect of HOHA-lactone on VEGF secretion by ARPE-19 cells. ARPE-19 cells grown in a 96-well plate (25,000 cells per well) were treated with serum-free culture media containing varying concentrations (0–25 µM) of HOHA-lactone. After 12 h of incubation, supernatants were collected and secreted VEGF was measured using a human VEGF-A ELISA kit. The data represent the mean ± SD (n = 8

We also found that HOHA-lactone induces VEGF secretion from ARPE-19 cells in a dose dependent manner. At sub-micromolar concentrations of HOHA-lactone, 0.1–1.0 µM, ARPE-19 cells secreted significantly increased levels of VEGF into the extracellular medium (Figure 9B). But at non-toxic higher levels, from 2.5–10 µM, HOHA-lactone did not induce increased secretion of VEGF. A high levels of HOHA-lactone, i.e., 25 µM, it significantly inhibited the secretion of VEGF.

DISCUSSION

In the present study, we demonstrated that the oxidation of both DHA containing phospholipids and non-esterified DHA produce HOHA-lactone, and that HOHA-lactone converts primary amino groups of biomolecules into CEP derivatives. Human red blood cells are usually exposed to high oxygen tension, which results in a high level of oxidative stress,²⁰ and they possess a significant amount of membrane-associated docosahexaenoic acid.²¹ We showed that peroxidation of red blood cell membrane lipids can produce HOHA-lactone, providing a biologic model in which HOHA-lactone is generated and released from membrane phospholipids during oxidative damage.

We also found that the spontaneous (nonenzymatic) deacylation of phospholipids to produce HOHA-lactone and its reactions to form CEPs occurs more readily than the reaction of HOHA-containing phospholipids to generate CEP esterified to the hydroxyl of a lyso phospholipid. These findings are likely to be biologically significant for several reasons. Whereas HOHA esterified to phospholipids is membrane bound and unesterified HOHA is anionic, lipophobic (CLogP: −4.19) and unable to penetrate cell membranes, HOHA-lactone is orders of magnitude less lipophobic (CLogP: −1.02) and is expected to be comparable to cortisone (CLogP: −0.93) in the ability to diffuse across cell membranes (Scheme 6). It is also likely to diffuse through confluent cell monolayers resulting in the generation of CEPs in locations remote from the site of phospholipid oxidation in vivo (Figure 10).

Scheme 6.

Figure 10.

Diffusion of HOHA-lactone through cell membranes might provide a route for the generation CEPs at locations that are remote, e.g. in the blood, from their photooxidative generation from docosahexaenoyl phospholipids that are especially abundant in retinal photoreceptor disk membranes.

CEPs are especially abundant in the retina and blood from individuals with age-related macular degeneration (AMD). Diffusion of HOHA-lactone generated in photoreceptor disk membranes in the eye into the blood may contribute to the elevated levels of CEP present in blood from AMD patients. Our previous studies found that exposure of rats to bright light generates elevated levels of CEP in their retinas and blood.²² It seems reasonable to postulate that light exposure produces HOHA-lactone in the retina and that this product of photooxidative injury in the eye migrates into the blood of these animals where it forms CEP derivatives of blood proteins.

In the present study, we demonstrated that exposure of ARPE-19 cells to HOHA-lactone increases secretion of VEGF from these cells in a concentration dependent manner. Similar to the previously reported effect of 4-HNE, the secreted VEGF can exert a paracrine effect on the surrounding endothelial cells, resulting in the proliferation of endothelial cells in the retina, and contributing to their migration and tube formation that ultimately promotes the pathologic processes of retinal neo-vascularization.

CONCLUSIONS

Owing to an extraordinary proclivity toward lactone formation, HOHA – both as the free acid and esterified within phospholipids – is readily converted under physiological conditions into HOHA-lactone. This product of non-enzymatic, free radical-induced, lipid oxidation is expected to readily diffuse from the site of oxidative injury and serve as an important precursor of CEPs. This may help to explain why CEPs in the blood are markers of oxidative injury associated with pathologies, such as age-related macular degeneration. Thus, although AMD involves oxidative injury in the retina, CEP levels are elevated in the blood of AMD patients. CEPs are also lipid-derived biologically active derivatives of proteins and ethanolamine phospholipids that contribute to disease progression, e.g., AMD and cancer. In view of the importance of HOHA-lactone in the generation of CEPs, this molecule may provide a new target for the design of therapeutic interventions to block its contribution to CEP formation and its pathological consequences. Besides serving as an important precursor of CEPs, HOHA-lactone itself may possess biological activities that contribute to the choroidal neovascularization that characterizes “wet” AMD.

ABBREVIATIONS

ABTS

2,2'-azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt

AMD

age-related macular degeneration

CEP

2-(ω-carboxyethyl)pyrrole

DHA

docosahexaenoic acid

DHA-PC

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

ECL

enhanced chemiluminescence

HOHA

4-hydroxy-7-oxo-hept-5-eonate

HOHA-PC

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

HRP

horse radish peroxidase

MES

2-(N-morpholino)ethanesulfonic acid

PLA2

phospholipase A2

PLD

phospholipase D.