Quantitative Analysis of Acrolein-Specific Adducts Generated during

Jun 20, 2012 - Ginga SHIMAKAWA , Mayumi SUZUKI , Eriko YAMAMOTO , Akiko NISHI , Ryota SAITO , Katsuhiko SAKAMOTO , Hiroshi YAMAMOTO , Amane ...
1 downloads 0 Views 715KB Size
Download PDF
Article pubs.acs.org/crt

Quantitative Analysis of Acrolein-Specific Adducts Generated during Lipid Peroxidation−Modification of Proteins in Vitro: Identification of Nτ-(3-Propanal)histidine as the Major Adduct Takuya Maeshima,† Kazuya Honda,† Miho Chikazawa,† Takahiro Shibata,† Yoshichika Kawai,† Mitsugu Akagawa,‡ and Koji Uchida*,† †

Graduate School of Bioagricultural Sciences, Nagoya University, Nagoya 464-8601, Japan Graduate School of Life and Environmental Sciences, Osaka Prefecture University, Sakai 599-8531, Japan

‡

S Supporting Information *

ABSTRACT: Acrolein, a ubiquitous pollutant in the environment, is endogenously formed through oxidation reactions and is believed to be involved in cytopathological effects observed during oxidative stress. Acrolein exerts these effects because of its facile reactivity with biological materials, particularly proteins. In the present study, we quantitatively analyzed the acrolein-specific adducts generated during lipid peroxidation−modification of proteins and identified the acrolein adduct most abundantly generated in the in vitro oxidized low-density lipoproteins (LDL). Taking advantage of the fact that the acrolein−lysine adducts, Nε-(3-formyl-3,4-dehydropiperidino)lysine (FDP-lysine) and Nε-(3-methylpyridinium)lysine (MP-lysine), have stable core structures resistant to the acid hydrolysis condition of proteins, we examined the formation of these adducts in proteins using high performance liquid chromatography with online electrospray ionization tandem mass spectrometry. However, only MP-lysine was detected as a minor product in the iron/ascorbate-mediated oxidation of polyunsaturated fatty acids in the presence of proteins and in the oxidized low-density lipoproteins (LDL). However, using a reductive amination-based pyridylamination method, we analyzed the acrolein-specific adducts with a carbonyl functionality and found that acrolein modification of the protein produced a number of carbonylated amino acids, including an acrolein−histidine adduct. On the basis of the chemical and spectroscopic evidence, this adduct was identified as Nτ-(3-propanal)histidine. More notably, Nτ-(3-propanal)histidine appeared to be one of the major adducts generated in the oxidized LDL. These data suggest that acrolein generated during lipid peroxidation may primarily react with histidine residues of proteins to form Nτ-(3-propanal)histidine.

■

as the major product.8 In a later study, Nε-(3-methylpyridinium)lysine (MP-lysine) (Figure 1), resulting from the Schiff base formation of acrolein with the ε-amino group of lysine, was identified.9 The formation of the pyridinium adducts is also a dominant pathway for the modification of the primary amine with 2-alkenals, such as crotonaldehyde, 2-hexenal, and 2octenal.10 In addition, using a specific monoclonal antibody, protein-bound acrolein has been shown to indeed constitute atherosclerotic lesions, in which the intense positivity is primarily associated with macrophage-derived foam cells.11 Because of the lack of specific and reliable methods for the determination of acrolein-specific adducts, very few studies have so far quantitatively demonstrated their formation in proteins to definitively show that acrolein can be generated as a lipid peroxidation product. As a result of these circumstances, lipid peroxidation has been underestimated as the source of acrolein by several researchers.12−14 In the current studies, taking advantage of the fact that the acrolein-specific lysine adducts, FDP-lysine and MP-lysine, have stable core structures resistant to the acid hydrolysis condition of proteins, we attempted to

INTRODUCTION Acrolein has been commercially produced since the 1940s. It is mainly used in the production of acrylic acid, a starting material for acrylate polymers. It also is used in the production of DLmethionine and as a herbicide and slimicide. Acrolein naturally occurs in foods and is formed during the combustion of fossil fuels, including engine exhausts, wood, and tobacco, and during the heating of cooking oils. Acrolein is known as a metabolite formed during the biotransformation of allyl compounds and the widely used anticancer drug cyclophosphamide. The conversion of threonine into acrolein by myeloperoxidase in the presence of hydrogen peroxide and chloride has also been reported.1 In addition, lipid peroxidation is believed to be an efficient source of acrolein.2 Because it was identified as one of the unnatural components of tobacco smoke,3 a number of reports have appeared describing the damaging effects of acrolein on the tracheal ciliary movement and the pulmonary wall.4,5 Moreover, acrolein has been suggested to initiate bladder cancer in rats under certain conditions.6 Acrolein, among the 2-alkenals, is by far the strongest electrophile.7 We have investigated the reaction of protein with acrolein and identified a novel acrolein−lysine adduct, Nε-(3formyl-3,4-dehydropiperidino)lysine (FDP-lysine) (Figure 1), © 2012 American Chemical Society

Received: February 24, 2012 Published: June 20, 2012 1384

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392

Chemical Research in Toxicology

Article

the presence of BSA was performed by incubating BSA (1 mg/mL) with 2 mM unsaturated fatty acids in the presence of 50 μM Fe2+ and 1 mM ascorbic acid in 1 mL of 50 mM sodium phosphate buffer, pH 7.2, in atmospheric oxygen at 37 °C. The reaction was terminated by the addition of 1 mM butylated hydroxytoluene (BHT) and 100 μM diethylenetriaminepentaacetic acid. Lipid Peroxidation−Modification of BSA. The metal-catalyzed oxidation of unsaturated fatty acids in the presence of BSA was performed by incubating BSA (1 mg/mL) with 2 mM unsaturated fatty acids in the presence of 50 μM Fe2+ and 1 mM ascorbic acid in 1 mL of 50 mM sodium phosphate buffer, pH 7.2, in atmospheric oxygen at 37 °C. In Vitro Oxidation of Low-Density Lipoprotein (LDL). LDL (1.019−1.063 g/mL) was prepared from the plasma of healthy humans by sequential ultracentrifugation and then extensively dialyzed 3 times against phosphate-buffered saline (PBS and 10 mM sodium phosphate buffer, pH 7.2, containing 150 mM NaCl) containing 0.01% ethylenediaminetetraacetic acid (EDTA) at 4 °C. LDL used for the oxidative modification by Cu2+ was dialyzed 5 times against a 1000-fold volume of PBS at 4 °C. The oxidation of LDL was performed by incubating 0.5 mg of LDL with CuSO4 (5 μM) in 1 mL of PBS for 24 h at 37 °C. The reaction was terminated by the addition of 1 mM of EDTA and then stored at 4 °C. Preparation of Isotope-Labeled FDP-Lysine and MP-Lysine. To prepare the isotope-labeled FDP-lysine and MP-lysine, 1 mM [U-13C6,15N2]-Nα-Fmoc-Nε-Boc-lysine was treated with 1 mL of TFA for 2 h at room temperature, applied to a C-18 Sep-Pak column, and eluted with 5 mL of methanol. The solvent was evaporated and the residue redissolved in 1 mL of 50 mM sodium phosphate buffer at pH 7.4. The obtained [U-13C6,15N2]-Nα-Fmoc-lysine was modified with 1 mM acrolein for 24 h at 37 °C and subsequently treated with 20% piperidine for 2 h at room temperature to remove the Fmoc moiety. The resulting mixture was purified by HPLC, carried out under the same HPLC conditions as already described. The LC/ESI/MS/MS analysis in the selected reaction monitoring (SRM) mode of the isotope-labeled FDP-lysines showed that that the standards contained no endogenous (nonlabeled) adducts. LC/ESI/MS/MS Analysis of Acrolein−Lysine Adducts. LC/MS/ MS analyses were carried out using the API 2000 triple quadrupole mass spectrometer (Applied Biosystems) through a TurboIonSpray source. Chromatography was carried out on a Develosil ODS-HG-5 column (2.0 × 75 mm) using an Agilent 1100 HPLC system. A discontinuous gradient was used by solvent A (H2O containing 0.1% formic acid) with solvent B (acetonitrile containing 0.1% formic acid) as follows: 5% B at 0 min, 5% B at 3 min, 100% B at 7 min, and 100% B at 10 min; flow rate = 0.2 mL/min. Mass spectrometric analyses were performed online using ESI-MS/MS in the positive ion mode and the SRM mode. The SRM transitions monitored were as follows: [U-13C6,15N2]FDP-lysine, m/z 249.1.1 → 90.0; FDP-lysine, m/z 241.3 → 84.1; [U-13C6,15N2]MPlysine, m/z 231.1 → 90.0; MP-lysine, m/z 223.1 → 84.1. The amount of acrolein−lysine adducts was quantified by the ratio of the peak area of the target adducts and of the FDP-lysine- and MP-lysine-stable isotopes. For the LC/ESI/MS/MS analysis of the acrolein−lysine adducts in vitro, the protein samples were treated with an equal volume of 20% trichloroacetic acid. After centrifugation at 5,000g for 10 min at 4 °C, the proteins were hydrolyzed in vacuo with 2 mL of 6 N HCl for 24 h at 110 °C. The internal standard, the FDP-lysine- and MP-lysine-stable isotopes, was added to the samples prior to acid-hydrolysis. After acidhydrolysis, the samples were partially separated using Oasis MCX cartridges (Waters, Milford, MA). After sample loading, the MCX cartridges were washed with 2 mL of 2% formic acid and 3 mL of methanol, and the acrolein−lysine adducts were eluted with 8 mL of 25% ammonia−water/methanol (5:95 v/v). The samples were then dried, dissolved in water, and subjected to an LC/ESI/MS/MS analysis. Pyridylamination-Based Analysis of Acrolein Adducts. The method is based on the following two-step reactions: the first step is the reaction between protein-bound aldehyde (Protein-CHO) and 2-AP (NH2-Pyr) to form the Schiff base adduct (Protein-CHN-Pyr) (Scheme 1); the second step is the reduction of the Schiff base with

Figure 1. Chemical structures of acrolein-specific adducts, FDP-lysine (top), MP-lysine (middle), and Nτ-(3-propanal)histidine (bottom), analyzed in this study.

establish a highly specific and sensitive method for the detection of these adducts using high performance liquid chromatography with online electrospray ionization tandem mass spectrometry (LC/ESI/MS/MS) coupled with a stable isotope dilution method. In addition, using a reductive amination-based pyridylamination method, we analyzed the acrolein-specific adducts with a carbonyl functionality. During the course of this study, we identified an acrolein-histidine adduct, Nτ-(3propanal)histidine (Figure 1) and established that this acrolein-specific product represented one of the major adducts generated during the lipid peroxidation−modification of proteins in vitro.

■

MATERIALS AND METHODS

Materials. Nα-acetyl-L-lysine, Nα-acetyl-L-histidine, and sodium cyanoborohydride (NaBH3CN) were obtained from Sigma-Aldrich (St. Louis, MO). Acrolein was obtained from Tokyokasei (Tokyo, Japan). [U-13C6, 15N2]-Nα-Fmoc-Nε-Boc-lysine (where Fmoc is 9fluorenylmethoxycarbonyl, and Boc is t-butoxycarbonyl) was obtained from the Cambridge Isotope Laboratories (Andover, MA). Bovine serum albumin (BSA), 2-aminopyridine (2-AP), and the fatty acids, such as docosahexaenoic acid, eicosapentaenoic acid, arachidonic acid, αlinolenic acid, γ-linolenic acid, linoleic acid, and palmitic acid, were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). 2-Thiobarbituric Acid-Reactive Substances. The amount of the 2-thiobarbituric acid-reactive substances (TBARS) was determined according to the method described by Masaki et al.15 The protein samples (0.1 mL) were treated with 0.5 mL of 2.8% (w/v) trichloroacetic acid and 0.5 mL of 1% 2-thiobarbituric acid in 0.05 N NaOH and then boiled for 20 min. After cooling, the sample was centrifuged (11,000g, 3 min), and the absorbance of the supernatant solution was measured at 534 nm. Malondialdehyde bis(dimethyl acetal) (Aldrich), which yields malondialdehyde by the treatment with HCl, was used as the standard. Preparation of Authentic Acrolein−Lysine Adducts. The reaction mixture (10 mL) containing 200 mM Nα-acetyllysine was incubated with 200 mM acrolein in 50 mM sodium phosphate buffer (pH 7.2). After incubation for 24 h at 37 °C, the authentic FDP-lysine and MP-lysine adducts were purified by reversed-phase HPLC as previously reported.8,9 Acrolein Modification of Protein. The acrolein modification of the protein was performed by incubating BSA (1.0 mg/mL) with 0−10 mM acrolein in 1 mL of 50 mM sodium phosphate buffer (pH 7.2) at 37 °C for 24 h. The Fe2+-catalyzed oxidation of unsaturated fatty acids in 1385

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392

Chemical Research in Toxicology

Article

Figure 2. LC/ESI/MS/MS analysis of authentic acrolein−lysine adducts. (A) Collision-induced dissociation of the [M + H]+ of FDP-lysine at m/z 241 at a collision energy of 25 V and the proposed structures of individual ions. (B) LC/ESI/MS/MS analysis of authentic isotope-labeled (upper) and nonlabeled (lower) FDP-lysines. The ion current tracings with SRM are shown. (C) Collision-induced dissociation of the [M + H]+ of MP-lysine at m/z 223 at a collision energy of 25 V and the proposed structures of individual ions. (D) LC/ESI/MS/MS analysis of authentic isotope-labeled (upper) and nonlabeled (lower) MP-lysines. The ion current tracings with SRM are shown. NaBH3CN to form the pyridylaminated derivatives (Protein-CH2−NHPyr) (Scheme 2). Protein‐CHO + NH 2 → Protein‐CHN‐Pyr

carried out under the same HPLC conditions and finally obtained 14.7 mg of P-1. The structure was characterized by LC/MS and 1H- and 13C NMR. The LC/MS was measured by an ACQUITY TQD system (Waters) equipped with an ESI probe and interfaced with a UPLC system (Waters). The sample injection volumes of 10 μL each were separated on a Waters BEH C18 1.7 μm column (150 mm × 2.1 mm) at the flow rate of 0.3 mL/min. A discontinuous gradient was used by solvent A (H2O containing 0.1% formic acid) with solvent B (acetonitrile containing 0.1% formic acid) as follows: 5% B at 0 min, 7% B at 5 min, and 95% B at 10 min. The mass spectrometric analyses were performed online using ESI-MS in the positive ion mode (cone potential = 30 eV). The NMR analyses were performed using a Bruker ARX600 (600 MHz) instrument. 1H NMR (D2O): δH 2.00 (3H, s), 2.25−2.30 (2H, tt), 3.11−3.33 (2H, m), 3.41−3.43 (2H, t), 4.32−4.35 (2H, t), 4.70−4.72 (1H, m), 6.90−6.92 (1H, dd, J = 9.2 Hz), 7.00−7.01 (1H, d, J = 9.0 Hz), 7.79−7.80 (1H, d, J = 9.0 Hz), 7.89−7.92 (1H, dd, J = 10.8 Hz), 7.40 (1H, s), 8.70 (1H, s); δC 22.25, 27.15, 28.58, 39.10, 47.33, 52.44, 113.44, 113.56, 120.39, 130.72, 135.09, 135.56, 144.51, 153.19, 174.02, 174.54.

(1)

Protein‐CHN‐Pyr + NaBH3CN → Protein‐CH 2‐NH‐Pyr (2) 16

The pyridylamination was carried out as previously described. In brief, the reaction mixtures containing 50 mM Nα-acetyl derivatives of the amino acids were incubated with 50 mM acrolein in 50 mM sodium phosphate buffer (pH 7.4). After incubation for 24 h at 37 °C, the reaction mixtures were treated with 340 mM 2-AP and 90 mM NaBH3CN for 24 h at 37 °C. The pyridylaminated acrolein-Nαacetylhistidine reaction mixture was hydrolyzed in vacuo with 6 N HCl for 24 h at 110 °C. The pyridylamination of the protein samples were performed by incubating with 340 mM 2-AP and 16 mM NaBH3CN for 24 h at 37 °C. After the incubation, the proteins were precipitated by the addition of the equal volume of 20% TCA, and then the resulting proteins were hydrolyzed in vacuo with 6 N HCl for 24 h at 110 °C. The hydrolysates were concentrated, dissolved in 0.1% TFA, and then analyzed by reverse phase HPLC on a Sunrise C28 column (4.6 × 250 mm inner diameter, ChromaNik, Japan). The samples were eluted with a gradient of water containing 0.1% TFA (solvent A) and acetonitrile containing 0.1% TFA (solvent B), (time = 0−20 min, 100% A; 20−40 min, 100 to 93% A; and 40−55 min, 93 to 0% A) at a flow rate of 0.8 mL/ min. The elution profiles were monitored by the resulting fluorescence intensity (excitation, 315 nm; emission, 380 nm). Identification of an Acrolein−Histidine Adduct with Carbonyl Functionality. The reaction mixtures, containing Nα-acetylhistidine and acrolein, were pyridylaminated, hydrolyzed, and analyzed by reverse phase HPLC as described above. The isolation and purification of the acrolein−histidine adduct with the carbonyl functionality (P-1) were

■

RESULTS Stable Isotope Dilution-Based LC/ESI/MS/MS Analysis of Acrolein−Lysine Adducts. Taking advantage of the fact that both FDP-lysine and MP-lysine have a stable core structure resistant to the acid-hydrolysis condition of proteins, we sought to establish a highly sensitive and specific method for measurement of these adducts using LC/ESI/MS/MS coupled with a stable isotope dilution method. The ionization and fragmentation of FDP-lysine and MP-lysine were evaluated in both the positive and negative ion modes. The negative ion mode provided an approximately 10-fold lower sensitivity than the 1386

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392

Chemical Research in Toxicology

Article

positive ion mode (data not shown), and hence, the positive ion mode was chosen for further evaluation. Collision-induced dissociation of standard FDP-lysine produced relevant product ions at m/z 83.9, m/z 98.2, and m/z 130.5 (Figure 2A). The product ions at m/z 84 and m/z 130 were suggested to be the product ions of a lysine moiety, and the ion at m/z 98 originated from the FDP-lysine. The identification of these product ions was supported by the observation that collision-induced dissociation of the standard isotope-labeled [U-13C6,15N2] FDP-lysine produced relevant product ions at m/z 89.8, m/z 103.9, and m/z 137.0 (Supporting Information, Figure S-1). Figure 2B demonstrates the result on the standard isotopelabeled [U-13C6,15N2] FDP-lysine (upper) and nonlabeled FDPlysine (lower) using SRM between the transition from the protonated precursor ion [M + H]+ to the characteristic product ion (m/z 249 → 90) and product ion (m/z 241 → 84), respectively, allowing detection of the FDP-lysine. The LOQ (limit of quantitation) was 100 fmol on column with linearity (r2 = 0.999 (10−500 nM)) and accuracy (105.9 ± 8.6%). However, the collision-induced dissociation of the standard MP-lysine produced relevant product ions at m/z 83.9, m/z 94.0, and m/z 130.2 (Figure 2C). Figure 1D demonstrates the result on the standard isotope-labeled [U-13C6,15N2] MP-lysine (upper) and nonlabeled MP-lysine (lower) allowing selective detection of the MP-lysine. The LOQ was 100 fmol on column with linearity (r2 = 0.994 (10−500 nM)) and accuracy (102.2 ± 8.2%). Quantitative Analysis of Acrolein−Lysine Adducts Generated in the Acrolein-Modified Protein. Using the LC/ESI/MS/MS technique, we evaluated the formation of FDPlysine and MP-lysine in the acrolein-modified BSA. To detect these adducts in the protein samples, we added isotope-labeled [U-13C6, 15N2] FDP-lysine or [U-13C6,15N2]MP-lysine to the samples prior to the acid hydrolysis of the protein, which were corrected for the potential analyte loss during the various stages of the sample preparation process. Figure 3A demonstrates the result for the acrolein-modified protein sample, along with the standard FDP-lysine and MP-lysine, in the positive ion mode using SRM between the transition from the protonated precursor ion [M + H]+ to the characteristic product ions (m/z 241 → 84) and (m/z 223 → 84), allowing detection of the FDP-lysine and MP-lysine, respectively. No adducts were detected in the native BSA, whereas the treatment of BSA with 0−2 mM acrolein in 50 mM sodium phosphate buffer (pH 7.2) for 24 h at 37 °C resulted in a dose-dependent formation of the FDP-lysine and MP-lysine (Figure 3B). It is of interest to note that even after the formation of the MP-lysine reached a plateau when the protein was treated with 2−4 mM acrolein, the formation of FDP-lysine continued to increase with further increases in the acrolein concentration. Upon incubation of BSA with the highest level of acrolein (10 mM) for 24 h at 37 °C, about 60 mol of lysine residues per mol of protein was lost (data not shown). These losses were accompanied by the formation of FDP-lysine (about 40 mol/ mol) and MP-lysine (about 14 mol/mol), accounting for about 67% and 23% of the lysine residues that had disappeared, respectively (Figure 3B). LC/ESI/MS/MS Analysis of MP-Lysine Generated during Lipid Peroxidation Modification of Protein. To examine the involvement of lipid peroxidation during the formation of the FDP-lysine and MP-lysine adducts in the protein, we attempted to detect these acrolein−lysine adducts in BSA exposed to lipid peroxidation and in Cu2+-oxidized LDL in vitro. The unsaturated fatty acids were incubated with an iron/ ascorbate-mediated free radical generating system in the

Figure 3. Quantitative analysis of acrolein−lysine adducts in the acrolein-modified BSA. (A) The ion current tracings of FDP-lysine (upper two tracings) and MP-lysine (lower two tracings) using LC/ ESI/MS/MS with SRM. (B) Dose-dependent formation of FDP-lysine (○) and MP-lysine (●) in the acrolein-modified BSA. BSA (1.0 mg/ mL) was incubated with acrolein (0−10 mM) in 50 mM sodium phosphate buffer (pH 7.2) for 24 h at 37 °C. The native and modified BSAs were analyzed by LC/ESI/MS/MS in the SRM mode followed by acid hydrolysis. Each point is the mean of duplicate determinations.

presence of BSA, and the acrolein−lysine adducts generated in the modified proteins were analyzed by LC/ESI/MS/MS following acid-hydrolysis. As shown in Figure 4A, the iron/ ascorbate-mediated oxidation of palmitic acid and polyunsaturated fatty acids, such as linoleic acid, γ-linolenic acid, and arachidonic acid, in the presence of the protein resulted in the formation of MP-lysine. The yields of the MP-lysine were correlated to the extent of the lipid peroxidation assessed by the formation of TBARS (Figure 4B). When the protein was incubated with docosahexaenoic acid (DHA) in the presence of iron/ascorbate for 24 h, the amount of MP-lysine was about 0.027 mol/mol BSA. However, the FDP-lysine was nearly undetectable in the protein exposed to the lipid peroxidation reactions (data not shown). We then attempted to detect the acrolein−lysine adducts in the Cu2+-oxidized LDL. The LC/ESI/ MS/MS analysis clearly showed that the MP-lysine was generated in a time-dependent manner (Figure 4C). The yield of the MP-lysine almost linearly increased as a function of the incubation time and reached 0.06 mol/mol LDL after 48 h (Figure 4D). Meanwhile, FDP-lysine was hardly detected in the oxidized LDL. 1387

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392

Chemical Research in Toxicology

Article

Figure 4. Quantitative analysis of MP-lysine generated during lipid peroxidation−modification of protein. (A) LC/ESI/MS/MS analysis of MP-lysine in the protein exposed to lipid peroxidation. The abbreviations used are PA, palmitic acid; LA, linoleic acid; α-LNA, α-linolenic acid; γ-LNA, γ-linolenic acid; AA, arachidonic acid; EPA, eicosapentaenoic acid; and DHA, docosahexaenoic acid. (B) Yields of TBARS and MP-lysine generated in the protein exposed to lipid peroxidation. Each bar is the mean of duplicate determinations. (C) LC/ESI/MS/MS analysis of MP-lysine in the oxidized LDL. LDL (0.5 mg) was incubated with 5 μM Cu2+ in 1 mL of PBS at 37 °C. The native LDL and oxidized LDL were analyzed by LC/ESI/MS/MS in the SRM mode followed by acid hydrolysis. (D) Time-dependent formation of MP-lysine in the oxidized LDL. Each point is the mean of duplicate determinations.

Identification of an Acrolein−Histidine Adduct with Carbonyl Functionality. However, we have recently established a reductive amination-based method for the quantitative determination of carbonylated amino acids using 2-AP as a fluorescent probe. Using this reductive amination-based approach, we examined the formation of the acrolein-specific adducts with a carbonyl functionality upon the reaction of the Nα-acetylated amino acids with acrolein. Our previous study showed that when BSA (1 mg/mL) was treated with 1 mM acrolein for 24 h at 37 °C, 26 molecules of lysine residues and 8 molecules of histidine residues per molecule of protein were lost.11 Thus, it is very likely that acrolein forms carbonylated adducts specifically with these amino acids. The reaction mixtures of acrolein/Nα-acetylhistidine and acrolein/Nα-acetyllysine were treated with 2-AP/NaBH3CN and then analyzed by the reverse-phase HPLC following the acid-hydrolysis. As shown in Figure 5A, the reaction of Nα-acetylhistidine with acrolein gave a single product (P-1) based on the reverse-phase HPLC analysis. In addition, Nα-acetyllysine upon reaction with acrolein also produced one major product, P-2. The same products were

detected in the hydrolysate of the pyridylaminated acroleinmodified protein (Figure 5B). The ESI-MS analysis of a putative acrolein−histidine adduct P-1 showed an [M + H]+ peak at m/z 290.1 and a [M + 2H]2+ peak at 145.5, which corresponded to the pyridylamino derivative of Nτ-(3-propanal)histidine (Figure 5C). This structure was finally confirmed by 1H- and 13C NMR spectrometry. The stoichiometry between the concentrations of Nτ-(3-propanal)histidine and the increase in the peak area of the products showed a linear correlation (data not shown). These data suggest that the propanalation could be characteristic of the reaction for acrolein with histidine residues in proteins. In the meantime, based on the cochromatography with an authentic sample on the reverse phase HPLC and LC/ MS analysis, P-2 was found to be identical to the pyridylamino derivative of the FDP-lysine. The detection of FDP-lysine as a major carbonylated adduct in the acrolein-modified BSA is in good agreement with the data obtained by LC/ESI/MS/MS (Figure 3). Determination of Nτ-(3-Propanal)histidine Generated during Lipid Peroxidation−Modification of Protein. We 1388

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392

Chemical Research in Toxicology

Article

Figure 5. Identification of a novel acrolein−histidine adduct with carbonyl functionality. (A) HPLC analysis of the pyridylaminated acrolein−histidine (upper) and acrolein−lysine (lower) adducts. The reaction mixtures of acrolein/Nα-acetylhistidine and acrolein/Nα-acetyllysine were pyridylaminated with 2-AP/NaBH3CN followed by acid hydrolysis. The hydrolysates were then analyzed by reverse-phase HPLC. (B) HPLC analysis of the pyridylaminated acrolein adducts in the acrolein-modified protein. BSA (1 mg/mL) was incubated with 2 mM acrolein for 24 h at 37 °C, and the acrolein adducts were analyzed by the pyridylamination method. (C) Mass spectrum of P-1 and the structure of pyridylaminated Nτ-(3-propanal)histidine.

Figure 6. Determination of Nτ-(3-propanal)histidine generated during lipid peroxidation−modification of protein. (A) HPLC analysis of the pyridylaminated products in the protein exposed to lipid peroxidation. The product profile of the lipid peroxidation-modified BSA was compared to that of the oxidized BSA, oxidized arginine, and oxidized lysine with Fe2+/ascorbate. The reaction mixtures were pyridylaminated with 2-AP/NaBH3CN followed by acid hydrolysis. The hydrolysates were then analyzed by reverse-phase HPLC. The peak detected right before P-4 in the chromatogram (second trace) of the Fe2+/ascorbate-oxidized BSA gave the same pseudomolecular ion peak as that of P-4. It was only detected when the protein was treated with iron/ascorbate. Thus, we have not rigorously identified the product. (B) HPLC analysis of the pyridylaminated products in the oxidized LDL. (C) Time-dependent formation of Nτ-(3-propanal)histidine in the oxidized LDL. Each point is the mean of duplicate determinations.

then examined the formation of Nτ-(3-propanal)histidine during lipid peroxidation−modification of the proteins. As shown in Figure 6A, the iron/ascorbate-mediated oxidation of linoleic acid in the presence of BSA resulted in the formation of a product, which was found to be identical to Nτ-(3-propanal)histidine by the coelution test with the reaction mixtures of acrolein/Nαacetylhistidine and by the LC/MS analysis. After a 24-h

incubation, the amount of Nτ-(3-propanal)histidine was approximately 17.5 mmol/mol BSA. In addition to the acrolein−histidine adduct, two pyridylaminated products, P-3 and P-4, were also detected as the carbonylated amino acids. However, they were generated even in the absence of linoleic acid, suggesting that they might have originated from the oxidative modification of protein by iron/ascorbate. We indeed 1389

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392

Chemical Research in Toxicology

Article

mice exposed to ferric nitrilotriacetate (Fe3+-NTA), an iron chelate that induces free radical-mediated oxidative tissue damage (Maeshima, T., Shibata, T., and Uchida, K., unpublished data). Thus, although the levels of the MP-lysine in the lipid peroxidation-modified proteins in vitro were extremely low, covalent modification of the proteins by acrolein, generating the pyridinium-containing lysine adducts, could have diagnostic applications. Moreover, the LC/ESI/MS/MS methods described in the present paper provide a means for the specific quantification of the acrolein−lysine adduct in proteins. The formation of MP-lysine has been explained by the mechanism involving the formation of a Schiff base derivative as the first intermediate. The Schiff base further reacts with a second acrolein molecule via a Michael addition to generate an imine derivative. The subsequent conversion of this imine derivative to the final product (MP-lysine) obviously requires two oxidation steps and intramolecular cyclization, but its detailed mechanism has not yet been clarified. MP-lysine was detected as a minor product in the reaction of acrolein with the lysine derivative, whereas the formation of the pyridinium adducts has been reported to be a dominant pathway for the modification of the primary amine with 2-alkenals, such as 2-hexenal and 2octenal.19−21 In our previous study, we examined the specificity of a monoclonal antibody that recognizes the acrolein-modified proteins and found that it more efficiently recognized the MPlysine than FDP-lysine.9 The preferential recognition of the antibody to MP-lysine can be explained by the structural characteristics in the side chain of these adducts. In contrast to FDP-lysine, MP-lysine contains a more fixed, positive charge on the pyridinium side chain, which may represent important immunological epitopes. Indeed, a monoclonal antibody raised against crotonaldehyde-modified proteins recognized a similar pyridinium adduct as the major epitope.10 In addition, Nagai et al.22 have raised a monoclonal antibody against a glycolaldehydemodified protein and found that a lysine pyridinium adduct constitutes an epitope of the antibody. A large number of proinflammatory and pro-atherogenic properties have been ascribed to the oxidized LDL and their components.23 Thus, the oxidized lipid adducts, including these lysine pyridinium adducts, generated during oxidative modification of the LDL may be involved in the process of macrophage transformation into the foam cells during atherogenesis. The failure to detect FDP-lysine during the lipid peroxidation−modification of proteins appears inconsistent with our previous study showing that FDP-lysine was generated in the oxidatively modified LDL with Cu2+ and that the adduct formation was correlated with the LDL peroxidation.8,11 This may be partly because we adopted a less sensitive approach using an amino acid analysis and a contaminating peak coeluted with the authentic FDP-lysine in the previous study. Even so, these data cast some doubt on the suggestion that the adduct(s) result mainly from “free” acrolein arising from lipid peroxidation. It is still likely that the adduct(s) could result from an unknown intermediate arising from lipid peroxidation, which eventually generates the same adducts as those generated upon reaction with acrolein. The introduction of carbonyl groups into the amino acid residues is a hallmark for oxidative damage to proteins.24,25 Carbonyl groups are introduced into proteins by a variety of modification pathways in vivo and in vitro, particularly the metalcatalyzed oxidation of specific amino acid residues,26−28 and also the adduction of lipid peroxidation-derived aldehydes10,29−32 or carbohydrate (i.e., the Maillard reaction or glycation).33

observed that the same products could be formed upon the iron/ ascorbate-mediated oxidation of amino acids, such as arginine and lysine. These data and the observation that the LC/MS analysis of P-3 and P-4 showed a pseudomolecular ion peak at m/z 210.1 (M + H) and m/z 224.2 (M + H), respectively, suggested that P-3 and P-4 might represent the pyridylamino derivatives of γ-glutamic semialdehyde (GGS) and α-aminoadipic semialdehyde (AAS), respectively. The formation of Nτ(3-propanal)histidine and these oxidized amino acids were also confirmed in the Cu2+-oxidized LDL (Figure 6B). The oxidation of LDL with Cu2+ for 24 h produced approximately one molecule of Nτ-(3-propanal)histidine per molecule of apoB (Figure 6C). These data suggest that acrolein generated during lipid peroxidation mainly mediates the propanalation of proteins (Figure 7).

Figure 7. Propanalation of histidine residues in protein by acrolein.

■

DISCUSSION Because of the fact that the core structures of the FDP-lysine and MP-lysine adducts are resistant to the conventional acid hydrolysis of proteins, we established a highly specific and sensitive method for the detection of these adducts using LC/ ESI/MS/MS coupled with a stable isotope dilution method. The sensitivity of the LC/ESI/MS/MS assay used in the present studies permitted quantification of one of the acrolein−lysine adducts, MP-lysine, in the iron/ascorbate-mediated oxidation of polyunsaturated fatty acids in the presence of protein. Of note, the peroxidation of polyunsaturated fatty acids in the presence of BSA generated MP-lysine in concert with the formation of TBARS during the course of metal-catalyzed oxidation (Figure 4). These data suggest that although the mechanism of the formation of acrolein during lipid peroxidation has not yet been experimentally resolved, acrolein could be ubiquitously generated under oxidative stress. MP-lysine was also detected in the in vitro oxidized LDL. The formation of MP-lysine in the oxidized LDL was first suggested by Itabe and his colleagues.17 They have also shown that using LC/ESI/MS/MS in combination with the on-membrane sample preparation procedure, MP-lysine might be generated at Lys-293 in the oxidized LDL.17 Our present data showed that the yield of the adduct was about 0.06 mol/mol LDL after 48 h, which corresponds to only about 1% of the lysine lost. However, the levels of the lysine adducts that had been quantitatively analyzed in the oxidized LDL, such as the HNE-lysine and MDA-lysine adducts, were almost comparable to those of the MP-lysine.18 Thus, the majority of the modified lysines generated in the oxidized LDL has not yet been chemically characterized. In the meantime, using the stable isotope dilution-based LC/ESI/MS/ MS method, we have also detected the MP-lysine in the kidney of 1390

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392

Chemical Research in Toxicology

Article

based methods are insufficient for quantification. Because of the fact that the core structure of the pyridinium-containing lysine adducts are resistant to the conventional acid hydrolysis of proteins, we established a highly sensitive method for the detection of FDP-lysine and MP-lysine using LC/ESI/MS/MS. In addition, we successfully established a method for the quantification of an acrolein−histidine adduct using a reductive amination-based approach. Because of their specificity and sensitivity, these methods may form the basis for the detection of acrolein-derived adducts generated in proteins in vitro and in vivo. In conclusion, the present work describes, for the first time, the chemical nature and concentration of acrolein adducts to lysine and histidine generated during the lipid peroxidation− modification of proteins, such as Cu2+-oxidized human LDL, and presents a method for quantifying these structures in the proteins. The quantification of the lipoxidation products such as acrolein-specific adducts in atherosclerotic plaques and in tissue proteins should help to clarify the role of lipid peroxidation in the pathogenesis of atherosclerosis and other disorders in which oxidative stress is implicated.

Carbonyl derivatives can be measured by convenient methods using 2,4-dinitrophenylhydrazine, which reacts with carbonyl groups to generate dinitrophenylhydrazones with characteristic absorbance maxima at 360−390 nm.34 Using the methods, it has been confirmed that carbonyl derivatives accumulate on tissue proteins during aging and disease development. Nevertheless, the methods are unfortunately limited to measuring the total carbonyl derivatives formed by various unspecific pathways, and information on the chemical structures and formation mechanisms is barely provided. Therefore, more specific methods for the determination of carbonyl derivatives are required in order to understand the chemical nature, oxidation pathway, and distribution level in vivo. In our previous study, we established a reductive amination-based method for the quantitative determination of carbonylated amino acids using 2-aminopyridine (2-AP) as a fluorescent probe and analyzed the Michael addition-type 4-hydroxy-2-nonenal (HNE)-cysteine and HNE-histidine adducts.16,35 In the present study, this method allowed us to identify the acrolein−histidine adduct, Nτ(3-propanal)histidine. Formation of this adduct in acroleinmodified proteins has been previously suggested.36,37 However, due to its instability, very few studies have so far quantitatively demonstrated its formation in proteins. Of note, using a reductive amination-based pyridylamination method, this adduct was detected as one of the major carbonyl adducts generated in the acrolein-modified BSA (Figure 5). Moreover, Nτ-(3propanal)histidine was detected in the Cu2+-oxidized LDL as the major product. The maximal yield of Nτ-(3-propanal)histidine in the oxidized LDL was 1 mol/mol LDL apoB. We have recently reported that the oxidation of LDL with Cu2+ for 24 h gave approximately 5 mol of the HNE-histidine adduct per mol of LDL apoB.16 Because about 10 mol of histidine residues per mol of LDL apoB is lost during the LDL oxidation, about 60% (6 mol) of the histidine lost can, therefore, be ascribed presently to the chemically characterized and assayable products, the HNEhistidine and acrolein−histidine adducts. On the basis of our previous study on the crotonaldehyde modification of protein,10 the primary target amino acid of acrolein may be cysteine, producing the Michael addition adduct, S-(3-propanal)cysteine. The adductions may be readily reversed in aqueous buffer. This may lead to the reversible binding of acrolein to the cysteine residues of the proteins. Likewise, acrolein was expected to form Nε-(3-propanal)lysine upon the reaction with lysine residues. Indeed, its formation was suggested by the detection using HPLC (Figure 6B); the LC-MS analysis of a single peak eluted at 32 min showed the [M + H]+ peak at m/z 281.2, corresponding to the pyridylamino derivative of Nε-(3propanal)lysine. However, our preliminary experiment has also shown that the pyridylamination of the malondialdehyde/lysine reaction mixture gave the same peak showing the same pseudomolecular ion peak in the LC/MS analysis. Thus, the formation of Nε-(3-propanal)lysine during the lipid peroxidation−modification of proteins remains to be established. Various methods for determination of lipid peroxidationderived products have been developed. A colorimetric assay using 2-thiobarbituric acid for detection of malondialdehyde is probably the easiest way of evaluating lipid peroxidation. An alternative and probably more popular approach for the detection of lipid peroxidation products in biological samples is the use of antibodies. Immunological detection is a powerful tool that can be used to evaluate the presence of a desired target and its subcellular localization. However, due to the significant number of interfering substances in biological samples, antibody-

■

ASSOCIATED CONTENT

S Supporting Information *

LC/ESI/MS/MS analysis of isotope-labeled FDP-lysine adduct. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Funding

This work was supported by Grant-in-Aid for Scientific Research on Innovative Areas (Research in a Proposed Research Area), Japan (to K.U.). Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS We thank Chika Wakita for her technical support. ABBREVIATIONS 2-AP, 2-aminopyridine; AAS, α-aminoadipic semialdehyde; ABA, p-aminobenzoic acid; apoB, apolipoprotein B; BSA, bovine serum albumin; DTT, dithiothreitol; GGS, γ-glutamic semialdehyde; HNE, 4-hydroxy-2-nonenal; LDL, low density lipoprotein; LOQ, limit of quantitation; MALDI-TOF MS, matrixassisted laser desorption and ionization time-of-flight mass spectrometry; NaBH3CN, sodium cyanoborohydride

■

REFERENCES

(1) Anderson, M. M., Hazen, S. L., Hsu, F. F., and Heinecke, J. W. (1997) Human neutrophils employ the myeloperoxidase-hydrogen peroxide-chloride system to convert hydroxy-amino acids into glycolaldehyde, 2-hydroxypropanal, and acrolein. J. Clin. Invest. 99, 424−432. (2) Uchida, K. (1999) Current status of acrolein as a lipid peroxidation product. Trends Cardiovasc. Med. 9, 109−113. (3) Johnstone, R. A. W., and Plimmer, J. R. (1959) The chemical constituents of tobacco and tobacco smoke. Chem. Rev. 59, 885−936. (4) Kensler, C. J., and Battista, S. P. (1963) Components of cigarette smoke with ciliary-depressant activity. Their selective removal by filters containing activated charcoal granules. N. Engl. J. Med. 269, 1161−1166. 1391

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392

Chemical Research in Toxicology

Article

(5) Izard, C., and Liberman, C. (1978) Acrolein. Mutat. Res. 47, 115− 138. (6) Cohen, S. M., Garland, E. M., St., Jhon, M., Okamura, T., and Smith, R. A. (1992) Acrolein initiates rat urinary bladder carcinogenesis. Cancer Res. 52, 3577−3581. (7) Esterbauer, H., Schaur, R. J., and Zollner, H. (1991) Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radical Biol. Med. 11, 81−128. (8) Uchida, K., Kanematsu, M., Morimitsu, Y., Osawa, T., Noguchi, N., and Niki, E. (1998) Acrolein is a product of lipid peroxidation reaction. Formation of acrolein and its conjugate with lysine residues in oxidized low-density lipoprotein. J. Biol. Chem. 273, 16058−16066. (9) Furuhata, A., Ishii, T., Kumazawa, S., Yamada, T., Nakayama, T., and Uchida, K. (2003) Nε-(3-methylpyridinium)lysine, a major antigenic adduct generated in acrolein-modified protein. J. Biol. Chem. 278, 48658−48665. (10) Ichihashi, K., Osawa, T., Toyokuni, S., and Uchida, K. (2001) Endogenous formation of protein adducts with carcinogenic aldehydes: Implication for oxidative stress. J. Biol. Chem. 276, 23903−23913. (11) Uchida, K., Kanematsu, M., Sakai, K., Matsuda, M., Hattori, N., Mizuno, Y., Suzuki, D., Miyata, T., Noguchi, N., Niki, E., and Osawa, T. (1998) Protein-bound acrolein. Potential markers for oxidative stress. Proc. Natl. Acad. Sci. U.S.A. 95, 4882−4887. (12) Tomitori, H., Usui, T., Saeki, N., Ueda, S., Kase, H., Nishimura, K., and Kashiwagi, K (2005) Polyamine oxidase and acrolein as novel biochemical markers for diagnosis of cerebral stroke. Stroke 36, 2609− 2613. (13) Saiki, R., Nishimura, K., Ishii, I., Omura, T., Okuyama, S., Kashiwagi, K., and Igarashi, K. (2009) Intense correlation between brain infarction and protein-conjugated acrolein. Stroke 40, 3356−3361. (14) Igarashi, K., and Kashiwagi, K. (2011) Protein-conjugated acrolein as a biochemical marker of brain infarction. Mol. Nutr. Food Res. 55, 1332−1341. (15) Masaki, N., Kyle, M. E., and Farber, J. L. (1989) tert-butyl hydroperoxide kills cultured hepatocytes by peroxidizing membrane lipids. Arch. Biochem. Biophys. 269, 390−399. (16) Wakita, C., Honda, K., Shibata, T., Akagawa, M., and Uchida, K. (2011) A method for detection of 4-hydroxy-2-nonenal adducts in proteins. Free Radical Biol. Med. 251, 1−4. (17) Obama, T., Kato, R., Masuda, Y., Takahashi, K., Aiuchi, T., and Itabe, H. (2007) Analysis of modified apolipoprotein B-100 structures formed in oxidized low-density lipoprotein using LC-MS/MS. Proteomics 7, 2132−2141. (18) Requena, J. R., Chao, C. C., Levine, R. L., and Stadtman, E. R. (2001) Glutamin and aminoadipic semialdehydes are the main carbonyl products of metal-catalyzed oxidation of proteins. Proc. Natl. Acad. Sci. U.S.A. 98, 69−74. (19) Alaiz, M., and Barragán, S. (1995) Changes induced in bovine serum albumin following interactions with the lipid peroxidation product E-2-octenal. Chem. Phys. Lipids 77, 217−223. (20) Baker, A., Zídek, L., Wiesler, D., Chmelík, J., Pagel, M., and Novotny, M. V. (1998) Reaction of N-acetylglycyllysine methyl ester with 2-alkenals: an alternative model for covalent modification of proteins. Chem. Res. Toxicol. 11, 730−740. (21) Baker, A., Wiesler, D., and Novotny, M. V. (1999) Tandem mass spectrometry of model peptides modified with trans-2-hexenal, a product of lipid peroxidation. J. Am. Soc. Mass. Spectrom. 10, 613−624. (22) Nagai, R., Hayashi, C. M., Xia, L., Takeya, M., and Horiuchi, S. (2002) J. Biol. Chem. 277, 48905−48912. (23) Glass, C. K., and Witztum, J. L. (2001) Cell 104, 503−516. (24) Dalle-Donne, I., Rossi, R., Giustarini, D., Milzani, A., and Colombo, R. (2003) Protein carbonyl groups as biomarkers of oxidative stress. Clin. Chim. Acta 329, 23−38. (25) Nyström, T. (2005) Role of oxidative carbonylation in protein quality control and senescence. EMBO J. 24, 1311−1317. (26) Stadtman, E. R. (1992) Protein oxidation and aging. Science 257, 1220−1224. (27) Stadtman, E. R. (2001) Protein oxidation in aging and age-related diseases. Ann. N.Y. Acad. Sci. 928, 22−38.

(28) Levine, R. L. (2002) Carbonyl modified proteins in cellular regulation, aging and disease. Free Radical Biol. Med. 32, 790−796. (29) Uchida, K., Szweda, L. I., Chae, H. Z., and Stadtman, E. R. (1993) Immunochemical detection of 4-hydroxy-2-nonenal modified proteins in oxidized hepatocytes. Proc. Natl. Acad. Sci. U.S.A. 90, 8742−8746. (30) Yamada, S., Kumazawa, S., Ishii, T., Nakayama, T., Itakura, K., Shibata, N., Kobayashi, M., Sakai, K., Osawa, T., and Uchida, K. (2001) Immunochemical detection of a lipofuscin-like fluorophore derived from malondialdehyde and lysine. J. Lipid Res. 42, 1187−1196. (31) Furuhata, A., Nakamura, M., Osawa, T., and Uchida, K. (2002) Thiolation of protein-bound carcinogenic aldehyde: an electrophilic acrolein−lysine adduct that covalently binds to thiols. J. Biol. Chem. 277, 27919−27926. (32) Uchida, K. (2003) 4-Hydroxy-2-nonenal: a product and mediator of oxidative stress. Prog. Lipid Res. 42, 318−343. (33) Liggins, J., and Furth, A. J. (1997) Role of protein-bound carbonyl groups in the formation of advanced glycation endproducts. Biochim. Biophys. Acta 1361, 123−130. (34) Levine, R. L., Williams, J. A., Stadtman, E. R., and Shacter, E. (1994) Carbonyl assays for determination of oxidatively modified proteins. Methods Enzymol. 233, 346−357. (35) Wakita, C., Maeshima, T., Yamazaki, A., Shibata, T., Ito, S., Akagawa, M., Ojika, M., Yodoi, J., and Uchida, K. (2009) Stereochemical configuration of 4-hydroxy-2-nonenal-cysteine adducts and their stereoselective formation in a redox-regulated protein. J. Biol. Chem. 284, 28810−28822. (36) Pocker, Y., and Janjic, N. (1988) Differential modification of specificity in carbonic anhydrase catalysis. J. Biol. Chem. 263, 6169− 6176. (37) Gan, J. C., Oandasan, A., and Ansari, G. A. (1991) In vitro modification of serum albumin by acrolein. Chemosphere 23, 939−947.

1392

dx.doi.org/10.1021/tx3000818 | Chem. Res. Toxicol. 2012, 25, 1384−1392