Evidence That Malondialdehyde-Derived Aminoenimine Is Not a

of lipid peroxidation, contribute to the fluorescence formation of lipofuscin. Although early studies proposed an aminoenimine structure (RNHCHdCHCHdN...
0 downloads 0 Views 41KB Size
Chem. Res. Toxicol. 2001, 14, 473-475

473

Communications Evidence That Malondialdehyde-Derived Aminoenimine Is Not a Fluorescent Age Pigment Koichi Itakura*,† and Koji Uchida‡ Faculty of Education, Aichi University of Education, Kariya 448-8542, Japan, and Laboratory of Food and Biodynamics, Nagoya University Graduate School of Bioagricultural Sciences, Nagoya 464-8601, Japan Received March 5, 2001

It has been suggested that protein modifications by malondialdehyde (MDA), a major product of lipid peroxidation, contribute to the fluorescence formation of lipofuscin. Although early studies proposed an aminoenimine structure (RNHCHdCHCHdNR) formed from MDA and the -amino groups of the lysine residues for the fluorophores, there has been considerable doubt as to whether the aminoenimine is fluorescent. To date, however, there is no conclusive evidence that the aminoenimine is nonfluorescent. This is because that it has not yet been isolated. In this study, we succeeded in isolating an aminoenimine, N,N′-bis[5-(tert-butoxycarboxamido)-5-carboxypentyl]-1-amino-3-iminopropene [(Boc-Lys)2MDA], formed from the reaction of MDA with a lysine derivative, NR-tert-butoxycarbonyl-L-lysine (Boc-Lys), at neutral pH, and confirmed that the purified (Boc-Lys)2MDA exhibited no fluorescence. This result demonstrates that aminoenimines formed from MDA and lysine residues do not contribute to the fluorescence formation of lipofuscin.

Introduction It is generally believed that the age-related accumulation of fluorescent lipofuscin pigments in tissues and cells involves protein modifications by lipid peroxidation (13). Malondialdehyde (MDA)1 is a major end-product of lipid peroxidation and is frequently used as an indicator of oxidative stress in vivo (4). Incubation of protein with MDA results in the formation of protein-bound fluorophores. It had long been considered that the MDAderived fluorophores may be responsible for the fluorescence of lipofuscin. MDA readily reacts with the -amino groups of the lysine residues in proteins, resulting in the formation of various MDA-lysine adducts. In early studies, Chio and Tappel (5, 6) proposed aminoenimines (RNHCHdCHCHdNR) as the MDA-derived fluorophores and suggested that the fluorescence of lipofuscin is due to the aminoenimines. The aminoenimines can also account for the MDA-derived protein cross-linkings. Recently, Requena et al. (7) quantified the aminoeniminederived cross-linkings in native and oxidized low-density lipoproteins. In both cases, however, the structural confirmation of the aminoenimines was conducted after reduction with NaBH4 to the nonfluorescent RNHCH2CH2CH2NHR forms. On the other hand, Kikugawa et al. * To whom correspondence should be addressed. Fax: (81)-566-262596. E-mail: [email protected]. † Aichi University of Education. ‡ Nagoya University Graduate School of Bioagricultural Sciences. 1Abbreviations: MDA, malondialdehyde; Boc-Lys, NR-tert-butoxycarbonyl-L-lysine; TMP, 1,1,3,3-tetramethoxypropane; (Boc-Lys)2MDA, N,N′-bis[5-(tert-butoxycarboxamido)-5-carboxypentyl]-1-amino-3-iminopropene.

(8, 9) obtained strong fluorescent 1,4-dihydropyridine3,5-dicarbaldehydes from the reaction of MDA with primary amines. Recently, we also identified a fluorescent MDA-lysine adduct which has dihydropyridine and pyridinium rings (10). To date, however, the fluorophores formed in the MDA-modified proteins and their contribution to the fluorescence of lipofuscin have not been established. One of the main reasons for this is that the aminoenimine formed from MDA and lysine has not yet been obtained in a nonreduced form and its spectral characteristics have remained obscure. We report here the first isolation of an aminoenimine formed from the reaction of MDA with a lysine derivative, NR-tert-butoxycarbonyl-L-lysine (Boc-Lys), at neutral pH and confirmed its structure by NMR spectroscopy and mass spectrometry.

Experimental Procedures Materials and General Procedures. Boc-Lys was obtained from Tokyo Kasei Kogyo Co., Ltd. (Tokyo, Japan). 1,1,3,3Tetramethoxypropane (TMP) was obtained from Aldrich (Milwaukee, WI). All other chemicals were of the highest grade commercially available. The preparative HPLC was carried out on a C18 reversed-phase column (5 µm particle size, 8.0 × 250 mm) at a flow rate of 2.0 mL/min. Preparation of (Boc-Lys)2MDA. TMP (0.4 mL) was mixed with 1 N HCl (24.6 mL), and the resulting solution was allowed to stand at room temperature for 1 h. Before mixing with BocLys, the solution was neutralized with 1 N NaOH. A solution (200 mL) of Boc-Lys (100 mM) and TMP hydrolysate (10 mM) in 0.1 M phosphate buffer (pH 7.4) was allowed to stand at room temperature for 2 months. The solution was applied on an

10.1021/tx010055u CCC: $20.00 © 2001 American Chemical Society Published on Web 05/01/2001

474

Chem. Res. Toxicol., Vol. 14, No. 5, 2001

Figure 1. HPLC analysis of the products formed from the reaction of Boc-Lys with MDA. The analysis was carried out on a C18 reversed-phase column (5 µm particle size, 4.6 × 150 mm) at a flow rate of 0.8 mL/min. Products were monitored with a multiwavelength (195-650 nm) detector.

Communications

Figure 2. Structure of aminopropenal derivative, 1,4-dihydropyridine-3,5-dicarbaldehyde derivative, and (Boc-Lys)2MDA. Table 1. Selected 13C and 1H NMR Data for (Boc-Lys)2MDAa C positionb 1, 3

Amberlite XAD-2 column equilibrated with H2O/CH3CO2H (100/ 1). The column was stepwise eluted with H2O/CH3CO2H (100/ 1), 20, 40, 60, and 80% CH3OH in H2O/CH3CO2H (100/1), and CH3OH (200 mL each). The fractions eluted with 60 and 80% CH3OH in H2O/CH3CO2H (100/1) were combined and concentrated. The concentrated solution was subjected to preparative HPLC; elution was with a linear gradient from 0 to 54% CH3OH in H2O/CH3CO2H (100/1) over 63 min. The HPLC fraction containing (Boc-Lys)2MDA was collected and concentrated. The concentrated solution was further subjected to preparative HPLC; elution was with a linear gradient from 0 to 25% CH3CN in H2O/CH3CO2H (100/1) over 80 min. The HPLC fraction containing (Boc-Lys)2MDA was collected, and concentration of the combined fraction afforded (Boc-Lys)2MDA (1.6 mg) as a mixture (ca. 2:1) of two diastereomers. The diastereomers could not be separated from each other due to their interconversion. UV [H2O/CH3CO2H (ca. 100/1), pH 2.8] λmax 299 nm; 13C NMR (100 MHz; D2O; CH3CO2D as δ 21.10) δ 22.25, 22.71, 27.17, 27.98, 29.00, 31.32, 43.50, 43.69, 49.14, 49.34, 55.34, 56.77, 81.42, 88.73, 90.57, 157.97, 160.71, 162.27, 164.06, 165.55, and 179.19 (Although each diastereomer has 25 carbons, the 13C NMR spectrum showed only 21 signals probably due to signal overlapping.); 1H NMR (400 MHz; D2O; HOD as δ 4.80); δ 1.42 (18H, s, Boc), 1.16-1.86 (12H, m, β,γ,δ-CH2), 3.20-3.45 (4H, m, -CH2), 3.80-4.00 (2H, m, R-CH), 5.54 (1H, t, J ) 12 Hz), 7.56 (1H, d, J ) 13 Hz), 7.65 (2/3H, d, J ) 12 Hz), 7.66 (1/3H, d, J ) 11 Hz); electrospray ionization LC/MS m/z 529 (MH+).

Results and Discussion Boc-Lys (100 mM) was incubated with TMP hydrolysate (10 mM) in 0.1 M phosphate buffer (pH 7.4) at room temperature. When the reaction mixture was analyzed by HPLC equipped with a multiwavelength detector, two major peaks (A and B) were observed (Figure 1). Their UV spectra and LC/MS analysis indicated that A and B corresponded to an aminopropenal derivative (UV λmax 282 nm, MH+ m/z 301) and a 1,4-dihydropyridine-3,5dicarbaldehyde derivative (UV λmax 396 nm, MH+ m/z 381), respectively (Figure 2). The former was nonfluorescent and the latter was highly fluorescent. Both of them are well-known as major MDA-lysine adducts (8, 9, 11). Between peaks A and B on the chromatogram, two minor peaks (C and D), each of which exhibited the

2 -NCH2, dNCH2 a

δC

δH

160.71, 162.27, 164.06, 165.55 88.73, 90.57 43.50, 43.69, 49.14, 49.34

7.56 (1H, d), 7.65 (2/3H, d), 7.66 (1/3H, d) 5.54 (1H, t) 3.20-3.45 (4H, m)

A mixture of two diastereomers. b See Figure 2 for numbering.

same UV absorption maxima around 300 nm, were observed. An LC/MS analysis revealed that both of them exhibited a MH+ ion at m/z 529. Because the m/z was identical with that of an aminoenimine derived from BocLys and MDA, the confirmation of the structures of C and D was undertaken. They were purified by Amberlite XAD-2 and repeated reversed-phase HPLC. During the purifications, C and D were found interconvertible. Therefore, we subjected them as a mixture to structure determination. On the basis of the NMR spectroscopy (1H, 13C, DEPT, 1H-1H COSY, HMQC, and HMBC), they were confirmed to be the two geometrical isomers (ca. 2:1) of (Boc-Lys)2MDA as shown in Figure 2. The isomerization probably originated from the 1,2-CdC bond or 3-CdN bond. Table 1 shows selected 13C and 1H NMR data for (Boc-Lys)2MDA. The 1H NMR spectrum showed four resonances, δH 5.54 (1H, t), 7.56 (1H, d), 7.65 (2/3H, d), and 7.66 (1/3H, d) in the low field region. From their relative peak areas, δH 5.54 and 7.56 probably merged the resonance of the corresponding minor isomer. The 1H-1H COSY spectrum revealed that δ 5.54 coupled with H δH 7.56, 7.65, and 7.66, which were not coupled to each other. This indicated that δH 5.54 corresponded to H-2. The other resonances, δH 7.56, 7.65, and 7.66, which could not be assigned, corresponded to H-1 or H-3. The stereochemistry (cis and trans) of the double bond between C-1 and C-2 could not be deduced from the coupling constant (J ) 11-13 Hz) between H-1 and H-2, because it was in the overlap range of Jcis and Jtrans. In the 13C NMR spectrum, complete signal resolution of C-1, C-2, and C-3 of the isomers was achieved. The HMQC correlations of δC 88.73 and 90.57 to δH 5.54 (H-2) suggested that they corresponded to C-2. The resonances of C-1 and C-3 appeared at δC 160.71, 162.27, 164.06, and 165.55, which could not be assigned. Among them, δC 160.71, 162.27, and 165.55 exhibited HMBC cross-peaks with δH 3.20-3.45 (4H, m, CH2NHCHdCHCHdNCH2), suggest-

Communications

Chem. Res. Toxicol., Vol. 14, No. 5, 2001 475

ing that two -amino groups were connected to C-1 and C-3, respectively. In this study, we succeeded in obtaining the aminoenimine [(Boc-Lys)2MDA] from the reaction of MDA with the lysine derivative (Boc-Lys) and unambiguously confirmed its structure. Although Chio and Tappel demonstrated that the aminoenimines derived from the reaction of MDA with primary amines exhibited fluorescence at ex/em 370-400/450-470 nm, the purified (Boc-Lys)2MDA exhibited no fluorescence. In addition, its UV absorption maximum was at 299 nm, which is quite different from those reported by them (6). They reported that an aminoenimine formed from NR-acetyl-L-lysine and MDA showed fluorescence at ex/em 395/470 nm and an absorption maximum at 395 nm. Kikugawa et al. later proposed 1,4-dihydropyridine-3,5-dicarbaldehydes as the MDAderived fluorophores. The dihydropyridines exhibit fluorescence around ex/em 390/470 nm, which is in good agreement with those of MDA-modified proteins. It has been fairly clear that the original conclusions by Chio and Tappel were incorrect and the fluorescence they observed was not due to the aminoenimines, but probably due to the dihydropyridines. However, there has been no conclusive evidence that the aminoenimines are nonfluorescent. Recently, there is increasing evidence that lipid peroxidation-derived aldehydes other than MDA are involved in the fluorophore formation of lipofuscin, and several models for the fluorophores have been reported (12-18). This is because that the fluorescence characteristics of the dihydropyridines are different from those of the fluorophores derived from the peroxidized systems (19, 20). It has been generally believed that aminoenimines are so labile to even mild acid hydrolysis that they could not be isolated without stabilization by reduction. However, we found that (Boc-Lys)2MDA was stable even under a mild acidic condition. In the early stage of the reaction of Boc-Lys with MDA, (Boc-Lys)2MDA could not be detected by HPLC. Therefore, it is conceivable that the failure to observe the formation of the aminoenimines from the reaction of MDA with simple lysine derivatives has not been due to their instability, but probably due to their low yields. In summary, the present study demonstrates that aminoenimines do not contribute to the fluorescence formation in lipofuscin. In addition, the finding that the aminoenimine is stable enough to be isolated provides a new avenue to probe the involvement of MDA in in vivo protein cross-links by lipid peroxidation. Supporting Information Available: 1H NMR, 13C NMR, DEPT (90° and 135°), 1H-1H COSY, HMQC, and HMBC spectra. These materials are available free of charge via the Internet at http://pubs.acs.org.

References (1) Kikugawa, K., and Beppu, M. (1987) Involvement of lipid oxidation products in the formation of fluorescent and cross-linked proteins. Chem. Phys. Lipids 44, 277-296. (2) Tsuchida, M., Miura, T., and Aibara, K. (1987) Lipofuscin and lipofuscin-like substances. Chem. Phys. Lipids 44, 297-325.

(3) Yin, D. (1996) Biochemical basis of lipofuscin, ceroid, and age pigment-like fluorophores. Free Radical Biol. Med. 21, 871-888. (4) 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. (5) Chio, K. S., and Tappel, A. L. (1969) Synthesis and characterization of the fluorescent products derived from malonaldehyde and amino acids. Biochemistry 8, 2821-2827. (6) Chio, K. S., and Tappel, A. L. (1969) Inactivation of ribonuclease and other enzymes by peroxidizing lipids and by malonaldehyde. Biochemistry 8, 2827-2833. (7) Requena, J. R., Fu, M. X., Ahmed, M. U., Jenkins, A. J., Lyons, T. J., Baynes, J. W., and Thorpe, S. R. (1997) Quantification of malondialdehyde and 4-hydroxynonenal adducts to lysine residues in native and oxidized human low-density lipoprotein. Biochem. J. 322, 317-325. (8) Kikugawa, K., Machida, Y., Kida, M., and Kurechi, T. (1981) Studies on peroxidized lipids. III. Fluorescent pigments derived from the reaction of malonaldehyde and amino acids. Chem. Pharm. Bull. 29, 3003-3011. (9) Kikugawa, K., and Ido, Y. (1984) Studies on peroxidized lipids. V. Formation and characterization of 1,4-dihydropyridine-3,5dicarbaldehydes as model of fluorescent components in lipofuscin. Lipids 19, 600-608. (10) Itakura, K., Uchida, K., and Osawa, T. (1996) A novel fluorescent malondialdehyde-lysine adduct. Chem. Phys. Lipids 84, 75-79. (11) Nair, V., Vietti, D. E., and Cooper, C. S. (1981) Degenerative chemistry of malondialdehyde. Structure, stereochemistry, and kinetics of formation of enaminals from reaction with amino acids. J. Am. Chem. Soc. 103, 3030-3036. (12) Chen, P., Wiesler, D., Chmelı´k, J., and Novotny, M. (1996) Substituted 2-hydroxy-1,2-dihydropyrrol-3-ones: fluorescent markers pertaining to oxidative stress and aging. Chem. Res. Toxicol. 9, 970-979. (13) Itakura, K., Osawa, T., and Uchida, K. (1998) Structure of a fluorescent compound formed from 4-hydroxy-2-nonenal and NR-hippuryllysine: a model for fluorophores derived from protein modifications by lipid peroxidation. J. Org. Chem. 63, 185187. (14) Xu, G., and Sayre, L. M. (1998) Structural characterization of a 4-hydroxy-2-alkenal-derived fluorophore that contributes to lipoperoxidation-dependent protein cross-linking in aging and degenerative disease. Chem. Res. Toxicol. 11, 247-251. (15) Xu, G., Liu, Y., and Sayre, L. M. (1999) Independent synthesis, solution behavior, and studies on the mechanism of formation of a primary amine-derived fluorophore representing cross-linking of proteins by (E)-4-hydroxy-2-nonenal. J. Org. Chem. 64, 57325745. (16) Tsai, L., Szweda, P. A., Vinogradova, O., and Szweda, L. I. (1998) Structural characterization and immunochemical detection of a fluorophore derived from 4-hydroxy-2-nonenal and lysine. Proc. Natl. Acad. Sci. U.S.A. 95, 7975-7980. (17) Itakura, K., Oya-Ito, T., Osawa, T., Yamada, S., Toyokuni, S., Shibata, N., Kobayashi, M., and Uchida, K. (2000) Detection of lipofuscin-like fluorophore in oxidized human low-density lipoprotein: 4-hydroxy-2-nonenal as a potential source of fluorescent chromophore. FEBS Lett. 473, 249-253. (18) Xu, G., Liu, Y., and Sayre, L. M. (2000) Polyclonal antibodies to a fluorescent 4-hydroxy-2-nonenal (HNE)-derived lysine-lysine cross-link: characterization and application to HNE-treated protein and in vitro oxidized low-density lipoprotein. Chem. Res. Toxicol. 13, 406-413. (19) Kikugawa, K., Kato, T., and Hayasaka, A. (1991) Formation of dityrosine and other fluorescent amino acids by reaction of amino acids with lipid hydroperoxides. Lipids 26, 922-929. (20) Alaiz, M., Beppu, M., Ohishi, K., and Kikugawa, K. (1994) Modification of delipidated apoprotein B of low density lipoprotein by lipid oxidation products in relation to macrophage scavenger receptor binding. Biol. Pharm. Bull. 17, 51-57.

TX010055U