Enantioselective Oxidation of trans-4-Hydroxy-2 ... - Semantic Scholar
Chem. Res. Toxicol. 2007, 20, 887-895
887
Enantioselective Oxidation of trans-4-Hydroxy-2-Nonenal Is Aldehyde Dehydrogenase Isozyme and Mg2+ Dependent Jiri Brichac,†,‡ Kwok Ki Ho,§ Ales Honzatko,† Rongying Wang,| Xiaoning Lu,| Henry Weiner,§ and Matthew J. Picklo, Sr.*,†,⊥ Departments of Pharmacology, Physiology, and Therapeutics, of Chemistry, and of Pathology and Proteomics Core Facility, UniVersity of North Dakota, Grand Forks, North Dakota 58202-9024, and Department of Biochemistry, Purdue UniVersity, West Lafayette, Indiana 47907-2063 ReceiVed February 12, 2007
trans-4-Hydroxy-2-nonenal (HNE) is a cytotoxic R,β-unsaturated aldehyde implicated in the pathology of multiple diseases involving oxidative damage. Oxidation of HNE by aldehyde dehydrogenases (ALDHs) to trans-4-hydroxy-2-nonenoic acid (HNEA) is a major route of metabolism in many organisms. HNE exists as two enantiomers, (R)-HNE and (S)-HNE, and in intact rat brain mitochondria, (R)-HNE is enantioselectively oxidized to HNEA. In this work, we further elucidated the basis of the enantioselective oxidation of HNE by brain mitochondria. Our results showed that (R)-HNE is oxidized enantioselectively by brain mitochondrial lysates with retention of stereoconfiguration of the C4 hydroxyl group. Purified rat ALDH5A enantioselectively oxidized (R)-HNE, whereas rat ALDH2 was not enantioselective. Kinetic data using (R)-HNE, (S)-HNE, and trans-2-nonenal in combination with computer-based modeling of ALDH5A suggest that the selectivity of (R)-HNE oxidation by ALDH5A is the result of the carbonyl carbon of (R)-HNE forming a more favorable Bu¨rgi-Duntiz angle with the active site cysteine 293. The presence of Mg2+ ions altered the enantioselectivity of ALDH5A and ALDH2. Mg2+ ions suppressed (R)-HNE oxidation by ALDH5A to a greater extent than that of (S)-HNE. However, Mg2+ ions stimulated the enantioselective oxidation of (R)-HNE by ALDH2 while suppressing (S)-HNE oxidation. These results demonstrate that enantioselective utilization of substrates, including HNE, by ALDHs is dependent upon the ALDH isozyme and the presence of Mg 2+ ions. Introduction trans-4-Hydroxy-2-nonenal (HNE)1 is a cytotoxic R,βunsaturated aldehyde implicated in the pathology of multiple diseases involving oxidative damage (1-3). In the central nervous system, elevated levels of HNE and HNE-protein adducts occur in Alzheimer’s disease, Parkinson’s disease, and cerebral ischemia (4-7). The chemical basis of HNE toxicity is related to its electrophilic character and the capability to alkylate cellular nucleophiles via formation of Michael adducts with cysteinethiols, histidinylamines, and lysylamines and by formation of Schiff bases with lysylamines (8-11). Detoxification of HNE proceeds through multiple routes (1214). Oxidation of HNE to trans-4-hydroxy-2-nonenoic acid (HNEA; Figure 1A) via aldehyde dehydrogenases (ALDHs) is a major route of metabolism in many systems including smooth muscle cells, hepatocytes, astrocytes, and mitochondria from liver and brain (12-14). In isolated brain and liver mitochondria, * To whom correspondence should be addressed. Phone: (701) 7772293. Fax: (701) 777-0387. E-mail: [email protected]. † Department of Pharmacology, Physiology, and Therapeutics, University of North Dakota. ‡ On leave from the Department of Analytical Chemistry, Charles University, Prague 12843, Czech Republic. § Purdue University. | Department of Pathology and Proteomics Core Facility, University of North Dakota. ⊥ Department of Chemistry, University of North Dakota. 1 Abbreviations: HNEA, trans-4-hydroxy-2-nonenoic acid; HNE, trans4-hydroxy-2-nonenal; 4-ONE, trans-4-oxo-2-nonenal; NEN, trans-2-nonenal; ALDHs, aldehyde dehydrogenases; SPE, solid-phase extraction; DTT, dithiothreitol; BSA, bovine serum albumin; MALDI, matrix-assisted laser desorption ionization.
ALDH2 and ALDH5A (succinic semialdehyde dehydrogenase) are known to oxidize HNE (15, 16). These enzymes are active as homotetramers of ∼54 kDa subunits and are NAD+ dependent. These enzymes have multiple differences regarding substrate specificity, cellular expression, and levels of expression, however. Rat ALDH2 and rat ALDH5A share only 52% amino acid similarity by BLAST2 analysis and have very different substrate specificities (17, 18). ALDH5A is expressed in neurons, whereas ALDH2 is expressed in astrocytes (19, 20). In brain mitochondria, ALDH5A activity compromises the majority (>90%) of HNE oxidation, whereas in hepatic mitochondria, ALDH2 compromises about 65% of HNE oxidation activity (16). Although ALDH5A activity is 1.5-fold greater in isolated rat brain mitochondria than in isolated rat hepatic mitochondria, ALDH2 activity in isolated brain mitochondria is less than 6% than that of isolated hepatic mitochondria (16). HNE possesses a chiral center at C4 and exists as two enantiomers, (R)-HNE and (S)-HNE, with potentially different biochemical reactivities (Figure 1B). The stereochemistry of HNE has biological significance. We and others have shown that antibodies are formed in vivo that distinguish adducts derived from the separate HNE enantiomers (21, 22). (S)-HNE enantioselectively inactivates glyceraldehyde-3-phosphate dehydrogenase (GAPDH), an enzyme important for glycolysis and the apoptotic cascade that is also a component of neurofibrillary tangles found in Alzheimer’s disease (23-25). In rat and guinea pig liver cytosol HNE is detoxified S-selectively by GSH conjugation (23, 26). In rat liver cytosol ALDH-mediated oxidation is not enantioselective, whereas oxidation is Rselective in guinea pig liver cytosol (23, 26). We have shown in respiring rat brain mitochondria, that (R)-HNE is metabolized
10.1021/tx7000509 CCC: $37.00 © 2007 American Chemical Society Published on Web 05/05/2007
888 Chem. Res. Toxicol., Vol. 20, No. 6, 2007
at significantly higher rates than (S)-HNE, mainly by enantioselective oxidation to HNEA (22). In this work, we further elucidated the basis of enantioselective oxidation of HNE by brain mitochondria utilizing brain mitochondrial lysate, recombinant ALDH5A, and recombinant ALDH2. Our results show that the enantioselective oxidation of (R)-HNE occurs by rat brain mitochondria without alteration of stereoconfiguration. Purified rat ALDH5A enantioselectively oxidized (R)-HNE, wheras rat ALDH2 was not enantioselective. However, the presence of Mg2+ ions altered the enantioselectivity of both enzymes. Computer-assisted modeling studies suggest that the higher activity of ALDH5A toward (R)-HNE is the result of a more favorable Bu¨rgi-Dunitz angle formed between the active site cysteine of ALDH5A and the carbonyl carbon of (R)-HNE.
Experimental Procedures Chemicals. Racemic HNE, individual HNE enantiomers, d11HNEA, and HNEA were synthesized as described previously (22, 27, 28). The optical purity of HNE enantiomers was 96% for (S)HNE and 98% for (R)-HNE. HNEA enantiomers were synthesized by oxidation of the individual, purified HNE enantiomers. The products were purified by flash chromatography or solid-phase extraction (SPE), and the structures were confirmed by NMR or mass spectrometry (data not shown). NAD+ and trans-2-nonenal were purchased from Sigma-Aldrich (St. Louis, MO). All other chemicals were analytical or HPLC grade. trans-4-oxo-2-nonenal (4-ONE) was obtained from Cayman Chemicals (Ann Arbor, MI). Preparation of Rat Brain Mitochondria. Sprague-Dawley rats (male, 250-300 g) were purchased from Charles River (Wilmington, MA). Experimental protocols were in accordance with the NIH guidelines for the use of live animals and were approved by the University of North Dakota Institutional Animal Care and Use Committee. Brain mitochondria were isolated as described previously, resulting in synaptic and nonsynaptic mitochondria (29). The final mitochondrial pellets were resuspended in lysis buffer (pH 7.4) containing 20 mM sodium phosphate, 0.5% Triton X-100 (Alfa Aesar, Ward Hill, MA), 10 mM dithiothreitol (DTT; Fisher Scientific, Fair Lawn, NJ), 1 mM diethylenetriaminepentaacetic acid (Acros Organics, Geel, Belgium), and 50 µM 2,6-di-tert-butyl-4methylphenol (Acros Organics) (16). The solution was then diluted 1:1 with glycerol and stored at - 80 °C. Determination of Apparent Kinetic Parameters of HNEA Formation by Rat Brain Mitochondrial Lysate. Mitochondria were lysed by freezing/thawing and passed over a Sephadex G25 (Amersham Biosciences, Uppsala, Sweden) column (15-fold volume over sample) equilibrated with lysis buffer without DTT to remove potential endogenous aldehydes (30), cofactors, glycerol, and DTT. The protein concentration was determined using protein assay reagent (Bio Rad, Hercules, CA) with BSA as a standard (31). Mitochondria at a final concentration 0.2 mg/mL were incubated in 40 mM phosphate buffer (pH 7.4) containing 1 mM NAD+ at 37 °C. The total volume of the reaction mixture was 150 µL. Assays were performed in triplicate using five different concentrations of either (R)-HNE or (S)-HNE. The reaction was stopped after 3 min by addition of a 1/10 sample volume of cold 3% (w/w) phosphoric acid, and the tube was immediately put on ice (16, 22, 29). The HNEA content was measured by HPLC with UV detection. No detectable endogenous HNE and HNEA concentrations were present in mitochondrial lysate. Determination of HNEA. HNEA formed by the mitochondrial lysates was measured by RP-HPLC (Schimadzu Class-VP 7.2 system, Kyoto, Japan) with UV detection at 210 nm (16, 29). Because of the low KM of ALDH2 for HNE (see the Results), we measured the enzyme activity via monitoring the formation of HNEA by LC-MS/MS when determining KM and Vmax parameters as described previously with some modifications (22). The LCMS/MS system consisted of an Agilent 1100 series HPLC system
Brichac et al.
Figure 1. (A) ALDH-mediated oxidation of racemic HNE. (B) Structures of R,β-unsaturated aldehydes used in these studies: aldehyde dehydrogenase (ALDH), trans-4-hydroxy-2-nonenal (HNE), trans-4hydroxy-2-nonenoic acid (HNEA), trans-4-oxo-2-nonenal (4-ONE), trans-2-nonenal (NEN).
(Agilent Technologies, Santa Clara, CA) coupled to an API 3000 LC-MS/MS system (MDS Sciex, Concord, Canada). A 10 µL portion of the sample was injected onto a chromatographic column, Luna, 3 µm, C8(2), 150 × 2.1 mm i.d. (Phenomenex, Torrance, CA). HNEA eluted at 4.4 min in isocratic mode in the mobile phase, ACN-30 µM ammonium acetate, pH 3.8 (45:55, v/v), at a flow rate of 150 µL/min. The limit of quantitation for HNEA was below 30 fmol. Analysis of HNEA Stereochemistry. Oxidation of 160 µM (R)HNE or (S)-HNE by mitochondrial lysate was performed as described above. However, the reaction time was 10 min, and the volume of the reaction mixture was 2.4 mL. Prior to the separation of HNEA enantiomers by HPLC, the samples were purified by SPE using Oasis HLB 6 cm3 (0.2 g) extraction cartridges (Waters, Milford, MA). The cartridges were conditioned with 7 mL of methanol (MeOH) and equilibrated with 7 mL of deionized water. The sample was loaded onto the cartridge and washed with 7 mL of MeOH-100 mM phosphate buffer, pH 2 (40:60, v/v), and with 7 mL of water. HNEA was eluted by MeOH-100 mM phosphate buffer, pH 7 (50:50, v/v). The first 0.8 mL of the eluent was discarded, and the next 2.5 mL fraction containing HNEA was collected. A second SPE was used to concentrate HNEA and to transfer the sample into MeOH. MeOH was evaporated by a nitrogen evaporator, and the residue was dissolved in 300 µL of the mobile phase. HNEA enantiomers were separated by HPLC using a chiral column, Chiralcel OB, 250 mm × 4.6 mm (Daicel, Osaka, Japan), and a mobile phase of hexane-2-propanol (97:3, v/v). The flow rate was 0.8 mL/min, 40 µL of sample was injected, and 210 nm was used for UV detection. The method was used for the determination of the HNEA enantiomer ratio only. RP-HPLC was used for quantification of HNEA. Baseline resolution was achieved (R ) 1.83), and the retention time was 11.0 min for (S)-HNEA and 12.9 min for (R)-HNEA. The limit of detection (LOD) was 4.8 µM (192 pmol) for (S)-HNEA and 5.4 µM (216 pmol) for (R)HNEA. Preparation of Recombinant Rat ALDH5A and ALDH2. Recombinant rat ALDH5A and rat ALDH2 containing N-terminal polyhistidine tags were prepared as described previously using Rosetta 2 (DE3) Escherichia coli cells (Novagen, Madison, WI) (32). Following nickel agarose affinity chromatography, the purified enzymes (ALDH2 or ALDH5A) were dialyzed at 4 °C versus two changes of buffer (1:1000), 0.15 M NaCl, 40 mM sodium phosphate buffer, pH 7.4, containing 100 µM DTT to preserve the enzyme activity. Proteins of just over 50 kDa and of >95% purity were obtained as judged by SDS-PAGE analysis. Determination of Kinetic Parameters for Rat ALDH5A and ALDH2. Enzyme activity, unless stated otherwise, was measured by monitoring NADH formation at 340 nm under the initial rate conditions (16). Assays were performed with dialyzed protein on a microplate spectrophotometer, SpectraMax plus 384 (Molecular Devices, Sunnyvale, CA). The absorption coefficient of NADH was 6220 M-1 cm1. Assays were performed in a 300 µL reaction mixture containing 1 mM NAD+ in 40 mM phosphate buffer (pH 7.4) at 37 °C (16). For experiments in which the effects of Mg2+ ions
EnantioselectiVe Oxidation of HNE were examined, HEPES-free acid was substituted for the sodium phosphate. Mass Determination for Recombinant ALDH2. A 10 µL portion of ALDH2 protein sample (1 mg/mL) in 40 mM sodium phosphate was loaded onto C4 ziptip columns (Millipore), washed with 0.1% TFA, and eluted with 5 µL of 70% ACN with 0.1% TFA. Typically, 0.4 µL of the eluted protein sample was mixed on a matrix-assisted laser desorption ionization (MALDI) plate with 0.4 µL of matrix solution, 50 mM R-cyano-4-hydroxycinnamic acid (CHCA) in 50% ACN containing 0.1% TFA. The mass measurement was carried out on a 4800 MALDI-TOF/TOF mass spectrometer (Applied Biosystems, Foster City, CA) equipped with a Nd:YAG 200 Hz laser. The instrument was operated with delayed extraction in linear positive ion mode. The mass spectra of the recombinant ALDH2 protein were acquired by averaging 20 subspectra of 500 total laser shots and internally calibrated using bovine serum albumin (BSA) as the standard reference. Modeling and Substrate Docking for ALDH5A. The 3D model for ALDH5A (SwissProtein database accession number P51650) was generated using Geno3D (33), an automated Web server for protein molecular modeling (http://geno3d-pbil.ibcp.fr). The ALDH5A model was built on the basis of the solved crystal structure of the sheep cytosolic aldehyde dehydrogenase (ALDH1) (Protein Data Bank ID 1BXS). The putative NAD+ binding site in the ALDH5A model was identified by superposing the sheep ALDH1 structure onto the ALDH5A model. The 3D atomic coordinates of the trans2-nonenal and the (R)- or (S)-HNE in pdb format were obtained using the Dundee Prodrg2 Server (http://davapc1.bioch.dundee.ac.uk/ programs/prodrg/prodrg.html) (34). For docking of trans-2-nonenal and (R)- or (S)-HNE into the ALDH5A model, we employed the ArgusLab software, which makes use of the AScore scoring function to determine the low-energy binding mode with a grid encompassing C293 as a binding group. Molecular graphics images were produced using the UCSF Chimera software (http://www. cgl.ucsf.edu/chimera/) (35). Statistical and Curve Fitting Analyses. Vmax and KM were determined by nonlinear regression analysis using the MichaelisMenten equation V ) (Vmax[S])/(KM + [S]). Statistical significance between data sets was tested by two-way ANOVA with the Bonferronni posttest and was achieved when p < 0.05 (Prism 4, GraphPad Software, San Diego, CA).
Results Previously, we had shown that (R)-HNE was oxidized to a greater extent than (S)-HNE by respiring rat brain mitochondria. In this first experiment, we determined whether the stereoselectivity of HNE oxidation was retained in lysed brain mitochondria passed through G25 Sephadex. We performed this step to provide a crude preparation of brain mitochondrial ALDHs so that we could monitor HNE oxidation in brain mitochondria without confounds such as substrate and cofactor availability. The apparent KM and Vmax for (R)-HNE were 27.8 ( 8.5 µM and 8.72 ( 0.90 (nmol/min)/mg, respectively, and those for (S)HNE 50.9 ( 16.4 µM and 4.4 ( 0.6 (nmol/min)/mg (Figure 2). Using a comparison of Vmax/KM as a measure of efficiency, these data show that the ALDHs present in rat brain mitochondria oxidize (R)-HNE 3.7-fold more efficiently than (S)-HNE. There was no consumption of (R)-HNEA or (S)-HNEA by lysed mitochondria supplemented with NAD+ (not shown), demonstrating that selective formation of (R)-HNEA was not an artifact resulting from enantioselective metabolism of (S)-HNEA. Racemization of chiral molecules can occur enzymatically and non-enzymatically in biological and non-biological systems (36-39). During the formation of HNE, racemization or retention of the stereoconfiguration of the hydroxyl group is dependent upon which carbon of the hydroperoxide precursor was present (40, 41). To test whether the absolute configuration of the HNE C4 carbon was retained during the enzymatic
Chem. Res. Toxicol., Vol. 20, No. 6, 2007 889
Figure 2. (R)-HNE is selectively oxidized in brain mitochondrial lysate. Mitochondrial lysate was incubated with increasing concentrations of either (R)-HNE or (S)-HNE in the presence of 1 mM NAD+ at 37 °C for 3 min. The amount of HNEA formed by HNE oxidation was measured by HPLC as described in the Experimental Procedures. The data are presented as the mean ( SEM of three experiments performed in triplicate on three separate days (n ) 3). The nonlinear regression lines for the determination of KM and Vmax are given. The apparent KM and Vmax for (R)-HNE were 27.8 ( 8.5 µM and 8.72 ( 0.90 (nmol/ min)/mg and those for (S)-HNE 50.9 ( 16.4 µM and 4.4 ( 0.6 (nmol/ min)/mg.
oxidation of HNE, we developed an HPLC method using the chiral column Chiralcel OB for the separation of HNEA enantiomers (Figure 3). Using this method, we determined that the optical purity of (S)-HNEA formed by (S)-HNE oxidation by brain mitochondria was 99% and that of (R)-HNEA formed from (R)-HNE was 98%. These data demonstrate that racemization of the C4 carbon did not occur during the oxidation step. We next examined the enantioselectivity of HNE oxidation of ALDH5A and ALDH2, the two known mitochondrial ALDHs that oxidize HNE, using recombinant rat ALDH2 and ALDH5A. The purity of the resulting ALDHs is shown in Figure 4A. The kinetic parameters of acetaldehyde and succinic semialdehyde utilization by ALDH2 and ALDH5A are presented in Table 1. Our previous data suggested that ALDH5A was the major source of HNE oxidation activity in brain mitochondria. Therefore, we tested the hypothesis that ALDH5A preferentially oxidized (R)HNE (Figure 4B). Rat ALDH5A oxidized (R)-HNE with a KM of 20.4 ( 3.3 µM, a Vmax of 1.5 ( 0.1 (µmol/min)/mg, and a kcat of 1.4 s-1. On the other hand, (S)-HNE had a KM of 32 ( 6.1 µM, a Vmax of 0.42 ( 0.03 (µmol/min)/mg, and a kcat of 0.38 s-1. A comparison of kcat/KM demonstrated that (R)-HNE (kcat/KM ) 68.6 × 103 s-1/M) was oxidized 5.8-fold more efficiently than (S)-HNE (kcat/KM ) 11.8 × 103 s-1/M). We used trans-2-nonenal as a substrate as it lacks the hydroxyl group at the C4 position (Figure 1B). Interestingly, trans-2-nonenal had kinetic characteristics similar to those of (S)-HNE with a KM of 43.3 ( 7.8 µM, a Vmax of 0.52 ( 0.04 (µmol/min)/mg, and a kcat of 0.47 s-1. We attempted to determine whether fixing the oxygen group at C4 by using trans-4-oxo-2-nonenal (4-ONE) as a substrate would affect the activity. Unfortunately, this substrate even at low concentrations (5 µM) inhibited the ALDH5A activity (not shown) likely as a result of irreversible inhibition as has been noted for 4-ONE and human ALDH2 (42). Our previous data suggested that ALDH2 played a minor role in HNE oxidation in rat brain mitochondria (16). However, the finding that purified rat ALDH5A had a higher selectivity toward (R)-HNE than in brain mitochondrial lysate prompted us to examine the kinetic parameters of recombinant rat ALDH2 with the enantiomers of HNE. Our initial experiments using UV spectrophotometric measurement of NADH formation demonstrated that saturation of enzyme activity with (R)-HNE and (S)-HNE had already occurred at a 5 µM concentration of
890 Chem. Res. Toxicol., Vol. 20, No. 6, 2007
Figure 3. The stereochemistry of the C4 hydroxyl is retained during HNE oxidation. (A) Separation of (R)-HNEA and (S)-HNEA by normalphase chromatography using a Chiralcel-OB chiral column. (B) Sample loading of HNE and HNEA onto the SPE cartridge. Note that both HNEA (open circles) and HNE (closed squares) are completely retained. Because HNEA and HNE enantiomers coeluted on the Chiracel OB column, selective elution (C) with pH 7 buffer from the SPE cartridge was able to separate HNEA (open circles) versus HNE (closed squares). (D) HNE enantiomers were oxidized separately in lysates of brain mitochondria supplemented with NAD+. The ratio of the HNEA enantiomers was determined by chiral-phase HPLC. (S)-HNEA was not obtained by oxidation of (R)-HNE, and (R)-HNEA was not obtained following oxidation of (S)-HNE. Data are the mean ( SEM of three independent experiments (n ) 3).
either enantiomer, suggesting very low KM values (not shown). To determine the kinetic parameters for (R)-HNE and (S)-HNE, we therefore monitored HNEA formation by LC-MS/MS. (R)HNE was oxidized with a KM of 1.0 ( 0.1 µM, a Vmax of 42 ( 2 (nmol/min)/mg, and a kcat of 0.039 s-1. (S)-HNE was oxidized with a KM of 0.70 ( 0.1 µM, a Vmax of 43 ( 1 (nmol/min)/mg, and a kcat of 0.040 s-1 (Figure 4C). The kcat/KM for (R)-HNE was 39 × 103 s-1/M and that for (S)-HNE 57 × 103 s-1/M for ALDH2. For calculation of kcat, a mass of 56.5 kDa per subunit was obtained by MALDI-TOF/TOF mass spectrometry for the recombinant, polyhistidine-tagged rat ALDH2.
Brichac et al.
Figure 4. Enantioselective oxidation of HNE by ALDH5A but not ALDH2. (A) SDS-PAGE analysis of recombinant rat ALDH5A (lane 1, 4 µg) and ALDH2 (lane 2, 4 µg). Lane 3 contained molecular weight markers with the pertinent weights provided. (B) Oxidation of (R)HNE, (S)-HNE, or NEN by ALDH5A. NADH formation was used as a measure of activity. A 3 µg portion of ALDH5A was used in the reactions. (C) Nonenantioselective oxidation of HNE by ALDH2. Owing to the low activity and low KM, the formation of HNEA was used as a measure of the activity and was quantified by LC-MS/MS. A 180 ng portion of ALDH2 was used in the reactions. Data are expressed as the mean ( SEM of assays performed in triplicate. The assays were performed in 40 mM sodium phosphate buffer with 1 mM NAD+ at pH 7.4 and 37 °C.
We and others have shown that the presence of Mg2+ ions alters the configuration of NAD(H) at the active site of ALDH1 and ALDH2 and thereby affects the activities of these enzymes (43, 44). To gain mechanistic insight into the interaction of HNE chirality and the active sites of ALDH5A and ALDH2, we tested whether the presence of Mg2+ ions would affect the oxidation rates of (R)-HNE and (S)-HNE (Figure 5). Vmax concentrations of the aldehydes were used. Mg2+ ions inhibited the oxidation
EnantioselectiVe Oxidation of HNE
Chem. Res. Toxicol., Vol. 20, No. 6, 2007 891
Figure 5. Mg2+ ions alter the enantioselectivity of HNE oxidation by ALDH5A and ALDH2. ALDH5A (A,B) or ALDH2 (C,D) was incubated with NAD+ (1 mM) and Mg2+ ions (as MgCl2) for 1 min prior to the addition of the aldehyde substrate. For comparison, plots of the total rates and percentage of the control activity (no Mg2+ ions added) are provided. Note that, for ALDH5A, (R)-HNE oxidation is more sensitive to inhibition by Mg2+ ions (p < 0.05 by two-way ANOVA) whereas Mg2+ ions stimulate ALDH2-mediated (R)-HNE oxidation over 5-fold. Data are expressed as the mean ( SEM of assays performed in triplicate. The assays were performed in 40 mM HEPES buffer, pH 7.4, at 37 °C. For ALDH5A, 120 µM (R)-HNE or (S)-HNE was used. For ALDH2, 50 µM (R)-HNE, (S)-HNE, or NEN was used. Preliminary experiments demonstrated that the KM of NEN for ALDH2 was less than 5 µM (not shown). For ALDH5A activities, 4 µg of protein was used. For ALDH2 activities, 15 µg of protein was used. Reaction volumes were 300 µL. Table 1. Kinetic Parameters of Recombinant ALDH2 and ALDH5A with Acetaldehyde and Succinic Semialdehyde enzyme
KM (µM)
ALDH2 succinic semialdehyde 780 ( 100 acetaldehyde 0.4 ( 0.1 ALDH5A succinic semialdehyde 3.5 ( 0.9 acetaldehyde 5800 ( 300
Vmax kcata [(µmol/min)/mg] (s-1) 0.116 ( 0.01 0.280 ( 0.01 38 ( 1.2 3.9 ( 0.1
kcat/KM (s-1/M)
0.11 140 0.26 6.5 × 105 34 3.5
9.7 × 106 603
a For calculation of k , a mass of 54.4 kDa for the recombinant 6×Hiscat tagged ALDH5A was used as determined by MALDI-TOF mass spectrometry (16). For recombinant 6×His-tagged ALDH2, a subunit mass of 56.5 kDa was used. Data are derived from assays performed in triplicate. KM and Vmax were determined by nonlinear regression analysis with the data fit to the Michaelis-Menten equation. KM and Vmax are presented as the best fit value ( the standard error. Assays were performed at 37 °C in 40 mM sodium phosphate buffer (pH 7.4) with a 50 µM concentration of either aldehyde and 1 mM NAD+.
of (R)-HNE and (S)-HNE by rat ALDH5A in a concentrationdependent manner; however, the oxidation of (R)-HNE was inhibited to a slightly, but significantly, greater extent than (S)HNE. On the other hand, the oxidation of HNE enantiomers by rat ALDH2 responded differently to the presence of Mg2+ ions (Figure 5C,D). Both (R)-HNE and (S)-HNE oxidation rates were greatly elevated by the presence of a low concentration of Mg2+ ions (0.1 mM), but at higher concentrations, the activity declined, consistent with what has been shown for human ALDH2 (43). The magnitude of the activity increase was much greater for (R)-HNE than (S)-HNE. Whereas 5 mM MgCl2 inhibited the oxidation of both HNE enantiomers by 25-30% for ALDH5A, different effects were observed for ALDH2 in
which (S)-HNE oxidation was suppressed to 50% of the control values and (R)-HNE oxidation was still 2-fold higher than the control values. The oxidation of trans-2-nonenal showed a similar biphasic response to Mg2+ ions; however, the rate of oxidation at 5 mM Mg2+ ions was still higher than the control (no Mg2+ ions) values. Last, we attempted to gain insight into the structural basis for ALDH5A enantioselectivity. Since the crystal structure for rat ALDH5A has not been solved, we modeled the structure of rat ALDH5A (Figure 6) on the solved crystal structure of sheep cytosolic ALDH1. We obtained a model with a “head” and “tail” that are similar to those observed for the tetrameric ALDH2 and ALDH1 and the dimeric rat ALDH3 (45-48). We then fit in (R)-HNE, (S)-HNE, and trans-2-nonenal with NAD+ present in the active site to discern structural interactions that could account for the differences in the utilization of these aldehydes (Figure 7). Our analyses indicate that the low-energy binding mode of trans-2-nonenal was similar to that of (S)HNE (not shown). Although the distances between the active site cysteinyl thiol (C293) and the carbonyl carbon were similar, the Bu¨rgi-Dunitz angle formed among the thiol, the carbonyl carbon, and the carbonyl oxygen was more favorable for (R)HNE (109°) than for trans-2-nonenal (121°) and (S)-HNE (134°) (Table 2). The calculated energy of the reaction for (R)-HNE (-7.4 kcal/mol) was lower than that of (S)-HNE and trans-2nonenal.
Discussion HNE is a toxic, chiral, R,β-unsaturated aldehyde derived from the oxidation of linoleic acid and arachidonic acid (49). (R)-
892 Chem. Res. Toxicol., Vol. 20, No. 6, 2007
Brichac et al.
Figure 6. Proposed model of rat ALDH5A with NAD+. A putative model of the rat ALDH5A was generated from the primary amino acid sequence for rat ALDH5A (minus the mitochondrial targeting sequences) using sheep ALDH1 as a template. The protein has a “head” and “tail” configuration similar to that of ALDH 1 and ALDH2. In this model, NAD+ is docked into the active site. Table 2. Calculated Angles, Distances, and Reaction Energies for ALDH5A and Aldehydesa
aldehyde
Bu¨rgi-Dunitz angle between C293 and CdO carbon (deg)
distance between C293 and CdO carbon (Å)
distance between C4-OH and R166 (Å)
energy of reaction (kcal/mol)
(R)-HNE (S)-HNE NEN
109 134 121
3.2 3.5 3.2
3.9 5.0 N/A
-7.4 -8.1 -9.0
a All aldehydes were modeled in the active site with NAD+ present. NEN ) trans-2-nonenal.
HNE undergoes selective oxidation by brain mitochondria (22). In this work, we further elucidated the basis for the enantioselective oxidation of HNE by brain mitochondria. First, we show that the R-selective oxidation of HNE by brain mitochondria occurs and that, as expected, oxidation occurs without racemization of the C4-hydroxyl group. We then demonstrate that recombinant ALDH5A, but not ALDH2, enantioselectively oxidizes (R)-HNE. Unexpectedly, the presence of increasing Mg2+ ion concentrations altered the enantioselectivity of both enzymes. Last, using computer-assisted modeling, we propose that the higher activity of ALDH5A toward (R)-HNE is the result of a more suitable bond angle being formed between the active site cysteine and the carbonyl carbon of (R)-HNE. Possible explanations for the enantioselectivity exhibited by ALDH5A is that a gating mechanism occurs such that (A) (R)HNE is able to pass to the active site of the enzyme as opposed to (S)-HNE in which the C4-hydroxyl is in the opposite direction or (B) (R)-HNEA is able to exit the active site more easily than (S)-HNEA. These explanations do not account for finding that trans-2-nonenal, lacking the C4-hydroxyl, has kinetic parameters similar to those of (S)-HNE. The data lead us to suggest that (R)-HNE is positioned in a more advantageous manner in the active site. Our modeling data support this hypothesis. These
Figure 7. (R)-HNE and (S)-HNE are positioned differently in the active site of rat ALDH5A. (R)-HNE (A) and (S)-HNE (B) were docked into the putative ALDH5A active site with NAD+ present. The HNE enantiomers have the dark blue fill. For orientation purposes, the carbonyl oxygens on C1 of the HNE enantiomers (pictured as the red ball) are near the T292 and N158 residues, and the C4 oxygens are positioned near M163. Oxygens are represented by the red balls, whereas nitrogens are blue and sulfurs are yellow. The carbonyl carbon of (S)-HNE is positioned closer to N158, the general acid equivalent to N169 in ALDH1 and ALDH2, thus creating a greater angle than that of (R)-HNE with respect to cysteine 293 (C293), which forms a thiohemiacetal with the carbonyl carbon and is equivalent to C302 in the other isozymes (56, 57). Note that the configuration of the (R)HNE molecule itself in the active site is different from that of (S)HNE.
modeling data predict that (R)-HNE forms a more favorable Burgi-Dunitz angle (109°) than that of (S)-HNE or trans-2nonenal. This angle is the angle that is formed among the nucleophile (the thiolate anion of cysteine) and the carbonyl carbon of the aldehyde and its oxygen atom. An angle of 110° is predicted to be the optimum angle of attack of the nucleophile on the carbonyl carbon owing to the presence of the carbonoxygen sp2 electron orbitals (50). A plot of experimentally derived kcat versus the computer-derived Bu¨rgi-Dunitz angle demonstrates a significant relationship between the rate of catalysis and the angle formed (Figure 8). Multiple reasons may exist as to why (R)-HNE is positioned in a more favorable manner with respect to the active site cysteine. A likely possibility is that there is hydrogen bonding between the C4-hydroxyl group of (R)-HNE and a positively charged amino acid at the active site. This also could explain why succinic semialdehyde is utilized highly efficiently by ALDH5A versus acetaldehyde as shown by our data (Table 1)
EnantioselectiVe Oxidation of HNE
Figure 8. The rate of catalysis is related to the Bu¨rgi-Dunitz angle. Experimentally derived kcat values for (R)-HNE, (S)-HNE, and NEN are plotted versus the computer-derived Bu¨rgi-Dunitz angle. The line shown is derived from linear regression analysis of the data and has an r2 of 0.85.
and those of others (18). On the basis of our modeling data, we speculate that the arginine residue at position 166 of ALDH5A is the likely positively charged amino acid. Although the (R)hydroxyl group oxygen is closer to R166 (3.9 Å) than (S)-HNE (5.0 Å), these distances are too long for significant hydrogen bonding to occur unless there is movement in the active site to allow a closer interaction. Interestingly, the residue analogous to R166 at this position in ALDH2 is a nonpolar tryptophan residue. Mutations of this R166 residue with subsequent kinetic analyses are needed. The crystal structure of ALDH5A has not been solved, so the modeling data presented cannot be verified. The known structures of other ALDH isozymes, including sheep ALDH1, human ALDH2, beef ALDH2, rat ALDH3, and cod ALDH9, all have essentially the same structure as some bacterial forms of the enzyme (45-48, 51). Even the structure of retinaldehyde dehydrogenase, like that of the sheep class 1 enzyme, was solved using the coordinates from the class 2 isozyme (46, 52). This near structural identity was found even though there are large differences in the amino acid sequences between family members (45-81% similarity of the above ALDHs when compared to sheep ALDH1). Thus, we are confident that the modeling data are very representative of what actually exists for ALDH5A. Supporting the modeling data is noting that there is a relationship between the kcat and the bond angle (Figure 8). The ALDH2 activity plays a minor role in the oxidation of HNE by brain mitochondria in contrast to liver mitochondria, where ALDH2 plays a major role in metabolizing HNE (16). The lack of enantioselectivity under our initial conditions (with no Mg2+ ions) may be the result of ALDH2 possessing a less hindered, lipophilic binding pocket. The KM values for HNE enantiomers are more than 10-fold lower with Vmax values more than 10-fold higher than those reported for racemic HNE by Mitchell and Petersen using ALDH2 purified from rat liver (15). This likely is the result of the fact that rapidly purified, recombinant enzyme was used. The presence of Mg2+ ions had modest effects upon enantioselective oxidation by ALDH5A (as compared to ALDH2) although (R)-HNE oxidation was inhibited to a significantly greater extent than (S)-HNE oxidation in the presence of Mg2+ ions. Previous work from our laboratories and others demonstrates that Mg2+ ions increase the rate of deacylation for ALDH2, thereby elevating the activity and increasing the binding affinity of NADH to the active site of ALDH1 and thereby decreasing the enzyme activity (43, 44). This observation and others were used to conclude that inhibition of the ALDH activity by the presence of Mg2+ ions indicated that NADH release was the rate-limiting step for enzyme activity.
Chem. Res. Toxicol., Vol. 20, No. 6, 2007 893
Although we did not explore all the enzymatic steps of ALDH5A activity, our current data point to NADH release becoming rate limiting for HNE oxidation since oxidation was suppressed at most concentrations of Mg2+ ions. The addition of a low concentration (100 µM) of Mg2+ ions increased the rates of oxidation of (R)-HNE and (S)-HNE by ALDH2. These data are consistent with previous work demonstrating that Mg2+ ions, at this concentration, increase the rate of the deacylation step by ALDH2 (43). What was not expected, however, was that Mg2+ ions elevated (R)-HNE oxidation to a much larger extent than for (S)-HNE. These data suggest that the presence of Mg2+ ions alters the active site such that (R)-HNE becomes more favorably positioned for formation of the thiohemiacetal with C302 of the active site or hydride transfer or that (R)-HNEA is able to leave the active site at a faster rate than (S)-HNEA during deacylation or a combination of these mechanisms. (R)-HNE oxidation decreased with increasing concentrations of Mg2+ ions but did not return to the control (no Mg2+ ions) activity, whereas (S)HNE oxidation was suppressed to 50% of the control activity. The decrease in the rate of oxidation for both enantiomers is consistent with the finding that Mg2+ ions form a complex with NADH that slows the rate of NADH release from the enzyme (43). The extent to which HNE is metabolized enantioselectively in vivo needs further study. The concentration of free Mg2+ in mitochondria is estimated to be between 0.4 and 1.0 mM (53, 54). Thus, it is possible that the in vivo oxidation of HNE would be R-selective by ALDH2 as well as by ALDH5A. On the other hand, it is not known whether ALDH3A (another ALDH that oxidizes HNE) is enantioselective (55). Although our data show the R-selectivity of HNE oxidation, we think it unlikely that (S)-HNE protein-adducts would then accumulate in cells given the S-selectivity of glutathione S-transferase catalyzed detoxification of HNE (23, 26). Nonetheless, the chirality of HNE adds another level of complexity since immunochemical data derived using (R)-HNE adduct and (S)-HNE adduct antibodies demonstrate that the adducts of the individual enantiomers are distributed differently in cells in vivo following oxidative damage (21). In conclusion, this work demonstrates that chirality plays an important role in the metabolism of HNE and that the enantiomers of HNE serve as useful tools for studying enzymesubstrate interaction. Our experiments show that the enantioselectivity of HNE oxidation by ALDH5A and ALDH2 can be regulated by the Mg2+ ion content. It is not known the extent to which Mg2+ availability alters aldehyde oxidation in vivo or even in intact mitochondria. While this might not be important in regions of brain where it appears that the bulk of the oxidation of HNE is catalyzed by ALDH5, it could be more important in tissue such as liver where the oxidation is primarily oxidized by ALDH2 (16). Regulation of the activity by the Mg2+ ion content may be of consequence in different regions of the central nervous system where these enzymes have different levels of expression and cell-specific expression. Further experiments are needed to determine the extent to which the chirality of HNE influences protein binding and metabolism and the role of Mg2+ ions in regulating HNE metabolism in situ. Acknowledgment. We gratefully acknowledge NIH Grants P20 RR17699-01 COBRE (M.J.P.) from the NCRR and AA15145-01 (M.J.P.) and AA05812 (H.W.) from NIAAA. The Proteomics Core Facility was supported by NIH Grant P20 RR016741 from the INBRE Program of the NCRR. X.L. is the recipient of a faculty initiation grant from NSF EPSCoR (EPS-
894 Chem. Res. Toxicol., Vol. 20, No. 6, 2007
0447679). We thank Mr. Darryl Mosely and Ms. Laura Leiphon for the preparation of the mitochondria.
Brichac et al.
(23)
References (1) 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. (2) Uchida, K. (2003) 4-Hydroxy-2-nonenal: a product and mediator of oxidative stress. Prog. Lipid Res. 42, 318-343. (3) Picklo, M. J., Montine, T. J., Amarnath, V., and Neely, M. D. (2002) Carbonyl toxicology and Alzheimer’s disease. Toxicol. Appl. Pharmacol. 184, 187-197. (4) McKracken, E., Graham, D. I., Nilsen, M., Stewart, J., Nicoll, J. A., and Horsburgh, K. (2001) 4-Hydroxynonenal immunoreactivity is increased in human hippocampus after global ischemia. Brain Pathol. 11, 414-421. (5) Montine, K. S., Olson, S. J., Amarnath, V., Whetsell, W. O., Jr., Graham, D. G., and Montine, T. J. (1997) Immunohistochemical detection of 4-hydroxy-2-nonenal adducts in Alzheimer’s disease is associated with inheritance of APOE4. Am. J. Pathol. 150, 437443. (6) Sayre, L. M., Zelasko, D. A., Harris, P. L., Perry, G., Salomon, R. G., and Smith, M. A. (1997) 4-Hydroxynonenal-derived advanced lipid peroxidation end products are increased in Alzheimer’s disease. J. Neurochem. 68, 2092-2097. (7) Yoritaka, A., Hattori, N., Uchida, K., Tanaka, M., Stadtman, E. R., and Mizuno, Y. (1996) Immunohistochemical detection of 4-hydroxynonenal protein adducts in Parkinson disease. Proc. Natl. Acad. Sci. U.S.A. 93, 2696-2701. (8) Sayre, L. M., Arora, P. K., Iyer, R. S., and Salomon, R. G. (1993) Pyrrole formation from 4-hydroxynonenal and primary amines. Chem. Res. Toxicol. 6, 19-22. (9) Amarnath, V., Valentine, W. M., Montine, T. J., Patterson, W. H., Amarnath, K., Bassett, C. N., and Graham, D. G. (1998) Reactions of 4-hydroxy-2(E)-nonenal and related aldehydes with proteins studied by carbon-13 nuclear magnetic resonance spectroscopy. Chem. Res. Toxicol. 11, 317-328. (10) Nadkarni, D. V., and Sayre, L. M. (1995) Structural definition of early lysine and histidine adduction chemistry of 4-hydroxynonenal. Chem. Res. Toxicol. 8, 284-291. (11) Petersen, D. R., and Doorn, J. A. (2004) Reactions of 4-hydroxynonenal with proteins and cellular targets, Free Radical Biol. Med. 37, 937-945. (12) Alary, J., Gueraud, F., and Cravedi, J. P. (2003) Fate of 4-hydroxynonenal in vivo: disposition and metabolic pathways. Mol. Aspects Med. 24, 177-187. (13) Kubatova, A., Murphy, T. C., Combs, C., and Picklo, M. J., Sr. (2006) Astrocytic Biotransformation of trans-4-Hydroxy-2-nonenal Is DoseDependent. Chem. Res. Toxicol. 19, 844-851. (14) Srivastava, S., Chandra, A., Wang, L. F., Seifert, W. E., Jr., DaGue, B. B., Ansari, N. H., Srivastava, S. K., and Bhatnagar, A. (1998) Metabolism of the lipid peroxidation product, 4-hydroxy-trans-2nonenal, in isolated perfused rat heart. J. Biol. Chem. 273, 1089310900. (15) Mitchell, D., and Petersen, D. (1987) The oxidation of a,b,unsaturated aldehydic products in lipid peroxidation by rat liver aldehyde dehydrogenases. Toxicol. Appl. Pharmacol. 87, 403-410. (16) Murphy, T. C., Amarnath, V., Gibson, K. M., and Picklo, M. J., Sr. (2003) Oxidation of 4-hydroxy-2-nonenal by succinic semialdehyde dehydrogenase (ALDH5A). J. Neurochem. 86, 298-305. (17) Chambliss, K. L., and Gibson, K. M. (1992) Succinic semialdehyde dehydrogenase from mammalian brain: subunit analysis using polyclonal antiserum. Int. J. Biochem. 24, 1493-1499. (18) Ryzlak, M. T., and Pietruszko, R. (1988) Human brain “high Km” aldehyde dehydrogenase: purification, characterization, and identification as NAD+-dependent succinic semialdehyde dehydrogenase. Arch. Biochem. Biophys. 266, 386-396. (19) Kang, T. C., Park, S. K., Hwang, I. K., An, S. J., Choi, S. Y., Cho, S. W., and Won, M. H. (2002) Spatial and temporal alterations in the GABA shunt in the gerbil hippocampus following transient ischemia. Brain Res. 944, 10-18. (20) Picklo, M. J., Olson, S. J., Markesbery, W. R., and Montine, T. J. (2001) Expression and activities of aldo-keto oxidoreductases in Alzheimer disease. J. Neuropathol. Exp. Neurol. 60, 686-695. (21) Hashimoto, M., Sibata, T., Wasada, H., Toyokuni, S., and Uchida, K. (2003) Structural basis of protein-bound endogenous aldehydes. Chemical and immunochemical characterizations of configurational isomers of a 4-hydroxy-2-nonenal-histidine adduct. J. Biol. Chem. 278, 5044-5051. (22) Honzatko, A., Brichac, J., Murphy, T. C., Reberg, A., Kubatova, A., Smoliakova, I. P., and Picklo, M. J., Sr. (2005) Enantioselective
(24) (25)
(26)
(27) (28)
(29) (30) (31) (32) (33) (34) (35)
(36) (37) (38)
(39) (40)
(41) (42) (43)
(44) (45)
(46)
metabolism of trans-4-hydroxy-2-nonenal by brain mitochondria. Free Radical Biol. Med. 39, 913-924. Hiratsuka, A., Hirose, K., Saito, H., and Watabe, T. (2000) 4-Hydroxy2(E)-nonenal enantiomers: (S)-selective inactivation of glyceraldehyde-3-phosphate dehydrogenase and detoxification by rat glutathione S-transferase A4-4. Biochem. J. 349, 729-735. Tatton, W. G., Chalmers-Redman, R., Brown, D. and Tatton, N. (2003) Apoptosis in Parkinson’s disease: signals for neuronal degradation. Ann. Neurol. 53 (Suppl. 3), S61-70, discussion S70-62. Wang, Q., Woltjer, R. L., Cimino, P. J., Pan, C., Montine, K. S., Zhang, J., and Montine, T. J. (2005) Proteomic analysis of neurofibrillary tangles in Alzheimer disease identifies GAPDH as a detergentinsoluble paired helical filament tau binding protein. FASEB J. 19, 869-871. Hiratsuka, A., Tobita, K., Saito, H., Sakamoto, Y., Nakano, H., Ogura, K., Nishiyama, T., and Watabe, T. (2001) (S)-preferential detoxification of 4-hydroxy-2(E)-nonenal enantiomers by hepatic glutathione S-transferase isoforms in guinea-pigs and rats. Biochem. J. 355, 237244. Chandra, A., and Srivastava, S. K. (1997) A synthesis of 4-hydroxy2-trans-nonenal and 4-(3H) 4-hydroxy-2-trans-nonenal. Lipids 32, 779-782. Alary, J., Bravais, F., Cravedi, J. P., Debrauwer, L., Rao, D., and Bories, G. (1995) Mercapturic acid conjugates as urinary end metabolites of the lipid peroxidation product 4-hydroxy-2-nonenal in the rat. Chem. Res. Toxicol. 8, 34-39. Murphy, T. C., Amarnath, V., and Picklo, M. J., Sr. (2003) Mitochondrial oxidation of 4-hydroxy-2-nonenal in rat cerebral cortex. J. Neurochem. 84, 1313-1321. Nguyen, E., and Picklo, M. J. (2003) Inhibition of succinic semialdehyde dehydrogenase activity by alkenal products of lipid peroxidation. Biochim. Biophys. Acta 1637, 107-112. Bradford, M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254. Leiphon, L. J., and Picklo, M. J. S. Inhibition of aldehyde detoxification in CNS mitochondria by fungicides. Neurotoxicology (in press). Combet, C., Jambon, M., Deleage, G., and Geourjon, C. (2002) Geno3D: automatic comparative molecular modelling of protein. Bioinformatics 18, 213-214. Schuttelkopf, A. W., and van Aalten, D. M. (2004) PRODRG: a tool for high-throughput crystallography of protein-ligand complexes. Acta Crystallogr., D: Biol. Crystallogr. 60, 1355-1363. Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S., Greenblatt, D. M., Meng, E. C., and Ferrin, T. E. (2004) UCSF Chimerasa visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605-1612. Yang, S. K., Bao, Z., and Shou, M. (1996) Stereoselective nucleophilic substitution of oxazepam and racemization in acidic methanol and ethanol. Chirality 8, 214-223. Aso, Y., Yoshioka, S., and Takeda, Y. (1990) Epimerization and racemization of some chiral drugs in the presence of human serum albumin. Chem. Pharm. Bull. (Tokyo) 38, 180-184. Gineyts, E., Cloos, P. A., Borel, O., Grimaud, L., Delmas, P. D., and Garnero, P. (2000) Racemization and isomerization of type I collagen C-telopeptides in human bone and soft tissues: assessment of tissue turnover. Biochem. J. 345 (Part 3), 481-485. Stadtman, E. R. (1988) Protein modification in aging, J. Gerontol. 43, B112-120. Schneider, C., Tallman, K. A., Porter, N. A., and Brash, A. R. (2001) Two distinct pathways of formation of 4-hydroxynonenal. Mechanisms of nonenzymatic transformation of the 9- and 13-hydroperoxides of linoleic acid to 4-hydroxyalkenals. J. Biol. Chem. 276, 20831-20838. Schneider, C., Porter, N. A., and Brash, A. R. (2004) Autoxidative transformation of chiral omega6 hydroxy linoleic and arachidonic acids to chiral 4-hydroxy-2E-nonenal. Chem. Res. Toxicol. 17, 937-941. Doorn, J. A., Hurley, T. D., and Petersen, D. R. (2006) Inhibition of human mitochondrial aldehyde dehydrogenase by 4-hydroxynon-2enal and 4-oxonon-2-enal. Chem. Res. Toxicol. 19, 102-110. Ho, K. K., Allali-Hassani, A., Hurley, T. D., and Weiner, H. (2005) Differential effects of Mg2+ ions on the individual kinetic steps of human cytosolic and mitochondrial aldehyde dehydrogenases. Biochemistry 44, 8022-8029. Dickinson, F. M., and Hart, G. J. (1982) Effects of Mg2+, Ca2+ and Mn2+ on sheep liver cytoplasmic aldehyde dehydrogenase. Biochem. J. 205, 443-448. Larson, H. N., Weiner, H., and Hurley, T. D. (2005) Disruption of the coenzyme binding site and dimer interface revealed in the crystal structure of mitochondrial aldehyde dehydrogenase “Asian” variant. J. Biol. Chem. 280, 30550-30556. Moore, S. A., Baker, H. M., Blythe, T. J., Kitson, K. E., Kitson, T. M., and Baker, E. N. (1998) Sheep liver cytosolic aldehyde dehydro-
EnantioselectiVe Oxidation of HNE
(47) (48)
(49) (50) (51) (52)
genase: the structure reveals the basis for the retinal specificity of class 1 aldehyde dehydrogenases. Structure 6, 1541-1551. Steinmetz, C. G., Xie, P., Weiner, H., and Hurley, T. D. (1997) Structure of mitochondrial aldehyde dehydrogenase: the genetic component of ethanol aversion. Structure 5, 701-711. Liu, Z. J., Hempel, J., Sun, J., Rose, J., Hsiao, D., Chang, W. R., Chung, Y. J., Kuo, I., Lindahl, R., and Wang, B. C. (1997) Crystal structure of a class 3 aldehyde dehydrogenase at 2.6 Å resolution. AdV. Exp. Med. Biol. 414, 1-7. Esterbauer, H., Schaur, R., and Zollner, H. (1991) Chemistry and Biochemistry of 4-hydroxynonenal, malondialdehyde, and related aldehydes. Free Radical Biol. Med. 11, 81-128. Burgi, H. B., and Dunitz, J. D. (1983) From crystal statics to chemical dynamics. Acc. Chem. Res. 16, 153-161. Johansson, K., El-Ahmad, M., Ramaswamy, S., Hjelmqvist, L., Jornvall, H., and Eklund, H. (1998) Structure of betaine aldehyde dehydrogenase at 2.1 Å resolution. Protein Sci. 7, 2106-2117. Lamb, A. L., and Newcomer, M. E. (1999) The structure of retinal dehydrogenase type II at 2.7 Å resolution: implications for retinal specificity. Biochemistry 38, 6003-6011.
Chem. Res. Toxicol., Vol. 20, No. 6, 2007 895 (53) Corkey, B. E., Duszynski, J., Rich, T. L., Matschinsky, B., and Williamson, J. R. (1986) Regulation of free and bound magnesium in rat hepatocytes and isolated mitochondria. J. Biol. Chem. 261, 25672574. (54) Lenzen, S., Hickethier, R., and Panten, U. (1986) Interactions between spermine and Mg2+ on mitochondrial Ca2+ transport. J. Biol. Chem. 261, 16478-16483. (55) Pappa, A., Estey, T., Manzer, R., Brown, D., and Vasiliou, V. (2003) Human aldehyde dehydrogenase 3A1 (ALDH3A1): biochemical characterization and immunohistochemical localization in the cornea. Biochem. J. 376, 615-623. (56) Farres, J., Wang, T. T., Cunningham, S. J., and Weiner, H. (1995) Investigation of the active site cysteine residue of rat liver mitochondrial aldehyde dehydrogenase by site-directed mutagenesis. Biochemistry 34, 2592-2598. (57) Sheikh, S., Ni, L., Hurley, T. D., and Weiner, H. (1997) The potential roles of the conserved amino acids in human liver mitochondrial aldehyde dehydrogenase. J. Biol. Chem. 272, 18817-18822.
TX7000509
887
Enantioselective Oxidation of trans-4-Hydroxy-2-Nonenal Is Aldehyde Dehydrogenase Isozyme and Mg2+ Dependent Jiri Brichac,†,‡ Kwok Ki Ho,§ Ales Honzatko,† Rongying Wang,| Xiaoning Lu,| Henry Weiner,§ and Matthew J. Picklo, Sr.*,†,⊥ Departments of Pharmacology, Physiology, and Therapeutics, of Chemistry, and of Pathology and Proteomics Core Facility, UniVersity of North Dakota, Grand Forks, North Dakota 58202-9024, and Department of Biochemistry, Purdue UniVersity, West Lafayette, Indiana 47907-2063 ReceiVed February 12, 2007
trans-4-Hydroxy-2-nonenal (HNE) is a cytotoxic R,β-unsaturated aldehyde implicated in the pathology of multiple diseases involving oxidative damage. Oxidation of HNE by aldehyde dehydrogenases (ALDHs) to trans-4-hydroxy-2-nonenoic acid (HNEA) is a major route of metabolism in many organisms. HNE exists as two enantiomers, (R)-HNE and (S)-HNE, and in intact rat brain mitochondria, (R)-HNE is enantioselectively oxidized to HNEA. In this work, we further elucidated the basis of the enantioselective oxidation of HNE by brain mitochondria. Our results showed that (R)-HNE is oxidized enantioselectively by brain mitochondrial lysates with retention of stereoconfiguration of the C4 hydroxyl group. Purified rat ALDH5A enantioselectively oxidized (R)-HNE, whereas rat ALDH2 was not enantioselective. Kinetic data using (R)-HNE, (S)-HNE, and trans-2-nonenal in combination with computer-based modeling of ALDH5A suggest that the selectivity of (R)-HNE oxidation by ALDH5A is the result of the carbonyl carbon of (R)-HNE forming a more favorable Bu¨rgi-Duntiz angle with the active site cysteine 293. The presence of Mg2+ ions altered the enantioselectivity of ALDH5A and ALDH2. Mg2+ ions suppressed (R)-HNE oxidation by ALDH5A to a greater extent than that of (S)-HNE. However, Mg2+ ions stimulated the enantioselective oxidation of (R)-HNE by ALDH2 while suppressing (S)-HNE oxidation. These results demonstrate that enantioselective utilization of substrates, including HNE, by ALDHs is dependent upon the ALDH isozyme and the presence of Mg 2+ ions. Introduction trans-4-Hydroxy-2-nonenal (HNE)1 is a cytotoxic R,βunsaturated aldehyde implicated in the pathology of multiple diseases involving oxidative damage (1-3). In the central nervous system, elevated levels of HNE and HNE-protein adducts occur in Alzheimer’s disease, Parkinson’s disease, and cerebral ischemia (4-7). The chemical basis of HNE toxicity is related to its electrophilic character and the capability to alkylate cellular nucleophiles via formation of Michael adducts with cysteinethiols, histidinylamines, and lysylamines and by formation of Schiff bases with lysylamines (8-11). Detoxification of HNE proceeds through multiple routes (1214). Oxidation of HNE to trans-4-hydroxy-2-nonenoic acid (HNEA; Figure 1A) via aldehyde dehydrogenases (ALDHs) is a major route of metabolism in many systems including smooth muscle cells, hepatocytes, astrocytes, and mitochondria from liver and brain (12-14). In isolated brain and liver mitochondria, * To whom correspondence should be addressed. Phone: (701) 7772293. Fax: (701) 777-0387. E-mail: [email protected]. † Department of Pharmacology, Physiology, and Therapeutics, University of North Dakota. ‡ On leave from the Department of Analytical Chemistry, Charles University, Prague 12843, Czech Republic. § Purdue University. | Department of Pathology and Proteomics Core Facility, University of North Dakota. ⊥ Department of Chemistry, University of North Dakota. 1 Abbreviations: HNEA, trans-4-hydroxy-2-nonenoic acid; HNE, trans4-hydroxy-2-nonenal; 4-ONE, trans-4-oxo-2-nonenal; NEN, trans-2-nonenal; ALDHs, aldehyde dehydrogenases; SPE, solid-phase extraction; DTT, dithiothreitol; BSA, bovine serum albumin; MALDI, matrix-assisted laser desorption ionization.
ALDH2 and ALDH5A (succinic semialdehyde dehydrogenase) are known to oxidize HNE (15, 16). These enzymes are active as homotetramers of ∼54 kDa subunits and are NAD+ dependent. These enzymes have multiple differences regarding substrate specificity, cellular expression, and levels of expression, however. Rat ALDH2 and rat ALDH5A share only 52% amino acid similarity by BLAST2 analysis and have very different substrate specificities (17, 18). ALDH5A is expressed in neurons, whereas ALDH2 is expressed in astrocytes (19, 20). In brain mitochondria, ALDH5A activity compromises the majority (>90%) of HNE oxidation, whereas in hepatic mitochondria, ALDH2 compromises about 65% of HNE oxidation activity (16). Although ALDH5A activity is 1.5-fold greater in isolated rat brain mitochondria than in isolated rat hepatic mitochondria, ALDH2 activity in isolated brain mitochondria is less than 6% than that of isolated hepatic mitochondria (16). HNE possesses a chiral center at C4 and exists as two enantiomers, (R)-HNE and (S)-HNE, with potentially different biochemical reactivities (Figure 1B). The stereochemistry of HNE has biological significance. We and others have shown that antibodies are formed in vivo that distinguish adducts derived from the separate HNE enantiomers (21, 22). (S)-HNE enantioselectively inactivates glyceraldehyde-3-phosphate dehydrogenase (GAPDH), an enzyme important for glycolysis and the apoptotic cascade that is also a component of neurofibrillary tangles found in Alzheimer’s disease (23-25). In rat and guinea pig liver cytosol HNE is detoxified S-selectively by GSH conjugation (23, 26). In rat liver cytosol ALDH-mediated oxidation is not enantioselective, whereas oxidation is Rselective in guinea pig liver cytosol (23, 26). We have shown in respiring rat brain mitochondria, that (R)-HNE is metabolized
10.1021/tx7000509 CCC: $37.00 © 2007 American Chemical Society Published on Web 05/05/2007
888 Chem. Res. Toxicol., Vol. 20, No. 6, 2007
at significantly higher rates than (S)-HNE, mainly by enantioselective oxidation to HNEA (22). In this work, we further elucidated the basis of enantioselective oxidation of HNE by brain mitochondria utilizing brain mitochondrial lysate, recombinant ALDH5A, and recombinant ALDH2. Our results show that the enantioselective oxidation of (R)-HNE occurs by rat brain mitochondria without alteration of stereoconfiguration. Purified rat ALDH5A enantioselectively oxidized (R)-HNE, wheras rat ALDH2 was not enantioselective. However, the presence of Mg2+ ions altered the enantioselectivity of both enzymes. Computer-assisted modeling studies suggest that the higher activity of ALDH5A toward (R)-HNE is the result of a more favorable Bu¨rgi-Dunitz angle formed between the active site cysteine of ALDH5A and the carbonyl carbon of (R)-HNE.
Experimental Procedures Chemicals. Racemic HNE, individual HNE enantiomers, d11HNEA, and HNEA were synthesized as described previously (22, 27, 28). The optical purity of HNE enantiomers was 96% for (S)HNE and 98% for (R)-HNE. HNEA enantiomers were synthesized by oxidation of the individual, purified HNE enantiomers. The products were purified by flash chromatography or solid-phase extraction (SPE), and the structures were confirmed by NMR or mass spectrometry (data not shown). NAD+ and trans-2-nonenal were purchased from Sigma-Aldrich (St. Louis, MO). All other chemicals were analytical or HPLC grade. trans-4-oxo-2-nonenal (4-ONE) was obtained from Cayman Chemicals (Ann Arbor, MI). Preparation of Rat Brain Mitochondria. Sprague-Dawley rats (male, 250-300 g) were purchased from Charles River (Wilmington, MA). Experimental protocols were in accordance with the NIH guidelines for the use of live animals and were approved by the University of North Dakota Institutional Animal Care and Use Committee. Brain mitochondria were isolated as described previously, resulting in synaptic and nonsynaptic mitochondria (29). The final mitochondrial pellets were resuspended in lysis buffer (pH 7.4) containing 20 mM sodium phosphate, 0.5% Triton X-100 (Alfa Aesar, Ward Hill, MA), 10 mM dithiothreitol (DTT; Fisher Scientific, Fair Lawn, NJ), 1 mM diethylenetriaminepentaacetic acid (Acros Organics, Geel, Belgium), and 50 µM 2,6-di-tert-butyl-4methylphenol (Acros Organics) (16). The solution was then diluted 1:1 with glycerol and stored at - 80 °C. Determination of Apparent Kinetic Parameters of HNEA Formation by Rat Brain Mitochondrial Lysate. Mitochondria were lysed by freezing/thawing and passed over a Sephadex G25 (Amersham Biosciences, Uppsala, Sweden) column (15-fold volume over sample) equilibrated with lysis buffer without DTT to remove potential endogenous aldehydes (30), cofactors, glycerol, and DTT. The protein concentration was determined using protein assay reagent (Bio Rad, Hercules, CA) with BSA as a standard (31). Mitochondria at a final concentration 0.2 mg/mL were incubated in 40 mM phosphate buffer (pH 7.4) containing 1 mM NAD+ at 37 °C. The total volume of the reaction mixture was 150 µL. Assays were performed in triplicate using five different concentrations of either (R)-HNE or (S)-HNE. The reaction was stopped after 3 min by addition of a 1/10 sample volume of cold 3% (w/w) phosphoric acid, and the tube was immediately put on ice (16, 22, 29). The HNEA content was measured by HPLC with UV detection. No detectable endogenous HNE and HNEA concentrations were present in mitochondrial lysate. Determination of HNEA. HNEA formed by the mitochondrial lysates was measured by RP-HPLC (Schimadzu Class-VP 7.2 system, Kyoto, Japan) with UV detection at 210 nm (16, 29). Because of the low KM of ALDH2 for HNE (see the Results), we measured the enzyme activity via monitoring the formation of HNEA by LC-MS/MS when determining KM and Vmax parameters as described previously with some modifications (22). The LCMS/MS system consisted of an Agilent 1100 series HPLC system
Brichac et al.
Figure 1. (A) ALDH-mediated oxidation of racemic HNE. (B) Structures of R,β-unsaturated aldehydes used in these studies: aldehyde dehydrogenase (ALDH), trans-4-hydroxy-2-nonenal (HNE), trans-4hydroxy-2-nonenoic acid (HNEA), trans-4-oxo-2-nonenal (4-ONE), trans-2-nonenal (NEN).
(Agilent Technologies, Santa Clara, CA) coupled to an API 3000 LC-MS/MS system (MDS Sciex, Concord, Canada). A 10 µL portion of the sample was injected onto a chromatographic column, Luna, 3 µm, C8(2), 150 × 2.1 mm i.d. (Phenomenex, Torrance, CA). HNEA eluted at 4.4 min in isocratic mode in the mobile phase, ACN-30 µM ammonium acetate, pH 3.8 (45:55, v/v), at a flow rate of 150 µL/min. The limit of quantitation for HNEA was below 30 fmol. Analysis of HNEA Stereochemistry. Oxidation of 160 µM (R)HNE or (S)-HNE by mitochondrial lysate was performed as described above. However, the reaction time was 10 min, and the volume of the reaction mixture was 2.4 mL. Prior to the separation of HNEA enantiomers by HPLC, the samples were purified by SPE using Oasis HLB 6 cm3 (0.2 g) extraction cartridges (Waters, Milford, MA). The cartridges were conditioned with 7 mL of methanol (MeOH) and equilibrated with 7 mL of deionized water. The sample was loaded onto the cartridge and washed with 7 mL of MeOH-100 mM phosphate buffer, pH 2 (40:60, v/v), and with 7 mL of water. HNEA was eluted by MeOH-100 mM phosphate buffer, pH 7 (50:50, v/v). The first 0.8 mL of the eluent was discarded, and the next 2.5 mL fraction containing HNEA was collected. A second SPE was used to concentrate HNEA and to transfer the sample into MeOH. MeOH was evaporated by a nitrogen evaporator, and the residue was dissolved in 300 µL of the mobile phase. HNEA enantiomers were separated by HPLC using a chiral column, Chiralcel OB, 250 mm × 4.6 mm (Daicel, Osaka, Japan), and a mobile phase of hexane-2-propanol (97:3, v/v). The flow rate was 0.8 mL/min, 40 µL of sample was injected, and 210 nm was used for UV detection. The method was used for the determination of the HNEA enantiomer ratio only. RP-HPLC was used for quantification of HNEA. Baseline resolution was achieved (R ) 1.83), and the retention time was 11.0 min for (S)-HNEA and 12.9 min for (R)-HNEA. The limit of detection (LOD) was 4.8 µM (192 pmol) for (S)-HNEA and 5.4 µM (216 pmol) for (R)HNEA. Preparation of Recombinant Rat ALDH5A and ALDH2. Recombinant rat ALDH5A and rat ALDH2 containing N-terminal polyhistidine tags were prepared as described previously using Rosetta 2 (DE3) Escherichia coli cells (Novagen, Madison, WI) (32). Following nickel agarose affinity chromatography, the purified enzymes (ALDH2 or ALDH5A) were dialyzed at 4 °C versus two changes of buffer (1:1000), 0.15 M NaCl, 40 mM sodium phosphate buffer, pH 7.4, containing 100 µM DTT to preserve the enzyme activity. Proteins of just over 50 kDa and of >95% purity were obtained as judged by SDS-PAGE analysis. Determination of Kinetic Parameters for Rat ALDH5A and ALDH2. Enzyme activity, unless stated otherwise, was measured by monitoring NADH formation at 340 nm under the initial rate conditions (16). Assays were performed with dialyzed protein on a microplate spectrophotometer, SpectraMax plus 384 (Molecular Devices, Sunnyvale, CA). The absorption coefficient of NADH was 6220 M-1 cm1. Assays were performed in a 300 µL reaction mixture containing 1 mM NAD+ in 40 mM phosphate buffer (pH 7.4) at 37 °C (16). For experiments in which the effects of Mg2+ ions
EnantioselectiVe Oxidation of HNE were examined, HEPES-free acid was substituted for the sodium phosphate. Mass Determination for Recombinant ALDH2. A 10 µL portion of ALDH2 protein sample (1 mg/mL) in 40 mM sodium phosphate was loaded onto C4 ziptip columns (Millipore), washed with 0.1% TFA, and eluted with 5 µL of 70% ACN with 0.1% TFA. Typically, 0.4 µL of the eluted protein sample was mixed on a matrix-assisted laser desorption ionization (MALDI) plate with 0.4 µL of matrix solution, 50 mM R-cyano-4-hydroxycinnamic acid (CHCA) in 50% ACN containing 0.1% TFA. The mass measurement was carried out on a 4800 MALDI-TOF/TOF mass spectrometer (Applied Biosystems, Foster City, CA) equipped with a Nd:YAG 200 Hz laser. The instrument was operated with delayed extraction in linear positive ion mode. The mass spectra of the recombinant ALDH2 protein were acquired by averaging 20 subspectra of 500 total laser shots and internally calibrated using bovine serum albumin (BSA) as the standard reference. Modeling and Substrate Docking for ALDH5A. The 3D model for ALDH5A (SwissProtein database accession number P51650) was generated using Geno3D (33), an automated Web server for protein molecular modeling (http://geno3d-pbil.ibcp.fr). The ALDH5A model was built on the basis of the solved crystal structure of the sheep cytosolic aldehyde dehydrogenase (ALDH1) (Protein Data Bank ID 1BXS). The putative NAD+ binding site in the ALDH5A model was identified by superposing the sheep ALDH1 structure onto the ALDH5A model. The 3D atomic coordinates of the trans2-nonenal and the (R)- or (S)-HNE in pdb format were obtained using the Dundee Prodrg2 Server (http://davapc1.bioch.dundee.ac.uk/ programs/prodrg/prodrg.html) (34). For docking of trans-2-nonenal and (R)- or (S)-HNE into the ALDH5A model, we employed the ArgusLab software, which makes use of the AScore scoring function to determine the low-energy binding mode with a grid encompassing C293 as a binding group. Molecular graphics images were produced using the UCSF Chimera software (http://www. cgl.ucsf.edu/chimera/) (35). Statistical and Curve Fitting Analyses. Vmax and KM were determined by nonlinear regression analysis using the MichaelisMenten equation V ) (Vmax[S])/(KM + [S]). Statistical significance between data sets was tested by two-way ANOVA with the Bonferronni posttest and was achieved when p < 0.05 (Prism 4, GraphPad Software, San Diego, CA).
Results Previously, we had shown that (R)-HNE was oxidized to a greater extent than (S)-HNE by respiring rat brain mitochondria. In this first experiment, we determined whether the stereoselectivity of HNE oxidation was retained in lysed brain mitochondria passed through G25 Sephadex. We performed this step to provide a crude preparation of brain mitochondrial ALDHs so that we could monitor HNE oxidation in brain mitochondria without confounds such as substrate and cofactor availability. The apparent KM and Vmax for (R)-HNE were 27.8 ( 8.5 µM and 8.72 ( 0.90 (nmol/min)/mg, respectively, and those for (S)HNE 50.9 ( 16.4 µM and 4.4 ( 0.6 (nmol/min)/mg (Figure 2). Using a comparison of Vmax/KM as a measure of efficiency, these data show that the ALDHs present in rat brain mitochondria oxidize (R)-HNE 3.7-fold more efficiently than (S)-HNE. There was no consumption of (R)-HNEA or (S)-HNEA by lysed mitochondria supplemented with NAD+ (not shown), demonstrating that selective formation of (R)-HNEA was not an artifact resulting from enantioselective metabolism of (S)-HNEA. Racemization of chiral molecules can occur enzymatically and non-enzymatically in biological and non-biological systems (36-39). During the formation of HNE, racemization or retention of the stereoconfiguration of the hydroxyl group is dependent upon which carbon of the hydroperoxide precursor was present (40, 41). To test whether the absolute configuration of the HNE C4 carbon was retained during the enzymatic
Chem. Res. Toxicol., Vol. 20, No. 6, 2007 889
Figure 2. (R)-HNE is selectively oxidized in brain mitochondrial lysate. Mitochondrial lysate was incubated with increasing concentrations of either (R)-HNE or (S)-HNE in the presence of 1 mM NAD+ at 37 °C for 3 min. The amount of HNEA formed by HNE oxidation was measured by HPLC as described in the Experimental Procedures. The data are presented as the mean ( SEM of three experiments performed in triplicate on three separate days (n ) 3). The nonlinear regression lines for the determination of KM and Vmax are given. The apparent KM and Vmax for (R)-HNE were 27.8 ( 8.5 µM and 8.72 ( 0.90 (nmol/ min)/mg and those for (S)-HNE 50.9 ( 16.4 µM and 4.4 ( 0.6 (nmol/ min)/mg.
oxidation of HNE, we developed an HPLC method using the chiral column Chiralcel OB for the separation of HNEA enantiomers (Figure 3). Using this method, we determined that the optical purity of (S)-HNEA formed by (S)-HNE oxidation by brain mitochondria was 99% and that of (R)-HNEA formed from (R)-HNE was 98%. These data demonstrate that racemization of the C4 carbon did not occur during the oxidation step. We next examined the enantioselectivity of HNE oxidation of ALDH5A and ALDH2, the two known mitochondrial ALDHs that oxidize HNE, using recombinant rat ALDH2 and ALDH5A. The purity of the resulting ALDHs is shown in Figure 4A. The kinetic parameters of acetaldehyde and succinic semialdehyde utilization by ALDH2 and ALDH5A are presented in Table 1. Our previous data suggested that ALDH5A was the major source of HNE oxidation activity in brain mitochondria. Therefore, we tested the hypothesis that ALDH5A preferentially oxidized (R)HNE (Figure 4B). Rat ALDH5A oxidized (R)-HNE with a KM of 20.4 ( 3.3 µM, a Vmax of 1.5 ( 0.1 (µmol/min)/mg, and a kcat of 1.4 s-1. On the other hand, (S)-HNE had a KM of 32 ( 6.1 µM, a Vmax of 0.42 ( 0.03 (µmol/min)/mg, and a kcat of 0.38 s-1. A comparison of kcat/KM demonstrated that (R)-HNE (kcat/KM ) 68.6 × 103 s-1/M) was oxidized 5.8-fold more efficiently than (S)-HNE (kcat/KM ) 11.8 × 103 s-1/M). We used trans-2-nonenal as a substrate as it lacks the hydroxyl group at the C4 position (Figure 1B). Interestingly, trans-2-nonenal had kinetic characteristics similar to those of (S)-HNE with a KM of 43.3 ( 7.8 µM, a Vmax of 0.52 ( 0.04 (µmol/min)/mg, and a kcat of 0.47 s-1. We attempted to determine whether fixing the oxygen group at C4 by using trans-4-oxo-2-nonenal (4-ONE) as a substrate would affect the activity. Unfortunately, this substrate even at low concentrations (5 µM) inhibited the ALDH5A activity (not shown) likely as a result of irreversible inhibition as has been noted for 4-ONE and human ALDH2 (42). Our previous data suggested that ALDH2 played a minor role in HNE oxidation in rat brain mitochondria (16). However, the finding that purified rat ALDH5A had a higher selectivity toward (R)-HNE than in brain mitochondrial lysate prompted us to examine the kinetic parameters of recombinant rat ALDH2 with the enantiomers of HNE. Our initial experiments using UV spectrophotometric measurement of NADH formation demonstrated that saturation of enzyme activity with (R)-HNE and (S)-HNE had already occurred at a 5 µM concentration of
890 Chem. Res. Toxicol., Vol. 20, No. 6, 2007
Figure 3. The stereochemistry of the C4 hydroxyl is retained during HNE oxidation. (A) Separation of (R)-HNEA and (S)-HNEA by normalphase chromatography using a Chiralcel-OB chiral column. (B) Sample loading of HNE and HNEA onto the SPE cartridge. Note that both HNEA (open circles) and HNE (closed squares) are completely retained. Because HNEA and HNE enantiomers coeluted on the Chiracel OB column, selective elution (C) with pH 7 buffer from the SPE cartridge was able to separate HNEA (open circles) versus HNE (closed squares). (D) HNE enantiomers were oxidized separately in lysates of brain mitochondria supplemented with NAD+. The ratio of the HNEA enantiomers was determined by chiral-phase HPLC. (S)-HNEA was not obtained by oxidation of (R)-HNE, and (R)-HNEA was not obtained following oxidation of (S)-HNE. Data are the mean ( SEM of three independent experiments (n ) 3).
either enantiomer, suggesting very low KM values (not shown). To determine the kinetic parameters for (R)-HNE and (S)-HNE, we therefore monitored HNEA formation by LC-MS/MS. (R)HNE was oxidized with a KM of 1.0 ( 0.1 µM, a Vmax of 42 ( 2 (nmol/min)/mg, and a kcat of 0.039 s-1. (S)-HNE was oxidized with a KM of 0.70 ( 0.1 µM, a Vmax of 43 ( 1 (nmol/min)/mg, and a kcat of 0.040 s-1 (Figure 4C). The kcat/KM for (R)-HNE was 39 × 103 s-1/M and that for (S)-HNE 57 × 103 s-1/M for ALDH2. For calculation of kcat, a mass of 56.5 kDa per subunit was obtained by MALDI-TOF/TOF mass spectrometry for the recombinant, polyhistidine-tagged rat ALDH2.
Brichac et al.
Figure 4. Enantioselective oxidation of HNE by ALDH5A but not ALDH2. (A) SDS-PAGE analysis of recombinant rat ALDH5A (lane 1, 4 µg) and ALDH2 (lane 2, 4 µg). Lane 3 contained molecular weight markers with the pertinent weights provided. (B) Oxidation of (R)HNE, (S)-HNE, or NEN by ALDH5A. NADH formation was used as a measure of activity. A 3 µg portion of ALDH5A was used in the reactions. (C) Nonenantioselective oxidation of HNE by ALDH2. Owing to the low activity and low KM, the formation of HNEA was used as a measure of the activity and was quantified by LC-MS/MS. A 180 ng portion of ALDH2 was used in the reactions. Data are expressed as the mean ( SEM of assays performed in triplicate. The assays were performed in 40 mM sodium phosphate buffer with 1 mM NAD+ at pH 7.4 and 37 °C.
We and others have shown that the presence of Mg2+ ions alters the configuration of NAD(H) at the active site of ALDH1 and ALDH2 and thereby affects the activities of these enzymes (43, 44). To gain mechanistic insight into the interaction of HNE chirality and the active sites of ALDH5A and ALDH2, we tested whether the presence of Mg2+ ions would affect the oxidation rates of (R)-HNE and (S)-HNE (Figure 5). Vmax concentrations of the aldehydes were used. Mg2+ ions inhibited the oxidation
EnantioselectiVe Oxidation of HNE
Chem. Res. Toxicol., Vol. 20, No. 6, 2007 891
Figure 5. Mg2+ ions alter the enantioselectivity of HNE oxidation by ALDH5A and ALDH2. ALDH5A (A,B) or ALDH2 (C,D) was incubated with NAD+ (1 mM) and Mg2+ ions (as MgCl2) for 1 min prior to the addition of the aldehyde substrate. For comparison, plots of the total rates and percentage of the control activity (no Mg2+ ions added) are provided. Note that, for ALDH5A, (R)-HNE oxidation is more sensitive to inhibition by Mg2+ ions (p < 0.05 by two-way ANOVA) whereas Mg2+ ions stimulate ALDH2-mediated (R)-HNE oxidation over 5-fold. Data are expressed as the mean ( SEM of assays performed in triplicate. The assays were performed in 40 mM HEPES buffer, pH 7.4, at 37 °C. For ALDH5A, 120 µM (R)-HNE or (S)-HNE was used. For ALDH2, 50 µM (R)-HNE, (S)-HNE, or NEN was used. Preliminary experiments demonstrated that the KM of NEN for ALDH2 was less than 5 µM (not shown). For ALDH5A activities, 4 µg of protein was used. For ALDH2 activities, 15 µg of protein was used. Reaction volumes were 300 µL. Table 1. Kinetic Parameters of Recombinant ALDH2 and ALDH5A with Acetaldehyde and Succinic Semialdehyde enzyme
KM (µM)
ALDH2 succinic semialdehyde 780 ( 100 acetaldehyde 0.4 ( 0.1 ALDH5A succinic semialdehyde 3.5 ( 0.9 acetaldehyde 5800 ( 300
Vmax kcata [(µmol/min)/mg] (s-1) 0.116 ( 0.01 0.280 ( 0.01 38 ( 1.2 3.9 ( 0.1
kcat/KM (s-1/M)
0.11 140 0.26 6.5 × 105 34 3.5
9.7 × 106 603
a For calculation of k , a mass of 54.4 kDa for the recombinant 6×Hiscat tagged ALDH5A was used as determined by MALDI-TOF mass spectrometry (16). For recombinant 6×His-tagged ALDH2, a subunit mass of 56.5 kDa was used. Data are derived from assays performed in triplicate. KM and Vmax were determined by nonlinear regression analysis with the data fit to the Michaelis-Menten equation. KM and Vmax are presented as the best fit value ( the standard error. Assays were performed at 37 °C in 40 mM sodium phosphate buffer (pH 7.4) with a 50 µM concentration of either aldehyde and 1 mM NAD+.
of (R)-HNE and (S)-HNE by rat ALDH5A in a concentrationdependent manner; however, the oxidation of (R)-HNE was inhibited to a slightly, but significantly, greater extent than (S)HNE. On the other hand, the oxidation of HNE enantiomers by rat ALDH2 responded differently to the presence of Mg2+ ions (Figure 5C,D). Both (R)-HNE and (S)-HNE oxidation rates were greatly elevated by the presence of a low concentration of Mg2+ ions (0.1 mM), but at higher concentrations, the activity declined, consistent with what has been shown for human ALDH2 (43). The magnitude of the activity increase was much greater for (R)-HNE than (S)-HNE. Whereas 5 mM MgCl2 inhibited the oxidation of both HNE enantiomers by 25-30% for ALDH5A, different effects were observed for ALDH2 in
which (S)-HNE oxidation was suppressed to 50% of the control values and (R)-HNE oxidation was still 2-fold higher than the control values. The oxidation of trans-2-nonenal showed a similar biphasic response to Mg2+ ions; however, the rate of oxidation at 5 mM Mg2+ ions was still higher than the control (no Mg2+ ions) values. Last, we attempted to gain insight into the structural basis for ALDH5A enantioselectivity. Since the crystal structure for rat ALDH5A has not been solved, we modeled the structure of rat ALDH5A (Figure 6) on the solved crystal structure of sheep cytosolic ALDH1. We obtained a model with a “head” and “tail” that are similar to those observed for the tetrameric ALDH2 and ALDH1 and the dimeric rat ALDH3 (45-48). We then fit in (R)-HNE, (S)-HNE, and trans-2-nonenal with NAD+ present in the active site to discern structural interactions that could account for the differences in the utilization of these aldehydes (Figure 7). Our analyses indicate that the low-energy binding mode of trans-2-nonenal was similar to that of (S)HNE (not shown). Although the distances between the active site cysteinyl thiol (C293) and the carbonyl carbon were similar, the Bu¨rgi-Dunitz angle formed among the thiol, the carbonyl carbon, and the carbonyl oxygen was more favorable for (R)HNE (109°) than for trans-2-nonenal (121°) and (S)-HNE (134°) (Table 2). The calculated energy of the reaction for (R)-HNE (-7.4 kcal/mol) was lower than that of (S)-HNE and trans-2nonenal.
Discussion HNE is a toxic, chiral, R,β-unsaturated aldehyde derived from the oxidation of linoleic acid and arachidonic acid (49). (R)-
892 Chem. Res. Toxicol., Vol. 20, No. 6, 2007
Brichac et al.
Figure 6. Proposed model of rat ALDH5A with NAD+. A putative model of the rat ALDH5A was generated from the primary amino acid sequence for rat ALDH5A (minus the mitochondrial targeting sequences) using sheep ALDH1 as a template. The protein has a “head” and “tail” configuration similar to that of ALDH 1 and ALDH2. In this model, NAD+ is docked into the active site. Table 2. Calculated Angles, Distances, and Reaction Energies for ALDH5A and Aldehydesa
aldehyde
Bu¨rgi-Dunitz angle between C293 and CdO carbon (deg)
distance between C293 and CdO carbon (Å)
distance between C4-OH and R166 (Å)
energy of reaction (kcal/mol)
(R)-HNE (S)-HNE NEN
109 134 121
3.2 3.5 3.2
3.9 5.0 N/A
-7.4 -8.1 -9.0
a All aldehydes were modeled in the active site with NAD+ present. NEN ) trans-2-nonenal.
HNE undergoes selective oxidation by brain mitochondria (22). In this work, we further elucidated the basis for the enantioselective oxidation of HNE by brain mitochondria. First, we show that the R-selective oxidation of HNE by brain mitochondria occurs and that, as expected, oxidation occurs without racemization of the C4-hydroxyl group. We then demonstrate that recombinant ALDH5A, but not ALDH2, enantioselectively oxidizes (R)-HNE. Unexpectedly, the presence of increasing Mg2+ ion concentrations altered the enantioselectivity of both enzymes. Last, using computer-assisted modeling, we propose that the higher activity of ALDH5A toward (R)-HNE is the result of a more suitable bond angle being formed between the active site cysteine and the carbonyl carbon of (R)-HNE. Possible explanations for the enantioselectivity exhibited by ALDH5A is that a gating mechanism occurs such that (A) (R)HNE is able to pass to the active site of the enzyme as opposed to (S)-HNE in which the C4-hydroxyl is in the opposite direction or (B) (R)-HNEA is able to exit the active site more easily than (S)-HNEA. These explanations do not account for finding that trans-2-nonenal, lacking the C4-hydroxyl, has kinetic parameters similar to those of (S)-HNE. The data lead us to suggest that (R)-HNE is positioned in a more advantageous manner in the active site. Our modeling data support this hypothesis. These
Figure 7. (R)-HNE and (S)-HNE are positioned differently in the active site of rat ALDH5A. (R)-HNE (A) and (S)-HNE (B) were docked into the putative ALDH5A active site with NAD+ present. The HNE enantiomers have the dark blue fill. For orientation purposes, the carbonyl oxygens on C1 of the HNE enantiomers (pictured as the red ball) are near the T292 and N158 residues, and the C4 oxygens are positioned near M163. Oxygens are represented by the red balls, whereas nitrogens are blue and sulfurs are yellow. The carbonyl carbon of (S)-HNE is positioned closer to N158, the general acid equivalent to N169 in ALDH1 and ALDH2, thus creating a greater angle than that of (R)-HNE with respect to cysteine 293 (C293), which forms a thiohemiacetal with the carbonyl carbon and is equivalent to C302 in the other isozymes (56, 57). Note that the configuration of the (R)HNE molecule itself in the active site is different from that of (S)HNE.
modeling data predict that (R)-HNE forms a more favorable Burgi-Dunitz angle (109°) than that of (S)-HNE or trans-2nonenal. This angle is the angle that is formed among the nucleophile (the thiolate anion of cysteine) and the carbonyl carbon of the aldehyde and its oxygen atom. An angle of 110° is predicted to be the optimum angle of attack of the nucleophile on the carbonyl carbon owing to the presence of the carbonoxygen sp2 electron orbitals (50). A plot of experimentally derived kcat versus the computer-derived Bu¨rgi-Dunitz angle demonstrates a significant relationship between the rate of catalysis and the angle formed (Figure 8). Multiple reasons may exist as to why (R)-HNE is positioned in a more favorable manner with respect to the active site cysteine. A likely possibility is that there is hydrogen bonding between the C4-hydroxyl group of (R)-HNE and a positively charged amino acid at the active site. This also could explain why succinic semialdehyde is utilized highly efficiently by ALDH5A versus acetaldehyde as shown by our data (Table 1)
EnantioselectiVe Oxidation of HNE
Figure 8. The rate of catalysis is related to the Bu¨rgi-Dunitz angle. Experimentally derived kcat values for (R)-HNE, (S)-HNE, and NEN are plotted versus the computer-derived Bu¨rgi-Dunitz angle. The line shown is derived from linear regression analysis of the data and has an r2 of 0.85.
and those of others (18). On the basis of our modeling data, we speculate that the arginine residue at position 166 of ALDH5A is the likely positively charged amino acid. Although the (R)hydroxyl group oxygen is closer to R166 (3.9 Å) than (S)-HNE (5.0 Å), these distances are too long for significant hydrogen bonding to occur unless there is movement in the active site to allow a closer interaction. Interestingly, the residue analogous to R166 at this position in ALDH2 is a nonpolar tryptophan residue. Mutations of this R166 residue with subsequent kinetic analyses are needed. The crystal structure of ALDH5A has not been solved, so the modeling data presented cannot be verified. The known structures of other ALDH isozymes, including sheep ALDH1, human ALDH2, beef ALDH2, rat ALDH3, and cod ALDH9, all have essentially the same structure as some bacterial forms of the enzyme (45-48, 51). Even the structure of retinaldehyde dehydrogenase, like that of the sheep class 1 enzyme, was solved using the coordinates from the class 2 isozyme (46, 52). This near structural identity was found even though there are large differences in the amino acid sequences between family members (45-81% similarity of the above ALDHs when compared to sheep ALDH1). Thus, we are confident that the modeling data are very representative of what actually exists for ALDH5A. Supporting the modeling data is noting that there is a relationship between the kcat and the bond angle (Figure 8). The ALDH2 activity plays a minor role in the oxidation of HNE by brain mitochondria in contrast to liver mitochondria, where ALDH2 plays a major role in metabolizing HNE (16). The lack of enantioselectivity under our initial conditions (with no Mg2+ ions) may be the result of ALDH2 possessing a less hindered, lipophilic binding pocket. The KM values for HNE enantiomers are more than 10-fold lower with Vmax values more than 10-fold higher than those reported for racemic HNE by Mitchell and Petersen using ALDH2 purified from rat liver (15). This likely is the result of the fact that rapidly purified, recombinant enzyme was used. The presence of Mg2+ ions had modest effects upon enantioselective oxidation by ALDH5A (as compared to ALDH2) although (R)-HNE oxidation was inhibited to a significantly greater extent than (S)-HNE oxidation in the presence of Mg2+ ions. Previous work from our laboratories and others demonstrates that Mg2+ ions increase the rate of deacylation for ALDH2, thereby elevating the activity and increasing the binding affinity of NADH to the active site of ALDH1 and thereby decreasing the enzyme activity (43, 44). This observation and others were used to conclude that inhibition of the ALDH activity by the presence of Mg2+ ions indicated that NADH release was the rate-limiting step for enzyme activity.
Chem. Res. Toxicol., Vol. 20, No. 6, 2007 893
Although we did not explore all the enzymatic steps of ALDH5A activity, our current data point to NADH release becoming rate limiting for HNE oxidation since oxidation was suppressed at most concentrations of Mg2+ ions. The addition of a low concentration (100 µM) of Mg2+ ions increased the rates of oxidation of (R)-HNE and (S)-HNE by ALDH2. These data are consistent with previous work demonstrating that Mg2+ ions, at this concentration, increase the rate of the deacylation step by ALDH2 (43). What was not expected, however, was that Mg2+ ions elevated (R)-HNE oxidation to a much larger extent than for (S)-HNE. These data suggest that the presence of Mg2+ ions alters the active site such that (R)-HNE becomes more favorably positioned for formation of the thiohemiacetal with C302 of the active site or hydride transfer or that (R)-HNEA is able to leave the active site at a faster rate than (S)-HNEA during deacylation or a combination of these mechanisms. (R)-HNE oxidation decreased with increasing concentrations of Mg2+ ions but did not return to the control (no Mg2+ ions) activity, whereas (S)HNE oxidation was suppressed to 50% of the control activity. The decrease in the rate of oxidation for both enantiomers is consistent with the finding that Mg2+ ions form a complex with NADH that slows the rate of NADH release from the enzyme (43). The extent to which HNE is metabolized enantioselectively in vivo needs further study. The concentration of free Mg2+ in mitochondria is estimated to be between 0.4 and 1.0 mM (53, 54). Thus, it is possible that the in vivo oxidation of HNE would be R-selective by ALDH2 as well as by ALDH5A. On the other hand, it is not known whether ALDH3A (another ALDH that oxidizes HNE) is enantioselective (55). Although our data show the R-selectivity of HNE oxidation, we think it unlikely that (S)-HNE protein-adducts would then accumulate in cells given the S-selectivity of glutathione S-transferase catalyzed detoxification of HNE (23, 26). Nonetheless, the chirality of HNE adds another level of complexity since immunochemical data derived using (R)-HNE adduct and (S)-HNE adduct antibodies demonstrate that the adducts of the individual enantiomers are distributed differently in cells in vivo following oxidative damage (21). In conclusion, this work demonstrates that chirality plays an important role in the metabolism of HNE and that the enantiomers of HNE serve as useful tools for studying enzymesubstrate interaction. Our experiments show that the enantioselectivity of HNE oxidation by ALDH5A and ALDH2 can be regulated by the Mg2+ ion content. It is not known the extent to which Mg2+ availability alters aldehyde oxidation in vivo or even in intact mitochondria. While this might not be important in regions of brain where it appears that the bulk of the oxidation of HNE is catalyzed by ALDH5, it could be more important in tissue such as liver where the oxidation is primarily oxidized by ALDH2 (16). Regulation of the activity by the Mg2+ ion content may be of consequence in different regions of the central nervous system where these enzymes have different levels of expression and cell-specific expression. Further experiments are needed to determine the extent to which the chirality of HNE influences protein binding and metabolism and the role of Mg2+ ions in regulating HNE metabolism in situ. Acknowledgment. We gratefully acknowledge NIH Grants P20 RR17699-01 COBRE (M.J.P.) from the NCRR and AA15145-01 (M.J.P.) and AA05812 (H.W.) from NIAAA. The Proteomics Core Facility was supported by NIH Grant P20 RR016741 from the INBRE Program of the NCRR. X.L. is the recipient of a faculty initiation grant from NSF EPSCoR (EPS-
894 Chem. Res. Toxicol., Vol. 20, No. 6, 2007
0447679). We thank Mr. Darryl Mosely and Ms. Laura Leiphon for the preparation of the mitochondria.
Brichac et al.
(23)
References (1) 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. (2) Uchida, K. (2003) 4-Hydroxy-2-nonenal: a product and mediator of oxidative stress. Prog. Lipid Res. 42, 318-343. (3) Picklo, M. J., Montine, T. J., Amarnath, V., and Neely, M. D. (2002) Carbonyl toxicology and Alzheimer’s disease. Toxicol. Appl. Pharmacol. 184, 187-197. (4) McKracken, E., Graham, D. I., Nilsen, M., Stewart, J., Nicoll, J. A., and Horsburgh, K. (2001) 4-Hydroxynonenal immunoreactivity is increased in human hippocampus after global ischemia. Brain Pathol. 11, 414-421. (5) Montine, K. S., Olson, S. J., Amarnath, V., Whetsell, W. O., Jr., Graham, D. G., and Montine, T. J. (1997) Immunohistochemical detection of 4-hydroxy-2-nonenal adducts in Alzheimer’s disease is associated with inheritance of APOE4. Am. J. Pathol. 150, 437443. (6) Sayre, L. M., Zelasko, D. A., Harris, P. L., Perry, G., Salomon, R. G., and Smith, M. A. (1997) 4-Hydroxynonenal-derived advanced lipid peroxidation end products are increased in Alzheimer’s disease. J. Neurochem. 68, 2092-2097. (7) Yoritaka, A., Hattori, N., Uchida, K., Tanaka, M., Stadtman, E. R., and Mizuno, Y. (1996) Immunohistochemical detection of 4-hydroxynonenal protein adducts in Parkinson disease. Proc. Natl. Acad. Sci. U.S.A. 93, 2696-2701. (8) Sayre, L. M., Arora, P. K., Iyer, R. S., and Salomon, R. G. (1993) Pyrrole formation from 4-hydroxynonenal and primary amines. Chem. Res. Toxicol. 6, 19-22. (9) Amarnath, V., Valentine, W. M., Montine, T. J., Patterson, W. H., Amarnath, K., Bassett, C. N., and Graham, D. G. (1998) Reactions of 4-hydroxy-2(E)-nonenal and related aldehydes with proteins studied by carbon-13 nuclear magnetic resonance spectroscopy. Chem. Res. Toxicol. 11, 317-328. (10) Nadkarni, D. V., and Sayre, L. M. (1995) Structural definition of early lysine and histidine adduction chemistry of 4-hydroxynonenal. Chem. Res. Toxicol. 8, 284-291. (11) Petersen, D. R., and Doorn, J. A. (2004) Reactions of 4-hydroxynonenal with proteins and cellular targets, Free Radical Biol. Med. 37, 937-945. (12) Alary, J., Gueraud, F., and Cravedi, J. P. (2003) Fate of 4-hydroxynonenal in vivo: disposition and metabolic pathways. Mol. Aspects Med. 24, 177-187. (13) Kubatova, A., Murphy, T. C., Combs, C., and Picklo, M. J., Sr. (2006) Astrocytic Biotransformation of trans-4-Hydroxy-2-nonenal Is DoseDependent. Chem. Res. Toxicol. 19, 844-851. (14) Srivastava, S., Chandra, A., Wang, L. F., Seifert, W. E., Jr., DaGue, B. B., Ansari, N. H., Srivastava, S. K., and Bhatnagar, A. (1998) Metabolism of the lipid peroxidation product, 4-hydroxy-trans-2nonenal, in isolated perfused rat heart. J. Biol. Chem. 273, 1089310900. (15) Mitchell, D., and Petersen, D. (1987) The oxidation of a,b,unsaturated aldehydic products in lipid peroxidation by rat liver aldehyde dehydrogenases. Toxicol. Appl. Pharmacol. 87, 403-410. (16) Murphy, T. C., Amarnath, V., Gibson, K. M., and Picklo, M. J., Sr. (2003) Oxidation of 4-hydroxy-2-nonenal by succinic semialdehyde dehydrogenase (ALDH5A). J. Neurochem. 86, 298-305. (17) Chambliss, K. L., and Gibson, K. M. (1992) Succinic semialdehyde dehydrogenase from mammalian brain: subunit analysis using polyclonal antiserum. Int. J. Biochem. 24, 1493-1499. (18) Ryzlak, M. T., and Pietruszko, R. (1988) Human brain “high Km” aldehyde dehydrogenase: purification, characterization, and identification as NAD+-dependent succinic semialdehyde dehydrogenase. Arch. Biochem. Biophys. 266, 386-396. (19) Kang, T. C., Park, S. K., Hwang, I. K., An, S. J., Choi, S. Y., Cho, S. W., and Won, M. H. (2002) Spatial and temporal alterations in the GABA shunt in the gerbil hippocampus following transient ischemia. Brain Res. 944, 10-18. (20) Picklo, M. J., Olson, S. J., Markesbery, W. R., and Montine, T. J. (2001) Expression and activities of aldo-keto oxidoreductases in Alzheimer disease. J. Neuropathol. Exp. Neurol. 60, 686-695. (21) Hashimoto, M., Sibata, T., Wasada, H., Toyokuni, S., and Uchida, K. (2003) Structural basis of protein-bound endogenous aldehydes. Chemical and immunochemical characterizations of configurational isomers of a 4-hydroxy-2-nonenal-histidine adduct. J. Biol. Chem. 278, 5044-5051. (22) Honzatko, A., Brichac, J., Murphy, T. C., Reberg, A., Kubatova, A., Smoliakova, I. P., and Picklo, M. J., Sr. (2005) Enantioselective
(24) (25)
(26)
(27) (28)
(29) (30) (31) (32) (33) (34) (35)
(36) (37) (38)
(39) (40)
(41) (42) (43)
(44) (45)
(46)
metabolism of trans-4-hydroxy-2-nonenal by brain mitochondria. Free Radical Biol. Med. 39, 913-924. Hiratsuka, A., Hirose, K., Saito, H., and Watabe, T. (2000) 4-Hydroxy2(E)-nonenal enantiomers: (S)-selective inactivation of glyceraldehyde-3-phosphate dehydrogenase and detoxification by rat glutathione S-transferase A4-4. Biochem. J. 349, 729-735. Tatton, W. G., Chalmers-Redman, R., Brown, D. and Tatton, N. (2003) Apoptosis in Parkinson’s disease: signals for neuronal degradation. Ann. Neurol. 53 (Suppl. 3), S61-70, discussion S70-62. Wang, Q., Woltjer, R. L., Cimino, P. J., Pan, C., Montine, K. S., Zhang, J., and Montine, T. J. (2005) Proteomic analysis of neurofibrillary tangles in Alzheimer disease identifies GAPDH as a detergentinsoluble paired helical filament tau binding protein. FASEB J. 19, 869-871. Hiratsuka, A., Tobita, K., Saito, H., Sakamoto, Y., Nakano, H., Ogura, K., Nishiyama, T., and Watabe, T. (2001) (S)-preferential detoxification of 4-hydroxy-2(E)-nonenal enantiomers by hepatic glutathione S-transferase isoforms in guinea-pigs and rats. Biochem. J. 355, 237244. Chandra, A., and Srivastava, S. K. (1997) A synthesis of 4-hydroxy2-trans-nonenal and 4-(3H) 4-hydroxy-2-trans-nonenal. Lipids 32, 779-782. Alary, J., Bravais, F., Cravedi, J. P., Debrauwer, L., Rao, D., and Bories, G. (1995) Mercapturic acid conjugates as urinary end metabolites of the lipid peroxidation product 4-hydroxy-2-nonenal in the rat. Chem. Res. Toxicol. 8, 34-39. Murphy, T. C., Amarnath, V., and Picklo, M. J., Sr. (2003) Mitochondrial oxidation of 4-hydroxy-2-nonenal in rat cerebral cortex. J. Neurochem. 84, 1313-1321. Nguyen, E., and Picklo, M. J. (2003) Inhibition of succinic semialdehyde dehydrogenase activity by alkenal products of lipid peroxidation. Biochim. Biophys. Acta 1637, 107-112. Bradford, M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254. Leiphon, L. J., and Picklo, M. J. S. Inhibition of aldehyde detoxification in CNS mitochondria by fungicides. Neurotoxicology (in press). Combet, C., Jambon, M., Deleage, G., and Geourjon, C. (2002) Geno3D: automatic comparative molecular modelling of protein. Bioinformatics 18, 213-214. Schuttelkopf, A. W., and van Aalten, D. M. (2004) PRODRG: a tool for high-throughput crystallography of protein-ligand complexes. Acta Crystallogr., D: Biol. Crystallogr. 60, 1355-1363. Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S., Greenblatt, D. M., Meng, E. C., and Ferrin, T. E. (2004) UCSF Chimerasa visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605-1612. Yang, S. K., Bao, Z., and Shou, M. (1996) Stereoselective nucleophilic substitution of oxazepam and racemization in acidic methanol and ethanol. Chirality 8, 214-223. Aso, Y., Yoshioka, S., and Takeda, Y. (1990) Epimerization and racemization of some chiral drugs in the presence of human serum albumin. Chem. Pharm. Bull. (Tokyo) 38, 180-184. Gineyts, E., Cloos, P. A., Borel, O., Grimaud, L., Delmas, P. D., and Garnero, P. (2000) Racemization and isomerization of type I collagen C-telopeptides in human bone and soft tissues: assessment of tissue turnover. Biochem. J. 345 (Part 3), 481-485. Stadtman, E. R. (1988) Protein modification in aging, J. Gerontol. 43, B112-120. Schneider, C., Tallman, K. A., Porter, N. A., and Brash, A. R. (2001) Two distinct pathways of formation of 4-hydroxynonenal. Mechanisms of nonenzymatic transformation of the 9- and 13-hydroperoxides of linoleic acid to 4-hydroxyalkenals. J. Biol. Chem. 276, 20831-20838. Schneider, C., Porter, N. A., and Brash, A. R. (2004) Autoxidative transformation of chiral omega6 hydroxy linoleic and arachidonic acids to chiral 4-hydroxy-2E-nonenal. Chem. Res. Toxicol. 17, 937-941. Doorn, J. A., Hurley, T. D., and Petersen, D. R. (2006) Inhibition of human mitochondrial aldehyde dehydrogenase by 4-hydroxynon-2enal and 4-oxonon-2-enal. Chem. Res. Toxicol. 19, 102-110. Ho, K. K., Allali-Hassani, A., Hurley, T. D., and Weiner, H. (2005) Differential effects of Mg2+ ions on the individual kinetic steps of human cytosolic and mitochondrial aldehyde dehydrogenases. Biochemistry 44, 8022-8029. Dickinson, F. M., and Hart, G. J. (1982) Effects of Mg2+, Ca2+ and Mn2+ on sheep liver cytoplasmic aldehyde dehydrogenase. Biochem. J. 205, 443-448. Larson, H. N., Weiner, H., and Hurley, T. D. (2005) Disruption of the coenzyme binding site and dimer interface revealed in the crystal structure of mitochondrial aldehyde dehydrogenase “Asian” variant. J. Biol. Chem. 280, 30550-30556. Moore, S. A., Baker, H. M., Blythe, T. J., Kitson, K. E., Kitson, T. M., and Baker, E. N. (1998) Sheep liver cytosolic aldehyde dehydro-
EnantioselectiVe Oxidation of HNE
(47) (48)
(49) (50) (51) (52)
genase: the structure reveals the basis for the retinal specificity of class 1 aldehyde dehydrogenases. Structure 6, 1541-1551. Steinmetz, C. G., Xie, P., Weiner, H., and Hurley, T. D. (1997) Structure of mitochondrial aldehyde dehydrogenase: the genetic component of ethanol aversion. Structure 5, 701-711. Liu, Z. J., Hempel, J., Sun, J., Rose, J., Hsiao, D., Chang, W. R., Chung, Y. J., Kuo, I., Lindahl, R., and Wang, B. C. (1997) Crystal structure of a class 3 aldehyde dehydrogenase at 2.6 Å resolution. AdV. Exp. Med. Biol. 414, 1-7. Esterbauer, H., Schaur, R., and Zollner, H. (1991) Chemistry and Biochemistry of 4-hydroxynonenal, malondialdehyde, and related aldehydes. Free Radical Biol. Med. 11, 81-128. Burgi, H. B., and Dunitz, J. D. (1983) From crystal statics to chemical dynamics. Acc. Chem. Res. 16, 153-161. Johansson, K., El-Ahmad, M., Ramaswamy, S., Hjelmqvist, L., Jornvall, H., and Eklund, H. (1998) Structure of betaine aldehyde dehydrogenase at 2.1 Å resolution. Protein Sci. 7, 2106-2117. Lamb, A. L., and Newcomer, M. E. (1999) The structure of retinal dehydrogenase type II at 2.7 Å resolution: implications for retinal specificity. Biochemistry 38, 6003-6011.
Chem. Res. Toxicol., Vol. 20, No. 6, 2007 895 (53) Corkey, B. E., Duszynski, J., Rich, T. L., Matschinsky, B., and Williamson, J. R. (1986) Regulation of free and bound magnesium in rat hepatocytes and isolated mitochondria. J. Biol. Chem. 261, 25672574. (54) Lenzen, S., Hickethier, R., and Panten, U. (1986) Interactions between spermine and Mg2+ on mitochondrial Ca2+ transport. J. Biol. Chem. 261, 16478-16483. (55) Pappa, A., Estey, T., Manzer, R., Brown, D., and Vasiliou, V. (2003) Human aldehyde dehydrogenase 3A1 (ALDH3A1): biochemical characterization and immunohistochemical localization in the cornea. Biochem. J. 376, 615-623. (56) Farres, J., Wang, T. T., Cunningham, S. J., and Weiner, H. (1995) Investigation of the active site cysteine residue of rat liver mitochondrial aldehyde dehydrogenase by site-directed mutagenesis. Biochemistry 34, 2592-2598. (57) Sheikh, S., Ni, L., Hurley, T. D., and Weiner, H. (1997) The potential roles of the conserved amino acids in human liver mitochondrial aldehyde dehydrogenase. J. Biol. Chem. 272, 18817-18822.
TX7000509
