Chem. Res. Toxicol. 1995,8, 34-39
34
Mercapturic Acid Conjugates as Urinary End Metabolites of the Lipid Peroxidation Product 4-Hydroxy-2-nonenalin the Rat Jacques Alary,* Fabienne Bravais, Jean-Pierre Cravedi, Laurent Debrauwer, Dinesh Rao, and Georges Bories Laboratoire des Xhobiotiques, I.N.R.A., BP 3, 31931 Toulouse Cedex, France Received June 27, 1994@ 4-Hydroxy-2-nonenal (HNE), a n aldehyde end product of lipid peroxidation in biological systems, is capable of producing a range of powerful biological effects. Despite its biological relevance, the metabolic fate of this aldehyde is unknown in vivo. This study examines the urinary excretion of HNE in the rat and the nature of metabolites formed. Following iv administration of L3H1HNE, the majority of the dose appeared in urine (67.1% after 48 h). The radio-HPLC metabolic profile showed that no unchanged parent compound was detected in urine whereas at least four metabolites were present, most of them corresponding t o mercapturic acid conjugates. Two major pathways were involved in the biotransformation of HNE in vivo: (i) reductiodoxidation of the aldehyde group, and (ii) conjugation to endogenous glutathione leading to mercapturic acid conjugates in urine. These end products were isolated by HPLC and identified by mass spectrometry as HNE mercapturic acid, 1,4-dihydroxynonene mercapturic acid, 4-hydroxynonenoic mercapturic acid, and the corresponding lactone.
Introduction
Materials and Methods
The reaction of oxygen species with membrane lipids leads to the formation of lipid hydroperoxides, the degradation of which generates several aldehydic compounds. Since HNEl is the major aldehyde formed from 0-6 unsaturated fatty acids (11, membrane phospholipids (Z),and lipoproteins (3), it is considered as a reliable index of free radical stimulated peroxidation. The level of endogenous HNE in tissues, including plasma, is in the range of 0.1-3.0 nmoVg and can increase to 10 nmoVg in conditions of oxidative stress (4). Owing to the high reactivity of HNE toward thiols ( 5 , 6 ) and amino groups (7,8), these levels seem sufficient to produce cytopathological effects. HNE has been shown to inhibit DNA (9) and protein synthesis (101,and to produce mutagenic (11, 12) and genotoxic (13) effects in various types of cells. Such damages, demonstrated in vitro, suggest that HNE may be considered as a potent cytotoxic compound. Despite its biological relevance, our knowledge of the biotransformation of HNE is still rather limited. In vitro studies conducted with rat liver subcellular fractions demonstrated that HNE can be metabolized to the corresponding alcohol and carboxylic acid or conjugated to glutathione (GSH)(14-16). These metabolic pathways were confirmed in isolated rat hepatocytes (17). However, there is no information concerning the metabolism of HNE in the whole animal. Nevertheless, Winter et a l . (18)have shown that the homologous 4-hydroxyhexenal injected t o the rat via the portal vein was partly excreted in urine as the corresponding mercapturic acid. The present study was designed to investigate the excretion of HNE biotransformation products in urine and to identify the metabolic pathways of this hydroxyalkenal in the rat after iv administration of PHIHNE.
Chemicals. HNE standard was kindly provided by Professor H. Esterbauer (Department of Biochemistry, University of Graz, Austria). It was supplied as HNE diethyl acetal dissolved in chloroform and was stored at -20 "C until required. Just prior to use, HNE was prepared from its diethyl acetal derivative by 1 mM HC1 hydrolysis during 1 h a t room temperature. For animal studies, HE diethyl acetal was synthesized according to Esterbauer et al. (29)and HNE liberated as for the standard. [4-3HlHNE diethyl acetal was synthesized a t CEA Service des Mol6cules Marquees CEN (Saclay, France), according to the method developed in our laboratory for the deuterated compound.2 The 4-hydroxy group of HNE diethyl acetal was oxidized into the corresponding ketone with activated dimethyl sulfoxide (20). The ketone was reduced by tritiated sodium borohydride into [4-3HlHNE diethyl acetal. The labeled compound showed the same chromatographic (TLC, HPLC) behavior as standard HNE diethyl acetal (data not shown). Its radiochemical purity, determined by HPLC, was 95%,and its specific activity was 222 GBq/mmol. Sodium borohydride, Raney nickel, (10%) palladium on activated carbon (PdK), sodium chlorite, sulfamic acid, and N-acetylcysteine were purchased from Aldrich (Saint Quentin Fallavier, France). Aldehyde dehydrogenase (EC 1.2.1.5) from bakers' yeast was purchased from Sigma Chimie (L'Isle d'Abeau Chesnes, France). All chemical solvents and reagents for the preparation of buffers and HPLC eluents were in the highest grade commercially available from Merck (Nogent-sur-Marne, France) or Carlo Erba (Rueil Malmaison, France). Ultrapure water from Milli-Q system (Millipore, Saint Quentin en Yvelines, France) was used for HPLC eluent preparation. Syntheses of Standards. HNE-MA was synthesized according to the procedure of Winter et al. (28)for 4-hydroxyhexenal mercapturic acid. HNE was reacted with a stoichiometric amount of N-acetylcysteine for 72 h at 50 "C in 1 mL of 0.1 M ammonium bicarbonate (pH 8.1). After addition of ammonium carbonate, the reaction mixture was extracted with ethyl acetate to remove unreacted HNE. The solution was then acidified to pH 2 with 1M phosphoric acid and extracted three times with 3 mL of ethyl acetate. The same procedure was used to prepare L3H1HNE-MA from L3H1HNE.
Abstract published in Advance ACS Abstracts, November 1,1994. Abbreviations: HNE, 4-hydroxy-2-nonenal;HNE-MA, 4-hydroxynonenal mercapturic acid; DHN, 1,4-dihydroxynonene;DHN-MA, 1,4dihydroxynonene mercapturic acid; HNA, 4-hydroxynonenoic acid; HNA-MA, 4-hydroxynonenoic mercapturic acid; HNA-lactone-=, hydroxynonenoic lactone mercapturic acid. @
0893-228x/95/2708-0034$09.00l0
D. Rao, F. Bravais, J. Alary, R. C. Rao, L. Debrauwer, G. Bories. In preparation.
0 1995 American Chemical Society
Chem. Res. Toxicol., Vol. 8, No. 1, 1995 35
Urinary Metabolites of 4-Hydroxy-2-nonenal
OH
I
H
C5H11-C -CH - C H 2 - C H 2 0 H I I /NH--CO-CH, H S- CH2-CH 'COOH
1,4-dihydroxynonane
t
NaBH,
OH I C5H, ]-C -CH =CH -CH,OH
I
I
NaBH,
C,H,
1-
OH I
7
- C H I CH
-
H
PH
c ~ H ,7~-CH H
goH
N-acetylcysteiz C ~ H I I 72h, 5OoC, pH 8.1
S- CHz-CH
'OH 4-hydroxynonanoic acid
CH,
HNE-MA
0
TCH
NH-CO-
'COOH
NaCQ 3h, pH 2.5
OH
I
C5Hii-C - C H -CH, 1 3 0 0 H I I / NH-CO-CH3
HNA-lactone-MA
Figure 1. Scheme of standards synthesis. DHN and DHN-MA were synthesized by sodium borohydride reduction of HNE and HNE-MA, respectively. HNA was synthesized by sodium chlorite oxidation of HNE using the procedure described by Lindgreen et al. (21) for the synthesis of aromatic acids. A mixture of [3HlHNE (as tracer) and HNE (13 pmol) in 500 pL of 1mM HC1 was added with 100 pL of a n aqueous solution of sulfamic acid (15 pmol) and 100 pL of sodium chlorite (15 pmol) in water, for 3 h a t room temperature. The solution was extracted three times with 2 mL of chloroform. The chloroform extract was treated with 5 mL of 0.1 M sodium bicarbonate solution. The organic layer was discarded, and then the aqueous phase was acidified to pH 2 with addition of 1 M phosphoric acid. This solution was extracted three times with 5 mL of chloroform. The yield of HNA was 25%, and the product was then characterized by F B I MS giving the pseudo molecular ion [M - HI- a t mlz 171. HNA-MA and HNA-lactone-MA were synthesized by reaction of HNA with N-acetylcysteine as follows: N-acetylcysteine (150 pmol) was added to 5 mL of 0.1 M sodium bicarbonate solution containing 85 pmol of HNA. The mixture was brought to pH 4.8 with 0.5 M phosphoric acid and stirred for 72 h at 55 "C. HPLC analysis (system 11)of the reaction mixture gave rise to three compounds characterized by FAB/MS. The major product (50%of the radioactivity) corresponded to unreacted HNA. A compound accounting for 15% of the radioactivity gave a pseudo molecular ion [M - HI- a t mlz 316 corresponding to HNAlactone-MA. The third isolated product, accounting for 17% of the radioactivity, gave [M - HI- at mlz 334 that corresponded to HNA-MA. The lactone of 4-hydroxynonanoic acid was obtained by catalytic hydrogenation of HNA. HNA (0.1 mg) in 10 mL of methanol was added with a large excess of PdlC and reduced with hydrogen at 40 psi in a Parr apparatus under magnetic stirring for 15 h. 4-Hydroxynonanoic acid was prepared by treatment of the corresponding lactone with 0.1 M NaOH at 60 "C for 15 h and subsequent extraction with ethyl acetate at pH 2. 1,4-Dihydroxynonane was obtained by hydrogenation of DHN in the presence of PdlC.
The procedures used for the synthesis of standards are summarized in Figure 1. Incubation of HNE and HNE-MA with Cytosolic Fractions. Rat liver cytosolic fractions obtained by differential centrifugation were incubated in the presence of NADH and [3H]HNE or €€NE-MA as previously described by Esterbauer (14). DHN or DHN-MA were determined by radio-HPLC (system I).
Incubation of HNE and HNE-MA with Aldehyde Dehydrogenase. A 2 mL aliquot of 100 mM phosphate buffer (pH 7.4) containing 1 mM (3.3 kBq) [3HlHNE or 1 mM (3.3 kBq) [3H]HNE-MA, 5 mM NAD+, 2 mM pyrazole, and 2 units of aldehyde dehydrogenase was incubated overnight at 25 "C. After removal of unchanged HNE by extraction with diethyl ether at pH 7.4, the reaction mixture was brought to pH 2 and extracted twice with 5 mL of ethyl acetate. The solvent was evaporated under Nz at room temperature. The residue was taken up with acetonitrile, and the respective oxidized products (HNA or HNAMA) were determined by radio-HPLC (system 11). HPLC. The HPLC system consisted of two 420 Kontron pumps with a gradient former 491 and a Spectra Physics UV 150 detector set at 223 nm. Radioactivity detection was carried out with a radiomatic FIo-one A-200 instrument (Radiomatic, La-Queue-lez-Yvelines,France) with Flo-scint I1 as scintillation cocktail (Packard Instrument Co, Downers Grove, IL). Collection of radioactive peaks for MS analysis was carried out with a Gilson Model 202 fraction collector (Gilson, Villiers-le-Bel, France). HPLC separations were performed on a Spherisorb ODS 2 reversed-phase column (5 pm, 25 x 0.46 cm) from Shandon (Eragny, France), protected by a precolumn (5pm, 1 x 0.46 cm Spherisorb ODS 2). Two gradient elution systems were developed: System I: Eluent A contained 90% 20 mM ammonium acetate in water, adjusted to pH 4.5 with acetic acid, and 10% of acetonitrile. Eluent B consisted of 80% aqueous buffer as described above and 20% acetonitrile. Solvents were delivered a t a flow rate of 0.8 mUmin as follows: 0-18 min 100% A, 18-19 min linear gradient from 100% A to 100% B; 19-60 min 100%B. System I1 consisted of a 20 min linear gradient from 100% A to 100%
Alary et al.
36 Chem. Res. Toxicol., Vol. 8, No. 1, 1995
aJ
m
U
I
a
U
-
t 1.15
?
i i
i
IT
I1
I\;
5
R e t e n t i o n Time ( m n )
Time ( h )
Figure 2. Cumulative urinary excretion of radioactivity following iv administration of [SHIHNE. Values are means f SD &om 3 rats.
acetonitrile at a flow rate of 1 mumin. Metabolites were quantitated using 3H activity monitoring. Animal Treatments. Four male Wistar rats, weighing about 200 g were used. They were weakly anesthetized with diethyl ether, and 3 rats were injected with 500 pL of Ringer's solution containing PHIHNE (1MBq = 700 ng) into the penis vein. To facilitate the production of larger quantities of metabolites needed for identification purposes by FABMS, the fourth rat was injected with a combination of 1.6 mg of HNE and L3H1HNE to a final specific activity of 0.62 MBq/mg. Following the injection, the rats were housed in individual metabolism cages for the collection of urine and feces for 48 h. Water was provided ad libitum. Urine was collected 2,8, 24, and 48 h after iv treatment. Urine volumes were measured, and samples were filtered through a Millex-HA 0.45pm Nter (Millipore)and stored at -20 "C until analysis. Radioactivity Determination. Aliquots (50-200 pL) of urine samples were directly counted in a Packard Tricarb scintillation counter (Model 4430) with Pico-Fluor 40 as the scintillation cocktail. The radioactivity of feces was measured after combustion in a Packard oxidizer (Model 306) with Monophase S as scintillation cocktail (Packard). Mass Spectrometry. Peak identification by FABMS was carried out using the 0-2 h urine from the rat dosed with L3H1HNE and HNE (1.6 mg). A 20 pL aliquot of urine was directly injected into HPLC to confirm that the radioprofile was similar to those obtained with 0-2 h urines from the three rats dosed with [3H]HNE alone. ARer concentration, three direct injections were carried out and the metabolites collected. The eluent was evaporated to dryness and each fraction residue (50-200 pg) dissolved in 50 pL of methanol and stored until mass spectrometry analysis. Mass spectra were obtained on a Nermag R-10-1OH (Delsi Nermag Instruments, Argenteuil, France) single quadrupole mass spectrometer working in the negative ion mode. FAB experiments were achieved with an M-Scan FAB gun (M-Scan Ltd., Ascot, U.K.), and xenon gas was used for bombardment at an accelerating voltage of 8 kV, with 1-2 mA as discharge current. Samples were prepared by mixing 1 pL of the sample solution (1pg/pL in methanol) with 1pL of the matrix Magic Bullet (dithiothreitol-dithioerythritol, 5:1).
Results Excretion of the administered radioactivity occurred rapidly in urine. As soon as 2 h after treatment 40.2% (f2.8%) of the radioactivity was eliminated via this route, this percentage reaching 67.1% (f7.0%) within 48 h (Figure 2). A minor part of the radioactivity was eliminated in feces (3.0 f 0.6% within 48 h, data not shown). HPLC analysis (system I) after direct injection of urine samples collected 2 h aRer treatment allowed the separa-
Figure 3. Typical HPLC profile of [3H]HNE urinary metabolites collected 2 h after iv administration of [3H]HNE. System I was used as chromatographic conditions (see text). A 20 pL aliquot of filtered 0-2 h urine was directly injected into HPLC.
tion of HNE metabolites (Figure 3). Total recovery of HPLC injected radioactivity was controlled for each urine analysis and averaged 96%. Unchanged HNE ( t =~77.7 min) was absent from the urine. A group of highly polar compounds accounting for 40% of the injected radioactivity (group 1)remained unresolved. The separation and characterization of this group containing more than four individual compounds are currently in progress. Metabolite 2 accounted for 3.2% of the administered dose. After cleavage of the thioether bond using Raney nickel treatment (22)the radioactivity was extracted with ethyl acetate at pH 2 only, suggesting the presence of an acidic group in the parent compound. HPLC (system 11)of the ethyl acetate extract gave a major peaks (80%) coeluting with standard 4-hydroxynonanoic acid ( t = ~ 13.6 min) and a minor peak (20%) coeluting with the corresponding standard lactone (tR = 19.8 min). Treatment of metabolite 2 with 0.1 M acetic acid at 60 "C for 15 h, yielded metabolite 5. In addition metabolite 2 coeluted with one of the compounds formed by reaction of HNA with N-acetylcysteine. These chemical data suggested that metabolite 2 should correspond to HNAMA. Further characterization of this metabolite was performed by FAB/MS that showed a pseudo molecular ion [M - HI-at mlz 334 (Figure 4). Metabolite 3 accounted for 12.4% of the injected dose. This compound had the same retention time as standard DHN-MA. f i r treatment with Raney nickel, metabolite 3 gave a compound extractable with chloroform at neutral pH which coeluted with standard 1,4-dihydroxynonane ( t =~ 13.1 min, system 11). Further characterization of metabolite 3 was performed by FAEVMS. The FAB spectrum exhibited an intense [M - HI- ion at mlz 320 (Figure 4), confirming that metabolite 3 was DHN-MA. Metabolite 4 which accounted for 5.9% of the administered dose showed the same retention time as standard HNE-MA. After reduction with sodium borohydride, metabolite 4 produced a single compound which coeluted with metabolite 3. Following the reduction step, a subsequent treatment by Raney nickel gave a compound with a retention time similar to that of standard 1,4dihydroxynonane, suggesting that metabolite 4 should correspond to HNE-MA. The identity of metabolite 4 was confirmed by FAB/MS showing a pseudo molecular ion [M - HI- a t mlz 318 (Figure 5). Metabolite 5 which represented 5.6%of the administered dose showed the same retention time as one of the compounds resulting from the reaction of HNA with
Chem. Res. Toxicol., Vol. 8, No. 1, 1995 37
Urinary Metabolites of 4-Hydroxy-2-nonenal -
-
1
OH
HNE-MA
HNA-MA
4joH
C A
,m-co-CH,
E- cH,-cn
\ COOH
1
I
DHN-MA I
l I ! i
Figure 4. Negative FAB mass spectra of [SHIHNE urinary metabolites. Molecular ions of HNA-MA (2)and DHN-MA (3) are located a t mlz 334 and 320,respectively. Ions due to the FAB matrix are present at mlz 273,305,and 307.
N-acetylcysteine. Following Raney nickel treatment metabolite 5 gave rise to a single compound completely extractable by organic solvent (chloroform-methanol, 9:1 v/v) at pH 7.2. Treatment of the formed compound with 25 mM NaOH, at 60 "C for 15 h, quantitatively yielded a product which coeluted (HPLC system 11) with the standard 4-hydroxynonanoic acid. After cleavage of the thioether bond, HPLC analysis (system 11) showed that the resulting compound coeluted with the lactone of 4-hydroxynonanoic acid (tR = 19.8 min). All these chemical data suggest that metabolite 5 could be HNA-lactoneMA. The nature of metabolite 5 was confirmed by FABI MS analysis that showed a [M - HI- ion at mlz 316 (Figure 5). Note that the FAB mass spectra presented in Figures 4 and 5 exhibit some peaks with a mass greater than [M - Hl-. Likely, these are due to endogenous material which gives rise to some additional peaks on the spectra. Nevertheless, when observing carefully the major peaks of the spectra, the [M - HI- ions of the metabolites are the unique peaks with an even mass number. Indeed, even mass numbers characterize deprotonated (even electron number) species containing one nitrogen atom, which is the case for the HNE mercapturic acid derivatives. Incubations of L3H1HNE with cytosolic fractions fortified with NADH resulted in the formation of DHN. In contrast, when [3HlHNE-MAwas used as substrate, no formation of [3H]DHN-MAoccurred. In the presence of aldehyde dehydrogenase and NAD+, L3H1HNEunderwent biotransformation to [3HlHNA. No formation of L3H1-
Figure 6. Negative FAB mass spectra of [3H]HNE urinary metabolites. Molecular ions of HNE-MA (4)and HNA-lactoneMA (5) are located at mlz 318 and 316,respectively. Ions due t o the FAB matrix are present at mlz 273,305,and 307. l5
! .w
m
t
T I
II
I 1
°
0
V
10
20
30
40
I 50
Time ( h )
Figure 6. Cumulative urinary excretion of metabolites of I3H1HNE: (A) DHN-MA, (0) HNE-MA; ( x ) HNA-lactone-MA, (0) HNA-MA. Values are measn f SD from 3 rats.
HNA-MA was observed when L3H1HNE-MAwas used as substrate (data not shown). Estimation of tritiated water was investigated using two experimental procedures. Liquidlsolid extraction using an SAX cartridge resulted in a total retention of urinary radioactivity, providing evidence for the absence of tritiated water. This result was confirmed by the lack of radioactive peak eluting a t the column void volume during HPLC analysis. The cumulative urinary excretion of the four metabolites is shown in Figure 6. The major part of each metabolite was excreted 2 h after treatment. It can be
38 Chem. Res. Toxicol., Vol. 8, No.1, 1995
/
OH
I
C5Hi 1-C -CH W H --CH,OH
1
" OH
I t
DHN
C5Hll-c
Alary et al.
OH I IJH =cH IJ
I
PH
H
I
C5Hl,--F-CH=CH
