Reactions of N-methyl-2-phenylindole with malondialdehyde and 4- hydroxyalkenals. Mechanistic aspects of the colorimetric assay of lipid peroxidation

Article abstract of DOI:10.1021/tx970180z

Under specific acidic conditions, both malondialdehyde (1, MDA) and 4- hydroxynonenal (2, 4-HNE) react with N-methyl-2-phenylindole (3) to give the same chromophoric cyanine 4 with maximal absorbance at 586 nm. Under such conditions, the reaction of 3 with 4-HNE (2) as well as with alkanals yields a second chromophoric cyanine 10 with maximal absorbance at 505 nm. The influence of different acids, iron(III), and oxygen on the reaction of 3 with such aldehydes was studied in detail. Under anaerobic conditions, the acid- induced reaction of 4-HNE with 3 afforded three rapidly interconverting intermediates, 5-7. Their subsequent fragmentation to 4 and hexanal in the presence of iron(III) and oxygen is consistent with the tandem β- fragmentation of an indolyl radical cation. 1-Indolylalkenes were identified as essential intermediates in the acid-induced reaction of 3 with alkanals. A very mild iron(III)catalyzed fragmentation of these intermediates afforded the corresponding 3-formylindole 11 as the direct precursor of the 505 nm chromophore 10. Such reactions were markedly influenced by the nature of the acid. Contrary to the rapid chromogenic reaction of 4-HNE which was observed in the presence of methanesulfonic acid, the HCl-induced reaction of 4-HNE with 3 did not afford the 586 nm chromophore. Furthermore, hexanal did not yield the 505 nm chromophore 10 upon reaction with 3 in the presence of HCl, again in contrast with the rapid chromogenic reaction which was observed in the presence of methanesulfonic acid. Comparison of the reaction mixtures under the two assay conditions confirmed that the same intermediates were formed. We conclude that the nature of the acid plays a crucial role in the oxidative fragmentation of intermediates into chromophores, allowing the selective assay of MDA in the presence of 4-HNE, using HCl acidic conditions.

Full text of DOI:10.1021/tx970180z

1184
Chem. Res. Toxicol. 1998, 11, 1184-1194
Rea ction s of N-Meth yl-2-p h en ylin d ole w ith
Ma lon d ia ld eh yd e a n d 4-Hyd r oxya lk en a ls. Mech a n istic
Asp ects of th e Color im etr ic Assa y of Lip id P er oxid a tion
Irene Erdelmeier,* Dominique Ge´rard-Monnier,† J ean-Claude Yadan, and
J ean Chaudie`re‡
Centre de Recherche Oxis International S.A., 94385 Bonneuil sur Marne, France
Received October 6, 1997
Under specific acidic conditions, both malondialdehyde (1, MDA) and 4-hydroxynonenal (2,
4-HNE) react with N-methyl-2-phenylindole (3) to give the same chromophoric cyanine 4 with
maximal absorbance at 586 nm. Under such conditions, the reaction of 3 with 4-HNE (2) as
well as with alkanals yields a second chromophoric cyanine 10 with maximal absorbance at
505 nm. The influence of different acids, iron(III), and oxygen on the reaction of 3 with such
aldehydes was studied in detail. Under anaerobic conditions, the acid-induced reaction of
4-HNE with 3 afforded three rapidly interconverting intermediates, 5-7. Their subsequent
fragmentation to 4 and hexanal in the presence of iron(III) and oxygen is consistent with the
tandem â-fragmentation of an indolyl radical cation. 1-Indolylalkenes were identified as
essential intermediates in the acid-induced reaction of 3 with alkanals. A very mild iron(III)-
catalyzed fragmentation of these intermediates afforded the corresponding 3-formylindole 11
as the direct precursor of the 505 nm chromophore 10. Such reactions were markedly influenced
by the nature of the acid. Contrary to the rapid chromogenic reaction of 4-HNE which was
observed in the presence of methanesulfonic acid, the HCl-induced reaction of 4-HNE with 3
did not afford the 586 nm chromophore. Furthermore, hexanal did not yield the 505 nm
chromophore 10 upon reaction with 3 in the presence of HCl, again in contrast with the rapid
chromogenic reaction which was observed in the presence of methanesulfonic acid. Comparison
of the reaction mixtures under the two assay conditions confirmed that the same intermediates
were formed. We conclude that the nature of the acid plays a crucial role in the oxidative
fragmentation of intermediates into chromophores, allowing the selective assay of MDA in
the presence of 4-HNE, using HCl acidic conditions.
be suppressed completely by using hydrochloric instead
of methanesulfonic acid, allowing the selective measure-
ment of MDA in the presence of 4-HNE (23).
These experimental findings prompted us to study the
reaction of MDA, 4-hydroxynonenal, and alkanals with
N-methyl-2-phenylindole in the presence of either acid
and with or without iron(III) salts.
In tr od u ction
In the preceding paper, we described a colorimetric
assay for the cytotoxic aldehydes malondialdehyde (MDA,1
1) and 4-hydroxynonenal (4-HNE, 2) as byproducts of
lipid peroxidation.
Ma ter ia ls a n d Meth od s
Ca u tion : 4-Hydroxynonenal (2) is reported to be cytotoxic (1).
Pentanal, hexanal, heptanal, and octanal are irritants; hydro-
chloric and methanesulfonic acid are strong corrosive liquids.
All these chemicals should be handled with protective gloves and
the reactions described carried out in a hood.
The principle of this assay is the measurement of the
absorbance at 586 nm, which is obtained with MDA as
well as with 4-HNE upon reaction with N-methyl-2-
phenylindole (3) under acidic conditions.
Moreover, the reaction of 4-HNE, which is accompanied
by the formation of a second chromophore at 505 nm, can
Ch em ica ls. All reagents used were of the highest grade
commercially available. 1,1,3,3-Tetramethoxypropane (TMP)
was used as a source of MDA. 4-Hydroxynonenal diethyl acetal
was prepared according to ref 2 and used for the in situ
preparation of 4-HNE. 3-Formyl-N-methyl-2-phenylindole (11)
was synthesized according to ref 3. For the investigations of
the influence of iron(III) salts, methanesulfonic acid of purissime
grade (>99%) was purchased from Fluka (St. Quentin-Fallavier,
France). Batches of methanesulfonic acid from other sources
were in general contaminated with trace amounts of iron(III).
Column chromatographic purifications were performed on silica
Merck Si60 F254 or alumina Merck Al2O3 90 (activity II-III),
and TLC plates were purchased from Macherey-Nagel (Poly-
gram Sil G/UV254, 0.25 mm). Melting points are uncorrected
and were measured on a Gallenkamp apparatus.
* Corresponding author: Centre de Recherche Oxis International
S. A., F-94385 Bonneuil sur Marne, France. Telephone: (33) 1 49 80
45 64. Fax: (33) 1 49 80 01 66. E-mail: 101511.3306@compuserve.com.
† Present address: INSERM, U 479, Faculte´ Xavier Bichat, 16 rue
Henri Huchard, 75018 Paris, France.
‡ Present address: UFR de Biologie, Universite´ de Paris VII, 2 place
J ussieu, 75251 Paris Cedex 05, France.
1
Abbreviations: MDA, malondialdehyde; 4-HNE, 4-hydroxy-2(E)-
nonenal; TMP, 1,1,3,3-tetramethoxypropane; TBME, tert-butyl methyl
ether; TBA, 2-thiobarbituric acid; MPLC, medium-pressure liquid
chromatography.
10.1021/tx970180z CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/30/1998

Colorimetric Assay of Lipid Peroxidation. 2
Chem. Res. Toxicol., Vol. 11, No. 10, 1998 1185
Sp ectr oscop y. 1H (200 MHz) and 13C NMR spectra (50
MHz) were recorded on a Varian Gemini-200 spectrometer and
are reported in parts per million downfield from TMS. Fast
atom bombardment mass spectrometry (FAB/MS), electron-
impact mass spectrometry (EI-MS), and chemical-ionization
mass spectrometry (CI-MS) were performed on a Nermag R10-
10B apparatus. UV/visible spectra were recorded on a Uvikon
941 spectrophotometer from Kontron Instruments, in quartz
microcuvettes with a 1 cm optical path length.
Ga s Ch r om a togr a p h y. GC analyses were conducted using
a Varian 3400 gas chromatograph equipped with a DB-1 column
of 30 m, with a 0.23 mm diameter (phase 0.25 µm), from J &W
Scientific and FID detection (detector temperature of 280 °C
and injector temperature of 250 °C). The vector gas was helium
(10 psi); the split divisor was set at 20 mL/min. The temper-
ature program was as follows: an initial temperature of 40 °C
for 10 min and a final temperature of 150 °C (5 °C/min). The
injection volume was 1-2 µL.
Isola tion of 586 n m Ch r om op h or e 4a . TMP (164 mg, 164
µL, 1 mmol) was dissolved in 1 mL of acetonitrile and the
mixture added to a solution of indole 3 (414 mg, 2 mmol) in 15
mL of hydrochloric acid (35%), 20 mL of water, and 65 mL of
acetonitrile/methanol (3:1) at 40-45 °C. The mixture, which
turned blue immediately, was initially stirred for 1 h at 40-45
°C and then for 1 h at 5 °C. After filtration, washing with 30
mL of TBME, and drying, 343 mg of 4a (71%) was obtained as
intensely green light crystals (4): mp 214 °C dec; 1H NMR (CD3-
OD) δ 8.20-7.40 (m, 21H), 3.74 (s, 6H, NCH3); 13C NMR (CD3-
OD) δ 158.9, 141.4, 132.7, 131.9, 130.9, 127.7, 126.9, 122.8,
122.4, 120.4, 113.9, 33.1; FAB/MS (3-nitrobenzyl alcohol matrix)
m/z (relative intensity) 451 (M - Cl-, 60); UV (3:1 CH3CN/
MeOH) λmax (log ꢀmax, M-1 cm-1) 586 nm (5.04).
Isola tion of 586 n m Ch r om op h or e 4b u p on Rea ction of
MDA w ith In d ole 3. TMP (164 mg, 164 µL, 1 mmol) was
dissolved in 1 mL of acetonitrile and the mixture added to a
solution of indole 3 (414 mg, 2 mmol) in 15 mL of methane-
sulfonic acid, 20 mL of water, and 65 mL of acetonitrile/
methanol (3:1) at 40-45 °C. The mixture, which turned blue
immediately, was stirred for 1 h at 40-45 °C. The organic
solvents were removed under reduced pressure, and 150 mL of
CH2Cl2 with 50 mL of water was added to the residue. The
organic phase was decanted and washed three times with 50
mL of water. Evaporation of the solvents after drying (MgSO4)
and filtration afforded the crude product. After trituration with
20 mL of TBME and filtration, 360 mg of 4b (67%) was obtained
as green powder: mp 204 °C dec; 1H NMR (CD3OD) δ 8.30-
7.40 (m, 21H), 3.78 (s, 6H, NCH3), 2.67 (s, 3H, CH3SO3-); 13C
NMR (CD3OD) δ 158.7, 141.1, 132.8, 132.0, 131.0, 127.8, 127.0,
(230 mg, 1 mmol) was dissolved in 1 mL of acetonitrile and the
mixture added under N2 to a degassed solution of indole 3 (414
mg, 2 mmol) in 15 mL of methanesulfonic acid, 20 mL of water,
and 65 mL of acetonitrile/methanol (3:1) at 40-45 °C. After
being stirred for 10 min at 40-45 °C, the mixture was cooled to
5 °C. Concentrated NaOH (25 mL) was added, and then 50 mL
of water and 50 mL of TBME. The organic phase was decanted,
washed three times with 50 mL of saturated brine, and dried
(K2CO3). TLC analysis of this solution (silica, 5:1 cyclohexane/
ethyl acetate) revealed four main products with Rf values of 0.78
(3), 0.46 (5), 0.26 (6), and 0.20 (7). The spot corresponding to 7
turned blue immediately at ambient oxygen pressure. After
removal of the solvents, intermediates 5-7 were isolated by
MPLC (silica, 9:1 cyclohexane/ethyl acetate).
1
5: colorless oil; H NMR (CDCl3, mixture of two diasteroiso-
mers, 70:30 A:B) δ 8.20-7.75 (m, 2H), 7.60-7.10 (m, 16H), 5.58
(t, 1H, J ) 7 Hz, 1-HB), 4.93 (dd, 1H, J ) 11 Hz, J ) 4.5 Hz,
1-HA), 4.43 (msym, 1H, 4-HB), 4.14 (td, 1H, J ) 7 Hz, J ) 3 Hz,
4-HA), 3.95-3.65 (m, 1H), 3.59 (s, 3H, NCH3B), 3.58 (s, 3H,
NCH3A), 3.57 (s, 3H, NCH3A), 3.56 (s, 3H, NCH3B), 3.07 (q, 1HA,
J ) 11 Hz), 2.75 (ddd, 1HB, J ) 13 Hz, J ) 8.5 Hz, J ) 7 Hz),
2.51 (ddd, 1HB, J ) 13 Hz, J ) 7.5 Hz, J ) 5 Hz), 2.15 (msym
,
1HA), 1.90-0.75 (m, 11H); 13C NMR2 (CDCl3) δ 83.6, 82.5, 76.1,
73.9, 41.5, 41.2, 39.8, 38.6, 33.7, 32.1, 31.5, 31.0, 27.2, 26.9, 22.9,
14.3, 14.2; CI-HRMS (isobutane) m/z (relative intensity) 553.3218
(MH+, 35; C39H41ON2, calcd m/z 553.3219).
6: white solid; mp 192-194 °C dec; 1H NMR (CDCl3) δ 7.70-
6.46 (m, 27H), 4.54 (m, 1H, 1-H), 3.44-3.34 (m, 1H, 4-H), 3.38
(s, 3H, NCH3), 3.30 (s, 3H, NCH3), 3.27 (s, 3H, NCH3), 3.04-
2.60 (m, 3H), 1.24-0.72 (m, 11H); 13C NMR (CDCl3)2 δ 74.8,
41.3, 36.0, 35.1, 32.5, 31.8, 30.5, 30.3, 30.1, 25.7, 22.5, 13.9; EI-
HRMS (70 eV) m/z (relative intensity) 759.4191 (M+, 100;
C54H53N3O, calcd m/z 759.4189).
1
7: colorless oil; H NMR (CD3CN) δ 7.83 (m, 2H), 7.48-7.04
(m, 16H), 6.62 (dd, 1H, J ) 16 Hz, J ) 8.5 Hz, 2-H), 6.22 (d,
1H, J ) 16 Hz, 1-H), 4.10 (m, 1H, 4-H), 3.49 (s, 3H, NCH3),
3.48 (s, 3H, NCH3), 3.37 (t, 1H, J ) 8.5 Hz, 3-H), 2.31 (d, 1H, J
) 3.5 Hz, OH), 1.70-1.00 (m, 8H), 0.81 (t, 3H, J ) 6.5 Hz); 13
C
NMR (CD3CN) δ 141.3, 141.1, 139.7, 139.5, 134.0, 133.2, 133.0,
132.8, 130.3, 130.2, 130.1, 129.9, 128.2, 127.4, 125.6, 123.9,
123.4, 122.6, 122.0, 121.8, 120.9, 114.7, 113.0, 111.8, 111.6, 74.7,
51.5, 36.9, 33.4, 32.1, 32.0, 26.7, 24.0, 15.0; EI-HRMS (70 eV)
m/z (relative intensity) 552.3124 (M+, 5; C39H40N2O, calcd m/z
552.3141); EI-MS (70 eV) m/z (relative intensity) 552 (M+, 3),
451 (100), 425 (45), 320 (35).
Syn th esis of Tetr a h yd r ofu r a n 5 fr om 6. 4-Hydroxy-
nonenal diethyl acetal (153 mg, 0.66 mmol) and indole 3 (414
mg, 2 mmol) were dissolved in 100 mL of methanol. After the
addition of 10 µL of methanesulfonic acid, the solution was
stirred for 24 h at room temperature. A white precipitate
formed, and the solution turned slightly blue. The precipitate
was filtered and washed with 25 mL of TBME to afford 6 (292
mg, 58%) as a white solid. Two hundred fifty milligrams (0.33
mmol) of this crude product was dissolved in 25 mL of dichlo-
romethane. After addition of 3 µL of methanesulfonic acid, the
mixture was stirred for 5 min at room temperature. The
solution was consecutively washed with 25 mL of saturated
bicarbonate solution and brine (25 mL) and dried (Na2SO4).
After removal of the solvent, the residue was purified by MPLC
(silica, 2:1 to 1:1 cyclohexane/dichloromethane) to afford 5 (117
mg, 64%) as a colorless oil.
122.8, 122.4, 120.4, 114.0, 39.8, 33.2; FAB/HRMS (3-nitrobenzyl
-
alcohol matrix) m/z (relative intensity) 451.2156 (M - CH3SO3
,
100; C33H27N2, calcd m/z 451.2174); UV (3:1 CH3CN/MeOH) λmax
586 nm.
Isola tion of 586 n m Ch r om op h or e 4b u p on Rea ction of
4-HNE w ith In d ole 3. 4-Hydroxynonenal diethyl acetal (230
mg, 1 mmol) was dissolved in 1 mL of acetonitrile and the
mixture added to a solution of FeCl3 (324 mg, 2 mmol) and indole
3 (414 mg, 2 mmol) in 15 mL of methanesulfonic acid, 20 mL of
water, and 65 mL of acetonitrile/methanol (3:1) at 40-45 °C.
The mixture, which turned blue immediately, was stirred for 1
h at 40-45 °C. Following the extraction procedure which was
described above, 586 nm chromophore 4b (350 mg, 64%) was
obtained as a green powder. The same reaction in the presence
of catalytic quantities of FeCl3 (40 mg, 0.25 mmol) afforded 312
mg of 4b (57%).
For GC analysis, 10 mL of the reaction mixture was quenched
under cooling, with 25 mL of NaOH (2 N), and extracted with
4 mL of a solution of heptanal (25 mM) in TBME. The organic
phase was dried (Na2SO4) and filtered. One to two microliters
of this solution was injected and the peak of hexanal (tR ) 9.4
min) compared to the peak of heptanal (tR ) 15.8 min).
Isola tion of 505 n m Ch r om op h or e 10a u p on Rea ction
of Hexa n a l w ith In d ole 3. Hexanal (100 mg, 120 µL, 1 mmol)
was added to a solution of FeCl3 (40 mg, 0.25 mmol) and indole
3 (414 mg, 2 mmol) in 15 mL of methanesulfonic acid, 20 mL of
water, and 65 mL of acetonitrile/methanol (3:1) at 40-45 °C.
Following the extraction procedure described for 4b, 10a [42
mg after 1 h (8%) and 172 mg after 14 h (33%)] was isolated as
a red powder (5): 1H NMR (CDCl3) δ 7.82 (m, 2H), 7.79 (s, 1H),
7.67-7.27 (m, 16H), 4.05 (s, 6H, NCH3), 2.73 (s, 3H, CH3SO3).
2
Isola tion of In ter m ed ia tes 5-7 fr om th e Rea ction of
4-HNE w ith 3 u n d er N2. 4-Hydroxynonenal diethyl acetal
Due to the complexity of the aromatic region and the superposition
of most of the sp2 carbons, only the aliphatic region is described.

1186 Chem. Res. Toxicol., Vol. 11, No. 10, 1998
Erdelmeier et al.
The reaction of octanal (128 mg, 1 mmol) under the same
conditions also afforded 10a [42 mg after 1 h (8%) and 129 mg
after 14 h (25%)].
C
23H27N, calcd m/z 317.2144). The spectrum shows the presence
of an impurity: m/z (relative intensity) 333 (C23H27NO, 4).
1-(N-Meth yl-2′-p h en ylin d ol-3-yl)h ex-1-en e (15). The re-
action of hexanal (400 mg, 4 mmol) as described for 14 afforded
15 (380 mg, 33%) as a pale yellow oil:3 1H NMR (CDCl3) δ 7.98
(m, 1H), 7.56-7.16 (m, 8H), 6.38 (d, 1H, J ) 16 Hz, 1-H), 6.21
(td, 1H, J ) 6.5 Hz, J ) 16 Hz, 2-H), 3.61 (s, 3H, NCH3), 2.17
(q, J ) 6.5 Hz, 3-H), 1.50-1.24 (m, 4H), 0.91 (t, 3H, J ) 7 Hz);
13C NMR (CDCl3) δ 140.0, 138.4, 132.8, 132.0, 130.0, 129.4,
129.0, 126.4, 124.0, 123.0, 121.5, 121.0, 113.0, 110.4, 34.2, 32.8,
31.2, 22.5, 14.2; EI-HRMS (70 eV) m/z (relative intensity)
289.1840 (M+, 82; C21H23N, calcd m/z 289.1830). The spectrum
shows the presence of an impurity: m/z (relative intensity) 305
(C21H23NO, 4).
Syn th esis of 505 n m Ch r om op h or e 10b Sta r tin g fr om
3-F or m ylin d ole 11. 3-Formylindole 11 (3) (235 mg, 1 mmol)
was added to a solution of indole 3 (414 mg, 2 mmol) in 15 mL
of methanesulfonic acid, 20 mL of water, and 65 mL of
acetonitrile/methanol (3:1) at 40-45 °C. After 3 h, the intensely
red mixture was similarly extracted, affording 10a (315 mg) as
a red powder, which was dissolved in 5 mL of methanol. After
being heated to reflux and addition of 6 mL of concentrated
perchloric acid to eliminate traces of (1-methyl-2-phenylindol-
3-yl)methane, the red solution was slowly cooled to room
temperature. Red crystals precipitated, which were filtered and
washed with 4 × 5 mL of TBME to afford 10b (232 mg, 44%):
1
2-P en tylfu r a n . 4-Hydroxynonenal diethyl acetal (690 mg,
3 mmol) in 1 mL of acetonitrile was added to a solution of 7.5
mL of concentrated HCl, 10 mL of water, and 32.5 mL of
acetonitrile/methanol (3:1) at 40-45 °C. After 10 min at 40-
45 °C, the reaction mixture was neutralized with 40 mL of
NaOH (1 N). Extraction with 10 mL of TBME, followed by
drying (Na2SO4) and evaporation, afforded a yellow oil, which
was identified as 2-pentylfuran (380 mg, 87%, purity of ∼95%):
1H NMR (CD3CN) δ 7.35 (m, 1H, 5-H), 6.30 (m, 1H, 4-H), 6.03
(m, 1H, 3-H), 2.61 (t, 2H, J ) 7 Hz, 1′-H), 1.62 (quint, 2H, J )
7 Hz, 2′-H), 1.31 (m, 4H, 3′-H, 4′-H), 0.90 (t, 3H, J ) 7 Hz, 5′-
H); 13C NMR (CD3CN) δ 157.9, 142.2, 111.5, 105.9, 32.2, 28.6,
28.5, 23.2, 14.4; EI-MS (70 eV) m/z (relative intensity) 138.1041
(M+, 100; C9H14O, calcd m/z 138.1045).
mp 128-129 °C; H NMR (CDCl3) δ 7.82 (s, 1H), 7.76 (m, 2H),
7.68-7.30 (m, 16H), 4.04 (s, 6H); 13C NMR (CDCl3) δ 160.8,
152.2, 141.1, 132.4, 131.9, 129.6, 127.7, 126.8, 126.2, 125.2,
124.8, 120.3, 113.3, 34.0; FAB/MS (glycerol/3-nitrobenzyl alcohol
matrix) m/z (relative intensity) 425 (M - ClO4-, 100), 185 (40),
93 (75); UV (3:1 CH3CN/MeOH) λmax 505 nm. 10 was described
as bromide (5).
1,1-Bis(N-m eth yl-2′-p h en ylin d ol-3-yl)octa n e (12). Octa-
nal (128 mg, 1 mmol) was added to a solution of indole 3 (1.03
g, 5 mmol) in 20 mL of acetonitrile. After the addition of 3 µL
of methanesulfonic acid, the mixture was stirred for 48 h at room
temperature. After neutralization with 5 mL of saturated
bicarbonate solution and evaporation, the residue was dissolved
in 50 mL of TBME and washed with 50 mL of water and
saturated brine. Drying, filtration, and evaporation afforded
978 mg of a yellow oil. Purification by MPLC (silica, 4:1 to 1:1
cyclohexane/dichloromethane) yielded 12 (380 mg, 72%) as a
highly viscous colorless oil: 1H NMR (CDCl3) δ 7.57 (d, 2H, J )
8 Hz), 7.38-7.10 (m, 10H), 6.94 (m, 6H), 4.40 (t, 1H, J ) 7 Hz,
1-H), 3.38 (s, 6H, NCH3), 2.18 (q, 2H, J ) 7 Hz, 2-H), 1.26-
0.94 (m, 10H), 0.80 (t, 3H, J ) 6.5 Hz, 8-H); 13C NMR (CDCl3)
δ 137.9, 136.9, 132.9, 131.0, 127.9, 127.8, 127.4, 121.2, 120.8,
118.8, 116.5, 108.9, 35.5, 34.7, 31.6, 30.3, 29.0, 28.8, 27.9, 22.3,
13.8; EI-MS (70 eV) m/z (relative intensity) 524 (M+, 20), 425
(100), 317 (17), 246 (17), 207 (18).
GC An a lysis. 4-Hydroxynonenal diethyl acetal (115 mg, 0.5
mmol) in 0.5 mL of acetonitrile was added to a solution of
concentrated acid (7.5 mL), 10 mL of water, and 32.5 mL of
acetonitrile/methanol (3:1) at 40-45 °C. Ten milliliters of the
reaction mixture was quenched while being cooled, with 25 mL
of NaOH (2 N). Four milliliters of a solution of 2-ethylfuran
(25 mM) in TBME was added, and the organic phase was
decanted, dried, and filtered. One to two microliters of this
solution was injected and the peak of 2-pentylfuran (tR ) 20.7
min) compared with the peak of 2-ethylfuran (tR ) 5.65 min).
1,1-Bis(N-m eth yl-2′-p h en ylin d ol-3-yl)h exa n e (13). Hexa-
nal (150 mg, 1.5 mmol) was added to indole 3 (621 mg, 3 mmol)
in 15 mL of acetonitrile. After addition of 4 µL of methane-
sulfonic acid, the mixture was stirred for 4 h. The white
precipitate was filtered and washed with 3 × 5 mL of cold
acetonitrile to afford 13 (460 mg, 61%): mp 145 °C; 1H NMR
(CDCl3) δ 7.60 (d, 2H, J ) 8 Hz), 7.40-7.10 (m, 10H), 7.06-
6.84 (m, 6H), 4.43 (t, 1H, J ) 8 Hz, H-1), 3.39 (s, 6H, NCH3),
2.21 (m, 2H, 2-H), 1.26-0.94 (m, 6H), 0.72 (t, 3H, J ) 7 Hz,
H-6); 13C NMR (CDCl3) δ 138.0, 136.9, 132.9, 131.1, 127.9, 127.8,
127.4, 121.2, 120.8, 118.8, 116.5, 108.9, 35.5, 34.9, 31.4, 30.3,
27.7, 22.2, 13.9; EI-MS (70 eV) m/z (relative intensity) 496 (M+,
15), 425 (100), 218 (30).
1-(N-Meth yl-2′-p h en ylin d ol-3-yl)oct-1-en e (14). Octanal
(512 mg, 4 mmol) was added under N2 to a solution of indole 3
(828 mg, 4 mmol) in 20 mL of TBME. After the addition of 3
µL of methanesulfonic acid, the mixture was stirred for 24 h at
room temperature under N2 and then washed with 2 × 20 mL
of a saturated bicarbonate solution. The solvent was removed
under reduced pressure after drying (Na2SO4) and filtration.
Purification of the crude product by MPLC (silica, 8:1 cyclohex-
ane/dichloromethane) afforded 14 (141 mg, 32%) as a pale yellow
oil:3 1H NMR (CDCl3) δ 8.01 (m, 1H), 7.60-7.20 (m, 8H), 6.42
(d, 1H, J ) 16 Hz, 1-H), 6.25 (td, 1H, J ) 7 Hz, J ) 16 Hz,
2-H), 3.61 (s, 3H, NCH3), 2.20 (q, J ) 7 Hz, 3-H), 1.56-1.20 (m,
8H), 0.92 (t, 3H, J ) 6.5 Hz); 13C NMR (CDCl3) δ 138.9, 137.8,
131.8, 131.1, 128.9, 128.4, 128.2, 125.9, 122.8, 122.1, 120.5,
120.2, 112.0, 109.5, 33.8, 31.6, 30.7, 29.8, 28.7, 22.5, 13.9; EI-
HRMS (70 eV) m/z (relative intensity) 317.2162 (M+, 44;
Resu lts
Acid -In d u ced Rea ction of MDA (1) w ith N-Meth -
yl-2-p h en ylin d ole (3). The acid-catalyzed reaction of
MDA with indole compounds is known to generate
intensely colored trimethincyanine dyes (3, 6). Hence,
we isolated as expected 586 nm chromophores 4a (70%)
and 4b (67%) upon reaction of 2 equiv of indole 3 with 1
equiv of MDA, generated in situ from tetramethoxypro-
pane (TMP), under the described assay conditions, e.g.,
in an aqueous solution of acetonitrile/methanol, in the
presence of either hydrochloric or methanesulfonic acid
at 40-45 °C for 1 h4 (eq 1).
3
14 and 15 decomposed slowly in solution and have to be stored
under nitrogen.

Colorimetric Assay of Lipid Peroxidation. 2
Chem. Res. Toxicol., Vol. 11, No. 10, 1998 1187
Id en tifica tion of th e F in a l P r od u cts of th e Acid -
In d u ced Rea ction of 4-HNE (2) w ith In d ole 3.
4-Hydroxynonenal reacts with an excess of indole 3 to
yield a concentration-dependent absorbance at 586 nm
when methanesulfonic acid is used in the presence of
catalytic amounts of iron(III) salts. Surprisingly, when
HCl was substituted for methanesulfonic acid, no sig-
nificant production of chromophore was observed (see
Figure 1B).
A closer investigation of the methanesulfonic acid-
induced reaction showed that no chromophore was pro-
duced in the absence of contaminating or added iron(III)
salts (Figure 1C). Under anaerobic conditions, complete
transformation of 4-HNE to the chromophore was ob-
tained upon addition of 2 equiv of iron(III) chloride to
the reaction mixture (Figure 1D).
Similarly, a crystalline dye could be isolated upon
stoichiometric reaction of 2 equiv of indole 3 with
4-hydroxynonenal, generated in situ from the diethyl
acetal instead of MDA. However, this was only possible
when methanesulfonic acid was used, in the presence of
iron(III) salts (64% with 2 equiv of FeCl3 and 57% with
0.25 equiv of FeCl3).
F igu r e 1. Effects of the acid type, iron salt, and oxygen on the
formation of the 586 nm chromophore from the reaction of
4-HNE (20 µM) with indole 3 (5 mM) for 30 min at 45 °C in
acetonitrile/methanol. Ten microliters of a solution of 4-HNE
in acetonitrile/methanol (3:1, 8 mM) was added to 4 mL of a
solution of indole 3 in acetonitrile/methanol/water/H+X-. The
586 nm absorbance was measured after 30 min at 45 °C: (A)
methanesulfonic acid, FeCl3 (0.2 equiv), and ambient oxygen
pressure, (B) hydrochloric acid, FeCl3 (0.2 equiv), and ambient
oxygen pressure, (C) methanesulfonic acid and ambient oxygen
pressure without FeCl3, and (D) methanesulfonic acid, FeCl3 (2
equiv), and an inert atmosphere.
essentially three products. Separation, isolation, and
characterization of these products allowed the attribution
of the structures 5-7 for these intermediates, confirming
that all three products were still bearing the alkyl side
chain of starting aldehyde 2.
The analytical characterization (1H NMR, 13C NMR,
MS, mp, and UV/vis) of the isolated dye established
complete structural identity with cyanine dye 4b, the
reaction product for reaction of indole 3 with MDA. The
1H and 13C NMR spectra especially showed the sole
presence of sp2 carbons in the molecule (besides the two
methyl groups at 39.8 and 33.2 ppm), implying a frag-
mentation of the alkyl chain of starting aldehyde 2 during
the reaction.
A GC analysis of TBME extracts of neutralized samples
from the highly colored reaction mixture of eq 2, taken
after 10-60 min, allowed us to identify hexanal as a
second product besides 586 nm chromophore 4, formed
by the stoichiometric reaction of 4-HNE with indole 3 in
about 45-60% yield (eq 2).
1
HRMS and H NMR data were crucial for the assign-
ment of structures 5-7 to the three reaction products.
The isotopic patterns of the M•+ for 6 and 7 (after HR-
EI) and MH+ for 5 (following HR-DCI) were consistent
with the molecular formulas C39H41N2O, C54H53N3O, and
C39H40N2O for 5-7, respectively (see Materials and
Methods). This corresponds to the addition of two
molecules of indole 3 to one molecule of 4-HNE for 5 and
7, versus three molecules of indole for 6.
1
The following key features in the H NMR spectrum
of 7 allowed the assignment of the olefinic structure: (i)
two olefinic protons at 6.22 (d, 1H, J ) 16 Hz) and 6.62
ppm (dd, 1H, J ) 16 and 8.5 Hz) assigned to H-1 and
H-2, (ii) a multiplet corresponding to one proton at 4.10
ppm, assigned to H-4, and (iii) finally two singlets (three
equivalent hydrogens integrated) at 3.48 and 3.49 ppm,
corresponding to the N-methyl groups.
Id en tifica tion of th e In ter m ed ia tes of th e Acid -
In d u ced Rea ction of 4-HNE w ith In d ole 3. Trying
to gain more insight into the course of this reaction, we
repeated the methanesulfonic acid-induced reaction of
4-HNE with indole 3, but in the absence of both oxygen
and iron(III) salts. Under these conditions, the homo-
geneous solution remained colorless upon addition of
either stoichiometric (2 equiv) or excess quantities (10
equiv) of indole 3. TLC examination of neutralized and
extracted samples of this colorless mixture (with either
2 or 10 equiv of indole 3) revealed the formation of
The identification of compound 5 as a tetrahydrofuran-
type addition product was based on the following. Its
molecular formula is the same as that of 7, but no olefinic
1
proton shows up in the H NMR spectrum. Two sets of
diastereoisomers were observed with a ratio of 70:30 (5A:
5B). Peaks observed at 5.58 ppm (0.3H, 5B) and 4.93
ppm (0.7H, 5A) were assigned to H-1 and those observed
at 4.14 ppm (0.7H, 5A) and 4.43 ppm (0.3H, 5B) to H-4.5
4
Despite the rapid and irreversible polymerization of MDA under
acidic conditions, a complete conversion of MDA to 586 nm chro-
mophore 4 is possible using a large excess of indole 3; see ref 23 for
further details.
1
The H NMR spectrum of 6 displayed three singlets
(three equivalent hydrogens integrated) for the three

1188 Chem. Res. Toxicol., Vol. 11, No. 10, 1998
Erdelmeier et al.
Sch em e 1
distinct N-methyl groups, whereas one proton (t, J ) 7
Hz) at 4.54 ppm was assigned to H-1. This chemical shift
is typical of H-1 in 1,1-diindolyl-substituted alkyl deriva-
tives. For example, it is observed in 12 at 4.40 ppm (t, J
) 7 Hz) and in 13 at 4.43 ppm (t, J ) 7 Hz).
Separate dissolution of each purified intermediate in
the acidic solvent mixture of the described (23) assay
results in the immediate formation of the three inter-
mediates 5-7 and the starting indole 3, as shown by TLC
analysis. This rapid equilibration of all intermediates
excluded the identification of the direct precursor for the
rapid and quantitative fragmentation to 586 nm chro-
mophore 4 and hexanal under the assay conditions in
the presence of iron(III) salts.
Whereas compounds 6 and 7 are the expected addition
products for addition of indole 3 to 4-HNE, an R,â-
unsaturated aldehyde, the tetrahydrofuran derivative 5,
could be formed via various pathways (see Scheme 1),
i.e., either by direct intramolecular addition to the
indolenium cation fragment in 9 (9 f 5H+) or by an
addition-elimination sequence via intermediate 8 (9 f
8 f 5H+). Actually, 9 is a direct intermediate of the acid-
induced addition of indole 3 to 4-HNE, but could also be
obtained by protonation of one indole nucleus in 6,
followed by elimination of indole 3 (see Scheme 1).
Interestingly, tetrahydrofuran derivative 5 is obtained
as the major product (along with indole 3) after addition
of catalytic amounts of an acid to a CH2Cl2 solution of
intermediate 6, and could conveniently be isolated in 64%
yield after purification (eq 3).
Rea ction of In ter m ed ia tes 5-7 w ith F eCl3. To
compare the reactivities of the three intermediates, we
studied their chromogenic transformation in the presence
of FeCl3 (0.5 equiv) in acetonitrile/methanol.
Air-saturated solutions (0.1 mM) of intermediate 5, 6,
or 7 in acetonitrile/methanol (3:1) were treated at 40-
45 °C with 0.5 equiv of FeCl3. The reactions were
followed by monitoring the absorbance at 586 nm. As
shown in Figure 2, the 586 nm chromophore was formed
in all reaction mixtures, even though it was formed much
more slowly than under the standardized assay condi-
tions. Moreover, significant differences in the rates of
chromophore formation could be observed in the first
hours, with the solution of 1-indolylalkene 7 turning blue
far more quickly than the solution of the tetrahydrofuran
derivative 5 and that of intermediate 6 (Figure 2).
Interestingly, no significant absorbance at 505 nm
could be detected in any of these reaction mixtures,
whereas the transformation of 4-HNE to the 586 nm
chromophore and hexanal is accompanied by the forma-
tion of a second chromophore at 505 nm.
Rea ctivity of 4-HNE w ith In d ole 3 in Hyd r och lo-
r ic Acid Med iu m . As shown above (see Figure 1), the
reaction of 4-HNE with indole 3 to form 586 nm chro-
mophore 4 and the corresponding saturated aldehyde
depends not only on the presence of iron salts but also
on the nature of the acid used; while the reaction is fast
and quantitative in the presence of methanesulfonic acid
(1 h at 45 °C), no significant production of the 586 nm
5
The formation of diastereomeric tetrahydofuran derivatives fol-
lowing Michael addition of nucleophiles such as n-butylamine and
N-acetylhistamine to 4-HNE was previously reported (7). The corre-
sponding chemical shifts for H-1 of 4.69-4.74 ppm for the former and
5.64-5.80 ppm for the latter are in good agreement with the figures
which were obtained for 5.

Colorimetric Assay of Lipid Peroxidation. 2
Chem. Res. Toxicol., Vol. 11, No. 10, 1998 1189
Ta ble 1. Com p a r ison of Hyd r och lor ic a n d
Meth a n esu lfon ic Acid -In d u ced F or m a tion of
2-P en tylfu r a n fr om 4-Hyd r oxyn on en a l (Gen er a ted in
Situ fr om th e Dieth yl Aceta l) in th e P r esen ce or Absen ce
of In d ole 3
acid
t
equiv of 3 % 2-pentylfurana
HCl
30 s
60 s
2 min
0
0
0
0
0
0
0
2
2
61 ( 7
87 ( 16
114 ( 2
9 ( 2
methanesulfonic acid 30 s
60 s
13 ( 2
20 ( 3
57 ( 9
3 ( 1
2 min
10 min
2 min
HCl
10 min
2 ( 2
F igu r e 2. Formation of the 586 nm chromophore upon reaction
of 5, 6, or 7 (0.1 mM) with FeCl3 (50 µM) in acetonitrile/methanol
(3:1) at 40-45 °C. Each aliquot of the reaction mixture was
diluted with 4 volumes of acetonitrile/methanol (3:1) before
measuring the 586 nm absorbance.
a
GC analysis. percentage of the peak area of 2-ethylfuran as
the internal standard, means of two experiments (see Materials
and Methods).
chromophore is observed when hydrochloric acid is used
instead of methanesulfonic acid.
This result was unexpected given the complete reaction
of MDA with indole 3 yielding 586 nm chromophore 4,
and therefore the stability of the chromophore under such
conditions.
To assess the stability of the starting aldehyde under
the assay conditions,6 a solution of 0.1 mmol of 4-hydroxy-
nonenal diethyl acetal in CD3CN/CD3OD/D2O (850 µL)
was treated with deuterated HCl (150 µL). The NMR
spectrum obtained after 5 min showed the rapid conver-
sion of 4-HNE (obtained upon hydrolysis of the acetal)
to a new product, which was identified as 2-pentylfuran
(eq 4).
F igu r e 3. Effects of the acid type, iron salt, and oxygen on the
formation of the 505 nm chromophore from the reaction of
hexanal (A-D, 10 µM) or octanal (E, 10 µM) with indole 3 (5
mM) for 30 min at 45 °C. Four microliters of a solution of
hexanal or octanal in acetonitrile/methanol (3:1, 10 mM) was
added to 4 mL of a solution of indole 3 in acetonitrile/methanol/
water/H+X-. The 505 nm absorbance was measured after 30
min at 45 °C: (A) methanesulfonic acid, FeCl3 (5 µM), and
ambient oxygen pressure, (B) hydrochloric acid, FeCl3 (5 µM),
and ambient oxygen pressure, (C) methanesulfonic acid, FeCl3
(20 µM), and an inert atmosphere, (D) methanesulfonic acid and
ambient oxygen pressure without FeCl3, and (E) methane-
sulfonic acid, FeCl3 (5 µM), and ambient oxygen pressure.
The formation of 2-pentylfuran was observed earlier by
Sayre et al. (10) upon refluxing ethanolic solutions of
4-HNE in the presence of phenethylamine chlorohydrate.
Interestingly, 2-pentylfuran was isolated by Chang et al.
(9) as a component of autoxidized soybean oil.
Interestingly, we observed that the transformation of
alkanals into the 505 nm chromophore upon reaction
with indole 3 was similarly affected by hydrochloric
versus methanesulfonic acid, as described in the following
section.
A rough estimation of the formation rate of this product
under the assay conditions in the absence of indole 3 was
obtained by means of gas chromatography, using 2-eth-
ylfuran as the internal standard. The results indicated
a faster transformation in the presence of HCl than in
the presence of methanesulfonic acid (see Table 1).
To test the hypothesis that rapid formation of pentyl-
furan in the presence of HCl competes efficiently with
chromophore formation, we repeated such measurements
in the presence of indole 3. Even in the presence of only
2 equiv of indole 3, essentially no pentylfuran was formed
(see Table 1). TLC examination of neutralized and
extracted samples indicated the same intermediates 5-7
as for the methanesulfonic acid-induced reaction.
Obviously, the difference between the two assay condi-
tions, using either HCl or methanesulfonic acid, concerns
the transformation of the intermediates to 586 nm
chromophore 4, which is practically inhibited in the
presence of HCl, allowing the selective measurement of
MDA under these conditions.
Id en tifica tion of th e F in a l P r od u cts of th e Acid -
In d u ced Rea ction of Alk a n a ls w ith In d ole 3. As
mentioned above, the acid-induced reaction of 4-HNE
with indole 3 results in the formation of 586 nm chro-
mophore 4, accompanied by the formation of a second
chromophore with a maximal absorbance wavelength of
505 nm. This second absorption was not a linear function
of the concentration of 4-HNE, and the maximal absor-
bance was somewhat variable.
Since hexanal was one byproduct of the acid-induced
reaction of 4-HNE with 2 equiv of indole 3 in the presence
of iron salts (and also a nonspecific byproduct of lipid
peroxidation), we compared the reaction of saturated
aldehydes with 3 under the assay conditions.
Interestingly, after the addition of hexanal or octanal
(10 µM) to a large excess (5 mM) of indole 3 in aqueous
acetonitrile/methanol containing methanesulfonic acid
and catalytic amounts of FeCl3 (5 µM), the reaction
mixture turned red immediately. This is a result of the
formation of a 505 nm chromophore (see Figure 3A,E),
but apparently in a nonlinear manner. Under the same
conditions, but using HCl instead of methanesulfonic
6 Hydroxynonenal was reported to dehydrate under acidic conditions
to trans,trans-2,4-nonadienal (8).

1190 Chem. Res. Toxicol., Vol. 11, No. 10, 1998
Erdelmeier et al.
acid, no significant chromophore formation took place
(Figure 3B). The same result was obtained either under
anaerobic conditions or in the absence of iron(III) salts
(Figure 3D). In the first case, even in the presence of 2
equiv of FeCl3, no significant absorbance at 505 nm could
be detected (Figure 3C), underlining the requirement for
both iron(III) and oxygen in the formation of the 505 nm
chromophore.
The corresponding diindolyloctane 12 and -hexane 13
as well as the indolyloctene 14 and -hexene 15, which
we synthesized and characterized independently, were
also identified by TLC analysis as major products of eq
5 in the absence of iron(III) salts.
To isolate and characterize the corresponding dye
compound, 1 equiv of hexanal was added to a solution of
2 equiv of indole 3 in methanesulfonic acid in the
presence of 0.25 equiv FeCl3. After 1 h at 45 °C, the
intensely red reaction mixture yields a red precipitate.
The 1H NMR spectroscopic analysis of this compound
(with a λmax of 505 nm in acetonitrile/methanol) indicates
the diindolylmethine structure 10a (8% at 1 h and 33%
at 14 h; eq 5).
As already mentioned for addition products 5-7,
products 12 and 14 just like products 13 and 15 are in
rapid equilibrium under acidic conditions, as shown by
separate dissolution of each product in the acidic reaction
medium and TLC analysis of the resulting mixture.
Accordingly, we studied the reaction of 1-indolylalkenes
14 and 15 in the presence of FeCl3 in acetonitrile/
methanol.
FeCl3 (20 mol %) was added to a 1 mM solution of
1-indolyloctene 14 in acetonitrile/methanol at 40-45 °C
under aerobic conditions. The TLC analysis of the
reaction mixture at 10 min revealed the transformation
of 14 to a new, more polar, UV-active product, the
reaction going to completion within 90 min. This product
was shown by 1H and 13C NMR analysis and comparison
with an authentic sample to be 3-formylindole 11,
isolated in 72-78% yield after evaporation of the solvents
and filtration of the residue over alumina. Using hexanal
dimethylacetal7 as the internal standard after completion
of the transformation, GC analysis showed the presence
of heptanal (as dimethylacetal) as the second major
product besides 3-formylindole 11, formed in 80-100%
yield (eq 7).
In a similar way, 10a (8% at 1 h and 25% at 14 h) was
isolated after treating octanal under the same conditions.
This structure was further confirmed by the independent
synthesis of 10b, upon reaction of 3-formylindole 11 with
indole 3 under acidic conditions and subsequent precipi-
tation of 10b as a perchlorate salt (eq 6).
The fact that the same 505 nm-absorbing structure is
obtained by the reaction of either hexanal, octanal, or
3-formylindole 11 indicates a one-carbon fragmentation
of the starting alkanal during the methanesulfonic acid-
induced reaction with indole 3 in the presence of iron-
(III) salts and oxygen.
In a similar way, the solution of 1-indolylhexene 15 in
acetonitrile/methanol reacted with FeCl3 to give 3-formyl-
indole 11 (78%) and pentanal (as dimethylacetal, 83%).
Again, 3-formylindole 11 is the direct precursor of 505
nm chromophore 10 via reaction with indole 3 under
acidic conditions.
R ea ct ion of 1-In d olyla lk en es 14 a n d 15 w it h
F eCl3. The acid-induced reaction of indole compounds
with saturated aldehydes is known to yield essentially
two products, 1,1-diindolylalkanes and 1-indolylalkenes
(11, 12).
Rea ctivity of Alk a n a ls w ith In d ole 3 in Hyd r o-
ch lor ic Acid Med iu m . As shown in Figure 3, 505 nm
7
Formed in situ by adding 3 mg of FeCl3 to 100 mL of a solution of
hexanal (1 mM) in acetonitrile/methanol (3:1).

Colorimetric Assay of Lipid Peroxidation. 2
Chem. Res. Toxicol., Vol. 11, No. 10, 1998 1191
Sch em e 2. P ossible Mech a n ism for th e F or m a tion
of 586 n m Ch r om op h or e 4 via â-F r a gm en ta tion of
a n In ter m ed ia te Alk oxyl Ra d ica l 16
Sch em e 3. Hyp oth etica l P a th w a ys for th e
F or m a tion of Alk oxyl Ra d ica l 16
chromophore 10 is not obtained by the reaction of hexanal
or octanal with indole 3 in the presence of HCl and iron
salts, whereas such a transformation is fast in the
presence of methanesulfonic acid.
Discu ssion
Nevertheless, in the presence of iron salts, the reaction
of hexanal or octanal with indole 3 yields essentially the
same products (12 and 14, and 13 and 15) using HCl or
methanesulfonic acid, as shown by TLC analysis of the
reaction mixture. Moreover, the formation of 505 nm
chromophore 10 via the reaction of the 3-formylindole 11
with indole 3 takes place in the presence of either HCl
or methanesulfonic acid.
Like the transformation of intermediates 5-7 into 586
nm chromophore 4, hydrochloric acid apparently inhibits
or slows the transformation of the 1-indolylalkenes 14
and 15 to 3-formylindole 11 as a precursor of the
chromophore 10. Both transformations (5-7 f 4 and
14 and 15 f 10) are induced by iron(III) salts and lead
to the fragmentation of the alkyl side chain.
An interesting feature of the colorimetric assay of MDA
and 4-HNE as byproducts of lipid peroxidation is the
formation of the same chromophore 4 at 586 nm. Start-
ing from MDA, this is a well-known acid-induced con-
densation reaction with indoles (6, 11, 12). Starting from
4-HNE, the formation of the same chromophore 4 obvi-
ously requires an oxidative fragmentation of the alkyl
side chain, as shown by the concomitant formation of
hexanal in the stoichiometric reaction (see eq 2). From
a mechanistic point of view, cleavage of the side chain
could be explained by a classical â-fragmentation (13), if
an alkoxyl radical 16 was formed by one-electron oxida-
tion of the rapidly interconverting adducts 5-7. This
hypothetical alkoxyl radical 16 would then afford hexanal

1192 Chem. Res. Toxicol., Vol. 11, No. 10, 1998
Erdelmeier et al.
Sch em e 4
Sch em e 5. P r op osed Mech a n ism for th e Ir on (III)-Med ia ted F or m a tion of Ch r om op h or e 10 (505 n m ) via
Oxid a tive Clea va ge of 14 or 15 to 11
and an allyl radical 17, which should be easily oxidized
to the final chromophore 4 (see Scheme 2). Identification
of the actual oxidant under the assay conditions would
require further investigation. Keeping in mind that
catalytic amounts of Fe(III) in the presence of oxygen as
well as 2 equiv of FeCl3 under anaerobic conditions are
sufficient for the reaction to be complete (see Figure 1),
we could envisage various mechanisms. One possibility
would be the addition of O2 to carbon-centered radical
17 followed by the extrusion of superoxide, most likely
intermediate 5 and from 1-indolylalkene 7, respectively
(see Scheme 3). While the oxidation of 5 via â-fragmen-
tation of the mesomeric form 18B would lead to the
alkoxyl radical 19, which represents the protonated form
of alkoxyl radical 16 (pathway A), an indirect pathway
can be envisaged for the transformation of 1-indolylal-
kene 7 to alkoxyl radical 16 (see Scheme 3, pathway B).
Despite the apparent greater complexity of the second
pathway, it is in better agreement with the experimental
facts.
Not only the greater reactivity of 1-indolylalkene 7 with
FeCl3 under model conditions but also the supposed lower
oxidation potential of the 1-indolylalkene fragment in 7
compared to that of the nonconjugated system in 5 favors
1-indolylalkene 7 as the precursor of chromophore 4 via
alkoxyl radical 16 (see Scheme 3).
•
in the form of HO2 which would easily recycle iron(III)
from iron(II).
The mechanism responsible for the formation of pos-
tulated alkoxyl radical 16 is less obvious. As described
above, intermediates 5-7 are all possible precursors of
chromophore 4, with 1-indolylalkene 7 reacting by far
more quickly in the presence of iron(III) chloride.
A one-electron oxidation (14) by iron(III) (15) would
afford radical cations 18 and 19 from the tetrahydrofuran
The formation of a second chromophore at 505 nm is
observed in the iron(III)-catalyzed reaction of indole 3
with 4-HNE, but also with alkanals. We therefore

Colorimetric Assay of Lipid Peroxidation. 2
Chem. Res. Toxicol., Vol. 11, No. 10, 1998 1193
assumed that hexanal produced upon â-fragmentation
reacted immediately under the assay conditions with
excess indole 3 to yield chromophore 10 and pentanal and
so on (see Scheme 4).
oxidation and subsequent transformations to chro-
mophores 4 and 10. The difference in the coordination
sphere of the iron(III) complex would then account for
the selective assay of MDA in the presence of 4-HNE in
hydrochloric acid medium.
In conclusion, this study outlines the chemical back-
ground of a new colorimetric assay of MDA and 4-HNE
as an example for 4-hydroxyalkenals. A detailed inves-
tigation of the reaction of these aldehydes with indole 3
showed that the oxidation of 1-indolylalkene intermedi-
ates by iron(III) to the corresponding radical cations is a
prerequisite of the formation of chromophores 4 and 10.
When the implications of both oxygen and iron(III) are
considered in the transformation of alkanals to 505 nm
chromophore 10 (see Figure 3), we suggest that 1-indolyl-
alkene 15 which is obtained from the reaction of hexanal
with indole 3 is oxidized by iron(III) to radical cation 20A
(Scheme 5). The mesomeric form 20B would be trapped
by oxygen, yielding a peroxyl radical intermediate 21
which could then yield indolyl dioxetane 22 after reduc-
tion and nucleophilic intramolecular addition. Dioxe-
tanes are known to yield carbonyl compounds by
electrocyclic ring opening (16), e.g., in this case the
3-formylindole 11 and heptanal starting from 14 and 11
and pentanal from 15. As shown before, 3-formylindole
11 reacts immediately with indole 3 to give 505 nm
chromophore 10 (see Scheme 5). A similar indolyldiox-
etane is proposed (17) to explain the fragmentation of
1-methyl-3-styrylindole to 3-formylindole and benzalde-
hyde upon reaction with singlet oxygen.
Besides the role of this reaction in the colorimetric
assay of byproducts of lipid peroxidation, the global
reaction represents a one-carbon fragmentation of a
saturated aldehyde under relatively mild conditions. This
rather rare reaction type was observed earlier by Read
et al. (18, 19) during the reaction of alkanals with
2-thiobarbituric acid (TBA) and their oxo analogues in
the presence of iron(III). Their proposed mechanism (18,
19) included the oxidation of an allylanion by iron(III),
followed by reaction of oxygen with the allyl radical and
anionic fragmentation of the hydroperoxide intermediate
to give the chromophore and an aldehyde which is one
methylene group shorter than the starting aldehyde.
Even if our hypothetical pathway, including a dioxe-
tane intermediate as proposed in Scheme 4, is not fully
demonstrated, the anionic pathway proposed by Read et
al. for the TBA reaction does not seem to be compatible
with the strong acidic conditions of our indole-based
assay.
A striking aspect of the colorimetric assay for 4-HNE
is the influence of the acid used, with HCl suppressing
the formation of both 586 nm chromophore 4 and 505
nm chromophore 10, in contrast to the complete trans-
formation observed in the presence of methanesulfonic
acid. Our experiments have shown that the same
intermediates 5-7 are produced from the reaction of
4-HNE with indole 3, in the presence of either methane-
sulfonic or hydrochloric acid. Similarly, the reaction of
hexanal and octanal with 3 afforded 1-indolylalkenes 14
and 15, respectively, under both conditions as precursors
for the formation of the 505 nm chromophore 10.
Obviously, the influence of the acid has some bearing
on the transformation of intermediates to final chro-
mophores 4 and 10. As proposed in Schemes 3 and 5,
the formation of 586 nm chromophore 4 from 5 or 7 and
that of 505 nm chromophore 10 from 14 or 15 require an
oxidation step. A difference in the coordination sphere
of the iron(III) complex in the reaction mixture could
possibly account for the experimentally observed differ-
ences. While the weakly coordinating (20, 21) and hence
more flexible methanesulfonato ligands would enable the
coordination of indole intermediates to the iron(III), the
chloride counterions, being more nucleophilic (22), could
inhibit or delay that coordination and therefore the
Refer en ces
(1) Esterbauer, H., Schaur, R. J ., and Zollner, H. (1991) Chemistry
and biochemistry of 4-hydroxynonenal, malonaldehyde and re-
lated aldehydes. Free Radical Biol. Med. 11, 81-128.
(2) Esterbauer, H., and Weger, W. (1967) Ueber die Wirkungen von
Aldehyden auf gesunde und maligne Zellen, 3. Mitt.: Synthese
von homologen 4-Hydroxy-2-alkenalen, II. Monatsh. Chem. 98,
1994-2000.
(3) Poot, A. L., Lemahieu, R. G., J anssens, W., and De J aeger, A. A.
(1973) Herstellung und Verwendung druckempfindlicher Aufze-
ichnungs- und Kopiermaterialien, Offenlegungsschrift 2362956.
(4) Galiullina, T. N., and Abramenko, P. I. (1983) Synthesis and
properties of substituted indolo-3,3′-trimethinecyanines. Khim.
Geterotsikl. Soedin. 12, 1652-1656.
(5) Galiullina, T. N., Abramenko, P. I., Kazymov, A. V., and Gor-
okhovski, V. M. (1981) Synthesis and properties of substituted
indolo-3-cyanine dyes. Khim. Geterotsikl. Soedin. 8, 1067-1071.
(6) Scherz, H., Stehlik, G., Bancher, E., and Kaindl, K. (1967)
2-Methylindole als Reagens auf Malondialdehyd. Mikrochim. Acta
5, 915-919.
(7) Nadkarni, D. V., and Sayre, L. M. (1995) Structural definition of
early lysine and histidine adduction chemistry of 4-hydroxyno-
nenal. Chem. Res. Toxicol. 8, 284-291.
(8) Witz, G., Lawrie, N. J ., Zaccaria, A., Ferran, H. E., J r., and
Goldstein, B. D. (1986) The reaction of 2-thiobarbituric acid with
biologically active R,â-unsaturated aldehydes. Free Radical Biol.
Med. 2, 33-39.
(9) Chang, S. S., Smouse, T. H., Krishnamurthy, R. G., Mookherjee,
B. D., and Reddy, R. B. (1966) Isolation and identification of
2-pentyl-furan as contributing to the reversion flavour of soybean
oil. Chem. Ind. (London), 1926-1927.
(10) 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.
(11) Remers, W. A. (1972) Properties and reactions of indoles. In The
Chemistry of Heterocyclic Compounds: Indoles. Part One (Houli-
han, W. J ., Ed.) pp 105-111, Wiley-Interscience, New York.
(12) Sundberg, R. J . (1970) Electrophilic substitution reactions on the
indole ring. Reactions with ketones and aldehydes. In The
Chemistry of Indoles (Blomquist, A. T., Ed.) pp 39-56, Academic
Press, New York.
(13) Curran, D. P. (1991) Radical cyclizations and sequential radical
reactions. In Comprehensive Organic Synthesis (Trost, B. M.,
Fleming, I., and Semmelhack, M. F., Eds.) Vol. 4, pp 779-832,
Pergamon Press, Oxford, U.K.
(14) Balogh-Hergovich, E., and Speier, G. (1986) The chlorination of
indoles by copper(II) chloride. J . Chem. Soc., Perkin Trans. 1,
2305-2308.
(15) Smith, L. R. (1972) Alkyl, alkenyl, and alkynyl indoles. 2.
Oxidation. In The Chemistry of Heterocyclic Compounds: Indoles.
Part Two (Houlihan, W. J ., Ed.) pp 78-80, Wiley-Interscience,
New York.
(16) Adam, W. (1983) Four membered ring peroxides: 1,2-dioxetanes
and R-peroxylactones. In The Chemistry of Functional Groups.
Peroxides (Patai, S., Ed.) pp 829-920, Wiley-Interscience, New
York.
(17) Matsumoto, M., and Kondo, K. (1977) Sensitized photooxygenation
of 1-methyl-3-vinylindoles. J . Am. Chem. Soc. 99, 2393-2394.
(18) Bigwood, T., Delve, R., and Read, G. (1990) A novel iron(III)
catalyzed degradation of aliphatic aldehydes to their lower
homologues with implications for lipid peroxidation chemistry.
J . Chem. Soc., Chem. Commun., 776-778.

1194 Chem. Res. Toxicol., Vol. 11, No. 10, 1998
Erdelmeier et al.
(19) Read, G. (1991) A perspective of catalytic oxidation. In Dioxygen
Activation and Homogeneous Catalytic Oxidation (Simandi, L. I.,
Ed.) pp 71-81, Elsevier Science, Amsterdam.
(20) Hathaway, B. J . (1987) Oxyanions. In Comprehensive Coordina-
tion Chemistry (Wilkinson, G., Gillard, R. D., and McCleverty, J .
A., Eds.) Vol. 2 (Ligands), pp 413-434, Pergamon Press, Oxford,
U.K.
(21) Brandsch, T., Schell, F.-A., Weis, K., Ruf, M., Mu¨ller, B., and
Vahrenkamp, H. (1997) On the ligating properties of sulfonate
and perchlorate anions towards zinc. Chem. Ber./ Recl., 283-289.
(22) Edwards, A. J . (1987) Halogens as ligands. In Comprehensive
Coordination Chemistry (Wilkinson, G., Gillard, R. D., and
McCleverty, J . A., Eds.) Vol. 2 (Ligands), pp 675-687, Pergamon
Press, Oxford, U.K.
(23) Ge´rard-Monnier, D., Erdelmeier, I., Re´gnard, K., Moze-Henry, N.,
Yadan, J .-C., and Chaudie`re, J . (1998) Chem. Res. Toxicol. 11,
1176-1183.
TX970180Z