Oxidation of serotonin by superoxide radical: Implications to neurodegenerative brain disorders

Article abstract of DOI:10.1021/tx970185w

Many new lines of evidence implicate both superoxide anion radical (O2^.-) and biogenic amine neurotransmitters in the pathological mechanisms that underlie neuronal damage caused by methamphetamine (MA), glutamate- mediated oxidative toxicity, ischemia-reperfusion, and other neurodegenerative brain disorders. In this investigation the oxidation of 5- hydroxytryptamine (5-HT, serotonin) by an O₂^.- -generating system (xanthine/xanthine oxidase) in buffered aqueous solution at pH 7.4 has been studied. The major product of the O₂^.- -mediated oxidation of 5-HT is tryptamine-4,5-dione (T-4,5-D). However, O₂^.- and H₂O₂, cogenerated by the xanthine oxidase-mediated oxidation of xanthine to uric acid, together react with trace levels of iron that contaminate buffer constituents to give a chemically ill-defined oxo-iron species. This species mediates the oxidation of 5-HT to a C(4)-centered carbocation intermediate that reacts with 5-HT to give 5,5'-dihydroxy-4,4'-bitryptamine (4,4'-D) and with uric acid to give 9-[3-(2-aminoethyl)-5-hydroxy-1H-indol-4-yl]-2,6,8-triketo- 1H,3H,7H-purine (7) as the major products. These products differ from those formed in the HO^.-mediated oxidation of 5-HT under similar conditions. When the reaction is carried out in the presence of the intraneuronal nucleophile glutathione (GSH), T-4,5-D is scavenged to give 7-(S-glutathionyl)tryptamine- 4,5-dione, whereas the putative carbocation intermediate is scavenged to give 4-(S-glutathionyl)-5-hydroxytryptamine. T-4,5-D also reacts with the sulfhydryl residues of a model protein, alcohol dehydrogenase, and inhibits its activity. Previous investigators have proposed that T-4,5-D is a serotonergic neurotoxin. This raises the possibility that T-4,5-D and perhaps other putative intraneuronal metabolites formed by the O₂^.-/H₂O₂/oxo- iron-mediated oxidations of 5-HT might be endotoxins that contribute to neurodegeneration in brain regions innervated by serotonergic neurons caused by MA, ischemia-reperfusion, and other neurodegenerative brain disorders.

Full text of DOI:10.1021/tx970185w

Chem. Res. Toxicol. 1998, 11, 639-650
639
Oxid a tion of Ser oton in by Su p er oxid e Ra d ica l:
Im p lica tion s to Neu r od egen er a tive Br a in Disor d er s
Monika Z. Wrona and Glenn Dryhurst*
Department of Chemistry and Biochemistry, University of Oklahoma, Norman, Oklahoma 73019
Received October 13, 1997
Many new lines of evidence implicate both superoxide anion radical (O2•-) and biogenic amine
neurotransmitters in the pathological mechanisms that underlie neuronal damage caused by
methamphetamine (MA), glutamate-mediated oxidative toxicity, ischemia-reperfusion, and
other neurodegenerative brain disorders. In this investigation the oxidation of 5-hydroxy-
tryptamine (5-HT, serotonin) by an O2•--generating system (xanthine/xanthine oxidase) in
buffered aqueous solution at pH 7.4 has been studied. The major product of the O2•--mediated
oxidation of 5-HT is tryptamine-4,5-dione (T-4,5-D). However, O2•- and H2O2, cogenerated by
the xanthine oxidase-mediated oxidation of xanthine to uric acid, together react with trace
levels of iron that contaminate buffer constituents to give a chemically ill-defined oxo-iron
species. This species mediates the oxidation of 5-HT to a C(4)-centered carbocation intermediate
that reacts with 5-HT to give 5,5′-dihydroxy-4,4′-bitryptamine (4,4′-D) and with uric acid to
give 9-[3-(2-aminoethyl)-5-hydroxy-1H-indol-4-yl]-2,6,8-triketo-1H,3H,7H-purine (7) as the
major products. These products differ from those formed in the HO•-mediated oxidation of
5-HT under similar conditions. When the reaction is carried out in the presence of the
intraneuronal nucleophile glutathione (GSH), T-4,5-D is scavenged to give 7-(S-glutathionyl)-
tryptamine-4,5-dione, whereas the putative carbocation intermediate is scavenged to give 4-(S-
glutathionyl)-5-hydroxytryptamine. T-4,5-D also reacts with the sulfhydryl residues of a model
protein, alcohol dehydrogenase, and inhibits its activity. Previous investigators have proposed
that T-4,5-D is a serotonergic neurotoxin. This raises the possibility that T-4,5-D and perhaps
other putative intraneuronal metabolites formed by the O2•-/H2O2/oxo-iron-mediated oxidations
of 5-HT might be endotoxins that contribute to neurodegeneration in brain regions innervated
by serotonergic neurons caused by MA, ischemia-reperfusion, and other neurodegenerative
brain disorders.
it is important not only to elucidate the mechanisms
responsible for oxidative stress but also to characterize
the actual reduced oxygen species generated and their
primary in vivo targets. It is widely believed, for
example, that cellular damage caused by ROS is sus-
tained directly by lipids, proteins, and nucleic acids (6).
However, a number of lines of evidence indicate that the
roles of ROS in neurodegenerative brain disorders might
be appreciably more complex. Potentially useful insights
into these complexities emerge from studies of MA
administration and ischemia-reperfusion using experi-
mental animals. To illustrate, cerebral ischemia evokes
a massive release of 5-hydroxytryptamine (5-HT) (7),
norepinephrine (NE) (8), and dopamine (DA) (9) in
addition to glutamate (Glu) and aspartate (Asp) (10).
Upon reperfusion (reoxygenation) serotonergic, noradr-
energic, and dopaminergic terminals degenerate (11) as
do cells in the CA1 region of the hippocampus and in
several layers of the cortex (7). Accumulating evidence
suggests that tissue injury occurs during reperfusion and
is mediated, directly or indirectly, by ROS (3, 12).
Because transgenic mice that overexpress cytoplasmic
Cu-Zn superoxide dismutase (SOD) are protected against
In tr od u ction
Oxidative stress has been implicated with the neuro-
degeneration that occurs in the brain in disorders such
as Alzheimer’s disease (AD)1 (1), Parkinson’s disease (PD)
(2), and ischemia-reperfusion (3) and as a consequence
of neurotoxic doses of methamphetamine (MA) (4). Oxi-
dative stress refers to a situation where the production
of reactive oxygen species (ROS), such as superoxide
anion radical (O2•-), hydroxyl radical (HO•), and hydrogen
peroxide (H2O2), exceeds the protective capacities of
endogenous cellular defense mechanisms (5).
To understand the pathologic processes underlying the
neurodegeneration that occurs in these brain disorders,
* Address correspondence to this author. Tel: (405) 325-4811. Fax:
(405) 325-6111. E-mail: gdryhurst@ou.edu.
1
Abbreviations: Alzheimer’s disease, AD; Parkinson’s disease, PD;
methamphetamine, MA; reactive oxygen species, ROS; superoxide
anion radical, O2•-; hydroxyl radical, HO•; 5-hydroxytryptamine, 5-HT;
norepinephrine, NE; dopamine, DA; glutamate, Glu; aspartate, Asp;
superoxide dismutase, SOD; xanthine oxidase, XOD; N-methyl-D-
aspartate, NMDA; 5,6-dihydroxytryptamine, 5,6-DHT; 6-hydroxy-
dopamine, 6-OHDA; 5-(hydroxyindolyl)-3-(ethylamino)-2-oxindole,
5-HEO; glutathione, GSH; tryptamine-4,5-dione, T-4,5-D; 7-(S-glu-
tathionyl)tryptamine-4,5-dione, 7-S-Glu-T-4,5-D; 4-(S-glutathionyl)-5-
hydroxytryptamine, 4-S-5-HT; 5,5′-dihydroxy-4,4′-bitryptamine, 4,4′-
D; 9-[3-(2-aminoethyl)-5-hydroxy-1H-indol-4-yl]-2,6,8-triketo-1H,3H,7H,-
purine, 7; saturated calomel reference electrode, SCE; alcohol
dehydrogenase, ALCD; potassium superoxide, KO2; dimethyl sulfoxide,
Me2SO; bovine serum albumin, BSA; perhydroxy radical, HO2‚; blood-
brain barrier, BBB.
•-
reperfusion injury (13), it follows that intraneuronal O2
plays a key role in the neurodegenerative mechanism-
(s). There are several possible sources of this O2•-. For
example, during the ischemic period ATP levels fall,
S0893-228x(97)00185-9 CCC: $15.00 © 1998 American Chemical Society
Published on Web 05/05/1998

640 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
Wrona and Dryhurst
xanthine accumulates (14), and a Ca2+ influx triggers
proteolytic cleavage of xanthine dehydrogenase to xan-
thine oxidase (XOD) (15). Thus, upon reoxygenation
XOD catalyzes the oxidation of xanthine to uric acid in
indolyl)-3-(ethylamino)-2-oxindole (5-HEO) (35). How-
ever, neurotoxic doses of MA do not evoke elevated levels
of this compound in any region of rat brain (34).
Taken together, the preceding information suggests
that neither MA nor ischemia-reperfusion mediates
abnormally high fluxes of HO• in the brain. Rather,
available evidence more strongly implicates O2•-, the
biogenic amine neurotransmitters, and, perhaps, their
resulting oxidation products (metabolites) in the neuro-
degenerative mechanisms. Accordingly, as a first step
toward exploring this idea, this article describes a study
•-
a reaction that generates O2 (and H2O2). Elevated
extracellular Glu and Asp levels can stimulate N-methyl-
D-aspartate (NMDA) receptor-mediated excitotoxicity (16)
in a process that also involves intraneuronal O2•- produc-
tion (17). MA also evokes a massive release of 5-HT, NE,
and DA and, with appropriate dose regimens, degenera-
tion of serotonergic and dopaminergic terminals in
certain brain regions (18-20) along with a subpopulation
of unidentified cell bodies in the somatosensory cortex
(21). The neurodegeneration evoked by MA does not
appear to be mediated by the drug itself or its known
metabolites (19, 20) but, directly or indirectly, by ROS
(4, 22). Cu-Zn SOD transgenic mice are protected
against MA-induced neurodegeneration (4) implying
of the oxidation of 5-HT by the xanthine/XOD system
•-
employed as a source of O2
.
Ma ter ia ls a n d Meth od s
Ch em ica ls. 5-Hydroxytryptamine hydrochloride (5-HT‚
HCl), xanthine, xanthine oxidase (EC 1.1.3.22), uric acid,
potassium superoxide (KO2), superoxide dismutase (SOD), cata-
lase, ethylenediaminetetraacetic acid disodium salt (Na2EDTA),
diethylamine, diethylenetriaminepentaacetic acid disodium salt
(Na2DTPA), nicotine adenine dinucleotide (NAD+), semicarba-
zide, bovine serum albumin, alcohol dehydrogenase (EC 1.1.1.1.),
mannitol, desferrioxamine mesylate (desferal), sodium octyl
sulfate, and glutathione (GSH, free base) were obtained from
Sigma (St. Louis, MO) and were used without further purifica-
tion. Citric acid monohydrate was obtained from Mallinckrodt
(St. Louis, MO). Tryptamine-4,5-dione (T-4,5-D) (37), 7-(S-
glutathionyl)tryptamine-4,5-dione (7-S-Glu-T-4,5-D) (38), 4-(S-
glutathionyl)-5-hydroxytryptamine (39), and 5,5′-dihydroxy-4,4′-
bitryptamine (4,4′-D) (40) were synthesized using published
procedures.
•-
again that O2 is the ROS primarily involved in the
neuropathological mechanisms.
However, the connection between elevated production
•-
of O2 or other ROS and neuronal damage caused by
ischemia-reperfusion and MA does not seem to be related
simply to elevated Glu and Asp release and NMDA
receptor-mediated excitotoxicity. For example, the pro-
tective effects of DA and 5-HT uptake inhibitors against
MA-induced neurodegeneration is not easily reconciled
with NMDA receptor-mediated excitotoxicity (23, 24).
Furthermore, surgical procedures or pharmacological
manipulations that deplete one or more biogenic amines
provide significant protection against the neurodegen-
eration caused by ischemia-reperfusion (25-27) and MA
(21, 28) in areas of the brain innervated by these
neurotransmitter systems. Taken together, such obser-
vations implicate both the biogenic amine and excitatory
High -P er for m a n ce Liqu id Ch r om a togr a p h y. Two HPLC
systems equipped with glassy carbon electrochemical detectors
were employed, one with the detector set in the oxidative mode
(HPLC-ECox) and the other with the detector set in the reductive
mode (HPLC-ECRed). A third preparative HPLC system em-
ployed a UV detector (HPLC-UV).
•-
amino acid neurotransmitters in addition to O2 and
HP LC-ECox: This system employed a Gilson (Middleton, WI)
model 307 pump, a Rheodyne (Cotati, CA) 7125 injector
equipped with a 5.0-µL sample loop, and a BAS (Bioanalytical
Systems, West Lafeyette, IN) LC-4C amperometric detector
equipped with a glassy carbon detector electrode set at +825
mV versus a silver/silver chloride (Ag/AgCl) reference electrode.
The detector filter was set at 0.1 Hz. A reversed-phase column
(BAS Phase II - ODS, 3 µm, 100 × 3.2 mm) and guard column
(BAS Phase II - ODS, 7 µm, 15 × 3.2 mm) were employed. The
mobile phase consisted of 0.085% (v/v) diethylamine, 0.63 mM
Na2EDTA, 0.258 mM sodium octyl sulfate, 0.1 M citric acid
dissolved in deionized water containing 5% (v/v) HPLC grade
acetonitrile (MeCN). The pH of this solution was adjusted to
2.14 with concentrated hydrochloric acid. After preparation, the
perhaps other ROS in mechanisms that mediate the
neurodegeneration caused by MA and ischemia-reperfu-
sion. A link between these three factors can be inferred
from a recent study using cultures of rat cortical and
hippocampal neurons demonstrating that the uptake and
intraneuronal oxidation of a biogenic amine neurotrans-
mitter are both essential steps associated with Glu
oxidative toxicity (29). The conclusion of this study was
that intraneuronal oxidation of the biogenic amine neu-
rotransmitter by an unknown enzyme is responsible for
generation of cytotoxic levels of ROS.
However, it seems equally plausible to suggest that
O2•-, generated by NMDA receptor activation, might
oxidize intraneuronal 5-HT, DA, or NE to endotoxic
metabolites that mediate Glu excitotoxicity. Similarly,
the neurodegeneration evoked in vivo by ischemia-
reperfusion, MA, or other neurodegenerative brain dis-
orders in which oxidative stress and biogenic amines are
implicated might also involve endotoxins formed by
intraneuronal oxidation of the latter neurotransmitters.
Indeed, it has been reported that MA induces formation
of the serotonergic neurotoxin 5,6-dihydroxytryptamine
(5,6-DHT) (30) and the catecholaminergic neurotoxin
6-hydroxydopamine (31) that are responsible for mediat-
ing neurodegeneration. However, efforts to reproduce
these results have been unsuccessful (32-34). Further-
more, 5,6-DHT and 6-OHDA appear to be products of only
the HO•-mediated oxidation of 5-HT (35) and DA (36),
respectively. The major and most stable product of the
HO•-mediated oxidation of 5-HT in vitro is 5-(hydroxy-
mobile phase was filtered through
membrane (Millipore, Bedford, MA) and degassed by stirring
under vacuum for 20 min. The flow rate was 0.6 mL min-1
a 0.22-µm type GVHP
.
HPLC-ECox was employed to detect uric acid, 4,4′-D, 4-S-Glu-
5-HT, 5-HT, and compound 7 in reaction product solutions.
HP LC-ECRed : A BAS 200B instrument was used equipped
with a Rheodyne 9125 injection valve with a 5.0-µL sample loop
and dual parallel thin-layer glassy carbon detector electrodes.
One electrode was set at -50 mV and the other at -550 mV vs
Ag/AgCl. The detector filter was set at 0.05 Hz. To remove
molecular oxygen the entire system was purged with ultrahigh
purity He gas for 2 h. The exhaust valve was then closed, and
the system was pressurized at 4 psi with He. The mobile phase,
detector oven, detector cell, and column were maintained at 35
( 0.1 °C. The analytical and guard columns were the same as
used for the HPLC-ECox system. The mobile phase was
prepared using 7.5% (v/v) MeCN in deionized water but was
otherwise identical to that used for the HPLC-ECox system. The
flow rate was 1.0 mL min-1
. The background current at the

Oxidation of Serotonin by Superoxide Radical
Chem. Res. Toxicol., Vol. 11, No. 6, 1998 641
detector electrode set at -50 mV vs Ag/AgCl was typically
between 1.3 and 6.0 nA, and that at the -550 mV detector
electrode was between 24 and 35 nA. The detector electrodes
were resurfaced when the background current (mobile phase
alone flowing) exceeded the latter limits. The glassy carbon
detector electrodes were resurfaced (polished) using the abrasion
method recommended by the manufacturer (BAS). The elec-
trodes were then electropolished at -2.0 V vs Ag/AgCl for 2 h
with the mobile phase flowing through the cell. Typically,
detector electrode resurfacing was necessary every 2-3 weeks.
The BAS 200B was controlled by BAS ChromGraph software,
and data were analyzed using ChromGraph Report software.
HPLC-ECRed was employed to detect T-4,5-D and 7-S-Glu-T-
4,5-D because these compounds gave no response using HPLC-
reaction solution was injected via a 10 mL sample loop into the
HPLC-UV system.
9-[3-(2-Am in oet h yl)-5-h yd r oxy-1H -in d ol-4-yl]-2,6,8-t r i-
k eto-1H,3H,7H-p u r in e (7). Compound 7 was initially sepa-
rated from other reaction products using preparative HPLC-
UV. The solution that eluted under the peak corresponding to
7 was desalted using the same HPLC-UV system but with
deionized water as the only mobile phase. The eluent containing
7 was collected, shell-frozen, and freeze-dried to give a white
solid that had very low solubility in water or common organic
solvents. Compound 7 gave a UV spectrum (pH 3.0) with λmax
) 294 nm. Attempts to obtain a fast atom bombardment mass
spectrum or matrix-assisted laser desorption ionization mass
spectrum on 7 using a variety of matrixes were unsuccessful.
However, a solution of 7 freshly collected in the HPLC-UV
mobile phase (5% MeCN in ammonium formate buffer, pH 4.7)
and diluted in 1% acetic acid in 1:1 methanol water gave an
electrospray mass spectrum with m/z ) 343 (MH+, 100%). 1H
NMR (400 MHz, D2O) gave δ 7.30 (d, J ) 8.8 Hz, 1H, C(7)-H),
7.08 (s, 1H, C(2)-H), 6.77 (d, J ) 8.8 Hz, 1H, C(6)-H), 2.80 (m,
2H, C(â)-H2), 2.60 (m, 2H, C(R)-H2). Compound 7 was also
synthesized by an independent electrochemical method. In a
typical experiment, 5-HT‚HCl (2 mg) and uric acid (2.4 mg) were
dissolved in 40 mL of pH 7.4 phosphate buffer (µ ) 0.2) and
electro-oxidized at a working electrode consisting of several
plates of pyrolytic graphite using an applied potential of 195
mV vs SCE [equipment and procedures are described elsewhere
(37)]. This applied potential was sufficient to electro-oxidize
5-HT but not uric acid. Under these conditions 7 was formed
in high yield and was separated and purified as described
earlier. The spectral and chromatographic properties of elec-
trochemically synthesized 7 were identical to those of the
compound formed in the xanthine/XOD-mediated oxidation of
5-HT.
ECox
. Quantitative determinations of T-4,5-D using HPLC-
ECRed were based on linear analytical calibration curves. The
procedures employed to prepare solutions of T-4,5-D of known
concentration by exhaustive controlled potential electro-oxida-
tion of 5-HT in 0.01 M HCl at 740 mV vs SCE (saturated calomel
electrode) have been described in detail elsewhere (37); λmax (nm)
(log ꢀmax, M-1 cm-1) of T-4,5-D in 0.01 M HCl was 540 (3.21).
P r ep a r a tive HP LC-UV: The equipment and chromato-
graphic conditions employed for preparative HPLC-UV were the
same as described previously (40). This system was employed
for identification of reaction products and for isolation of 7.
Sp ectr oscop y. NMR spectra were recorded on a Varian
(Palo Alto, CA) Inova-400 spectrometer. Electrospray mass
spectra were obtained with a PE Sciex model API III instrument
with the orifice voltage set at 45 V. UV-visible spectra were
recorded on a Hewlett-Packard (Palo Alto, CA) 8452A diode
array spectrophotometer.
Oxid a tion of 5-HT by O2•- (Xa n th in e/Xa n th in e Oxid a se
System ). The xanthine oxidase (XOD)-mediated oxidation of
xanthine to uric acid in buffered aqueous solution at pH 7.4 was
Alcoh ol Deh yd r ogen a se (ALCD) Activity Mea su r em en t.
Typically, 250 µL of ALCD (250 µg mL-1 in pH 7.4 phosphate
buffer) was initially incubated for a predetermined period of
time (0-22 h) with 250 µL of pH 7.4 phosphate buffer (µ ) 0.2;
control) or 250 µL of T-4,5-D (20-110 µM in pH 7.4 phosphate
buffer) at 25 °C. Then, to a 250-µL aliquot of this solution was
added 390 µL of phosphate buffer, 350 µL of 2 mM Na2EDTA,
500 µL of 10 mM NAD+, and 500 µL of 280 mM semicarbazide,
all reactants being dissolved in pH 7.4 phosphate buffer. The
reaction was initiated by addition of 10 µL of ethanol (96%).
Activity of ALCD was based on the absorbance measured at 340
nm (NADH) after 10 min.
•-
used to generate O2 (47, 48). A typical reaction solution
consisted of 400 µL of 1.0 mM xanthine in pH 7.4 phosphate
buffer (µ ) 0.2), 200 µL of 1.0 mM Na2EDTA (in water), 600 µL
of pH 7.4 phosphate buffer (µ ) 0.2), 740 µL of deionized water,
and 40 µL of 10 mM 5-HT (in water). After this solution was
thoroughly mixed the oxidation reaction was initiated by
addition of 20 µL of XOD suspension [in 2.3 M (NH4)2SO4, 10
mM pH 7.8 phosphate buffer containing 1 mM Na2EDTA, and
1 mM sodium salicylate as supplied by the manufacturer].
Thus, the final reaction solution (2000 µL) initially contained
200 µM xanthine, 100 µM Na2EDTA, 200 µM 5-HT, and 79 mg
mL-1 XOD (0.087 unit mL-1). When xanthine or XOD was
omitted from this reaction solution 5-HT was not oxidized.
Under the preceding reaction conditions but substituting the
5-HT solution with water (40 µL), it was found that the
concentration of uric acid increased with time up to 8-10 min
when it reached a constant level of 186.0 ( 10.9 µM (mean (
standard deviation; n ) 3). Thus, under the experimental
conditions employed xanthine was almost quantitatively oxi-
Resu lts
•-
P r od u cts of th e O2 (Xa n th in e/XOD)-Med ia ted
Oxid a tion of 5-HT. Preliminary experiments estab-
lished that under the reaction conditions employed a 200
µM solution of xanthine was almost quantitatively (g93%)
oxidized to uric acid in e10 min in the presence of XOD
(0.087 unit mL-1). Chromatograms of the product solu-
tion formed following incubation of 5-HT (200 µM) with
xanthine (200 µM) and XOD (0.087 unit mL-1) for 10 min
are presented in Figure 1. HPLC-ECox (Figure 1A)
showed formation of two major electro-oxidizable prod-
ucts, 5,5′-dihydroxy-4,4′-bitryptamine (4,4′-D) and 7.
Additional peaks in this chromatogram corresponded to
unreacted 5-HT and uric acid. HPLC-ECRed (detector set
at -50 mV vs Ag/AgCl) showed formation of T-4,5-D as
the only electrochemically reducible product (Figure 1B).
Product identities were confirmed by injecting reaction
solutions into the preparative HPLC-UV system and
collecting solutions that eluted under the peaks corre-
sponding to each product. T-4,5-D and 4,4′-D were
identified on the basis of their UV-visible spectra and
electrochemical and chromatographic properties (37, 40).
•-
dized to uric acid in e10 min. Because the generation of O2
parallels uric acid formation in this reaction, a maximum
incubation time of 10 min was normally employed for oxidation
of 5-HT by the xanthine/XOD system. The progress of this
reaction was monitored by periodic withdrawal of samples for
analysis by HPLC-ECox (to measure 5-HT consumed and forma-
tion of 4,4′-D) and by HPLC-ECRed (to measure T-4,5-D formed).
Aliquots of the reaction solution (5 µL) were injected directly
into the HPLC-ECRed system. However, prior to HPLC-ECox
analysis an aliquot (40 µL) of the reaction solution was
transferred into 360 µL of ice-cold 0.1 M HCl. The resulting
solution was vortexed and immediately injected into the HPLC-
ECox system.
The preparative HPLC-UV system was employed on occasion
to monitor the decrease of 5-HT concentration and the appear-
ance of various products. However, the total reaction solution
volume was then increased to 10-mL keeping the concentrations
of all reactants the same as described previously. The entire

642 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
Wrona and Dryhurst
F igu r e 1. (A) HPLC-ECox (detector potential 825 mV vs Ag/
AgCl) and (B) HPLC-ECRed (detector potential -50 mV) chro-
matograms of the product solution obtained following incubation
of 5-HT (200 µM) with xanthine (200 µM) and XOD (0.087 unit
mL-1) in pH 7.4 phosphate buffer for 10 min at 25 °C.
Procedures employed to structurally characterize 7 are
presented in Materials and Methods.
The time course for oxidation of 5-HT by the xanthine/
XOD system is shown in Figure 2. Thus, during the
initial 8 min of the reaction the concentration of 5-HT
decreased rapidly and that of T-4,5-D and 4,4′-D cor-
respondingly increased. At longer times the consumption
of 5-HT and formation of 4,4′-D became very much
slower. Under the experimental conditions employed, the
results of several replicate experiments (n ) 9) demon-
strated that after 10 min the xanthine/XOD-mediated
oxidation of 141 ( 27 nmol (mean ( SD) of 5-HT
generated 50 ( 4 nmol of T-4,5-D and 32 ( 3 nmol of
4,4′-D. Thus, 1 mol of 5-HT was oxidized to 0.36 mol of
T-4,5-D and 0.23 mol of 4,4′-D. The remaining 5-HT
oxidized was accounted for primarily as 7. However, it
was noted throughout the course of this investigation that
different batches of XOD had different activities and,
F igu r e 2. Time course for the oxidation of 5-HT (200 µM) by
xanthine (200 µM)/XOD (0.087 unit mL-1) in pH 7.4 phosphate
buffer at 25 °C: (A) consumption of 5-HT; (B) formation of T-4,5-
D; (C) formation of 4,4′-D. 5-HT and 4,4′-D were determined by
HPLC-ECox; T-4,5-D was determined by HPLC-ECRed
.
•-
hence, somewhat variable rates of O2 (and H2O2)
production that in turn were reflected in the slightly
variable product distributions reported in this article.
Using very high detector sensitivities, HPLC-ECox analy-
sis of product solutions showed peaks for several very
minor electro-oxidizable products. These peaks, however,
did not correspond to 5-HEO, 5,6-DHT, or other products
characteristic of the HO•-mediated oxidation of 5-HT (35).
T-4,5-D accumulated only during the initial 8-10 min
when 5-HT was oxidized by the xanthine/XOD system
•-
(Figure 2B), i.e., during the period when O2 was
generated. Subsequently, T-4,5-D concentrations de-
clined. Incubation of T-4,5-D with all of the constituents
of the reaction mixture except XOD resulted in the loss
(decomposition) of less than 5% of the dione after 60 min
(Figure 3). However, in the presence of XOD but not
xanthine, the concentration of T-4,5-D decreased with
time to 74% of its initial value after 60 min (Figure 3).
These results suggested that T-4,5-D might bind co-
valently to XOD, perhaps by addition to nucleophilic
F igu r e 3. Stability of T-4,5-D (7.3 µM) when incubated with
(A) 200 µM xanthine or (B) 0.087 unit mL-1 XOD in phosphate
buffer at 25 °C. The concentration of T-4,5-D was measured
using HPLC-ECRed (-50 mV).
residues of the enzyme. However, such binding did not
affect the activity of XOD. To illustrate, following a 60-

Oxidation of Serotonin by Superoxide Radical
Chem. Res. Toxicol., Vol. 11, No. 6, 1998 643
Ta ble 1. In flu en ce of Su p er oxid e Dism u ta se (SOD),
Ca ta la se, Hyd r oxyl Ra d ica l Sca ven ger s, a n d Ir on
Ch ela tor s on th e Oxid a tion of 5-HT by th e Xa n th in e/
Xa n th in e Oxid a se System a
Ta ble 2. In flu en ce of Glu ta th ion e on th e Oxid a tion of
5-HT by th e Xa n th in e/Xa n th in e Oxid a se System a
GSH
concn
(µM)
5-HT
oxidized
(nmol)
nmol of
T-4,5-D/nmol
of 5-HT oxidized
nmol of
4,4′-D/nmol of
5-HT oxidized
nmol of
nmol of
5-HT
oxidized
(nmol)
T-4,5-D/nmol 4,4′-D/nmol
0
20
50
100
500
1000
5000
134
190
186
164
64
0.34
0.10
0.10
0.14
0.23
0.21
0.14
0.26
0.23
0.25
0.23
0.11
0.06
0.02
of 5-HT
oxidized
of 5-HT
oxidized
added compd (concn)
SOD
(0)
134
204
174
100
128
82
0.34
0.06
0.04
0.04
0.41
0.57
0.91
0.39
0.36
0.44
0.29
0.43
0.53
0.56
0.77
0.25
0.30
0.29
0.43
0.23
0.18
0.10
0.29
0.29
0.33
0.22
0.28
0.27
0.12
0.10
(25 µg mL-1
)
(102 µg mL-1
)
)
62
44
(255 µg mL-1
catalase (0)
(1 µg mL-1
(20 µg mL-1
mannitol (0)
a
5-HT (200 µM) was incubated in pH 7.4 phosphate buffer (µ
) 0.2) at 25 °C with xanthine (200 µM), XOD (0.087 unit mL-1),
and the indicated concentration of GSH in a total volume of 2000
µL. Ten minutes after addition of XOD, the resulting product
solution was analyzed for T-4,5-D (HPLC-ECRed) and 4,4′-D
(HPLC-ECox).
)
)
22
122
128
110
158
142
134
136
102
(10 mM)
(10 mM)
0
(100 µM)
(200 µM)
(1000 µM)
(1850 µM)
formate
desferal
of T-4,5-D at the expense of 4,4′-D. High concentrations
of the HO• scavengers mannitol and formate had no
significant effects on the oxidation of 5-HT by the
xanthine/XOD system. The powerful iron (Fe3+)-com-
plexing agent desferal appeared to have relatively little
influence on the extent of oxidation of 5-HT when
included in the normal incubation mixture with xanthine
and XOD, although with increasing concentrations it
shifted the reaction toward formation of T-4,5-D at the
a
5-HT (200 µM) was incubated in pH 7.4 phosphate buffer (µ
) 0.2) at 25 °C with xanthine (200 µM) and XOD (0.087 unit mL-1
)
in a total solution volume of 2.0 mL. Other additions are shown
in the table. Ten minutes after addition of XOD, the resulting
product solution was analyzed for T-4,5-D (HPLC-ECRed) and
4,4′-D (HPLC-ECox).
expense of 4,4′-D. Heat-denatured catalase (74 µg mL-1
)
min incubation of XOD (0.087 unit mL-1, 7.9 mg of
protein mL-1) with T-4,5-D (50 µM) in pH 7.4 phosphate
buffer at 25 °C, HPLC-ECRed analysis revealed that
approximately 26% of the initially added dione had been
lost. After addition of xanthine (200 µM), the concentra-
tion of uric acid measured (HPLC-ECox) after 10 min was
identical to that measured when XOD was preincubated
for 60 min in the absence of T-4,5-D prior to addition of
xanthine. However, even in the absence of XOD, T-4,5-D
slowly decomposed in pH 7.4 phosphate buffer. For
example, after 24 h the concentration of a solution
initially 50 µM in T-4,5-D had declined to 27.3 ( 2.8 µM.
In the presence of XOD (0.087 unit mL-1, 7.9 mg of
protein mL-1) an otherwise identical solution of T-4,5-D
declined to 10.6 ( 5.6 mM after 6 h and was undetectable
(HPLC-ECRed) after 24 h.
or bovine serum albumin (BSA; 473 µg mL-1) had no
significant effects on the oxidation of 5-HT by the
xanthine/XOD system in terms of reaction rate or identity
and yields of products. Similarly, Na2EDTA (0-500 µM)
and Na2DTPA (0-500 µM) had no effect on the reaction
rate and products. Quantitative analytical data for 7
were not obtained. However, factors that caused an
increase in the yield of 4,4′-D always evoked a parallel
increase in the chromatographic peak height (HPLC-
ECox) for 7 and a corresponding decrease in that for uric
acid.
In flu en ce of Glu ta th ion e on th e Oxid a tion of
5-HT by O2•- (Xa n th in e/XOD). Low concentrations of
GSH (20-100 µM) appeared to potentiate the oxidation
of 5-HT by the xanthine/XOD system (Table 2). However,
higher concentrations (g500 µM) inhibited the reaction,
although even 5 mM GSH was unable to completely block
the oxidation of 5-HT. Low concentrations of GSH
inhibited formation of T-4,5-D, whereas high concentra-
tions inhibited formation of 4,4′-D. A comparison of the
chromatograms (HPLC-ECRed, detector set at -50 mV vs
Ag/AgCl) of the product solutions obtained when 5-HT
was oxidized by the xanthine/XOD system in the absence
(Figure 4A) and presence (Figure 4B) of 30 µM GSH
indicated that the decreased yield of T-4,5-D under the
latter conditions was primarily due to formation of 7-S-
Glu-T-4,5-D. However, HPLC-ECRed with the detector
electrode set at -550 mV vs Ag/AgCl revealed that
additional unknown electrochemically reducible products
(identified as Unk 1 and Unk 2 in Figure 4C) were formed
when 5-HT was oxidized by xanthine/XOD in the pres-
ence of GSH. The chromatogram (HPLC-ECRed) shown
in Figure 4D was recorded 60 s after addition of GSH
(30 µM) to the product solution analyzed in Figure 4A
and demonstrates the rapid reaction of GSH with T-4,5-D
to give 7-S-Glu-T-4,5-D along with minor yields of Unk
1 and Unk 2. However, incubation of 7-S-Glu-T-4,5-D
(14 µM) with the xanthine (200 µM)/XOD (0.087 unit
mL-1) system at pH 7.4 for 10 min resulted in its almost
Oxid a t ion of 5-H T w it h P ot a ssiu m Su p er oxid e
(KO2). Addition of KO2 (1.8 mg, 14.4 µmol) to a vigor-
ously stirred solution of 5-HT (2 mM, 6 µmol) in dimeth-
ylsulfoxide (Me2SO; 3 mL) exposed to the atmosphere
caused the colorless solution to initially turn pink and
then bright green. Addition of 7 mL of mobile phase A
(pH 4.7), used for preparative HPLC-UV (37), caused this
solution to become bright purple. The resulting solution
was injected into the preparative HPLC-UV system. The
only product formed in the reaction based on its chro-
matographic retention time, cyclic voltammetry, UV-
visible spectrum, and subsequent analysis by HPLC-
ECRed was T-4,5-D.
In flu en ce of An tioxid a n ts, Ra d ica l Sca ven ger s,
•-
a n d Ir on Ch ela tor s on th e Oxid a tion of 5-HT by O2
(Xa n th in e/XOD System ). The most obvious effect of
SOD on the oxidation of 5-HT by the xanthine/XOD
system was a dramatic decrease in the formation of
T-4,5-D along with a tendency to increased yields of
4,4′-D (Table 1). However, increasing SOD concentra-
tions first stimulated and then inhibited oxidation of
5-HT. By contrast, catalase inhibited oxidation of 5-HT
but simultaneously shifted the reaction toward formation

644 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
Wrona and Dryhurst
F igu r e 5. HPLC-ECRed chromatograms of (A) 7-S-Glu-T-4,5-D
(14 µM) and (B) the same solution 10 min after adding xanthine
(200 µM) and XOD (0.087 unit mL-1). Detector electrode set at
-550 mV vs Ag/AgCl.
F igu r e 6. HPLC-ECox chromatogram of the product solution
formed after oxidation of 5-HT (200 µM) by xanthine (200 µM)/
XOD (0.087 unit mL-1) in pH 7.4 phosphate buffer (µ ) 0.2) for
10 min in the presence of GSH (1000 µM).
relatively high concentrations of GSH (Table 2) were in
part related to increased formation of 4-S-Glu-5-HT
(Figure 6).
Role of Oxo-Ir on Sp ecies in th e Oxid a tion of 5-HT
by O2•- (Xa n th in e/XOD). Heat-denatured catalase and
BSA had no significant effect on the oxidation of 5-HT
by the xanthine/XOD system. Thus, the inhibitory effect
of active catalase and the shift of the reaction toward
T-4,5-D formation (Table 2) imply that a major route
leading to 4,4′-D requires H2O2 and is not related to
oxygen radical-scavenging properties of the protein.
Furthermore, the phosphate buffer system employed in
this investigation was certainly contaminated with trace
(micromolar) concentrations of iron and other transition-
metal ions. Under the reaction conditions employed, iron
chelated by EDTA or phosphate would initially be in the
Fe(III) oxidation state which is readily reduced to Fe(II)
F igu r e 4. HPLC-ECRed chromatograms of the product solutions
formed after oxidation of 5-HT (200 µM) by the xanthine (200
µM)/XOD (0.087 unit mL-1) system in pH 7.4 phosphate buffer
(µ ) 0.2) for 10 min: (A) in the absence of GSH; (B and C) in
the presence of GSH (30 µM); detector electrode set at (A) -50
mV and (B and C) -550 mV. In panel D GSH (30 µM) was added
to the solution in panel A and analyzed after 60 s (-550 mV).
complete (>90%) transformation into Unk 1 and Unk 2
(Figure 5). Because free GSH was not present in this
reaction, it appears that 7-S-Glu-T-4,5-D may be oxidized
by the xanthine/XOD system. In the absence of free GSH
or exposure to the xanthine/XOD system, solutions of 7-S-
Glu-T-4,5-D were moderately stable in pH 7.4 phosphate
buffer. To illustrate, the concentration of a 46.5 µM
solution of 7-S-Glu-T-4,5-D declined to 41.0 ( 4.0 and
12.1 ( 6.0 µM after 1 and 14 h, respectively.
by O2 that, in turn, is capable of generating HO• from
•-
H2O2 by Fenton chemistry or from H2O2 and O2•- together
by the Haber-Weiss reaction (41). This raised the
possibility that 4,4′-D (and 7) is formed by oxidation of
5-HT by HO• generated by Fe(II) and O2•-/H2O2 that are
the byproducts of the xanthine/XOD reaction (42). How-
ever, HO• scavengers had no significant influence on the
The decreased yields of 4,4′-D observed when 5-HT was
oxidized by the xanthine/XOD system in the presence of

Oxidation of Serotonin by Superoxide Radical
Chem. Res. Toxicol., Vol. 11, No. 6, 1998 645
Ta ble 3. Oxid a tion of 5-HT w h en Ad d ed a fter
Com p letion (10 m in ) of th e XOD-Med ia ted Oxid a tion of
Xa n th in ea
Ta ble 4. Oxid a tion of 5-HT w h en Ad d ed a fter
Com p letion (10 m in ) of th e XOD-Med ia ted Oxid a tion of
Xa n th in e in th e P r esen ce of Desfer a la
uric
acid
nmol of
nmol of
peak
uric
acid
nmol of
T-4,5-D/nmol 4,4′-D/nmol
of 5-HT
oxidized
nmol of
peak
height
5-HT
T-4,5-D/nmol 4,4′-D//nmol height for
of 5-HT
oxidized
5-HT
time concn oxidized
(min) (µM)
of 5-HT
oxidized
dimer 7b
(nA)
time concn oxidized
(min) (µM)
of 5-HT
oxidized
for dimer
(nmol)
(nmol)
7b (nA)
10
14
30
55
185
165
120
98
10
14
30
55
190
178
128
131
106
270
318
0.04
0.03
0.03
0.32
0.25
0.23
12.4
34.6
40.4
182
298
298
0.62
0.59
0.52
0.10
0.09
0.10
10.6
21.4
22.0
a
a
Xanthine (200 µM) and XOD (0.087 unit mL-1) were incubated
in pH 7.4 phosphate buffer at 25 °C for 10 min. The uric acid
concentration was measured (HPLC-ECox) at 10 min, and simul-
taneously, 5-HT (200 µM) was added. The resultant product
solution was analyzed at the times indicated for T-4,5-D (HPLC-
Xanthine (200 µM), XOD (0.087 unit mL-1), and desferal (1.85
mM) were incubated in pH 7.4 phosphate buffer at 25 °C for 10
min. The uric acid concentration was measured (HPLC-ECox) at
10 min, and simultaneously, 5-HT (200 µM) was added. The
resultant product solution was analyzed at the times indicated
for T-4,5-D (HPLC-ECRed) and uric acid, 5-HT, 4,4′-D, and 7
b
ECRed) and uric acid, 5-HT, 4,4′-D, and 7 (HPLC-ECox). Chro-
matographic peak height for 7 using HPLC-ECox
b
.
(HPLC-ECox). Chromatographic peak height for 7 using HPLC-
ECox
.
xanthine/XOD-mediated oxidation of 5-HT (Table 1).
Furthermore, 5-HEO, the major product of the in vitro
oxidation of 5-HT by HO• (35), was not observed in the
xanthine/XOD-mediated reaction. Conversely, 4,4′-D is
not a product of the HO•-mediated oxidation of 5-HT (35).
The fact that only the C(4)-position of 5-HT is activated
(to form an intermediate that is the precursor of 4,4′-D,
7, and 4-S-Glu-5-HT) in the xanthine/XOD-mediated
oxidation also suggests that the reaction must involve a
much more selective oxidant than HO•. Because incuba-
tions of 5-HT with XOD and H2O2 in pH 7.4 phosphate
buffer for 10 min in the absence of xanthine, and
therefore O2•-, resulted in no significant oxidation of the
neurotransmitter, it appeared that a major reaction
pathway leading to 4,4′-D (and 7) depends on O2•-, H2O2,
and trace concentrations of iron species. Indeed, in this
respect, the complex effects of SOD, catalase, desferal
(Table 1), and other reducing agents such as GSH (Table
2) are similar to those reported for iron-catalyzed lipid
completed, 5-HT was rather extensively oxidized. Very
little T-4,5-D was formed under these conditions, whereas
yields of 4,4′-D were about the same as when 5-HT was
initially included in the incubation mixture with xanthine
and XOD. However, it was notable that uric acid
concentrations decreased appreciably with time and,
correspondingly, the yields of 7 increased. Control
experiments confirmed that O2•- was essential for these
reactions. Thus, when xanthine (200 µM) and H2O2 (220
µM) were incubated for 10 min or XOD (0.087 unit mL-1
)
and H2O2 (220 µM) were incubated for 10 min, subse-
quent addition of 5-HT (200 µM) resulted in no significant
oxidation of the neurotransmitter. Together, the preced-
ing results indicated that in the absence of 5-HT, the
xanthine/XOD system generates an oxidant dependent
•-
on O2 and H2O2 that is subsequently responsible for
oxidation of the neurotransmitter primarily to 4,4′-D and
7. Because of their instability in aqueous solution,
neither O2 nor HO• can be this oxidant species.
A
•-
•-
peroxidation reactions (43, 44). In these reactions O2
or other reductants serve to reduce Fe(III) to Fe(II), and
the subsequent oxidation of Fe(II) by H2O2 together serve
to generate a chemically ill-defined oxo-iron species that
is a strong oxidizing agent (43, 44). This oxidant requires
both Fe(III) and Fe(II), and optimal oxidizing activity
occurs when these species are present in a 1:1 ratio. The
Fe(III)-desferal complex cannot be reduced to the Fe-
plausible interpretation of these results is that in the
initial absence of 5-HT more O2 is available to reduce
•-
trace levels of Fe(III) to Fe(II) with the same concentra-
tions of H2O2 required for oxidation of Fe(II) to an oxo-
iron species; i.e., conditions are favorable for formation
of the species that oxidizes 5-HT primarily to 4,4′-D and
7. Because the concentrations of iron in the reaction
solutions must be very low (although these were not
measured), the oxidation of 5-HT to the latter compounds
must involve catalytic regeneration of the putative oxo-
iron species.
•-
(II) oxidation state by O2 or other species with similar
reducing properties (41). Furthermore, desferal shifts the
xanthine/XOD-mediated oxidation of 5-HT toward forma-
tion of T-4,5-D at the expense of 4,4′-D (Table 1). These
results, therefore, support a role for Fe(II) and thence
an oxo-iron species in a major reaction pathway that
leads to 4,4′-D (and 7) (see also later discussion for the
influence of desferal on the O2•--mediated oxidation of
5-HT).
To explore a role for oxo-iron species in the xanthine/
XOD-mediated oxidation of 5-HT, additional experiments
were performed. Thus, xanthine (200 µM) and XOD
(0.087 unit mL-1) were incubated in pH 7.4 phosphate
buffer for 10 min initially in the absence of 5-HT. At
the end of this time, i.e., when the generation of O2•- had
terminated, an aliquot (5 µL) of the solution was analyzed
by HPLC-ECox to measure the concentration of uric acid.
Simultaneously, 5-HT (200 µM) was added to the solution
that was then periodically analyzed for T-4,5-D (HPLC-
ECRed) and uric acid, 5-HT, 4,4′-D, and 7 (HPLC-ECox)
with the results presented in Table 3. Thus, despite the
fact that the XOD-mediated oxidation of xanthine was
The Fe(III)-desferal complex cannot be reduced by
•-
O2 (41). Accordingly, the putative oxo-iron oxidant
activity should be suppressed in the presence of desferal,
and indeed, this was observed when 5-HT is oxidized by
the xanthine/XOD system in the presence of this com-
plexing agent (Table 1). However, it is also known that
•-
desferal can react with O2 to form a rather strongly
oxidizing nitroxide radical having a lifetime of several
minutes (45). To investigate the possible role of the latter
nitroxide radical, the xanthine/XOD system was incu-
bated with desferal (1.85 mM) for 10 min at 25 °C in the
absence of 5-HT. After this time an aliquot of the
reaction solution (5 µL) was analyzed by HPLC-ECox to
measure uric acid formation, and simultaneously, 5-HT
(200 µM) was added to the solution. The resulting
product mixture was then analyzed at various times with
the results shown in Table 4. Thus, at a stage of the
reaction (g10 min) when the XOD-mediated oxidation

646 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
Wrona and Dryhurst
Discu ssion
In Me2SO solution the O2•--mediated oxidation of 5-HT
generated T-4,5-D as the only detected reaction product.
Furthermore, in buffered aqueous solution at pH 7.4 SOD
profoundly, although not completely, inhibited formation
of T-4,5-D when 5-HT was incubated with the xanthine/
XOD system as a source of O2•- (Table 1). Although the
xanthine/XOD system generates both O2•- and H2O2 (41,
42), the latter observations support the conclusion that
T-4,5-D is predominantly formed from 5-HT in a reaction
•-
mediated by O2•-. However, O2 is unable to directly
oxidize organic molecules with acidic protons such as
•-
5-HT (47, 48). Rather, it is the strong tendency of O2
to disproportionate via abstraction of protons from such
substrates that accounts for its ability to indirectly
mediate oxidation reactions (47, 48). Thus, it is probable
that O2•- abstracts a proton from 5-HT to generate anion
1a /1b, HO2-, and molecular oxygen (Scheme 1). By
analogy with oxidations of related hydroxyindolamines,
it is probable that 1b is the principal electron donor to
molecular oxygen giving the free-radical-superoxide
complex 2 (49, 50). Recombination of the superoxide
residue of 2 with the incipient C(4)-centered free radical
yields the hydroperoxy anion 3 that, upon protonation,
generates hydroperoxide 4 that then decomposes to
T-4,5-D (48, 49).
•
F igu r e 7. (A) Time-dependent and (B) concentration-dependent
inhibition of ALCD by T-4,5-D. (A) ALCD (125 µg mL-1) was
incubated with T-4,5-D (70.9 µM) for indicated times, and
activity was then assayed. (B) ALCD (125 µg mL-1) and T-4,5-D
at concentrations indicated in pH 7.4 phosphate buffer (µ ) 0.2)
were incubated for 20 h, and then activity was assayed. The
assay for the ALCD-mediated oxidation of ethanol is described
in Materials and Methods.
The pKa of the perhydroxyl radical (HO2 ) in water is
4.69 (48), and hence its concentration at pH 7.4 is low
•
compared to the concentration of O2•-. However, HO2
•-
is also formed by deprotonation of H2O2 by O2 (eq 1)
(51), both of which are byproducts of the XOD-mediated
oxidation of xanthine.
-
O2•- + H2O2 f HO2• + HO2
(1)
of xanthine to uric acid was complete and O2•- generation
had ceased, 5-HT was oxidized for several minutes. The
major product of this reaction was T-4,5-D. Much lower
yields of 4,4′-D were formed. Furthermore, the presence
of desferal in the O2•--generating reaction prior to addi-
tion of 5-HT resulted in significantly lower uric acid
consumption and hence lower yields of 7 (Table 4) than
in the absence of this iron chelator (Table 3). In a control
experiment, desferal (1.85 mM) and XOD (0.087 unit
mL-1) were incubated for 10 min in the absence of
xanthine prior to addition of 5-HT (200 µM). Under these
conditions O2•- is not generated, and subsequently, 5-HT
was not oxidized. These results strongly suggest that the
influence of desferal on the xanthine/XOD-mediated
oxidation of 5-HT (Table 1) is due to its reaction with
Thus, it is likely that significant albeit transient fluxes
of HO2 are formed. Accordingly, it is conceivable that
•
HO2• directly oxidizes (one-electron abstraction) 5-HT (or
1) to the C(4)-centered radical 5 (Scheme 1). Dimeriza-
tion of 5, therefore, represents a plausible although minor
reaction leading to 4,4′-D. Alternatively, disproportion-
ation of 5 to carbocation 6 (and 1) followed by nucleophilic
addition of 5-HT (or 1) could also lead to 4,4′-D. Nucleo-
philic addition of uric acid or water to 6 provides minor
routes to 7 and 4,5-dihydroxytryptamine (4,5-DHT) that
is easily autoxidized to T-4,5-D (37).
SOD both potentiates (low concentrations) and inhibits
(high concentrations) the xanthine/XOD-mediated oxida-
tion of 5-HT (Table 1) but shifts the reaction toward
formation of 4,4′-D (and 7) at the expense of T-4,5-D.
Catalase, however, inhibits the oxidation of 5-HT but
shifts the reaction that does occur toward T-4,5-D at the
expense of 4,4′-D (Table 1). These effects appear to
reflect the influence of O2•- and H2O2 on formation of an
oxo-iron species that has also been implicated in lipid
peroxidation reactions (43, 44). Indeed, the experiments
described in connection with Table 3 strongly implicate
O2•- and H2O2 together in formation of such an oxo-iron
species that, after exhaustion of the XOD-mediated
oxidation of xanthine, oxidizes 5-HT predominantly to
4,4′-D and 7. The chemically ill-defined oxo-iron species
therefore selectively activates the C(4)-position of 5-HT.
It is proposed therefore that the oxo-iron complex ab-
stracts an electron from 5-HT to give radical 5 and thence
carbocation 6, the precursors of 4,4′-D and 7 (Scheme 1).
•-
O2 to give a nitroxide radical (45) that then oxidizes
the neurotransmitter primarily to T-4,5-D with cor-
respondingly lower yields of 4,4′-D and 7.
Bin d in g of T-4,5-D w ith P r otein s. T-4,5-D reacted
both with XOD (Figure 3) and, more rapidly, with GSH
(Figure 4D). These observations suggested that reactions
of T-4,5-D with protein nucleophiles, particularly sulf-
hydryl residues, might contribute to its putative neuro-
toxic properties (see later discussion). To explore the
binding of T-4,5-D to proteins, its effects on a model
enzyme, alcohol dehydrogenase (ALCD) that contains 24
sulfhydryl residues (46), was investigated. Indeed, T-4,5-D
evoked a time-dependent (Figure 7A) and concentration-
dependent (Figure 7B) inhibition of the ALCD-mediated
oxidation of ethanol.

Oxidation of Serotonin by Superoxide Radical
Chem. Res. Toxicol., Vol. 11, No. 6, 1998 647
Sch em e 1
At relatively low concentrations, GSH potentiates
oxidation of 5-HT by the xanthine/XOD system, but at
higher concentrations it inhibits the reaction (Table 2).
These effects are probably related to the ability of GSH
to influence the Fe(III)/Fe(II) concentration ratio and
hence the reaction mediated by oxo-iron species. De-
creased yields of T-4,5-D are, in part, accounted for by
its reaction with GSH to give 7-S-Glu-4,5-DHT that is
readily autoxidized to 7-S-Glu-T-4,5-D (Scheme 1). How-
ever, that latter conjugate can be further oxidized to give
unknown products (Figure 5). Decreased yields of 4,4′-D
result, at least in part, from nucleophilic addition of GSH
on putative carbocation 6 to give 4-S-Glu-5-HT (Scheme
1), a reaction favored by high concentrations of the
tripeptide (Table 2).
taneous generation of O2•- and H2O2. The HO•-mediated
oxidation of 5-HT differs significantly from the xanthine/
XOD-mediated reaction forming 5-HEO, 5,6-DHT, and
8 (Scheme 2) as major reaction products (35). Although
T-4,5-D is formed in the HO•-mediated oxidation of 5-HT,
it reacts rapidly with the carbanion of 5-HEO to give 8,
and hence the dione does not appear as a free product
(35). In the presence of free GSH, a major constituent
of nerve terminals and axons (52), both the O2•-- and HO•-
mediated oxidations of 5-HT form 7-S-Glu-T-4,5-D as a
result of reaction of the tripeptide with T-4,5-D. How-
ever, the xanthine/XOD-mediated reaction also forms
4-S-Glu-5-HT, a product not observed in the HO•-medi-
ated oxidation of 5-HT. In principle, therefore, 5-HEO
and 5,6-DHT should be formed in regions of the brain
innervated by serotonergic neurons under pathological
conditions where HO• is generated. While there are
reports that 5,6-DHT is formed in rat brain following a
neurotoxic dose of MA and has, therefore, been proposed
to mediate serotonergic terminal degeneration (30), ef-
forts to confirm this or to detect 5-HEO or 6-OHDA have
been unsuccessful (20, 32-34). The protection provided
by elevated levels of cytoplasmic Cu-Zn SOD against the
The most important conclusion to be drawn from this
investigation is that oxidation of 5-HT in aqueous solu-
tion at pH 7.4 by the xanthine/XOD system generates
two major products, T-4,5-D and 4,4′-D. T-4,5-D is
•-
primarily formed by an oxidation reaction driven by O2
.
By contrast 4,4′-D is primarily formed by oxidation of
5-HT by an unknown oxo-iron species present in very low
concentrations and dependent for its formation on simul-

648 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
Wrona and Dryhurst
Sch em e 2
neurodegeneration evoked by MA (4) and ischemia-
MA (21, 28), ischemia-reperfusion (25-27), and Glu-
mediated oxidative toxicity (29). It is also of possible
relevance that blood platelets, the major cellular struc-
ture within an acute cerebral thrombus (58), store high
concentrations of 5-HT (59). Furthermore, not only is
the integrity of the blood-brain barrier (BBB) disrupted
at the site of cerebral ischemia and becomes permeable
to biogenic amines (60), but also platelets secrete a
number of substances at this site. These include 5-HT
and a factor that stimulates glial cell proliferation (61,
62) and activation during postischemic injury (63) with
reperfusion (13) points to O2 rather than HO• playing
•-
a role in the neurotoxic mechanisms. Both an experi-
mentally induced ischemic insult (7-9) and MA (16-18)
evoke a massive release of 5-HT, other biogenic amines,
Glu, and Asp and generation of O2•-, probably in the
•-
neuronal cytoplasm (4, 13). Such O2 might be gener-
ated by interaction of released Glu and Asp (10) with
NMDA receptors (16, 17) or by activation of the xanthine/
XOD system (14, 15). Subsequent reuptake of 5-HT, DA,
and NE into their parent and indeed other neurons (29)
would therefore necessarily expose elevated cytoplasmic
•-
resultant O2 generation (64). Thus, at the site of
•-
levels of these neurotransmitters to O2 and its dismu-
ischemic injury, 5-HT is released into brain parenchyma
from both sides of the BBB and activated microglia and
intraneuronal mechanisms described earlier generate
O2•-, i.e., conditions that should oxidize 5-HT to T-4,5-D
and 4,4′-D. Indeed, recent studies have demonstrated
that activated macrophages and brain microglia oxidize
5-HT to 4,4′-D (65, 66) which strongly suggests that the
O2•-/H2O2/oxo-iron system is the oxidant and, more
importantly, that the putative neurotoxin T-4,5-D should
also be a major reaction product. Because of its reactiv-
ity, however, T-4,5-D will undoubtedly be very difficult
to detect in the free state in brain tissue. However, the
facile reaction of T-4,5-D with the sulfhydryl residues of
proteins and GSH might indicate that the putative
neurotoxic properties of this dione might be related to
this property. Indeed, modification of the sulfhydryl
residues of brain guanine nucleotide-binding regulatory
proteins has been proposed to underlie the serotonergic
neurotoxicity of T-4,5-D (67). Interestingly, the activities
of phospholipase C (68), the R-ketoglutarate dehydroge-
nase complex (69), and protein kinase C (70), enzymes
with vulnerable sulfhydryl residues, have all been found
to be significantly reduced in AD brains.
In summary, the xanthine/XOD system generates
T-4,5-D, by oxidation of 5-HT in a reaction mediated by
O2•-, and 4,4′-D, in a reaction that appears to be mediated
catalytically by an ill-defined oxo-iron species. Strong
evidence implicates 5-HT (and other biogenic amine
neurotransmitters) and O2•- in pathological mechanisms
that underlie the neurodegeneration caused by MA,
ischemia-reperfusion, and other neurodegenerative brain
disorders. Accordingly, it is conceivable that T-4,5-D and
perhaps other products of the O2•-/H2O2/oxo-iron oxida-
tion of 5-HT might be endotoxins that contribute to the
neurodegeneration in these brain disorders.
tation product H2O2. Furthermore, the normally very low
cytoplasmic levels of low-molecular-weight iron would be
augmented by its mobilization from ferritin by O2•- (53).
Thus, both MA and ischemia-reperfusion might result in
intraneuronal conditions in which O2•-, H2O2, and trace
levels of oxo-iron species are simultaneously present. This
raises the possibility that intraneuronal oxidation of
elevated cytoplasmic levels of the biogenic amine neu-
rotransmitters by the O2•-/H2O2/oxo-iron system might
generate significant levels of endotoxic metabolites that
contribute to neurodegenerative processes. Preliminary
experiments have demonstrated that all the biogenic
amines are oxidized by the xanthine/XOD system, al-
though 5-HT is the most vulnerable to this reaction. To
illustrate, incubation of xanthine (200 µM) and XOD
(0.087 unit mL-1) in pH 7.4 phosphate buffer with these
neurotransmitters (200 µM) for 10 min resulted in the
oxidation of 138 ( 5, 44 ( 4, and 37 ( 5 nmol of 5-HT,
DA, and NE, respectively.
Several reports suggest T-4,5-D might be a serotonergic
neurotoxin and that unusual unidentified oxidized forms
of 5-HT are present in the cerebrospinal fluid of AD
patients (54-56). Furthermore, an in vitro oxidation
product of DA speculated to form in the cytoplasm of
dopaminergic cell bodies in the substantia nigra of PD
patients has recently been demonstrated to evoke ir-
reversible inhibition of mitochondrial respiration at the
complex I level (57). Thus, the massive release of 5-HT,
DA, and NE followed by subsequent reuptake into their
parent and other proximate neurons and cytoplasmic
oxidation to endotoxic metabolites might explain why
monoamine uptake inhibitors (23, 24) or prior manipula-
tions that deplete these neurotransmitters provide pro-
tection against the neurodegenerative consequences of

Oxidation of Serotonin by Superoxide Radical
Chem. Res. Toxicol., Vol. 11, No. 6, 1998 649
(22) De Vito, M. J ., and Wagner, G. C. (1989) Methamphetamine-
induced neuronal damage: A possible role for free radicals.
Neuropharmacology 28, 1145-1150.
Ack n ow led gm en t. This work was supported by a
grant from the National Institutes of Health (GM-32367)
and additional support provided by the Vice President
for Research and Research Council at the University of
Oklahoma.
(23) Schmidt, C. J ., and Gibb, J . W. (1985) Role of the dopamine uptake
carrier in the neurochemical response to methamphetamine:
Effects of amfonelic acid. Eur. J . Pharmacol. 109, 73-80.
(24) Ricaurte, G. A., Fuller, R. W., Perry, K. W., Seiden, L. S., and
Schuster, C. R. (1983) Fluoxetine increases long-lasting neostri-
atal dopamine depletion after administration of d-methamphet-
amine and d-amphetamine. Neuropharmacology 22, 1165-1169.
(25) Busto, R., Harik, S. I., Yoshida, S., Scheinberg, P., and Ginsberg,
M. D. (1985) Cerebral norepinephrine depletion enhances recovery
after brain ischemia. Ann. Neurol. 18, 329-336.
(26) Weinberger, J ., Nieves-Rosa, J ., and Cohen, G. (1985) Nerve
terminal damage in cerebral ischemia: protective effects of alpha-
methyl-para-tyrosine. Stroke 16, 864-870.
(27) Globus, M. Y.-T., Ginsberg, M. D., Dietrich, W. D., Busto, R., and
Scheinberg, P. (1987) Substantia nigra lesion protects against
ischemic damage to the striatum. Neurosci. Lett. 80, 251-256.
(28) Axt, K., Commins, D. L., Vosmer, G., and Seiden, L. S. (1990)
R-Methyl-p-tyrosine pretreatment partially prevents metham-
phetamine-induced neurotoxin formation. Brain Res. 515, 269-
276.
(29) Maher, P., and Davis, J . B. (1996) The role of monoamine
metabolism in oxidative glutamate toxicity. J . Neurosci. 16, 6394-
6401.
(30) Commins, D. L., Axt, K. J ., and Seiden, L.-S. (1987) 5,6-
Dihydroxytryptamine, a serotonergic neurotoxin, is formed en-
dogenously in the rat brain. Brain Res. 403, 7-14.
(31) Seiden, L. S., and Vosmer, G. (1984) Formation of 6-hydroxy-
dopamine in caudate nucleus of the rat brain after a single large
dose of methylamphetamine. Pharmacol. Biochem. Behav. 21, 29-
31.
(32) Rollema, H., De Vries, J . B., Westerink, B. H. C., Van Putten, F.
M., and Horn, A. S. (1986) Failure to detect 6-hydroxydopamine
in rat striatum after the dopamine releasing drugs dexamphet-
amine, methylamphetamine and MPTP. Eur. J . Pharmacol. 132,
65-69.
(33) Karoum, F., Chrapusta, S. J ., Egan, M. F., and Wyatt, R. J . (1993)
Absence of 6-hydroxydopamine in the rat brain after treatment
with stimulants and other dopaminergic agents: a mass frag-
mentographic study. J . Neurochem. 61, 1369-1375.
(34) Yang, Z., Wrona, M. Z., and Dryhurst, G. (1997) 5-Hydroxy-3-
(ethylamino) -2-oxindole is not formed in rat brain following a
neurotoxic dose of methamphetamine: Evidence that metham-
phetamine does not induce hydroxyl radical-mediated oxidation
of serotonin. J . Neurochem. 68, 1929-1941.
(35) Wrona, M. Z., Yang, Z., McAdams, M., O’Connor-Coates, S., and
Dryhurst, G. (1995) Hydroxyl radical-mediated oxidation of
serotonin: Potential insights into the neurotoxicity of metham-
phetamine. J . Neurochem. 64, 1390-1400.
(36) Slivka, A., and Cohen, G. (1985) Hydroxyl radical attack on
dopamine. J . Biol. Chem. 260, 15466-15472.
Refer en ces
(1) Markesbery, W. R. (1997) Oxidative stress hypothesis in Alzhe-
imer’s disease. Free Radical Biol. Med. 23, 134-137.
(2) J enner, P., Dexter, D. T., Sian, J ., Schapira, A. H. V., and Marsden
C. D. (1992) Oxidative stress as a cause of nigral cell death in
Parkinson’s disease and incidental Lewy body disease. Ann.
Neurol. 32, S83-S87.
(3) Oliver, C. N., Starke-Reed, P. E., Stadtman, E. R., Liu, G. J .,
Carney, J . M., and Floyd, R. A. (1990) Oxidative damage to brain
proteins, loss of glutamine synthetase activity, and production
of free radicals during ischemia-reperfusion-induced injury in
gerbil brain. Proc. Natl. Acad. Sci. U.S.A. 87, 5144-5147.
(4) Hirata, H., Ladenheim, B., Rothman, R. B., Epstein, C., and
Cadet, J . L. (1995) Methamphetamine-induced serotonin neuro-
toxicity is mediated by superoxide radicals. Brain Res. 677, 345-
347.
(5) Simonian, N. A., and Coyle, J . T. (1996) Oxidative stress in
neurodegenerative diseases. Annu. Rev. Pharmacol. Toxicol. 36,
83-106.
(6) Halliwell, B. (1992) Oxygen radicals as key mediators in neuro-
logical disease: Fact or fiction? Ann. Neurol. 32, S10-S15.
(7) Globus, M. Y.-T., Wester, P., Busto, R., and Dietrich, W. D. (1992)
Ischemia-induced extracellular release of serotonin plays a role
in CA1 neuronal cell death in rats. Stroke 23, 1595-1601.
(8) Globus, M. Y.-T., Busto, R., Dietrich, W. D., Martinez, E., Valde´s,
I., and Ginsberg, M. D. (1989) Direct evidence for acute and
massive norepinephrine release in hippocampus during transient
ischemia. J . Cereb. Blood Flow Metab. 9, 842-846.
(9) Phebus, L. A., and Clemens, J . A. (1989) Effects of transient global
cerebral ischemia on striatal extracellular dopamine, serotonin
and their metabolites. Life Sci. 44, 1335-1342.
(10) Benveniste, H., Drejer, J ., Schouboe, A., and Diemer, N. H. (1984)
Elevation of extracellular concentrations of glutamate and as-
partate in rat hippocampus during transient cerebral ischemia
monitored by intracerebral microdialysis. J . Neurochem. 43,
1369-1374.
(11) Weinberger, J ., Cohen, G., and Nieves-Rosa, J . (1983) Nerve
terminal damage in cerebral ischemia: Greater susceptibility of
catecholamine nerve terminals relative to serotonin nerve ter-
minals. Stroke 14, 986-989.
(12) Sakamoto, A., Ohnishi, S. T., Ohnishi, T., and Ogawa, R. (1991)
Relationship between free radical production and lipid peroxida-
tion during ischemia-reperfusion injury in rat brain. Brain Res.
554, 186-192.
(13) Yang, G. Y., Chan, P. H., Chen, J ., Carlson, E., Chen, S. F.,
Weinstein, P., Epstein, C. J ., and Kamii, H. (1994) Human
copper-zinc superoxide dismutase transgenic mice are highly
resistant to reperfusion injury after focal cerebral ischemia. Stroke
25, 165-170.
(14) Onodera, H., Iijima, K., and Kogure, K. (1986) Mononucleotide
metabolism in the rat brain after transient ischemia. J . Neuro-
chem. 46, 1704-1710.
(15) Kinuta, Y., Kimura, M., Itokawa, Y., Ishikawa, M., and Kikuchi,
H. (1989) Changes in xanthine oxidase in ischemic rat brain. J .
Neurosurg. 71, 417-420.
(16) Choi, D. W. (1988) Glutamate neurotoxicity and diseases of the
nervous sytem. Neuron 1, 623-634.
(37) Singh, S., Wrona, M. Z., and Dryhurst, G. (1992) Synthesis and
reactivity of the putative neurotoxin tryptamine-4,5-dione. Bioorg.
Chem. 20, 189-203.
(38) Wong, K.-S., Goyal, R. N., Wrona, M. Z., Blank, C. L., and
Dryhurst, G. (1993) 7-S-Glutathionyl-tryptamine-4,5-dione:
A
possible aberrant metabolite of serotonin. Biochem. Pharmacol.
46, 1637-1652.
(39) Wrona, M. Z., Singh, S., and Dryhurst, G. (1995) Influence of
glutathione on the electrochemical and enzymatic oxidation of
serotonin. J . Electroanal. Chem. 382, 41-51.
(40) Wrona, M. Z., and Dryhurst, G. (1990) Electrochemical oxidation
of 5-hydroxytryptamine in aqueous solution at physiological pH.
Bioorg. Chem. 18, 291-317.
(41) Halliwell, B., and Gutteridge, J . M. C. (1985) Free Radical in
Biology and Medicine, Clarendon Press, Oxford.
(42) Kappusamy, P., and Zweier, J . L. (1989) Characterization of free
radical generation by xanthine oxidase. J . Biol. Chem. 264, 9880-
9884.
(43) Minotti, G., and Aust, S. D. (1989) The role of iron in oxygen
radical mediated lipid peroxidation. Chem.-Biol. Interact. 71,
1-19.
(44) Minotti, G., and Aust, S. D. (1992) Redox cycling of iron and lipid
peroxidation. Lipids 27, 219-226.
(45) Davies, M. J ., Donkor, R., Dunster, C. A., Gee, C. A., J onas, S.,
and Willson, R. L. (1987) Desferrioxamine (desferal) and super-
oxide free radicals: Formation of an enzyme-damaging nitroxide.
Biochem. J . 246, 725-729.
(17) Lafon-Cazal, M., Pletri, S., Calcusi, M., and Bockaert J . (1993)
NMDA-dependent superoxide production and neurotoxicity. Na-
ture 364, 535-537.
(18) Bowyer, J . F., and Holson, R. R. (1995) Methamphetamine and
amphetamine neurotoxicity. In Handbook of Neurotoxicology
(Chang, L. W., and Dyer, R. S., Eds.) pp 845-870, Marcel Dekker,
New York.
(19) Gibb, J . W., Hanson, G. R., and J ohnson, M. (1994) Neurochemical
mechanisms of toxicity. In Amphetamine and Its Analogues:
Psychopharmacology, Toxicology and Abuse (Cho, A. K., and
Segal, D. S., Eds.) pp 269-295, Academic Press, New York.
(20) Seiden, L. S., and Sabol, K. E. (1995) Neurotoxicity of metham-
phetamine-related drugs and cocaine. In Handbook of Neurotoxi-
cology (Chang, L. W., and Dyer, R. S., Eds.) pp 825-843, Marcel
Dekker, New York.
(21) Commins, D. L., and Seiden, L.-S. (1986) R-Methyltyrosine blocks
methamphetamine-induced degeneration in the rat somatosen-
sory cortex. Brain Res. 365, 15-20.
(46) Li, T., and Vallee, B. L. (1965) Reactivity and function of
sulfhydryl groups in horse liver alcohol dehydrogenase. Biochem-
istry 4, 1195-1202.

650 Chem. Res. Toxicol., Vol. 11, No. 6, 1998
Wrona and Dryhurst
(47) Nanni, E. J ., Stallings, M. D., and Sawyer, D. T. (1980) Does
superoxide ion oxidize catechol, R-tocopherol, and ascorbic acid
by direct electron transfer? J . Am. Chem. Soc. 102, 4481-4485.
(48) Sawyer, D. T., and Valentine, J . S. (1981) How super is super-
oxide? Acc. Chem. Res. 14, 393-400.
(49) Sinhababu, A. K., and Borchardt, R. T. (1985) Mechanism and
products of autoxidation of 5,7-dihydroxytryptamine. J . Am.
Chem. Soc. 107, 7618-7627.
(60) Hervonen, H., Steinwall, O., Spatz, M., and Klatzo, I. (1980)
Behavior of the blood-brain barrier towards biogenic amines in
experimental cerebral ischemia. Adv. Exp. Med. Biol. 131, 295-
305.
(61) J oseph, R., Tsering, C., Grunfeld, S., and Welch, K. M. A. (1992)
Platelet secretory products may contribute to neuronal injury.
Stroke 22, 1448-1451.
(62) Besnard, F., Perraud, F., Sensenbrenner, M., and Labourdette,
G. (1987) Platelet-derived growth factor is a mitogen for glial but
not for neuronal rat brain cells in vitro. Neurosci. Lett. 73, 287-
292.
(50) Tabatabaie, T., Wrona, M. Z., and Dryhurst, G. (1990) Autoxi-
dation of the serotonergic neurotoxin 5,7-dihydroxytryptamine.
J . Med. Chem. 33, 667-672.
(51) Afanas' ev, I. B. (1989) Superoxide Ion: Chemistry and Biological
Implications, Vol. I, pp 202-209, CRC Press, Boca Raton, FL, pp
202-209.
(52) Slivka, A., Mytilineou, C. A., and Cohen, G. (1987) Histochemical
evaluation of glutathione in brain. Brain Res. 409, 275-284.
(53) Halliwell, B. (1992) Reactive oxygen species and the central
nervous system. J . Neurochem. 59, 1609-1623.
(54) Volicer, L., Chen, J .-C., Crino, P. B., Vogt, B. A., Fishman, J .,
Rubins, J ., Schnepper, P. W., and Wolfe, N. (1989) Neurotoxic
properties of a serotonin oxidation product: Possible role in
Alzheimer’s disease. In Alzheimer’s Disease and Related Disorders
(Iqbal, K., Wisniewski, H. M., and Winblad, B., Eds.) pp 453-
465, A. R. Liss, Inc., New York.
(55) Crino, P. B., Vogt, B. A., Chen, J .-C., and Volicer, L. (1989)
Neurotoxic effects of partially oxidized serotonin: Tryptamine-
4,5-dione. Brain Res. 504, 247-257.
(56) Chen, J .-C., Schnepper, P. W., To, A., and Volicer, L. (1992)
Neurochemical changes in the rat brain after intraventricular
administration of tryptamine-4,5-dione. Neuropharmacology 31,
215-219.
(57) Li, H., and Dryhurst, G. (1997) Irreversible inhibition of mito-
chondrial complex I by 7-(2-aminoethyl)-3,4-dihydro-5-hydroxy-
2H-1,4-benzothiazine-3-carboxlic acid (DHBT-1) : A putative
nigral endotoxin of relevance to Parkinson’s disease. J . Neuro-
chem. 69, 1530-1541.
(63) Banati, R. B., Schubert, P., Rothe, G., Gehrmann, J ., Rudolphi,
K., Valet, G., and Kreutzberg, G. W. (1994) Modulation of
intracellular formation of reactive oxygen intermediates in peri-
toneal macrophages and microglia/brain macrophages by pro-
tentofylline. J . Cereb. Blood Flow Metab. 14, 145-149.
(64) Colton, C., and Gilbert, D. (1987) Production of superoxide anions
by a CNS macrophage, the microglia. FEBS Lett. 223, 284-288.
(65) Schuff-Werner, P., Splettsto¨sser, W., Schmidt, F., and Huether,
G. (1995) Serotonin acts as a radical scavenger and is oxidized to
a dimer during the respiratory burst of human mononuclear and
polymorphonuclear phagocytes. Eur. J . Clin. Invest. 25, 477-484.
(66) Huether, G., Fettko¨tter, I., Keilhoff, G., and Wolf, G. (1997)
Serotonin acts as a radical scavenger and is oxidized to a dimer
during the respiratory burst of activated microglia. J . Neurochem.
69, 2096-2101.
(67) Fishman, J . B., J ordon, J . B., Chen, J .-C., Dickey, B. F., and
Volicer, L. (1991) Modification of brain guanine nucleotide-binding
regulatory proteins by tryptamine-4,5-dione, a neurotoxic deriva-
tive of serotonin. J . Neurochem. 56, 1851-1854.
(68) Matsushima, H., Shimohama, S., Fujimoto, S., Takenawa, T., and
Kimura, J . (1995) Changes in platelet phopholipase C protein
level in Alzheimer’s disease. Neurobiol. Aging 16, 895-899.
(69) Mastrogiacomo, F., Bergeron, C., and Kish, S. J . (1993) Brain
R-ketoglutarate dehydrogenase complex activity in Alzheimer’s
disease. J . Neurochem. 61, 2007-2014.
(58) J oseph, R., Welch, K. M. A., D′Andrea, G., and Riddle, J . M. (1989)
Evidence for the presence of red and white cells within “platelet”
aggregates formed in whole blood. Thromb. Res. 53, 485-491.
(59) Holmsen, H. (1987) Platelet secretion. In Thrombosis and Hemo-
stasis. 2nd ed. (Colman, R. W., Hirsch, J ., Marder, V. J ., and
Salzman, E. W., Eds.) pp 606-617, J . B. Lippincott Co., PA.
(70) Grammas, P., Moore, P., Botchlet, T., Hanson-Painton, O., Cooper,
D. R., Ball, M. J ., and Roher, A. (1995) Cerebral microvessels in
Alzheimer’s have reduced protein kinase C activity. Neurobiol.
Aging 16, 563-569.
TX970185W