ANTIOXIDANT AND PRO-OXIDANT NATURE OF CATECHOLAMINES Arno Garakhanian Siraki A thesis submitted in confomiity with the requirements for the degree of Master of Science Graduate Department of Pharmacology University of Toronto O Copyright by Amo Garakhanian Siraki 2000
109
Embed
ANTIOXIDANT AND PRO-OXIDANT NATURE OF CATECHOLAMINES€¦ · The author has granted a non- exclusive licence allowing the National Librv of Canada to reproduce, loan, distribute or
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ANTIOXIDANT AND PRO-OXIDANT NATURE OF CATECHOLAMINES
Arno Garakhanian Siraki
A thesis submitted in confomiity with the requirements for the degree of Master of Science
Graduate Department of Pharmacology University of Toronto
O Copyright by Amo Garakhanian Siraki 2000
National Library I*l .,,da Bibliothèque nationale du Canada
Acquisitions and Acquisitions et Bibliographie Services services bibliographiques
395 Wellington Street 395, rue Wellington ûttawaON K1AON4 OttawaON K1AON4 Canada Canada
Yaur iik Votre referena,
Our Ne Notre ?eletenCa
The author has granted a non- exclusive licence allowing the National L i b r v of Canada to reproduce, loan, distribute or sel1 copies of this thesis in microform, paper or electronic formats.
L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microficheffilm, de reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.
ANTIOXIDANT AND PRO-OXLDANT NATURE OF CATECHOLAMINES Master of Science, 2000 Arno Garakhanian S i f i Department of Pharmacology University of Toronto
ABSTRACT
This study compares the antioxidant versus pro-oxidant potential of catecholamines.
Firstly, catecholamines scavenged superoxide, and also prevented hypoxia-reoxygenation injury.
The femc complexes of catecholarnines were much more effective and cytoprotective. This
could prove useful in stroke therapy. Secondly, catecholamines were shown to mediate the
oxidation of ascorbate and NADH, which was directly related to their O-quinone half-life.
Glutathione prevented NADH oxidation, and glutathione-conjugates were formed, indicating that
O-quinones were the metabolites responsibie for the oxidation. Lastly, dopamine cytotoxicity
was potentiated by catalytic manganese(Iï) concentrations which resulted in toxic dopamine O-
quinone formation, offering insight into the Parkinson's-like disorder found in manganese
miners. Dopamine was metabolized by hepatocyte P450 peroxygenase to a cytotoxic product
(possibly dopamine O-quinone), utilizing H202 formed by monoamine oxidase, and possibly
relevant to idiopathic Parkinson's disease. In concIusion, the balance of catecholamines as anti-
or pro-oxidant wiii depend on dose and arnbient cellular conditions.
ACKNOWLEDGMENTS
1 wish to thank my parents, Thomas and Alvart, for their support in my acadernic endeavors,
without which this degree, or my previous one, would not have been possible. My family,
including my grandparents in Toronto, Arsham and Zaghik, my aunt Rima, and my siblings,
Arby and Anita, were supportive and understanding of my objectives. Aiso, the latter two didn't
get in my way, so 1 actually could get some work done.
1 hold the highest respect and admiration for m y supervisor, Dr. Peter J. O'Brien, who
gave me a chance to show my capabilities in a research environment. It's no secret that 1 didn't
achieve high marks in our systematic grading system, but Dr. O'Brien's guidance facilitated my
development in scientific research. 1 regard him as a wizard and myself as a wizard's apprentice,
in the tradition of leaming by example and leadership, and not simply by words, but by actions.
Furthemore, he kept his promise for the duration of rny M.Sc. as a part-time student-I did
finish on time, as we agreed in 1998. His support and undying enthusiasm is inspirational and
motivational, which facilitated progress in al1 aspects of lab activities. Indeed, it's with his
support and encouragement that 1 currentiy wish to pursue a Ph.D.-something I was opposed to
a year ago-for what 1 believe are the right reasons: the need to know and to let others know.
Dr. O'Brien, 1 salute you.
1 wish to thank my defense cornmittee, who accepted the responsibilities that 1 requested.
My interna1 appraiser, Dr. Denise Tomkins, came through on short notice in my time of need.
My extemal appraiser, Dr. Peter Pennefather, was available for consultation long before this
event. 1 could discuss topics related and unrelated to my research area. His insight and
ingenuity was remarkable and always of interest. Cheers! My thanks to Dr. Jose Nobrega, who
participated as the additionai voting member. Also thanks to Dr. Aiian Okey for volunteering to
iii
chair the defense. I thank everyone for their understanding and Bexibiiity. September is a
difficult month for both professors as well as M.Sc. students since the students want to avoid
paying tuition and professors are busy with gants. In this respect, I am really appreciative of the
efforts of my defense committee.
1 wish to mention the members of my lab whom I got dong with ... al1 of them. Itys quite
surprising that such a large and diverse group of people could get dong without making plans to
poison one another. Specifically, I'd like to thank Dr. Majid Moridani for his help and
suggestions in research and academic choices. My most sincere "whauap" to everyone in the
lab. Although these fine peopie are my coileagues, I also regard them as my friends. Speaking
of which, 1 wish to mention ail my fnends across the spectnim. Actually. 1 will not mention
them in verbatim. They know who they are, as do 1. 1 think a mere mention doesn't do justice to
such a bond. If one of you is reading this, and if you are indeed a true fnend, then you know
what I am thinking and what 1 mean, (and no, I'm not BS-ing).
My sincerest thanks to Ms. Angela Moy who hired me at GlaxoWellcome Canada, and
allowed a flexible working schedule with my studies as the fmt priority. By giving me a chance
to prove my capabilities, I became part of a cohesive working team that brought out my best
qualities. It goes without saying, that 1 have Dr. Hira Kazarians to thank for sponsoring me as a
summer snident at GW. Also, thanks to Mr. Naresh Persaud who allowed me to continue
balancing my studies with work. My heilo's to everyone in the QC & QA teams at GW.
Also 1 wish to thank Dr. Tigran V. Chalikian who offered his guidance and help at any
time. Even though our research areas were different, he offered guidance in research and career
objectives.
1 have made the decision to pursue a Ph.D. In case things don't go my way, however, 1
have the foilowing request: that this research be continued, particularly the section of dopamine
metabolic activation by P450 peroxygenase activity. It is my instinct, with the direct aid and
supervision of Dr. O'Brien. that this system is somehow involved in the etiology of idiopathic
Parkinson's disease. No one has thought of this. Therefore, if 1 cannot pursue it, 1 request that
the reader of this work take into account this uncharted territory and let people know about it.
Nothing is more noble than the search for knowledge and tmth.
"Ail good is knowledge. AU evil is ignorance" - Socrates.
TABLE OF CONTENTS
Abstract
Acknowledgements
Ab breviations
List of Tables
List of Figures
List of Schemes
List of Publications, Abstracts, and Posters
Page
. . 11
iii
viii
ix
X
xi
xii
Generd Introduction 1
Chapter 1 : Superoxide radical scavenging and attenuation of 13
hypoxia-reoxygenation injury by femc complexes in isolated rat hepatocytes.
Abstract
Introduction
Materials & Methods
Resul ts
Discussion
Chapter 2: Catecholamine O-quinones mediate ascorbic acid and
NADH oxidation, which is prevented by GSH: the relationship
between O-quinone stabiiity and catecholamine cyclization.
Abstract
Introduction
MateriaIs & Methods
R e d ts
Discussion
Chapter 3: Dopamine metabolic activation by P450 peroxygenase
activity versus manganese (II):
DA O-quinone as the mediator of cytotoxicity
Abstract
Introduction
Materials & Methods
Discussion
General Conclusions and Future Expenments
Re ferences
C.4
DA
DOPA
DOPAC
ErnA
EPI
GSH
GSSG
H202
HRP
H x
HVA
MAO
NAcDA
~ a . 1 0 ~
NE
NQO ROS
SOD
ozb XO
XTr
catec holamine
dopamine
3,4-dihydroxyphenyldanine
3,4-dihydroxyphen y lacetic acid
ethylenediaminetetraacetic acid
epinephnne
glutathione
glutathione disulfide (oxidized)
hydrogen peroxide
horse radish peroxidase
hypoxanthine
homovanillic acid
monoamine oxidase
N-ace tyldopamine
sodium periodate
norepinephnne
NAD(P)H:Quinone Oxidoreductase
reactive oxygen species
superoxide dismutase
superoxide
xanthine oxidase
2 , 3 - b i s [ 2 - m e t h o x y - 4 - n i t r o - 5 - s u l f o p h e n y p
viii
LIST OF TABLES
Table 1 . 1 ICso values for OzC scavenging activity of neurotransrnitters
and neurotransmitter-iron(m) complexes in the HX/XO system.
Table 1.3 Cytotoxicity of hepatocytes upon hypoxia-reoxygenation by
neurotransmitters and neurotransmitter: iron(iII) complexes.
Table 1.3 O2 uptake with 2: 1 neurotransmitter:metaI complexes.
Table 2.1 Cyclization rates and O-quinone half-life of catecholamines.
Table 2.2 Ascorbate CO-oxidation with Substrates Using WM202 &
Tyrosinase.
Table 2.3 NADH Oxidation In The Presence Of Ascorbate or GSH
by HRP/H20t.
Table 2.4 GSH depletion in microsornai preparation.
TabIe 3.1 IntracelluIar [GSH] and [GSSG] after incubation with DA or
tyramine.
LIST OF FIGURES
Figure 1. The Catecholamine Biosynthetic Cascade.
Figure II. Proposed Oxidative Pathways for DOPA and CAS.
Fig III. Structures of compounds used in this study.
Figure 2.1 Percent NADH Oxidized by Catechol(amine)s by
HRP/H202 and the Inhibitory Effect of Ascorbate or GSH.
Figure 2.2 Products Found by Mass Spectroscopy with NAcDA + Tyrosinase
or HRP/H202 + GSH.
Figure 3.1 Cornparison of Dopamine Cytotoxicity Cataiyzed by Different
Metals.
Figure 3.2 Cytoprotection Against DA:M~" by GSH, Ascorbate, and Xylitol.
acid ( W A ) however, showed no detectable Oz' scavenging activity. A marked increase in
SOD-mimicking ability was however found when the phenolic neurotransmitters were
cornplexed with femc iron. Serotonin showed only marginal activity, and HVA showed no
detectable OzC scavenging activity when complexed with femc ions. Femc iron alone showed
no detectable O; scavenging activity. The m k order for the OzC scavenging of the
neurotransmitter-~e'+ complexes was: a-methyiDOPA > NE > 3.4-dihydroxyphenylacetic acid >
EPI > DA > DOPA > 5-hydroxyindole acetic acid > tyrosine > serotonin. It is interesting to note
that some of the weakest Oc scavenging substrates became much more active when complexed
with iron. Any direct inhibition of XO was mled out since the iron complexes did not affect uric
acid production even at much higher concentrations.
As shown in Table 1.2, SOD, DA, and serotonin protected hepatocytes from hypoxia-
reoxygenation injury. Furthemiore, when the DA was complexed with femc iron, 40pM of the
D A : F ~ ~ + was as effective as 150pM DA in protecting the hepatocytes from hypoxia-
reoxygenation injury. DA was much more cytoprotective as a femc iron complex yet ferric iron
alone was not cytoprotective. A similar enhancement of the cytoprotective effectiveness of NE
and serotonin was also found when these catecholic/phenoLic neurotmsrnitters were complexed
with ferric iron.
The iron complexes formed a purple colour upon addition of FeC13,
by the addition of excess EDTA to reform DA with no evidence of
which was reversed
quinone formation.
Furthemore, very little oxygen uptake was observed over a 1-hour incubation period when 1 mM
D A - F ~ ~ + was incubated in contrat to DA-CU", (Table 3). The copper complexes however,
gradually fomed a deep yellow coloured product, which was not reversed by EDTA.
DISCUSSION
The results descnbed show that femc iron complexes of catecholic neurotransmitters
were much more effective at scavenging Oz' radicals than the uncomplexed neurotransmitter.
The femc iron catecholic complexes most potent at scavenging 02- radicals were a-
r n e t h y ~ 0 ~ ~ - ~ e 3 + (IC50 - ISpM), NE-F~)+ (& - 4pM), DOPAC-F~~+ (1Cso - 15pM), EPE
~ e ~ + (ICSo - 17CLM), and DA-F~)' (IC 50 - 19p.M). Serotonin-Fe '+ was the least potent 0;-
scavenger because hydroxyindoles have a much lower affinity for l?e3+ than catechols (79).
For the uncomplexed neurotransmittes, DA and NE were the most potent Oz*
scavengers (IcSa - 55p.M and - 1 IOpM, respectively). The Oz* scavenging ability of DA and
NE may be refiected by their low reduction potential, however this does not inàicate why
serotonin is not an equal, if not a better scavenger by virtue of its reduction potential (Table 1).
It is likely that the catechol moiety yields a more potent 0; scavenging activity, since serotonin
is an indoleamine containing a phenol group. Although the reduction potentials (E/mV) partly
correlate to 02 scavenging, pK. values did not show such a direct correlation (Table 1). The
les t effective scavenger was EPI whereas HVA, a methylated metaboiite of DOPAC, w u
ineffective probably because of its lack of a catechol moiety.
Hepatocytes in our hypoxia-reoxygenation model were significantly protected against
reoxygenation injury by the femc complexes. The DA-F~~ ' complex conferred the most potent
cytoprotection, equivalent to approximately 4 times DA alone. Interestingly, serotonin was more
potent in vitro in cornparison to the enzymatic 0;- scavenging systern, indicating that sorne
other mode of cytoprotection may be provided. XO inhibition can be ruled out since no
enzymatic interference was observed with the femc complexes. However, it has been reported
that DA-F~~ ' complex at 1mM can undergo autoxidation if 5mM cysteine is present (80).
However, we have found that DA-F~'+ was not cytotoxic to isolated hepatocytes even at
concentrations 20-fold higher than those used here (see Chapter 3, Fig. 3.1 for details).
Ischernic brain injury following stroke ensues over a penod of hours, during which a
cascade of cellular and biochemical events inevitably Ieads to destruction of brain tissue.
Synthetic SODkatalase mimetics such as saien-manganese complexes (known as EUK-8 and
EUK-134), showed substantial neuroprotective effects in a rat stroke model (4). Although the
involvement of ROS in ischemic brain injury is not the only mechanism of damage, it is one of
the primary elements. Thus, it would be interesting to investigate whether DA-iron is more
effective than DA at protecting dopaminergic neurons against OzC mediated reoxygenation
injury.
The relevant antioxidant contribution of the substrates tested may be significant in the
light of neurotransmitter concentrations found in various brain regions. For exarnple, the DA
concentration in a dopaminergic nerve terminal has been reported to be approximately 50mM.
although mostly stored in vesicles (81, 82). Furthemore, it has been reported that astrocyte
miiochondria sequester redox-active iron in nigral astroglia (83).
In summary, neurotransmitter-iron(III) complexes were shown to be potent ozk radical
scavengers, and DA-iron was shown to convey potent cytoprotection when hepatocytes were
challenged with hypoxia-reoxygenation injury. Our model could reflect an in vivo mechanism
since hypoxia-reoxygenation has been shown to be as damaging as ischemia-repemision to the
liver (68). Further research is required to determine whether catecholamine-iron complexes are
also cytoprotective in a stroke model and could have therapeutic potential.
22
Table 1.1 ICso values for 0 1 scavenging activity of neurotransmitters and neurotransrnitter- iron(m) complexes
Neuro transmi tt e r
Ligand
DA
DOPAC
HVA
Serotonin
5-Hydrox yindole- acetic acid
Tyrosine
DOPA
a-methylDOPA
NE
EPI
in the HXXO system.
O?* scavenging was assessed spectrophotornetncaUy by following formation of the reduction product of XTT at h = 470nm as descnbed above. * Ref. (84); ** Refs. (85), (74).
Table 1.2 Cytotoxicity of hepatocytes upon hypoxia-reoxygenation by neurotransmitters and iron(III) complexes.
--
Neurotransmîtter % Cytotoxicity (70 min)
Control
SOD (lûûU/mL)
serotonin ( 1SOp.M) 35.0 & 3.6*
serotonin (40p.M) 59.3 + 4.0
serotonin-~e)+ (40:20FLM) 48.3 k 3.5* Isolated hepatocytes (10' cellslml) were incubated under a hypoxic atmosphere of 5%C02/95%N2 with neurotransrnitter/ion complexes followed by reoxygenation at 70 minutes with 1 %02/5%COt/94%N2 as described above. Cytotoxicity represents the percentage of dead cells assessed by %an blue uptake at 70 min. Results are the means of three separate expenments (I SEM). * p < 0.05
Table 1.3 O2 uptake with 2: 1 neurotransrnitter:metal complexes.
DA
a-methylDOPA
NE
DOPAC
EPI 1
Neuro transmitter
1 I
Oxygen uptake was measured using a Clarke-type O2 electrode in a 2m.L charnber containing O.1M Tris-HCl buffer pH 7.4 and neurotransrnitter (ImM) at 20 OC. The reaction was initiated by the addition of either OSmM femc or cupnc metals. Concentrations shown are finai in a 2mL volume.
O2 uptake (nmol O f i n ) FeC13 CuSO4
Chapter 2
Ascorbic acid, glutathione, and NADH oxidation by catecholarnine o-quinones: the relationship between O-quinone stability and catecholamine cyclization.
ABSTRACT
The toxicity of various CAS has been suggested as contributing to neurodegenerative
diseases, adrenal carcinogenesis, or repemision injury. In this study, the ability of oxidized CAS,
specifically CA O-quinones, to oxidize and/or deplete ascorbic acid, GSH, and NADH. have been
compared. The half-life of their respective O-quinones, before forming the irreversible
cyclization products, aminochromes, have been measured. It was found that the various CAS
were less likely to oxidize ascorbate and NADH if they have a short O-quinone half-life as a
result of rapid cyclization. even though quinoid end-products were formed. The results of this
study suggest that CA O-quinones mediate the oxidation of ascorbate or NADH, and depletion of
GSH catalyzed by microsornesMADPH. Therefore, CA O-quinones c m be regarded as
Conditions: 250pM substrate was added to a 2mL quartz cuvette containing 50mM phosphate buffer pH 4.0. 250pM Nd04 was added Iast to initiate the reaction. The kinetics of the reaction were followed at the specific aminochrome wavelength (L) for each substrate. The O-quinone half-life was measured kineticaily between 390400 nm. *Not detectable.
N/ A 470 485 475 475 487
N/ A 11.7 27.5 71 24 1 690
39.6 13.2 5 -9 5.1 3.2 -*
Table 2.2 Ascorbate Co-oxidation with Substrates Using KRi?/H20 & Tyrosinase.
Substrate ( 1 0
NAcDA DOPAC NE DA a-methylDOPA DOE4 EPI Adrenochrome alone* Conditions: 0.1M Tris-HCl pH 7.4 with 2mM DETAPAC was added f i t to a 2 mL cuvette.
Rate of Ascorbate oxidaüon, k (mine1)
50pM of ascorbaie was then added and the absorbante was recorded. 10pM substrate and H202
W/H207 0.665
were added and 0.lpM HRP was added to initiate the reaction. For tyrosinase-mediated oxidation, 70U/mL was added to initiate the reaction. Ascorbate oxidation was followed at k266 nm.
Tyrosinase 1.1
* Adrenochrome was added in the absence of either enzyme or H2O2.
Figure 2.1 Percent NADH Oxidized by Catechol(amine)s by HRP/H202 and the Inhibitory Effect of Ascorbate or GSH. 50pM substrate and 50pM Hz02 was added to a 2 mL quartz cuvette containing 0.1M Tris-HC1 buffer pH 7.4 with 2mM DETAPAC. 50p.M of NADH, with or without 50pM ascorbate or GSH, was added followed by O.1pM HRP to initiate the reaction. NADH oxidation (solid bars), in the presence of ascorbate (diagonal bars) or GSH (clear bars) was measured at a fixed wavelength (h = 340nm). Bars represent the average percent of NADH
Table 23 NADH Oxidation In The f resence Of Ascorbate or GSH bv HRP/H,O?. Substrate (50p.M)
NAcDA DOPAC DA NE DOPA EPI
I - - NADH oxidized in 10 minutes
NADH 48.1 f 5.2 41.6 1: 3.8 21.2 f 1.5 12.4 + 0.9 1.9 f 0.2 ,*
Ascorbate, W H 23.1 f 1.2 31.8f 1.7 15.8 I 1.0 8.8 10.3 0.8 =t 0.04 -*
GSH, NADH 6.7 k 0.5 1.3 f 0.08 0.7 f 0.04 0.4 I 0.03 0.3 f 0.01 -*
50p.M substrate and 50pM H202 was added to a 2 rnL quartz cuvette containing 0.1M Tris-HC1 buffer pH 7.4 with 2mM DETAPAC. 50pM of NADH, with or without 50pM ascorbate or GSH, was added and the absorption was followed kinetically at h340nm. O . 1 p M HRP was added last to initiate the reaction. Results are the mean of three separate experiments (IS.D.). * Not detectable. ** Adrenochrorne was added in the absence of HRP/&û2.
NAcDA
HO
û-HO-NAcDA quinone (orlho or pan,)
NAcDA
quinone
GSH 308
O 1 O0 200 300 400 500 600
mhc
Figure 2.2 Products Found by Mass Spectroscopy with NAcDA + Tyrosinase or HRPM202 + GSH. Conditions: ImM of NAcDA was added to a 1.5mL en end off via1 contalning purified water. 1mM GSH was added before or afier initiating the reaction. 20 U/mL of tyrosinase or O. 1pM HRPllmM H202 were added to initiate the reaction. Formation of the 6-hydroxylated NAcDA products was only accomplished by adding 200pM NaOH just before injection into the mass spectrometer.
Table 2.4 GSH depletion in microsomal preparation
% GSH depletion (rnicrosomes)
NAcDA
15 minutes
+ pheny Iimidazole + SOD
30 minutes
DOPAC
+ pheny limidazole + SOD
+ pheny Iimidazole -c SOD
DOPA
+ phenylimidazole + SOD
-
-
-
-
O I O Conditions: Microsorne = lmg/mL, NADPH = lmM, GSH = 200ph4, phenylimidazole =
300pM, and test compound = ImM in a total volume of 1mL (0.1 M Tris-HCl ImM DETAPAC). Percentages are derived from a standard c w e for GSH over the concentration range used at 412nrn (TNB peak absorbance). Phenylirnidazole or SOD was preincubated 5 minutes before initiating the reaction.
EPI
+ phenyIimidrrzole + SOD
+ phenyIimidazoIe + SOD
DISCUSSION
Our findings show that the CA-mediated ascorbate and NADH oxidation catalyzed by
HRP/H202 or tyrosinase is fastest with catechol(amine)s that form stable O-quinones and
suggests that the O-quinone metabolites mediate ascorbate or NADH oxidation and GSH
depletion in microsomesMADPH. The order for ascorbate oxidation by HRP/H202 ( N A c D b
DOPAC>NE>Dba-methylDOPA>DOPA=EPI) was aimost identical to the order found for
CAS with microsomes~ADPH depleted GSH-with one exception-in a similar order, (EPb
NAcDPUDOPAC>DA>DOPA>NE). The stability of O-quinones at pH 4 (NAcDA>DOPAC>
DA>NE>DOPba-methyDOPA>EPI) seems to be inversely related to the cyciization rate
(EPba-methylDOPA>DOPA>NE>DA). EPI O-quinone was undetectable presumably because
of its very rapid cyclization rate. Our findings for CA cyclization is in agreement with a
previous study using DA, NE, and EPI at an alkaline pH using periodate (17). NAcDA O-
quinone or DOPAC were the most stable O-quinones as they did not cyclize, and were most
effective at oxidizing ascorbate or NADH.
The cyclization property of the CAS causes them to form aminochrome end-products with
characteristic absorption spectra. As described in the General introduction (see Fig. LI), in order
to form these end-products, the CA must h t f o m its corresponding O-quinone. Our findings
show a clear inverse order between the stability of the O-quinone versus the cyclization rate
(aminochrome formation). If the O-quinone of a given CA has a long half-life, it will cyclize to
its minochrome more slowly than a CA O-quinone with a relatively short half-Me. Since this
cyclization occurs readily at physiological pH, an acidic buffer was used to study this
relationship for the CAS. NAcDA, therefore, was useful as a mode1 of a stable DA O-quinone, as
it did not form an aminochrome because it lacks the amino terminus present in CAS. Therefore,
in Eqn. 3, NAcDA O-quinone could represent a long-lived CA O-quinone intermediate, without
the event of forming its aminochrome end-product:
[O1 ~ a s t at T p~ CA CA oquinone b Arninoc hrome
Slow at 4 pH
Ascorbate is an important antioxidant in the brain that has a relatively high concentration.
In both rat and human striatum, ascorbate concentration is 1 to 2 mM (109, 1 IO), and
approximately 10 rnM in isolated nerve terminais (1 11). In order to determine what reactive CA
metabolite is involved in oxidizing ascorbate, Le., CA O-quinone or aminochrome, two enzyme
systems were used. HRP is a good mode1 peroxidase for endogenous prostaglandin H synthase.
another peroxidase (100). HRP catdyzes the 1-electron oxidation of a catechol substrate to a
serni-quinone intermediate. This semiquinone then disproportionates spontaneously to form an
O-quinone and OzC, as depicted in Eqn. 4:
HzOz + CA CA semi-quinone + H20 O2 + CA O-quinone + 4' (4)
The other enzyme used to oxidize CAS was tyrosinase. The latter uses O? as a cofactor and
catalyzes the 2 electron oxidation of a catecholamine to an O-quinone (Eqn. S), without the semi-
quinone intermediate:
Recent evidence suggests that tyrosinase could be expressed in the brain (1 12, 1 13), but others
disagree ( 1 8).
Dehydroascorbate r-
Asc
Scheme 2.1 Biochemical pathways involved in the oxidation of catechols to O-quinones and the mechanisms of ascorbate or NADH oxidation and GSH conjugation. (See text for details).
In order to determine whether CA O-quinone or aminochrome was responsible for the oxidation
of ascorbate, it was necessary to correlate O-quinone stability or cyclization rate with the
ascorbate oxidation rate. If the cyclization rate was to directly conelate with ascorbate oxidation
rate, this would indicate that the less stable a CA O-quinone is, the more tikely it is to oxidize
ascorbate (rnost probably due to the formation of its aminochrome endproduct). On the contrary,
we found that the O-quinone half-iife was directly related to the oxidation rate of ascorbate (the
longer O-quinone half-life would yield the faster rate of ascorbate oxidation). This implies that
the aminochrome end-product is not responsible for the oxidation of ascorbate. Indeed, it was
shown that adrenochrome, the EPI oxidation end-product, did not oxidize ascorbate, even though
it contains a quinoid structure. However, in another reaction system, 40pM adrenochrome and
20mM ascorbate showed that under these conditions, adrenochrome reduction (therefore
ascorbate oxidation) occurred ( 1 14). Therefore, it cannot be categorically concluded that
adrenochrome doesn't oxidize ascorbate; it simply occun at much higher concentration than for
CA oquinones.
Since we used two enzymatic systems to f ~ s t oxidize the catechol(amine) which then
oxidized ascorbate, the factor of substrate specificity was unavoidable. In the HRPR1202 system
for ascorbate oxidation, NE had a faster rate of ascorbate oxidation than DA. If O-quinone half-
Me is to correlate with ascorbate oxidation, then DA should have a higher rate. NE is oxidized
by HRP almost 5 fold faster than DA (98). The next discrepancy in this series was that
a-methylDOPA oxidized ascorbate faster than DOPA. HRP also catalyzed the oxidation of DA
five-fold higher faster than DOPA. One study showed that a charged amino group couid inhibit
access to the peroxidase heme group, (1 15). Possibly, the presence of a methyl group could give
a-methylDOPA easier access to the peroxidase active site. In the tyrosinase series, DOPA
oxidized ascorbate better than NE. Since DOPA cyclizes faster than NE, it should oxidize
ascorbate at a slower rate than NE. In fact, DOPA is about a 4 fold better tyrosinase substrate
than NE, which could explain this observation (98). Furthemore, the potency of NAcDA and
DOPAC in CO-oxidizing ascorbate could be explained by the presence of an acetyl group on the
alkyl c h a h N-acetyltyrosine was a 19 times better HRP substrate than tyrosine (1 15).
With respect to ascorbate oxidation, CA cyclization may be considered as a pathway ihat
would prevent ascorbate oxidation. However, the aminochromes themselves may participate in
other biochemicaily relevant pathways not studied here. NADH is a vital cofactor in various
enzyrnatic reactions. NAD(P)H:quinone oxidoreductase, which reduces quinones back to their
parent dihydroxy molecule, uses NADH preferentially as a cofactor (1 16). In fact, semi-
dehydroascorbate reductase. responsible for the reduction of the ascorbyl radical, also uses
NADH as an electron donor (117). and NADH dehydrogenase is a key component of the
mitochondrial respiratory chah that utilizes NADH ( 1 18). Interestingly, the quinones formed in
this study affect both ascorbate and NADH. NADH oxidation was studied to determine if a
sirnilar trend would be seen as with ascorbate oxidation, i.e., direct involvement of the o-
quinone. NADH undergoes a two-electron oxidation by quinones with a 1 : 1 stoichiometry (Eqn.
6) (108), therefore without forming a serni-quinone intermediate:
CA o-quinone + NADH + H) -+ CA + NAD+ (0)
The catalytic activity of the various catechol(amine)s for oxidiYng NADH in the W&02
system was identical to the stability of their respective oquinones with the most rapid oxidation
of NADH occuming with the most stable O-quinones. By definition, therefore, the CAS with the
fastest cyclization rate would be the least effective at oxidizing NADH. Adrenochrome again
did not oxidize NADH Qust as with ascorbate above). In both ascorbate and NADH oxidation,
the lack of evidence for the oxidation of either compound by adrenochrome most likely reflects
its low redox potential (E = -0.253) compared with its precursor, EPI O-quinone (E = 0.38) (1 19).
Furthemore, both ascorbate and GSH prevented the oxidation of NADH, implicating the o-
quinone as the oxidiYng agent. It must be noted, however, that GSH provided substantially
more prevention of NADH oxidation than ascorbate. This probably reflects the different
mechanism of quinone reduction as depicted in Scheme 2.1. The "futile cycle" could occur with
ascorbate because of its two-step reduction of the O-quinone. Since quinones readily form GSH
conjugates (120), we have identified, for the first time, the NAcDA-GSH conjugate.
HRP and tyrosinase both produced GSH conjugates of NAcDA that were identified by
mass spectroscopy. Although NAcDA has only been found in the periphery (23), this result
could be extrapolated for the formation of the potent neurotoxin. 6-hydroxy-DA, since 6-
hydroxylated NAcDA was identified. Our finding could mean that the formation of 6-hydroxy-
DA could occur by two steps: an initial oxidation reaction foliowed by nucleophilic addition of
hydroxyl (water) to the 6-position of the benzene ring. The formation of 6-hydroxy-NAcDA was
demonstrated biochemically in the W/H202 system. however excess HzOl was required to
carry out the reaction (106). Although the 6-hydroxy NAcDA products were formed at an
aikaline pH, the 6-hydroxylation also occured at a physiologicai pH (106). Although
adrenochrome was shown to be relatively non-reactive in our system, one group studied which
glutathione-S-transferases would best conjugate GSH with the aminochromes themselves, but
did not study GSH conjugation to the correspondhg O-quinone precursors (93). It was shown
that the specific isofom of glutathione-S-transferase (Ml-1) that cataiyzed GSH conjugate
formation ffom adrenochrome was different fiom the isoforms that catalyzed GSH conjugate
fonnation with other aminochromes (93). Since this specific isofom is expressed in the rat Liver,
it could account for the GSH depletion observed by EPI (121). In our study, however, the GSH
conjugate of NAcDA was formed in the absence of glutathione-S-transferase.
Our assay for GSH depletion by CAS catalyzed by microsomesMADPH was similar to
the CA order for ascorbate oxidation catalyzed by tyrosinase system. Since tyrosinase produced
an O-quinone metabolite directly, it is most probable that this same O-quinone metabolite was
formed and caused the observed GSH depletion. Al1 catechol(arnine)s depleted GSH in an order
directly related to their O-quinone half-lives. One important exception. however, was EPI which
caused the most GSH depletion, although it was the least effective at oxidizing ascorbate. Since
EPI is the fastest CA to form its aminochrome (adrenochrome), and has a very short-lived
o-quinone, it is unlikely that EPI would deplete GSH catalyzed by microsornes via its EPI O-
quinone metabolite. It is therefore likely, that microsorna1 glutathione-S-transferase (122)
catalyzed adrenochrome:GSH conjugate formation.
The role of CYP 2El in generating O?' (Eqn. 7) that catalyzes the autoxidation of the
e' RH-(F~*C)-O~ --+ R H - ( F ~ ~ - O Z ' - ~ RH-(FelC)-02" + 02' 2, CA O-quinone (7)
2 8 + 202' SoD O2 + HZ02 (8)
CAS to form the CA O-quinone (Eqn. 8) was demonstrated since phenylirnidazole (CYP 2E1
inhibitor) and SOD both completely prevented the depletion of GSH by al1 compounds tested.
Oxidation of CAS have been shown to be mediated by 02- generated by P-450 or XO (43, 97,
123, 124). It must be noted, however, that OzC c m disproportionate to H202, which can also
oxidize GSH, but at a slower rate if not accompanied by GSH peroxidase catalysis. The
prevention of GSH depletion with SOD, therefore, niles out GSH oxidation by HzOz.
In addition to other cpotoxic events, we have established that catechol(amine) o-
quinones have the potential to oxidize ascorbate and NADH, as well as deplete GSH in a
microsorne/N ADPH catal yzed reac tion. Furthermore, as a mode1 arninoc home, adrenochrome
was shown to be relatively benign in this system. If the O-quinone, therefore, is to be considered
as a cellular hazard, how cm the ceii prevent oxidative injury? Various reductants exist in the
cell, but this discussion will focus on the enzymatic processes that can provide detoxification.
Quinones cm be reduced by one- or two-electron transfer enzymes. NADPH-cytochrome P-450
reductase catdyzes a one-electron reduction of a quinone to a semi-quinone. However, as was
shown above with the example of the ascorbate "fittile cycle," the semi-quinones are unstable
and will disproportionate to reform the quinone, concomitantly producing O?' (125).
Another enzyme that would offer better elirnination of quinones is NAD(P)H:Quinone
Oxidoreductase (NQO, EC 1.6.5.5.; also referred to as DT-diaphorase and quinone reductase).
This enzyme cataiyzes a two-electron reduction of the quinone to form its hydroquinone (126).
NQO would catalyze the reduction of CA O-quinone to its parent CA. In fact, one group has
performed extensive research on the aminochromes of DOPA, DA, and NE, to which end they
found that if NQO catalyzed the reduction of any one of these aminochromes to its hydroquinone
variety (referred to as leucoaminochrome), this molecuie would quickly autoxidize to reform the
aminochrome and form (127-129). This could represent a more downstream "futile cycle"
than that described above, in that the redox cycle is occumng with the aminochrome rather than
with its precursor (Scheme 2.2). Therefore, caution must be exercised in order to determine the
possible outcome of quinone reduction. The non-cyclized CA O-quinones have not yet been
shown to be substrates for NQO, although it is iikely that they are.
In summary, our findings show that CA O-quinones oxidize ascorbate, or NADH. They
aiso deplete GSH by forming GSH conjugates. The CA cyclization rate is inversely related to
the O-quinone half-life and the pro-oxidant activity of the O-quinone. Further work should be
canied out to study the in vitro and in vivo possibility of O-quinone toxicity contributing to the
pathological consequences of Parkinson's disease, as weli as the cardiotoxicity induced by CA
administration.
R1
HO O 0 (relatively slow autoxidation) t
HO
I /NH R2 /NH
CA
R2
CA o-quinone
NADH NAD+
Leucoaminochrome Aminoc hrome or reduced aminochrome (relatively stable) (unstable)
. L
O2
Scheme 2.2 Reduction of a CA oquinone and aminochrome. Although two-electron quinone reduction is generaiiy thought to be a protective event for the ce11 (as with the reduction of the CA O-quinone), reduction of the aminochrome results in the formation of an unstable hydroquinone that WU spontaneously re-oxidue to its CA precursor with concomitant Oz' production. If this "futile cycle" occurs repeatedly, it could Iead to oxidative stress.
Chapter 3
Dopamine rnetabolic activation by Pa50 peroxygenase activity versus manganese (II): DA O-quinone as the mediator of cytotoxicity.
ABSTRACT
DA depletion in the dopaminergic neurons of the substantia Riga is the biochemical basis
for Parkinson's disease. The incidence of a Parkinson's-type disease is markediy increased in
manganese minen. The purpose of this study was to compare the cytotoxic mechanisms of
manganesesatalyzed DA oxidation, versus normal cellular DA oxidation. Isolated hepatocytes
from Sprague-Dawley rats were used as a mode1 ce11 system since they are the ce11 choice for
studying dnig metabolism. Cytotoxicity for 700p.M DA:25pM ~ n ' + was 79.7 f 4.0 % compared
to 700 pM DA aione (- 1 %), and was prevented by GSH, ascorbate, and xylitol. The antioxidant
enzymes SOD, cataiase, and the antioxidants Trolox, Tempol, or butylated hydroxy toluene
(BHT) had no effect, indicating that ROS were not involved. Cytotoxicity was markedly
potentiated by the NQO inhibitor, dicumarol. It is concluded that DA O-quinone is responsible
for the cytotoxicity of DA:M~? The next set of experiments was aimed at determining the
cellular metabolic activation mechanism of DA in the absence of ~ n ' + . Al1 three MAO
inhibitors showed complete protection against DA induced cytotoxicity. Phenylimidazole
(CYPZE I inhibitor) d so prevented DA toxicity. However, dicumarol (NQO inhibitor) and azide
(catalase inhibitor) potentiated the toxicity of DA alone. The results of this study suggest that in
the hepatocyte DA is oxidized to the O-quinone via P450 peroxygenase activity which utilizes the
dicumarol. glucose, glucose oxidase (EC.1.1.3.4), and phenylimidazole were obtained
commercially, (Sigma Aldrich, Oakville, Ont.). Al1 chernicals were resuspended in ~ i l l i ~ @
purified water.
Hepatocyte Isolation and Preparation
Adult male Sprague-Dawley rats, 250-300 g, were obtained from Charles River Canada
Laboratories (Montreal, P.Q.), fed ad libitum and were allowed to acclimatize for 1 week on clay
chip bedding. Freshly isolated hepatocytes were prepared by collagenase perfusion of the Lver
as described by Moldeus et al., (135). Damaged cens, debris, and Kupffer cells were removed
by centrifugation with Percoll(136). The cells were preincubated in Krebs-Hensleit bicarbonate
buffer (pH 7.4) supplemented with 12.5mM HEPES for 30 minutes in a carbogen atmosphere, in
continuously rotating 50ml round bottom flasks at 37 O C before addition of chemicals.
Hepatocyte viability was assessed by the trypan blue (O. 1 % w/v) exclusion assay.
For statistical cornparison. at least n = 3 (flasks) was used for each condition tested. The
paired r-test was used since the experimental conditions for control and test flasks were identicai.
DA :~n'+ cytotoxicity experimen ts
M e r a 30 minute incubation period, DA was added to the cells immediately followed by
~n", (or other metals). Substrates added to modulate cytotoxicity. e.g., ascorbate or dicumarol,
were preincubated 30 minutes pnor to the addition of DAM&
DA Metabo lic Activation by CYP2 E I Peroxygenase Activiiy
Enzyme inhibitors or antioxidants were added as descnbed in the previous section, (prior
to the addition of DA alone). Since DA is rnetabolized by MAO-A in the rat (137), only
clorgyline was used to specifically inhibit that enzyme. The non-specific or irrevenible MAO
inhibitors pargyiine and phenelzine, respectively, were used for comparative purposes.
Similariy, both phenylirnidazole and metyrapone were used to inhibit different cytochrome P-
450 isoforms, 2E 1 and 2B 1, respectively.
lntracellular GSH and GSSG measurements
The total amount of GSH and oxidized glutathione (GSSG) in isolated hepatocytes were
measured by the HPLC analysis of deproteinized samples (5% meta phosphoric acid) after
derivatization with iodoacetic acid and fluoro-2,4-dinitrobenzene ( 13 8), using a Waters HPLC
system (model 510 pumps, WISP 710B auto injecter, and model 410 UVfvisible detector)
equipped with a Water rn Bondpack NH2 (10mM) 3.9 x 300 mm column. These methods were
used previously in our lab, and were repeated for our study, (1 39).
RESULTS
DA toxiciiy is potentiared by iWn2+, cu2+, but not ~ e ' +
The presence of 25p.M ~n'+ markedly enhanced DA cytotoxicity (approximately 80-
fold) (see Fig. 3.1). In the absence of ~n", a 21nM DA concentration was required to cause a
similar degree of cytotoxicity. The same concentration of CU" had a similar cataiytic effect and
enhanced DA cytotoxicity 46 fold. However, ~ e ~ + had no such effect on DA cytotoxicity, and
showed control cytotoxicity levels. Interestingly, ~ n " alone showed no cytotoxicity up to
ImM. However, 25p.M CU" alone caused some cytotoxicity in our system ( 140).
Cytoprotectiun against D A : M ~ ~ + by GSH, Ascorbate, and Xy litol
700p.M DA:SOpM ~ n ' + was used to cause 100% ce11 death since it was found to be a
lethal dosage. SOD, catalase, or ROS scavengers (TROLOX, TEMPOL, BHT) were ineffective
in preventing cytotoxicity, (Fig. 3.2). Only 1mM GSH, 1OmM ascorbate, or lOmM xylitol were
effective in preventing cytotoxicity. Ascorbate attenuated cytotoxicity presumably by reducing
the DA O-quinone to the serni-quinone, which subsequently reacts with ascorbyl radical,
resulting in the re-forming the parent compound. GSH is known to potentiy conjugate reactive
O-quinones such as DA O-quinone. Xylitol is a glycolytic substrate, which is oxidized to form
xylulose, using NAD' as an electron donor and producing NADH (141). The latter is the
cofactor for NQO and is utilized in quinone reduction.
NQO Inhibition and GSH Depletion Promo te DA: ~ n ~ + toxicity
Dicumarol, an inhibitor of NQO, potentiated DA:M~'+ toxicity (Fig. 3.3). Taken together
with Fig. 3 2, a potential mechanism of ~n'+-catal~zed DA toxicity is shown in Scheme 3.1.
Figure 3.1 Cornparison of Dopamine Cytotoxicity Catalyzed by Different Metals. Rasks containhg lOmL of hepatocyte suspension (106 cells/ml) were acclimated to a carbogen atmosphere before addition of corn ounds. In each case, DA was added first, followed by the metal (exeept with ImM Mn'+ alone). See text for details. * indicates significant difference from 7ûûp.M DA alone by paired t-test (p < 0.005).
Figure 3.2 Cytoprotection Against DA:M~'+ by GSH, Ascorbate, and Xylitol. Antioxidants or antioxidant enzymes were preincubated for 30 minutes before the addition of 7OO @A DA 150 pM Mn(II). GSH (ImM), ascorbate (IOmM), and xylitol(1OrnM) were the only antioxidants to significantly prevent ce11 death induced by 700m DA5Op.M ~ n " . * Indicates significant difference from 700m DA:SOpM hAn2+ by paired t-test @ c 0.005)
DA: Mn2+ 500: 10pM
Figure 3.3 NQO Inhibition Promotes DAIM^" toxicity. Dicumarol was preincubated for 30 minutes before the addition of DA / ~ n " to the hepatoc ytes. * Indicates significant difference from 500pM DA I 10pM ~ n ~ + by paired t-test @ < 0.005)
M n2+
NQO GS-DA
- P Y
NAD+ JT\ii iiicumarot A m NAD+
Xylulose Xylitol
Scheme 3.1 Proposed rnechanism of ~n~'-catal~zed DA cytotoxicity. See text for further details. DA-sQ, DA semi-quinone radical; DA-oQ, DA O-quinone; NQO, NAD(P)H:Quinone Oxidoreductase; GS-DA, DA-GSH conjugate; AA, ascorbic acid; A h , ascorbyl radical; DHAA, dehydro ascorbic acid.
M A O In hibitors Preven t DA Cyto toxicity
2mM DA alone is sufficient to cause significant ce11 death in 2 hours. The use of 20@l MAO-A
selective (clorgyline) and non-selective MAO inhibitors (pargyline, phenelzine) were effective at
protecting the ceils from DA-induced cytotoxicity, (Fig. 3.4).
H2 O2 is Involved in DA Metabolie Activation
IrnM DA alone does not affect hepatocyte viability, (Fig. 3.5). However, DA was highly
cpotoxic to NQO inhibited hepatocytes (with 20pM dicumarol). Inactivation of the MAO
activity of these hepatocytes prevented DA cytotoxicity. However, inactivation of catalase with
4mM azide rnarkedly increased DA cytotoxicity (100% ce11 death in 2 hours). At this
concentration, &de did not affect hepatocyte respiration or ce11 viability.
In the absence of ~n'+, 2mM DA was required to cause extensive ceIl death. As shown
in Fig. 3.6, the ROS scavengen (TEMPOL, TROLOX, TEMPO) did not confer any protection
against DA. Only 300pM phenylimidazole, a CYP2El inhibitor, was cytoprotective. The result
implicates the combined contribution of Hz02 and CYP2E1 in the peroxygenase-mediated
oxidation of DA. To contrast this finding with a sirnilar MAO substrate, 2mM tyramine was
used together with 300pM phenylimidazole and TEMPOL, (Fig. 3.7). There was no protective
effect of phenylirnidazole, but TEMPOL (ROS scavenger) significantly protected hepatocytes
from ce11 death, thus indicating a difierent mechanisrn of toxicity with this substrate.
Intracellular Levels of GSH and GSSG Refect Reactive Metabolites Formed
A cytotoxic concentration of DA almost completely depleted the hepatocytes of GSH,
with some GSSG formation (Table 3.1). Tyramine also caused the depletion of GSH, as a result
of GSH oxidation to GSSG. The DA O-quinone readily formed a GSH-conjugate (with GSH).
Figure 3.4 MAO Inhibitors Prevent DA Cytotoxicity. The MAO inhibitors (20pM) were preincubated for 30 minutes pnor to the addition of DA (2mM). * Indicates significant difference from 2mM DA alone, paired t-test @ < 0.005).
Figure 3.6 CYP2E 1 is hvolved in the Peroxygenase-mediated Activation of DA. The antioxidants and cytochrome P450 inhibiton, phenylirnidazole (CYP 2E 1 inhibitor), and metyrapone (CYP 2B 1 inhibitor), were preicubated for 30 minutes before the addition of DA. * Indicates significant difference from 2mM DA alone, paired t-test @ c 0.005).
Figure 3.7 Tyramine, uniike DA Cytotoxicity, is Inhibited by a MAO Inhibitor and a ROS scavenger, but not by a CYP 2E 1 Inhibitor. Pargyline, phenylirnidazole or TEMPOL were preincubated for 30 minutes before the addition of tyramine. * Indicates significant ciifference from ImM DA alone, paired t-test @ < 0.05).
Table 3.1 Intracellular [GSH] and [GSSG] after incubation with DA or tyrarnine.
2mM Tyramine il Percentages are compared to control values. Values shown are the result of three separate experiments (S.E.M.). Isolated hepatocytes were allowed to acclimatize for 30 minutes in rotating round-bottom flasks at 37 O C . Either DA or tyramine was added to the flask, and at 3 hours incubation, GSH or GSSG concentrations were determined by HPLC. See text for details.
27.3 + 3.0 % 144.3 f 16.2 %
DISCUSSION
Our study was composed of two main parts: a) bln2+ catdyzed DA cytotoxicity, and b)
P450 peroxygenase activation of DA to a cytotoxic metabolite. Although the two approaches
focus on different etiology, i.e., exogenous/environmentai vs. endogenous, we believe that both
meet at the same juncture: the DA O-quinone, (Scheme 3.2,3.3).
Several theories have been presented to outline a mechanism for the Parkinson's-like
syndrome among manganese rninea. Manganese toxicity exhibits similar behavioural effects
seen in Parkinson's disease patients, with the exception of dystonia being a manganism
associated effect (13 1). One of the latesi reviews proposed a hypothesis for the mechanism of
~ n " toxicity in a scheme where the higher manganese oxidation states are thought to cataiyze
DA autoxidation and form H202, where higher rnanganese oxidation states are considered to be
the cytotoxic mediator (142). Another group proposed that DA could form toxic products,
possibly 6-hydroxy-DA via interaction of DA with free radicals produced by ~ n " (21). These
theories address the cataiyst, manganese, as the cytotoxic reactive species but the evidence
provided here shows that cytotoxichy was much more dependent on the DA concentration than
the ~ n " concentration. Furthemore, high concentrations of MnClz in the absence of DA were
not cytotoxic. In fact, manganese alone has relevant cytoprotective properties, such as ROS
scavenging and prevention of lipid peroxidation ( 132).
Previously, using electron spin resonance experiments. Mn2+ was shown to form a suong
but highly reactive complex with DA, which produced DA O-quinone and released Mn2+ (134).
Note that the manganese is in the same valence at the end of the ceaction, inaicating that only a
smail, hence catalytic, arnount of would be needed to oxidize DA (scheme 3.2). The
marked increase in DA induced cytotoxicity if hepatocyte NQO (EC L .6.5.5.) was inactivated
Scheme 3.2 Mechanism of manganesesatalyzed DA O-quinone formation, (adapted from Lloyd (1995), see ref. 129).
MAOl's (e.g., pargyIIne)
CYTOTOXICITY
Scheme 3.3 Proposed pathway for DA metabolic activation by P450 peroxygenase ac tivity/H202.
with dicumarol beforehand suggests that DA quinoid metabolites and not manganese higher
oxidation States, were responsible for the cytotoxicity.
Further evidence suggesting DA O-quinone involvement is the prevention of DA
cytotoxicity by xylitol. The latter is a sugar alcohol that is oxidized by xylitol dehydrogenase
(EC 1.1.1.9) to D-xylulose. The cofactor NAD+ is utilized as the electron acceptor in this
reaction, which forms NADH (141). We believe that NQO in the hepatocyte utilized the NADH
in order to reduce the quinone of DA. Microsomal NQO was recently characterized in the rat
liver and compared with its cytosolic counterpart. It was found to be more resistant to dicumarol
than the cytosolic NQO (1 16), indicating that the DA O-quinone is formed in the cytosol.
DA:M~'+ cytotoxicity was dso prevented by ascorbate or GSH. The mechanism shown in
scheme 3.1 iilustrates the roles of ascorbate and GSH in detoxiQing the DA O-quinone.
However, SOD, catalase, or hydroxyl radical scavengen did not protect against DA cytotoxicity.
Taken together, these results suggest that the interaction of ~ n " with DA results in the
formation of a toxic DA O-quinone that is mainly responsible for ce11 death.
As a cornparison of the mechanism of DA:M~'+-induced cytotoxicity, we investigated a
possible pathway in which DA in the absence of exogenous agents (e.g., metals) could cause a
similar event. We found that DA induced cytotoxicity was prevented by MAO inhibitors, and
by a cytochrome P450 2E1 inhibitor, phenylimidazole, but not by a CYP 2Bl inhibitor
(metyrapone). Furthemore, DA-induced cytotoxicity was potentiated by NQO inhibition (by
dicumarol) and by catalase inhibition @y azide). ROS scavengers were not protective in this
system. This result is in conflict with other investigators who c l a h that DA cytotoxicity is due
to Hz02 formation resulting from MAO metabolism (143). Therefore, we investigated the
possibility that P450 peroxygenase catalyzed the metabolic activation of DA to a reactive
metabolite.
To assess the role of HzOz produced endogenously by MAO, we preincubated isolated
hepatocytes with clorgyline (MAO-A inhibitor) and pargyline or phenelzine (MAO-A and B
inhibitors) and added DA to these cells. We showed that al1 of the MAO inhibitors were
successful in preventing the toxicity exerted by DA. Presumably, the HzOz produced normally
by DA turnover could not cause ceii death by either ROS or P450 peroxygenase activity. This
result is in conflict with previous reports which assumed that DA or DOPA cytotoxicity is due to
ROS formation as a result of DA or DOPA autoxidation (87, 143). Since the MAO is inhibited
in our system, this leaves DA unmetabolized and subject to supposed autoxidation. However,
little cytotoxicity was detected when MAO was inhibited. The prevention of DA cytotoxicity
seen with clorgyline, a MAO-A inhibitor, is due to the preferentiai metabolism of DA by this
isoform in the rat, whereas in humans MAO-B is preferred (137, 144). To investigate the role of
Hz02 formed by MAO-rnediated DA deamination, we preincubated isolated hepatocytes with
dicumarol, with and without pargyline, and with azide (catalase inhibitor). Dicumarol greatly
potentiated an othenvise non-toxic concentration of DA. indicating that NQO detoxification
provides significant cytoprotection.
The presence of pargyline with dicumarol significantly reduced the cytotoxicity seen with
dicumarol alone, indicating that MAO metabolism is stdl required for the toxicity of DA in NQO
inactivated hepatocytes. Azide with dicumarol were synergistic in increasing DA cytotoxicity,
indicating that the presence of H202 in combination with DA initiates a cytotoxic reaction. Since
the concentration of DA used was non-toxic to hepatocytes, these results suggest that Hz02 is
involved in DA oxidative activation.
CYP 2E1 inhibited hepatocytes were resistant to DA, indicating that CYP2El is chiefly
involved in the bioactivation of DA. The CYP2Bl inhibitor, metyrapone, was not
cytoprotective. These results in combination lead us to believe that CYP2E1 peroxygenase
activity is involved in the metabolic activation of DA to a cytotoxic O-quinone. Hydroperoxides
are beiieved to bypass the rate limiting step of the monooxygenase system of P450, Le., femc
P450 reduction by NADPHP450 reductase (145). Previously, tert-butyl-hydroperoxide or H a 2
was shown to enhance the cytotoxicity and metabolic activation of a variety of phenolic
xenobiotics which was prevented by the CYP2El inhibitor, phenylirnidazolc. It was concluded
that physiological hydroperoxides can be used by P450 to support the bioactivation of these
xenobiotics (139, 146).
The only difference between tyramine and DA is the presence of a 3-hydroxy group on
the benzene ring, hence DA is also referred to as 3-hydroxytyramine. Oxidation of tyramine
would not therefore be able to fonn a quinone. Phenylimidazole did not protect isolated
hepatocytes against a toxic dose of tyramine, indicating that tyramine was not oxidized to a
cytotoxic species or that tyramine was not a substrate for P450 peroxygenase. The ROS
scavenger TEMPOL or MAO inhibitor, pargyline, however prevented tyramine cytotoxicity
suggesting that the cytotoxicity was caused by H202 generated by tyramine metabolism by
MAO.
An analysis of intracellular GSH and GSSG level in isolated hepatocytes exposed to
tyramine or DA shows that GSH is depleted by almost 100% when incubated with DA, whereas
tyramine incubation resuited in GSN oxidation to GSSG, probably by the HzOr generated by the
MAO catalyzed oxidation of tyramine. These data further provides evidence that a reactive DA
O-quinone is formed since quinones readily react with GSH to form a covalent GSH conjugate
(120).
In conclusion, what similarities and contrasts can be drawn between the manganese
catalyzed oxidation of DA versus its P450 peroxygenase bioactivation? Firstly, the keystone of
both mechanisms hinges on the formation of the DA O-quinone, (Scherne 3.1, 3.2, and 3.3).
Manganese acts as a tnie catalyst in that it fonns DA O-quinone and is retumed to its same
valence state (2+). The presence of oxygen is required in order to activate the reactive DA:M~'+
complex. The P450 peroxygenase metabolism of DA, however, requires H202 Presumably, a
hydroperoxide could substitute for MO2, but this was not studied here since the endogenous
production of the latter is physiologically linked to DA.
The hepatocyte has proved to be a very usehl mode1 ce11 to study CA cytotoxic
mechanisms, since it contains the relevant biotransforming enzymes, namely MAO and CYP2E 1
(147, 148). Hepatocytes aiso lack CA synthesizing ability, which allowed us to know
beforehand how much DA is present in the system to start with. However, it is now important to
study these CA cytotoxic mechanisms in cultured neuronal cells of substantia n i p . CYP2E1 is
localized in dopaminergic neurons and is inducible (149, 150). Also, these celk contain NQO
(DT-diaphorase), presumably to detoxiQ DA O-quinone that could be formed (151). Further
research in this area could reveal a new mechanism of Parkinson's disease and provide further
evidence for the therapeutic efficacy of MAO inhibition.
GENERAL CONCLUSIONS
Based on the duaiity of their nature, the CAS can be regarded as agents that possess both
antioxidant and pro-oxidant properties. This cornes as no surprise, since many antioxidant
substances also have their toxic components. The widely used antioxidant, ascorbate, for
example, has been shown to be both mutagenic and toxic to Chinese hamster ovary cells (152).
and has been used in chernotherapy to kill tumor cells (153). GSH is also responsible for the
metabolic activation of 1,2-dibromoethane to a reactive intermediate responsible for DNA
damage ( 154).
In Chapter 1, CAS and related compounds were shown to be highly effective at
scavenging 02* and furthemore, they prevented hepatocyte hypoxia-reoxygenation injury. It
was shown for the fmt time that the femc complexes of these compounds were in fact, much
more efficacious than CAS alone.
In chapter 2, CAS were shown to fom O-quinones that could mediate the depletion of
ascorbate, GSH, and NADH. Since the depletion of antioxidants is regarded as a toxic event, the
implications of such findings are that the CAS also have pro-oxidant activity. An interesting
finding was the rate of CA O-quinone cyclization was inversely proportional to the rate of
ascorbateMADH oxidation. We therefore propose that CA cyclization could be an inherent
antioxidant mechanism built into the chemistry of CAS. This holds for EPI, as the fastest
cyclizing CA, whereas DA was a slow cyclizer, implying that it had more toxic potential.
In Chapter 3, DA cytotoxicity was shown to be mediated by the O-quinone catalyzed by
endogenous P-450 peroxygenase activity, utilizing H202 generated by MAO rnetabolism of DA.
However, the addition of cataiytic amounts of MI?+ increased DA cytotoxicity more than 2-fold.
This suggests that M.n2+ in the environment could catalyze DA autoxidation and thereby activate
CAS.
Altogether, the results of this thesis have demonstrated that CAS can act as both
antioxidants and pro-oxidants. A key factor in differentiating cytotoxicity from cytoprotection
was the concentration of CA used. Cytoprotection by CAS in the ischemia-repemision injury
model occurred at micromolar concentration, whereas approximately 10-fold this concentration
was required to induce cytotoxicity. Their biphasic nature, therefore, depends on their
concentration.
F'UTUm EXPERLMENTS
L. As mentioned in Chapter 1, it would be of interest to inject a DA-Fe(Q complex
intrathecally in vivo to investigate whether it is protective in a rat stroke model. The two
questions to be asked regarding the CA-Fe(m) complexes are: a) would they confer similar
protection in a neuronal culture, and b) do they possess any therapeutic advantage in an in
vivo study? The complexes would be injected intrathecdly, since they may not pass the
blood brain barrier, so it's partition coefficient would need to be calculated. It should also be
tested for its stability in the blood. Also, it would be useful to find out the redox potential of
the iron complexes versus their CA ligand alone. This may shed light on the potential for
cytoprotection versus cytotoxicity of the given CA-Fe(Q complex, and allow for
predictability. For comparison, they shodd be tested with the synthetic manganese-sden
SOD-mimic complexes in a rat stroke model, as a standard for comparison of efficacy (4).
2. There should be more in vitro work perfonned in comparing the cytotoxic mechanisms and
effectiveness of CA O-quinones with their aminochrome metaboiites. Preliminary resuits
with isolated rat hepatocytes suggests that dicumarol (the NQO inhibitor), potentiates EPI
toxicity (presumably because the EPI O-quinone is not being reduced), but prevents
adrenoc hrome toxicity .
Compound % Cpotoxicity at 3 hrs
2mM EPI 8.4
2mM EPI + 20pM Dicumarol 70.7
2m.M Adrenochrome 56.0
2rn.M Adrenochrome + 20p.M Dicumarol 34.5
(Pilot data. See text for explanation)
This is contrary to the belief that NQO is a cytoprotective quinone reductase.
Adrenochrome, the most stable quinone intermediate formed dunng EPI metabolsim was
found to be actually activated by NQO, (see Chapter 2, Scheme 2.2). Segura-Aguilar's group
showed that NQO in vitro catalyzed the two-electron reduction of aminochromes to their
reduced aminochromes which caused oxygen activation. However, a cellular mode1 of this
concept has not been tested. Of coune, one must address the issue of whether adrenochrome
foms in vivo. Perhaps noradrenochrome (since the locus ceruieus is pigmented) or
doparninochrome (precuaor of neuromelanin of substantia nigra) would be more relevant for
study of cytotoxic mechanism.
3. The P-450 peroxygenase activation of DA (Chapter 3) should be investigated in
dopaminergic neurons as they contain CYP 2El (149, 150), NQO (15 l), and MAO (155).
To accomplish this objective, it is necessary to isolate and culture neurons of the substantia
nigra or striatai neurons.
4. Genetic polymorphisms may provide additional insights into the etiology of Parkinson's
disease. Genetic polymorphisms have been associated with the poor metabolizer phenotype
of CYP2D6 and CYPlAl in relation to Parkinson's disease. The data compiled by
Checkoway et al., shows that while some researchea have found an increased risk of
Parkinson's disease linked with these genetic polymorphisms, others have not found such an
association ( 156). Furthemore, researchers have investigated genetic pol ymorphisms in
MAO leading to less enzymatic activity. Although previous studies showed a specific
genetic polymorphism (intron 2, GT repeat) not to be associated with Parkinson's disease, a
recent study shows the opposite result with Chinese patients (157). Other genetic
polymorphisms have also been identified, but differ in the exact ailelic variant (156). MAO-
B knockouts were found to be resistant to MPTP (137, 158), corroborating previous findings
of MAO-B inhibition when challenged with this Parkinson's disease-simulating neurotoxin.
It would be useful to study DA-ergic toxicity in the MAO-B knockout mouse since DA is
preferentiaily metabolized by MAO-B in the mouse (and human). Since the system used in
Chapter 3 shows that MAO metaboiism of DA is a prerequisite for cytotoxicity, the MAO-B
knockout wouid allow for investigation of other pathways of DA toxicity.
5. The apparent role of NQO in detoxification of the DA O-quinone shows that in the case of
manganese toxicity andlor high DA concentration, this enzyme would be a critical reductase.
In fact, a polymorphism exists for NQO (159). if other genetic factors corne into play, it may
be possible to put together the ailele combination that could predispose an individual to a
higher nsk of Parkinson's disease.
6. Many researchea beiieve mitochondnal dysfunction to be involved in Parkinson's disease,
as Complex 1 (NADK-Q reductase) bas been shown to be markedly reduced in Parkinson's
disease (160, 161). Genetic polymorphisms have been identified in Complex 1, but more
research is required to provide evidence that wodd predispose an individual to Parkinson's
disease (156). Although there is some progress in our understanding of the genetic bais for
Parkinson's disease, many factors seem to contribute to this neurodegenerative disease
thereby making one specific cause difficult to isolate. The ideal combination for research
could involve a union of biochemical and genetic mechanisms to better our knowledge of the
etiology of Parkinson's disease in hopes for treatment.
REFERENCES
Alam, 2. I., Jenner, A., Daniel, S. E., Lees, A. J., Cairns, N., et al. 1997. Oxidative DNA
damage in the parkinsonian brain: an apparent selective increase in 8-hydroxyguanine
levels in substantia nigra. J Neurochem 69: 1 196-203
Alam, 2. L, Halliwell, B., Jenner, P. 2000. No Evidence for hcreased Oxidative Damage
to Lipids, Proteins, or DNA in Huntington's Disease. J Neurochem 75840-846
Rigattieri. S., Buffon, A., Ramazzotti, V., Mordente, A., Crea, F., et al. 2000. Oxidative
stress in ischemia-reperfusion injury: assessrnent by three independent biochemical
markers. Irai Heurt J 1:68-72
Baker, K., Marcus, C. B., Huffman, K., Kruk, H., Malfroy, B., et al. 1998. Synthetic
combined superoxide dismutase/catalase mimetics are protective as a delayed treatment
in a rat stroke model: a key role for reactive oxygen species in ischemic brain injury. J
Phanacol Exp Ther 284:2 15-2 1
Offen, D., Hachman, A., Gorodin, S., Ziv, L, Shirvan, A., et al. 1999. Oxidative stress
and neuroprotection in Parkinson's disease: implications from studies on dopamine-
induced apoptosis. Adv Neurol80:265-9
Lyrer, P., Landolt, H., Kabiersch, A., Langemann, H.. Kaeser, H. 199 1. Levels of low
molecular weight scavengen in the rat brain during focal ischemia. Brain Res 567:3 17-20
Rice, M. E., Russo-Mema, 1. 1998. Differential compartmentalization of br in ascorbate
and glu tathione between neurons and giia. Neuroscience 82: 1 2 1 3-23
HalliwelI, B. 1992. Reactive oxygen species and the centrai nervous system. J
Neurochem 59: 1609-23
9. Forster, C. 1998. Autonornic Nervous System Neurotransmitters. In Principles of
Medical Phamacology, ed. H . Kalant, W. H. E. Roschiau, pp. 135-148. New York:
Oxford University Press. 6th / ed.
10. King, R. A., Olds, D. P. 1984. Tyrosine uptake in normal and albino hairbulbs. Arch
Dennatoi Res 276:3 13-6
1 1. Blaschko, H. 1939. The specific action of L-dopa decarboxylase. J Physiol (Lund) 9650-
5 1
12. Nagatsu, T.. Levitt, M., Udenfriend, S. 1964. Conversion of L-tyrosine to 3,4-
dihydroxyphenylalanine by cell-free preparations of bnin and sympatheticaily innervated
tissues. Biochem Biophys Res Commun 14543-9
13. Rios, M., Habecker, B., Sasaoka, T., Eisenhofer, G., Tian, H., et al. 1999. Catecholamine
synthesis is mediated by tyrosinase in the absence of tyrosine hydroxylase. J Neurosci
1935 19-26
14. Raichle, M. E., Hartman, B. K., Eichling, I. O., Sharpe, L. G. 1975. Central
noradrenergic regdation of cerebral blood flow and vascular permeability. Proc Nat1
Acad Sci U S A 72:3726-30
15. Laduron, P. M. 1975. Evidence for a localization of dopamine-beta-hydroxylase within
the chromaffin granules. FEBS Len 52: 132-4
16. Bindoli, A., RigobelIo, M. P., Gaizigna, L. 1989. Toxicity of aminochromes. Toxicol Len
48:3-20
17. Graham, D. G. 1978. Oxidative pathways for catecholamines in the genesis of
neuromelanin and cytotoxic quinones. Mol Phannacol 14:633-43
d'Ischia, M., Prota, G. 1997. Biosynthesis, structure, and function of neuromelanin and its
relation to Parkinson's disease: a critical update. Pigment Cell Res 10:370-6
Okun, M. R. 1996. The role of peroxidase in mammalian melanogenesis: a review.
Physiol Chem Phys Med NMR 28:91-100
Landas, S. K., Leigh, C., Bonsib, S. M., Layne, K. 1993. Occurrence of melanin in
pheochromocytoma. Mod Pathol6: 175-8
Donaidson, J., LaBella, F. S., Gesser, D. 1980. Enhanced autoxidation of dopamine as a
possible basis of manganese neurotoxicity. Advances in neurotoxicology : proceedings of
the International Congress on Neurotoxicology.. Varese, Italy, 27-30 September 1979
Arango, V., Underwood, M. D.. Mann, J. J. 1994. Fewer pigmented neurons in the locus
coeruleus of uncomplicated alcoholics. Brain Res 650: 1-8
Elchisak, M. A., Hausner, E. A. 1984. Demonstration of N-acetyldoparnine in human
kidney and urine. Life Sci 35956 1-9
Tyce, G. M. 197 1. Metabolism of 3PdihydroxyphenylaIanine by isolated perfused rat
liver. Biochem Pharmacol 20:3447-62
Tanz, R. D., Marcus, S. M. 1966. Observations on responses of the heart to
catecholamine-depletion produced by reserpine. froc Soc Erp Bi01 Med 12 1 : 853-7
Armitage, A. K. 1965. Effects of nicotine and tobacco smoke on blood pressure and
release of catechol amines from the adrend glands. Br 1 Pharmacol 2 5 5 15-26
Janssen, L. J., Daniel, E. E. 199 1. Classification of postjunctional beta adrenoceptors
mediating relaxation of canine brouchi. J Phamcol E ip nier 256:670-6
Wenke, M . 1966. Effects of catecholamines on lipid metabolism. Adv Lipid Res 4:69-105
Samanin, R., Bemasconi, S., Garattini. S. 1975. The effect of selective lesioning of brain
catecholamine-containing neurons on the activity of various anorectics in thr rat. Eur J
Phannacol 34:373-5
Sudhof, T. C. 1995. The synaptic vesicle cycle: a cascade of protein-protein interactions.
Nature 375645-53
Abu-Jayyab, A., el-Denshary, E. E., Ageel, A. M., Dakkak, M. R. 1987. Role of cyclic
AMP in the action of doparninergic D2 receptors of some endocrine glands in rats. Biosci
Rep 7:75 1-5
Lanca, A. J. 1998. Functional Organization of the Centrai Nervous System. In Principles
of Medicul Phamacology, ed. H. Kalant, W. H. E. Roschlau, pp. 217-240. New York:
Oxford University Press. 6th / ed.
Liu, J., Mori, A. 1993. Monoamine metabolism provides an antioxidant defense in the
brain against oxidant- and free radical-induced damage. Arch Biochern Biophys 302: 1 18-
27
Yen, G. C., Hsieh, C. L. 1997. Antioxidant effects of dopamine and related compounds.
Biosci Biotechnol Biochem 6 1 : 1646-9
Miura, T., Muraoka, S., Ogiso, T. 1998. Antioxidant activity of adrenergic agents derived
from catechol. Biochem Phannacol 55:2ûû 1-6
Smythies. 1. 1997. The biochemical basis of synaptic plasticity and neurocomputation: a
new theory. Proc R Soc Lond B Biol Sci 264575-9
Smythies, 1. 1999. The neurotoxicity of glutamate, dopamine, iron and reactive oxygen
species: functional interrelationships in health and disease. A review-discussion.