Retrorsine, but not monocrotaline, is a mechanism-based inactivator of P450 3A4

Article abstract of DOI:10.1016/j.cbi.2009.10.001

Retrorsine (RTS) and monocrotaline (MCT) cause severe toxicities via P450-mediated metabolic activation. The screening of mechanism-based inhibitors showed RTS inactivated 3A4 in the presence of NADPH. Unlike RTS, MCT failed to inhibit P450 3A4 and other enzymes tested. Further studies showed the loss of P450 3A4 activity occurred in a time- and concentration-dependent way, which was not recovered after dialysis. Dextromethorphan, a P450 3A4 substrate, protected the enzyme from the inactivation. Exogenous nucleophile glutathione (GSH) and reactive oxygen species scavengers catalase and superoxide dismutase did not protect P450 3A4 from the inactivation. GSH trapping experiments showed both P450 3A4 and 2C19 converted RTS and MCT to the corresponding electrophilic metabolites which could be trapped by GSH to form 7-GSH-DHP conjugate. We conclude that RTS and MCT are metabolically activated by P450 3A4 and 2C19, and that RTS, but not MCT, is a mechanism-based inactivator of P450 3A4.

Full text of DOI:10.1016/j.cbi.2009.10.001

Chemico-Biological Interactions 183 (2010) 49–56
Contents lists available at ScienceDirect
Chemico-Biological Interactions
journal homepage: www.elsevier.com/locate/chembioint
Retrorsine, but not monocrotaline, is a mechanism-based inactivator
of P450 3A4
Jieyu Dai, Fan Zhang, Jiang Zheng∗
Center for Developmental Therapeutics, Seattle Children’s Research Institute; Division of Gastroenterology, Department of Pediatrics,
University of Washington, 1900 9th Ave, Seattle, WA 98101, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Retrorsine (RTS) and monocrotaline (MCT) cause severe toxicities via P450-mediated metabolic activa-
tion. The screening of mechanism-based inhibitors showed RTS inactivated 3A4 in the presence of NADPH.
Unlike RTS, MCT failed to inhibit P450 3A4 and other enzymes tested. Further studies showed the loss
of P450 3A4 activity occurred in a time- and concentration-dependent way, which was not recovered
after dialysis. Dextromethorphan, a P450 3A4 substrate, protected the enzyme from the inactivation.
Exogenous nucleophile glutathione (GSH) and reactive oxygen species scavengers catalase and superox-
ide dismutase did not protect P450 3A4 from the inactivation. GSH trapping experiments showed both
P450 3A4 and 2C19 converted RTS and MCT to the corresponding electrophilic metabolites which could
be trapped by GSH to form 7-GSH-DHP conjugate. We conclude that RTS and MCT are metabolically acti-
vated by P450 3A4 and 2C19, and that RTS, but not MCT, is a mechanism-based inactivator of P450 3A4.
Received 17 July 2009
Received in revised form
30 September 2009
Accepted 1 October 2009
Available online 8 October 2009
Keywords:
Retrorsine
Monocrotaline
Pyrrolizidine alkaloid
P450 3A4
Mechanism-based inactivation
© 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
groups in DNA and proteins to form 6,7-dihydro-7-hydroxy-1-
hydroxymethyl-5H-pyrrolizine (DHP) modified DNA or protein
Retrorsine (RTS) and monocrotaline (MCT) are two retronecine-
type pyrrolizidine alkaloids (PAs), present in Senecio and Crotalaria
species, respectively (Fig. 1). RTS is a 12-membered macrocyclic
diester PA with an ␣,␤-unsaturated double bond linked to the ester
group at the C-7 position of the necine base, which exhibits hepato-
toxicity [1,2], whereas MCT, an 11-membered macrocyclic diester
PA without ␣,␤-unsaturated double bond linked to the ester, pre-
dominantly induces pneumotoxicity [3,4]. The major routes of
human exposure to PAs are accidental intake of PA-contaminated
foods and intentional ingestion of PA-containing herbal tea, herbal
medicine, and dietary supplements [5,6]. Previous studies have
implied that metabolic activation of PAs is associated with their
toxicities [7,8]. Retronecine-type PAs are metabolically converted
to electrophilic dehydrogenated metabolites by cytochromes P450,
and the resulting pyrrole metabolites have two electrophilic sites
at C-7 and C-9 positions that are readily attacked by nucleophilic
adducts, and interstrand DNA–DNA and DNA–protein crosslinking,
possibly leading to various toxicities [9,10]. Several DHP-derived
DNA adducts have been observed in in vitro experiments when
incubating RTS or MCT with rat liver microsomes in the presence of
calf thymus DNA, and were proposed to be biomarkers of tumori-
genicity induced by retronecine-type PAs [11,12]. Moreover, the
target proteins of MCT in pulmonary artery endothelial cells have
been identified and were considered to be associated with MCT-
induced pulmonary hypertension [13,14].
Given the reported alkylation of DNA and proteins, we specu-
lated the reactive metabolites of PAs might bind covalently to the
nucleophilic amino acids located in active sites of P450s, acting as
suicide or mechanism-based inactivators (Fig. 1). Currently, infor-
mation regarding the interaction of PAs with cytochromes P450
is limited. It has been reported that the metabolic activation of
retronecine-type PAs is mainly catalyzed by P450 3A and 2B iso-
forms [8,10]. However, the specific P450 isoenzymes participating
in RTS metabolic activation have not been identified, though one or
more isoenzymes were suggested to be involved in RTS metabolism
[15]. MCT is mainly metabolized by P450 3A [16] and was found
to inhibit the liver drug-metabolizing enzymes in rats [17], but
the inhibitory effect on the individual P450 isoenzymes remains
unknown.
Abbreviations: DHP, 6,7-dihydro-7-hydroxy-1-hydroxymethyl-5H-pyrrolizine;
7,9-diGSH-DHP, 7,9-diglutathionyl-6,7-dihydro-1-hydroxymethyl-5H-pyrrolizine;
GSH, glutathione; 7-GSH-DHP, 7-glutathionyl-6,7-dihydro-1-hydroxymethyl-5H-
pyrrolizidine; KI,
a concentration of inactivator required for half-maximal
inactivation; kinact, a maximal rate constant for inactivation; kobs, initial rate con-
stants for inactivation; MRM, multiple reaction monitoring; P450, cytochrome P450;
PA, pyrrolizidine alkaloid; SIM, selected ion monitoring.
In the present studies, we examined the effects of RTS and
MCT metabolism on the enzymatic activities of eight P450 isoen-
zymes, including P450s 1A2, 2A6, 2B6, 2C9, 2C19, 2D6, 2E1, and
∗
Corresponding author. Tel.: +1 206 884 7651; fax: +1 206 987 7660.
E-mail address: jiang.zheng@seattlechildrens.org (J. Zheng).
0009-2797/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.cbi.2009.10.001

50
J. Dai et al. / Chemico-Biological Interactions 183 (2010) 49–56
Fig. 1. Mechanisms for the metabolic activation of RTS and MCT.
3A4. Then, the kinetic characteristics of mechanism-based inacti-
vation of P450 3A4 by RTS were studied. Additionally, the P450
isoenzymes involved in metabolic activation of RTS and MCT were
also investigated by GSH trapping experiments.
dissolved in chloroform (5.0 mL), and a solution of o-chloranil
(0.10 mmol) in chloroform (5.0 mL) was added at room tempera-
ture. After 5 min, the mixture was shaken vigorously with a cooled
solution (1.0 mL) containing KOH (70%) and sodium borohydride
(2%) for 10–15 s. The organic phase was separated, immediately
dried over anhydrous Na2SO4, and concentrated under reduced
pressure to give the nearly pure dehydromonocrotaline, which
was further purified by recrystallization from the mixture of ben-
zene and petroleum ether. Glutathione conjugates 7-GSH-DHP and
7,9-diGSH-DHP were synthesized from dehydromonocrotaline.
Dehydromonocrotaline (0.03 mmol) dissolved in DMSO (100 ␮L)
was added into a solution of glutathione (0.045 mmol) in water
(4.0 mL), and the mixture was stirred overnight at room tem-
perature. The resulting GSH conjugates were purified by HPLC.
7-GSH-DHP: ESI-MS m/z 443.4 ([M+H]+, 100), 367.3 (75). 7,9-
diGSH-DHP: ESI-MS m/z 732.1 ([M+H]+, 100), 424.8 (25).
2. Materials and methods
2.1. Chemicals
Retrorsine (RTS), monocrotaline (MCT), testosterone, phena-
cetin, coumarin, tolbutamide, dextromethorphan, 4-nitrophenol,
acetaminophen, nirvanol, 7-hydroxycoumarin, 4-hydroxytolbuta-
mide, 4ꢀ-hydroxymephenytoin, dextrorphan, 4-nitrocatechol, 2-
fluoro-4-nitrophenol, (+)-camptothecin, l-glutathione reduced
(GSH), NADPH, trifluoroacetic acid, o-chloranil, sodium borohy-
dride, ammonium acetate, and HPLC-grade formic acid were
purchased from Sigma–Aldrich Chemical Co. (Milwaukee, WI).
6␤-Hydroxytestosterone and 11␤-hydroxytestosterone were from
Steraloids Inc. (Pawling, NY) and Ikapharm (Ramat-Gan, Israel),
respectively. S-Mephenytoin was purchased from BD Biosciences
(St. Louis, MO). HPLC-grade methanol, ethanol, and acetonitrile
(ACN), and Tris, DMSO, chloroform, anhydrous Na2SO4, and potas-
sium hydroxide (KOH) were from EMD Biosciences (La Jolla, CA).
Human P450 SupersomesTM enzymes (P450s 1A2, 2A6, 2B6, 2C9,
2C19, 2D6. 2E1, and 3A4) were purchased from BD Biosciences (St.
Louis, MO). Catalase from bovine liver and superoxide dismutase
from bovine liver were obtained from Sigma Chemical Co. (Mil-
waukee, WI).
2.3. Mechanism-based inhibitor screening
The primary reaction mixtures contained 0.1 ␮M individual
P450s (100 pmol), 3.2 mM MgCl2, and 100 ␮M RTS or MCT in 0.1 M
Tris–HCl or potassium phosphate. Control incubation was con-
ducted in the absence of RTS and MCT. The organic solvent was
typically methanol at a final concentration of 0.5% (v/v). After equi-
libration of the reaction mixtures at room temperature for 3 min,
the reactions were initiated by the addition of NADPH to a final con-
centration of 0.45 mM, and the mixtures were incubated at 37 ◦C
for 0, 5, and 30 min. At indicated time points, aliquots of the pri-
mary reaction mixtures were transferred at a dilution of 1:5 to a
secondary reaction mixture containing the corresponding probe
substrate (Table 1), 0.45 mM NADPH, and 3.2 mM MgCl2 in 0.1 M
Tris–HCl (pH 7.4). The secondary incubation mixtures were fur-
ther incubated at 37 ◦C for various time periods (20 min for P450s
2.2. Synthesis
Dehydromonocrotaline was synthesized according to a pro-
cedure described previously [18]. In brief, MCT (0.06 mmol) was

J. Dai et al. / Chemico-Biological Interactions 183 (2010) 49–56
51
Table 1
Substrates and products involved in mechanism-based inactivation screening studies.
Isoenzyme
Probe substrate
Concentration (␮M)
Metabolite
Detection
254 nmb
CYP 1A2
CYP 2A6
CYP 2B6
CYP 2C9
CYP 2C19
CYP 2D6
CYP 2E1
CYP 3A4
Phenacetin
Coumarin
200a
400a
100 [26]
100 [27]
100a
Acetaminophen
7-Hydroxycoumarin
Nirvanol
4-Hydroxytolbutamide
4-Hydroxymephenytoin
Dextrorphan
m/z 163 → 107c
s-Mephenytoin
Tolbutamide
s-Mephenytoin
Dextromethorphan
4-Nitrophenol
Testosterone
m/z 205c
m/z 287c
m/z 235c
10 [28]
m/z 258c
500a
4-Nitrocatechol
6␤-Hydroxyltestosterone
350 nmb
200a
m/z 305c
a
The concentrations were determined according to the protocols provided by the manufacturer.
Products were monitored by HPLC/UV.
Products were monitored by LC/MS.
b
c
1A2, 2A6, 2B6, 2C9, 2D6, and 3A4; 35 min for 2C19 and 2E1),
absence of a mixture of catalase and superoxide dismutase (each
800 unit/mL). The residual activities were monitored as described
below.
followed by addition of equal volume ice-cold ethanol contain-
ing selective internal standards (1.0 ␮g/mL camptothecin for 1A2,
2A6, 2B6, 2D6, 2C9, and 2C19; 1.5 ␮g/mL 11-hydroxytestosterone
for 3A4; and 10 ␮g/mL 2-fluoro-4-nitrophenol for 2E1). Samples
were centrifuged at 14,000 rpm for 20 min, and the supernatants
were analyzed directly by LC/MS or HPLC/UV. The residual enzyme
activities were determined by monitoring the amounts of the cor-
responding products formed in the secondary incubation mixtures.
2.7. Irreversibility of inactivation
Primary incubations containing 200 ␮M RTS were performed in
the presence of NADPH at 37 ◦C. The control incubations lacked RTS.
Aliquots of the mixtures were withdrawn at 0 and 25 min for enzy-
matic activity determination (non-dialyzed sample). At 25 min time
point, aliquots of the control and inactivated samples containing
100 pmol P450 3A4 were transferred respectively to Slide-A-Lyzer
membranes (molecular weight cut off 3500 Da; Pierce, Rockford, IL)
and dialyzed at 4 ◦C for 4 h against 2 L potassium phosphate buffer
(0.1 M, pH 7.4) supplemented with 0.5 mM EDTA. The dialyzed sam-
ples were brought to room temperature and added to the secondary
incubation mixture for the determination of the residual enzyme
activities. The remaining activities of the dialyzed and non-dialyzed
samples were measured as described below.
2.4. Time-, concentration-, and NADPH-dependent inactivation
of P450 3A4 by RTS
The primary incubation mixtures contained 0.1 ␮M P450 3A4
(100 pmol) and RTS at various concentrations (0, 5, 10, 25, 50,
100, and 200 ␮M) in 0.1 M potassium phosphate buffer (pH 7.4).
To determine the dependency of NADPH in enzyme inactivation,
100 ␮M RTS and P450 3A4 were incubated in the absence of NADPH
as negative control. The reactions were initiated by the addition of
NADPH, and aliquots of the primary mixtures (100 ␮L) containing
10 pmol of P450 3A4 were transferred at 0, 4, 8, and 12 min to the
secondary incubation mixtures containing 200 ␮M testosterone
and 0.45 mM NADPH in 0.1 M potassium phosphate buffer (pH 7.4)
to a total reaction volume of 0.5 mL. After 20 min of incubation, the
reaction was stopped by adding 0.5 mL ice-cold ethanol containing
11-hydroxytestosterone (1.5 ␮g/mL), followed by centrifugation at
14,000 rpm for 20 min. The supernatants were collected and ana-
lyzed by LC/MS as described below.
2.8. Enzymatic reaction product analysis
The resulting metabolites of each probe substrates (except
acetaminophen and 4-nitrocatechol) were analyzed by LC/MS. The
LC/MS system consists of Agilent 1100 HPLC (Agilent Technolo-
gies, Palo Alto, CA) coupled to an API 2000TM triple quadrupole
mass spectrometry (Applied Biosystems, Foster City, CA). The
chromatographic separations were carried out on an Alltima
HP C18 column (100 × 2.1 mm i.d., 3 ␮m, Alltech Associates Inc.,
Deerfield, IL). For 7-hydroxycoumarin, 4-hydroxytolbutamide,
4ꢀ-hydroxymephenytoin, demethylmephenytoin, nirvanol and
2.5. Substrate protection
Substrate protection from RTS-induced inactivation of P450 3A4
was determined by including 200 ␮M dextromethorphan together
with 100 ␮M RTS in the primary reaction mixture. The reaction
was initiated with NADPH, and aliquots of the mixtures (100 ␮L)
containing 10 pmol P450 3A4 were transferred at 0, 4, and 12 min
time points to the secondary incubation mixture for the deter-
mination of testosterone 6␤-hydroxylase activities of P450 3A4.
Control incubations lacking both RTS and dextromethorphan, incu-
bations only containing dextromethorphan, and incubations with
RTS alone were also performed in parallel.
dextrorphan,
a linear 2.0 min-gradient from 90:10 to 10:90
(water:ACN supplemented with 0.1% formic acid) was applied.
6␤-Hydroxytestosterone and 11␤-hydroxytestosterone were sep-
arated with a solvent system consisting of solvent A (0.1% formic
acid in water) and solvent B (50% ACN, 45% methanol, and 5% water
supplemented with 0.1% formic acid) using linear gradients from
30% B to 80% B over 5 min. Quantification was performed using
either multiple reaction monitoring (MRM) or selected ion moni-
toring (SIM) in positive-ion electrospray ionization mode (Table 1).
Acetaminophen and 4-nitrocatechol were monitored by a Shi-
madzu LC-10Avp series HPLC system (Kyoto, Japan), consisting of
LC-10ADvp pump, DGU-14A degasser, SIL-10ADvp auto-injector,
CTO-10Avp column oven, and SPD-M10Avp diode array UV detec-
tor. The chromatographic separation was also performed on the
Alltima HP C18 column with the mobile phase composed of a mix-
ture of water and methanol. The mobile phase for acetaminophen
analysis included 10 mM ammonium acetate, and a gradient elu-
tion started at 40% methanol and graded to 100% methanol in
9 min. The UV detector wavelength was set at 254 nm. For anal-
ysis of 4-nitrocatechol, the mobile phase was supplemented with
2.6. Effects of GSH, catalase, and superoxide dismutase on the
inactivation
GSH (2 mM) was included in the primary reaction mixture
together with 0.1 ␮M P450 3A4 (100 pmol), 100 ␮M RTS, and
0.45 mM NADPH. Aliquots were removed for determining the
remaining enzyme activities. In control samples, an equal volume
of phosphate buffer was added in place of GSH. In a separate study,
P450 3A4 was incubated with RTS and NADPH in the presence or

52
J. Dai et al. / Chemico-Biological Interactions 183 (2010) 49–56
Fig. 2. The remaining enzyme activities of individual P450s after exposure to vehicle (ꢀ), 100 ␮M MCT (᭹), or 100 ␮M RTS (ꢁ) in the presence of NADPH for 0, 5, and 30 min.
Each data represents the mean SD of duplicate incubations in two separate experiments. *p < 0.05.

J. Dai et al. / Chemico-Biological Interactions 183 (2010) 49–56
53
Fig. 3. (A) Time, concentration and NADPH-dependent inactivation of P450 3A4 by RTS. P450 3A4 was incubated with 0 ␮M (ꢁ), 5 (ꢀ), 10 (ꢂ), 25 (×), 50 (᭹), 100 ( ), and
200 ␮M (ꢁ) RTS, respectively, in the presence of NADPH or incubated with 100 ␮M RTS without NADPH (ꢃ) at 37 ◦C for 0, 4, 8 and 12 min. Aliquots of incubation mixtures
were transferred to the secondary incubation mixtures for the determination of residual enzymatic activity. Each data represents the average of four separate experiments.
(B) Double reciprocal plot of the rates of inactivation as a function of RTS concentrations. (C) Substrate protection against inactivation of P450 3A4 by RTS. P450 3A4 was
incubated with vehicle (᭹) and 100 ␮M RTS in the presence (ꢀ) or absence (ꢁ) of 200 ␮M dextromethorphan. Each data represents the mean SD of duplicate incubations
in two separate experiments.
0.1% trifluoroacetic acid, and a linear 7-min gradient from 50% to
80% methanol was applied. The UV detector wavelength was set at
350 nm.
in the presence of NADPH. The plots of percentage of remaining
enzyme activities versus preincubation time are shown in Fig. 2.
RTS resulted in a loss of about 30% P450 3A4 enzyme activity after
30 min of incubation, whereas MCT failed to inactivate 3A4. In the
screening of the other enzymes, including P450 1A2, 2A6, 2B6,
2C9, 2C19, 2D6, and 2E1, no such enzyme inactivation observed
in P450 3A4 incubations with RTS was found in comparison with
the corresponding spontaneous enzyme activity loss after exposure
to vehicle.
2.9. GSH conjugate trapping
RTS (200 ␮M) or MCT (200 ␮M), and 5 mM GSH were incu-
bated with 0.1 ␮M individual P450 enzymes, including 1A2, 2A6,
2B6, 2C9, 2C19, 2D6, 2E1, or 3A4 (50 pmol for each) in the
presence or absence of 1.0 mM NADPH at 37 ◦C for 60 min.
GSH conjugates 7-glutathionyl-6,7-dihydro-1-hydroxymethyl-5H-
pyrrolizidine (7-GSH-DHP) and 7,9-diglutathionyl-6,7-dihydro-1-
hydroxymethyl-5H-pyrrolizine (7,9-diGSH-DHP) were detected by
LC/MS [19–21]. For analysis of GSH-DHP, the mobile phase con-
sisted of water and ACN supplemented with 0.1% trifluoroacetic
acid. The gradient was maintained at 10% ACN for 0.5 min, fol-
lowed by a linear increase to 90% ACN in 5 min and then kept at
90% ACN for 5.5 min. Then the column was equilibrated at 10% ACN
for 7 min. Substrates and GSH conjugates were monitored in posi-
tive SIM mode, including MCT (m/z 326), RTS (m/z 352), GSH-DHP
(m/z 443), and diGSH-DHP (m/z 732). The spray voltage was set at
5 kV; source temperature set at 350 ◦C; and the dwell time for each
ion was 200 ms.
3.2. Time-, concentration-, and NADPH-dependent inactivation
of P450 3A4 by RTS
The remaining testosterone hydroxylase activities of P450 3A4
were measured and a semilogarithmic plot of percent remaining
activity versus preincubation time was performed (Fig. 3A). P450
3A4 retained the catalytic activity in control incubations (without
RTS) and incubations with RTS lacking NADPH over 12-min incu-
bation, whereas, in the presence of NADPH, RTS produced a time
course inhibition of P450 3A4 activity. Since there was no signifi-
cant loss of P450 3A4 activity in control incubations over 12 min of
incubation, the remaining enzymatic activity of the control incu-
bations at time 0 was normalized to 100%. When P450 3A4 was
preincubated with various concentrations of RTS (5–200 ␮M) in
the presence of NADPH, the loss of enzyme activity increased pro-
gressively with increasing concentrations of RTS, and the enzyme
inactivation followed pseudo first-order kinetics. The initial rate
constants for inactivation (kobs) were determined by linear regres-
sion analysis of the time course data, and a double reciprocal plot
of the values for kobs and the RTS concentrations was performed to
give a maximal rate constant for inactivation (kinact) of 0.025 min−1
and a concentration of inactivator required for half-maximal inacti-
vation (KI) of 7.7 ␮M (Fig. 3B). Therefore, the inactivation efficiency
3. Results
3.1. Mechanism-based inhibitor screening
One of the objectives of the present study was to examine
whether RTS and MCT produced mechanism-based inactivation
of those human cytochromes P450 mainly responsible for drug
metabolism. The study was initiated by examination of the remain-
ing activities of P450s 1A2, 2A6, 2B6, 2C9, 2C19, 2D6, 2E1, and
3A4 at 0, 5, and 30 min after exposure to 100 ␮M RTS or MCT
represented by kinact/KI was 3.25 min−1 mM−1
.

54
J. Dai et al. / Chemico-Biological Interactions 183 (2010) 49–56
Table 2
3.3. Substrate protection
Irreversibility of the inactivation of P450 3A4 by RTS.
An alternate substrate, dextromethorphan [22], was included
in the primary incubations along with RTS at a molar ratio of
2:1 (dextromethorphan:RTS). The remaining P450 3A4 activity
was monitored and normalized to control activity at time 0. After
correcting for the inhibitory effect of dextromethorphan on testos-
terone 6␤-hydroxylation (56.8 3.4%), the remaining enzyme
activity in samples coincubated with RTS and dextromethorphan
was 94.8 0.2% at 12 min, higher than samples incubated in the
presence of RTS (77.7 4.1%), indicating that dextromethorphan
reduced the rate of P450 3A4 inactivation produced by RTS (Fig. 3C).
Preincubation
Pre-dialysis
0 min
Post-dialysis
25 min
25 min
Vehicle
RTS
100 2.7
96.7 4.4
92.6 5.7
68.1 3.9
50.2 4.0
38.6 2.6
% Remaining activity
% of Control
96.7
73.5
76.9
Data are mean SD (n = 3).
RTS, and the remaining enzyme activities were 87.1 5.0% and
78.3 5.9% at 4 and 12 min respectively.
3.5. Irreversibility of inactivation
3.4. Effects of GSH and catalase/superoxide dismutase
P450 3A4 was incubated with RTS (200 ␮M) in the pres-
ence of NADPH, followed by 4-h dialysis. The testosterone
6␤-hydroxylation activities of control samples and those pre-
treated with RTS were determined before and after dialysis, and
normalized to control activity at time 0 (Table 2). The percent
remaining P450 3A4 activities measured after dialysis were cor-
rected by the volume change due to dialysis. No recovery in activity
was observed for inactivated samples.
After 4 and 12 min of preincubation with 100 ␮M RTS and
NADPH, the remaining P450 3A4 activity were 88.6 1.4% and
76.8 5.1%, respectively. The inclusion of 2 mM GSH, an elec-
trophile trapping agent, in the primary reaction mixture produced
a 5% increase in the remaining enzyme activities (92.9 2.7% at
4 min, and 81.6 3.6% at 12 min). In addition, a mixture of catalase
and superoxide dismutase, scavengers of reactive oxygen species,
had no protective effect against the inactivation of P450 3A4 by
Fig. 4. LC/MS chromatograms of (A) GSH conjugate standards and (B) 7-GSH-DHP (m/z 443) formed in supersome incubation mixtures in the presence or absence of NADPH.

J. Dai et al. / Chemico-Biological Interactions 183 (2010) 49–56
55
3.6. GSH conjugate trapping
than dehydromonocrotaline when they were incubated with 4-(p-
nitrobenzyl)pyridine, a model nucleophile as a trapping reagent
[23]. This supports the scenario that dehydromonocrotaline is
not reactive enough to modify nucleophilic amino acid residue(s)
before it is released from the active site of the enzyme. However,
we cannot exclude other factors, for example, (1) the distance of the
electrophilic centers of the corresponding dehydropyrrolizidine to
the nucleophile(s) in the active site of the enzyme; and (2) an appro-
priate conformational orientation required for the formation of
covalent bond between reactive metabolites of pyrrolizidine alka-
loids and the target amino acid residues.
Besides P450 3A4, P450s 1A2, 2A6, 2C9, 2C19, 2D6, and 2E1
are the most abundant P450 isoenzymes in human liver and most
important in the metabolism of clinically relevant drugs as well
[24]. Additionally, P450 2B family was reported to be involved
in the metabolic activation of senecionine, another retrornecine-
type PA [25]. Therefore, we examined whether RTS and MCT
showed mechanism-based inactivation of P450s 1A2, 2A6, 2B6,
2C9, 2C19, 2D6, and 2E1. Our results showed neither RTS nor
MCT exhibited time-dependent inhibition on P450 1A2, 2A6, 2B6,
2C9, 2C19, 2D6, and 2E1 in the presence of NADPH. In order to
clarify whether these enzymes participate in the metabolic acti-
vation of RTS and MCT, we incubated MCT or RTS with individual
P450s using GSH as a trapping agent. Among the enzymes, only
P450s 3A4 and 2C19 were found to catalyze the formation of
the corresponding dehydrogenated products. However, it appears
that neither MCT nor RTS showed time-dependent inhibition on
P450 2C19.
RTS or MCT was incubated with individual P450s, including 1A2,
2A6, 2B6, 2C9, 2C19, 2D6, 2E1, and 3A4, in the presence or absence
of NADPH for 60 min and the resulting reactive metabolites were
trapped with GSH in situ, followed by LC/MS analysis (Fig. 4). The
retention times of 7-GSH-DHP and 7,9-GSH-DHP (refer to Fig. 1
for their structures) were 11 and 10 min, respectively. The peak
responsible for 7-GSH-DHP (m/z 443) as the major GSH conjugate
was observed in the incubation of either RTS or MCT with P450s
3A4 or 2C19 in the presence of NADPH, but such peak was nei-
ther found in incubations without NADPH nor incubations of any
other P450 enzymes tested. In P450 3A4 incubation mixtures, the
formation of 7-GSH-DHP from RTS was 7.86 pmol/min/pmol P450,
and that from MCT was 9.52 pmol/min/pmol P450. In P450 2C19
incubations, the GSH conjugate formed from RTS and MCT were
15.6 pmol/min/pmol P450 and 6.34 pmol/min/pmol P450, respec-
tively. Nonetheless, bis-GSH conjugate 7,9-diGSH-DHP was not
found in the incubations of all P450s carried out.
4. Discussion
Although P450-mediated metabolic activation has been sug-
gested to play a critical role in the toxicities of retronecine-type PAs,
little attention has so far been paid to the interaction of PAs with
individual P450 isoenzymes. The present study started with screen-
ing tests which allowed us to determine rapidly whether RTS and
MCT showed time-dependent inhibition of selective cytochromes
P450, including P450s 1A2, 2A6, 2B6, 2C9, 2C19, 2D6, 2E1, and 3A4.
RTS was found to inactivate P450 3A4 but not the other isoen-
zymes tested. The subsequent kinetic studies further demonstrated
that RTS produced a time- and concentration-dependent inactiva-
In summary, pyrrolizidine alkaloids RTS and MCT are metabol-
ically activated by P450 3A4 and 2C19 to form chemically reactive
dehydrogenated intermediates. RTS is a mechanism-based inacti-
vator of P450 3A4 but not 2C19. Unlike RTS, MCT failed to show the
mechanism-based inactivation of P450 3A4.
tion of P450 3A4 with a KI of 7.7 ␮M and a kinact of 0.025 min−1
.
The enzyme inactivation was also found to be NADPH-dependent.
This implies the enzyme inhibition by RTS is mediated by bio-
transformation. The enzyme inactivation reached a plateau at the
concentration of approximately 100 ␮M. The inactivation of P450
3A4 could not be prevented by the addition of GSH in the prein-
cubation system, indicating that the inactivation by electrophilic
metabolites of RTS took place before escaping from the active site
of the enzyme. In addition, catalase and SOD, scavengers of reactive
oxygen species, failed to protect P450 3A4 against the inactiva-
tion by RTS, implying P450 3A4 inactivation was not caused by
the formation of reactive oxygen species during metabolism. How-
ever, an excess of dextromethorphan, an alternate substrate of
P450 3A4, protected P450 3A4 against the inactivation, suggesting
that dextromethorphan competes with RTS binding to the active
site of P450 3A4 and therefore reduces the formation of the reac-
tive metabolite responsible for the modification of the enzyme in
situ. As expected, the dialysis did not reverse the enzyme inhibi-
tion by RTS. The observed irreversibility of the enzyme inhibition
suggests that the reactive metabolite generated from metabolic
activation may covalently modify the enzyme responsible for the
bioactivation of RTS. Taken together, these results support the
conclusion that RTS inhibited P450 3A4 as a mechanism-based
inactivator.
Conflict of interest
These authors declare that there are no conflicts of interest.
References
[1] A.F. Dimande, C.J. Botha, L. Prozesky, L. Bekker, G.M. Rösemann, L. Labuschagnel,
E. Retief, The toxicity of Senecio inaequidens DC, J. S. Afr. Vet. Assoc. 78 (2007)
121–129.
[2] L. Muller, P. Kasper, G. Kaufmann, The clastogenic potential in vitro of
pyrrolizidine alkaloids employing hepatocyte metabolism, Mutat. Res. 282
(1992) 169–176.
[3] D.W. Wilson, H.J. Segall, L.C. Pan, M.W. Lamé, J.E. Estep, D. Morin, Mecha-
nisms and pathology of monocrotaline pulmonary toxicity, Crit. Rev. Toxicol.
22 (1992) 307–325.
[4] A.L. Zaiman, M. Podowski, S. Medicherla, K. Gordy, F. Xu, L. Zhen, L.A. Shi-
moda, E. Neptune, L. Higgins, A. Murphy, S. Chakravarty, A. Protter, P.B. Sehgal,
H.C. Champion, R.M. Tuder, Role of the TGF-beta/Alk5 signaling pathway in
monocrotaline-induced pulmonary hypertension, Am. J. Respir. Crit. Care Med.
177 (2008) 896–905.
[5] E. Roeder, Medicinal plants in China containing pyrrolizidine alkaloids, Phar-
mazie 55 (2000) 711–726.
[6] A.S. Prakash, T.N. Pereira, P.E.B. Reilly, A.A. Seawright, Pyrrolizidine alkaloids in
human diet, Mutat. Res. 443 (1999) 53–67.
[7] S.F. Zhou, C.C. Xue, X.Q. Yu, G. Wang, Metabolic activation of herbal and dietary
constituents and its clinical and toxicological implications: an update, Curr.
Drug Metab. 8 (2007) 526–553.
[8] P.P. Fu, Q. Xia, G. Lin, M.W. Chou, Pyrrolizidine alkaloids—genotoxicity,
metabolism enzymes, metabolic activation and mechanisms, Drug Metab. Rev.
36 (2004) 1–55.
Interestingly, MCT, unlike RTS, failed to inhibit the activity of
P450 3A4, although it has the same core structure (necine base
and diester) as RTS. The GSH trapping experiments showed that
both MCT and RTS were metabolized by P450 3A4 to the corre-
sponding dehydropyrrolizidines. We believe that the differences
in the reactivity of the electrophonic metabolites of RTS and MCT is
a factor for the modification of nucleophilic amino acid residues in
the active site of P450 3A4, leading to the loss of enzyme activity.
Kinetic studies showed that dehydroretrorsine was more reactive
[9] M.W. Chou, P.P. Fu, Formation of DHP-derived DNA adducts in vivo from
dietary supplements and Chinese herbal plant extracts containing carcinogenic
pyrrolizidine alkaloids, Toxicol. Ind. Health 22 (2006) 321–327.
[10] P.P. Fu, Q. Xia, G. Lin, M.W. Chou, Genotoxic pyrrolizidine alkaloids—
mechanisms leading to DNA adduct formation and tumorigenicity, Int. J. Mol.
Sci. 3 (2002) 948–964.
[11] Y.P. Wang, P.P. Fu, M.W. Chou, Metabolic activation of the tumorigenic
pyrrolizidine alkaloid, retrorsine, leading to DNA adduct formation in vivo, Int.
J. Environ. Res. Public Health 2 (2005) 74–79.

56
J. Dai et al. / Chemico-Biological Interactions 183 (2010) 49–56
[12] Y.P. Wang, J. Yan, R.D. Beger, P.P. Fu, M.W. Chou, Metabolic activation of the
tumorigenic pyrrolizidine alkaloid, monocrotaline, leading to DNA adduct for-
mation in vivo, Cancer Lett. 226 (2005) 27–35.
[21] G. Lin, Y.Y. Cui, E.M. Hawes, Characterization of rat liver microsomal metabolites
of clivorine, an hepatotoxic otonecine-type pyrrolizidine alkaloid, Drug Metab.
Dispos. 28 (2000) 1475–1483.
[13] M.W. Lame, A. Daniel Jones, D.W. Wilson, S.K. Dunston, H.J. Segall, Protein tar-
gets of monocrotaline pyrrole in pulmonary artery endothelial cells, J. Biol. Med.
275 (2000) 29091–29099.
[14] M.W. Lame, A. Daniel Jones, D.W. Wilson, H.J. Segall, Monocrotaline pyrrole tar-
gets proteins with and without cysteine residues in the cytosol and membranes
of human pulmonary artery endothelial cells, Proteomics 5 (2005) 4398–4413.
[15] G.J. Gordon, W.B. Coleman, J.W. Grisham, Induction of cytochrome P450
enzymes in the livers of rats treated with the pyrrolizidine alkaloid retrorsine,
Exp. Mol. Pathol. 69 (2000) 17–26.
[16] M.J. Reid, M.W. Lame, D. Morin, D.W. Wilson, H.J. Segall, Involvement
of cytochrome P450 3A in the metabolism and covalent binding of 14C-
monocrotaline in rat liver microsomes, J. Biochem. Mol. Toxicol. 12 (1998)
157–166.
[17] R.R. Dalvi, Dose-related inhibition of the drug-metabolizing enzymes of rat liver
by the pyrrolizidine alkaloid, monocrotaline, J. Pharm. Pharmacol. 39 (1987)
386–388.
[18] A.R. Mattocks, R. Jukes, J. Brown, Simple procedures for preparing putative toxic
metabolites of pyrrolizidine alkaloids, Toxicon 27 (1989) 561–567.
[19] C.C. Yan, R.A. Cooper, R.J. Huxtable, The comparative metabolism of the
four pyrrolizidine alkaloids, seneciphylline, retrorsine, monocrotaline, and tri-
chodesmine in the isolated, perfused rat liver, Toxicol. Appl. Pharmacol. 133
(1995) 277–284.
[20] G. Lin, Y.Y. Cui, E.M. Hawes, Microsomal formation of a pyrrolic alcohol glu-
tathione conjugate of clivorine. Firm evidence for the formation of a pyrrolic
metabolite of an otonecine-type pyrrolizidine alkaloid, Drug Metab. Dispos. 26
(1998) 181–184.
[22] L.L. Von Moltke, D.J. Greenblatt, J.M. Grassi, B.W. Granda, K. Venkatakrish-
nan, J. Schmider, J.S. Harmatz, R.I. Shader, Multiple human cytochromes
contribute to biotransformation of dextromethorphan in-vitro: role of CYP
2C9, CYP 2C19, CYP 2D6, and CYP 3A, J. Pharm. Pharmacol. 50 (1998) 997–
1004.
[23] R.A. Cooper, R.J. Huxtable, The relationship between reactivity of metabolites
of pyrrolizidine alkaloids and extrahepatic toxicity, Proc. West Pharmacol. Soc.
42 (1999) 13–16.
[24] S. Rendic, F.J. Di Carlo, Human cytochrome P450 enzymes: a status report sum-
marizing their reactions, substrates, inducers, and inhibitors, Drug Metab. Rev.
29 (1997) 413–580.
[25] W.G. Chung, C.L. Miranda, D.R. Buhler, A cytochrome P450 2B form is the major
bioactivation enzyme for the pyrrolizidine alkaloid senecionine in guinea pig,
Xenobiotica 25 (1995) 929–939.
[26] J. Ko, Z. Desta, D.A. Flockhart, Human N-demethylation of (S)-mephenytoin
by cytochrome P450S 2C9 and 2B6, Drug Metab. Dispos. 26 (1998) 775–
778.
[27] J. Palamanda, W. Feng, C. Lin, A.A. Nomeir, Stimulation of tolbutamide hydrox-
ylation by acetone and acetonitrile in human liver microsomes and in a
cytochrome P-450 2C9-reconstituted system, Drug Metab. Dispos. 28 (2000)
38–43.
[28] A. Yu, R.L. Haining, Comparative contribution to dextromethorphan
metabolism by cytochrome P450 isoforms in vitro: can dextromethor-
phan be used as a dual probe for both CYP 2D6 and CYP 3A activities? Drug
Metab. Dispos. 29 (2001) 1514–1520.