Carcinogenesis vol.22 no.5 pp.805–811, 2001
Mrp2-deficiency in the rat impairs biliary and intestinal excretion
and influences metabolism and disposition of the food-derived
carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine
(PhIP)
Christoph G.Dietrich2, D.Rudi de Waart, Roel Ottenhoff,
Albert H.Bootsma1, Albert H.van Gennip1 and
Ronald P.J.Oude Elferink3
Laboratory for Experimental Hepatology, Department of Gastroenterology
and 1Department of Genetic Metabolic Diseases, Academic Medical Centre,
F0-116, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands
2Present
address: Med. Klinik III, Universitätsklinikum der RWTH,
Pauwelsstraβe 30, 52057 Aachen, Germany
3To
whom correspondence should be addressed
Email: [email protected]
While metabolism of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), the most abundant food-derived
heterocyclic amine and carcinogen, has been studied extensively in several species, transport of this compound and its
metabolites has not been defined yet. Therefore we studied
metabolism and disposition of PhIP in Wistar and Mrp2deficient TR– rats to determine the role of Mrp2 in the
defence against this compound. In the first 2 h after
intravenous dosing, total excretion of PhIP and its metabolites in bile was >4-fold reduced in TR– rats compared with
Wistar rats, while excretion in the urine of the TR– rat
was 1.8-fold higher. This difference was the result of an
almost complete absence of secretion of glucuronidated
metabolites but also a reduced level of secretion of
unchanged PhIP into bile of the TR– rat. Direct intestinal
excretion of unmetabolized PhIP was 3-fold higher in
Wistar versus TR– rats. As a consequence, PhIP tissue
levels in the liver were 1.7-fold higher in TR– rats, and
tissue binding of PhIP, determined after ethanol extraction,
was elevated by a similar magnitude. Mrp2-mediated
transport of the parent compound PhIP is glutathione
(GSH)-dependent, because GSH depletion by L-buthionine[S,R]-sulfoximine (BSO) treatment in Wistar rats reduced
intestinal secretion to the same level as that in TR– rats.
TR– rats produced less glucuronides and 4’-OH-PhIP in
the 2 h following PhIP administration. We conclude that
Mrp2 protects against the carcinogen PhIP by biliary
excretion of the parent compound and all major phase-II
metabolites, but, more importantly, also by direct extrusion
of the parent compound from the gut mucosa.
Introduction
2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) is
the most abundant heterocyclic amine formed during the
cooking, frying and grilling of meat (1). It was first described
in 1986 (2) and was shown to induce colon and breast
carcinomas in rats (3). Data from several animal models led
to the conclusion that PhIP may be involved in human
carcinogenesis (4).
The metabolism of this compound has been studied extensively in different species (5–14). In general, after intake, PhIP
is metabolized in the liver to its active form N-OH-PhIP
mainly by CYP1A1 and CYPA2 (15). N-OH-PhIP is further
metabolized to N-acetoxy-PhIP (by N-acetyltransferase) and
N-sulfoxy-PhIP (by sulfotransferases). The latter two compounds can form DNA adducts (16) and are thought to be the
main mutagenic metabolites of PhIP. N-OH-PhIP can also be
further metabolized to N-OH-PhIP-glucuronide [depending on
the species, mainly N2- or N3-glucuronide (17)], which is a
detoxification metabolite (8). The main detoxification pathway
though is the conversion of PhIP to 4’-OH-PhIP, which
subsequently can be glucuronidated and sulfated (5). Conjugation with glutathione was found in PCB pretreated hepatocytes
but was never observed in vivo (18).
Much less is known about the transport of PhIP and its
metabolites. Transport of both the parent compound and/or its
metabolites is necessary for absorption, disposition and for
uptake in the cells of peripheral organs. In a very recent
publication it was shown that there might be active secretion
of PhIP from Caco-2 cells (19); inhibition studies indirectly
implied MRP2 as the most probable transporter involved.
MRP2 was initially identified as the canalicular multispecific
organic anion transporter (cMOAT) in the canalicular membrane of hepatocytes (20). First studies showed that it transports a wide range of conjugated compounds, including
glucuronidated, sulfated molecules and molecules conjugated
to glutathione (21). Uncharged compounds have not been
shown to be substrates for MRP2 yet and such a finding would
need clarification of the transport mechanisms involved. The
homologous transporter MRP1, which is expressed in a variety
of organs except liver, has been shown to transport uncharged
compounds in co-transport with glutathione (GSH) (22), a
mechanism which could account for MRP2-mediated PhIP
transport as well. Since PhIP is, after hydroxylation, subsequently glucuronidated and sulfated, MRP2 is a candidate
for biliary excretion of the parent compound and several of
its metabolites.
We therefore studied metabolism, disposition and excretion
of PhIP in Mrp2-deficient TR– rats and GSH-depleted Wistar
rats in comparison to control animals, which were untreated
Wistar rats. We show here that Mrp2 mediates biliary excretion
of PhIP and all its major phase-II metabolites and excretion
of PhIP into the lumen of the small intestine. Furthermore we
show that deficiency of Mrp2 in the rat, resulting in impaired
biliary secretion, influences metabolism and leads to higher
tissue binding of PhIP and its metabolites in the liver.
Materials and methods
Abbreviations: BSO, buthionine-sulfoximine; CYP, cytochrome P 450; GSH,
glutathione; HPLC, high performance liquid chromatography; Mrp, multidrug
resistance protein; OH, hydroxy; PhIP, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine.
© Oxford University Press
Chemicals
[14C]PhIP (specific activity 10 mCi/mmol) and unlabeled Nitro-PhIP were
obtained from Toronto Research Chemicals (North York, Ontario, Canada).
805
C.G.Dietrich et al.
Unlabeled PhIP was bought from ICN (Costa Mesa, CA, USA). All batches
were found to be ⬎98% pure by HPLC.
Sulfatase H-5, saccharo-1,4-lactone, L-buthionine-[S,R]-sulfoximine (BSO)
and glucuronidase (Escherichia coli Type VII-A and bovine liver Type B3)
were from Sigma (St Louis, MO, USA). All solvents used for HPLC separation
(HPLC or analysis grade) were from Merck (Darmstadt, Germany).
Animals
Female Wistar rats were obtained from Harlan-CPB (Zeist, The Netherlands).
Age- and weight-matched female TR– rats were from our own breeding colony
which has been characterized previously (23). Rats were housed in cages with
a 12 h light/dark cycle and given access to food and water ad libitum. Rats
were used for experiments at age 12–16 weeks (200–250 g weight). All
animal experiments in this work have been carried out in accordance with the
Declaration of Helsinki.
Meriden, CT, USA). The mobile phase (0.4 ml/min) consisted of (A) 10%
methanol in 0.1% diethylamine pH 5.0, 6.0 or 7.0 and (B) 90% methanol in
0.1% diethylamine with the same pH. After 5 min isocratic flow with 10%
(B) the gradient was linearly adjusted to 25% (B) at 15 min, then to 75% (B)
at 30 min, where it was held constant until 35 min. Fractions for fluorescence
spectra and mass spectroscopy (MS) identification were collected in 30 s
intervals.
Metabolism and excretion studies (four rats each group)
After anesthesia [50 mg pentobarbital sodium (Nembutal)/kg i.p.], the jugular
vein, the common bile duct and the urine bladder were cannulated. After
cannulation a continuous infusion of tauroursodeoxycholate (100 nmol/min/
100 g rat) in the jugular vein was started to maintain bile flow. [14C]PhIP
(2 µCi/200 g rat) was injected in the jugular vein and bile and urine samples
were obtained over a period of 2 h after injection.
Metabolite identification
HPLC buffers at three different pHs (pH 5.0, 6.0 and 7.0) were used to
determine the relative retention times of all metabolites under different
conditions; these data were compared with previously reported metabolite
characteristics at these pH values (7,14,25,26). Treatment with different
enzymes as described above already led to a tentative identification of most
metabolites. We used highly concentrated bile and urine samples from rats
exposed intravenously to very high doses of PhIP (at the limit of solubility
at pH 5.0, where PhIP is best soluble) to determine UV and fluorescence
spectra of each metabolite after separation by HPLC. Our spectra were
compared with those in the literature (13,17,18,25,27–29). In a last step all
metabolites identified with the methods described were subjected to MS–MS
to confirm this identification. For further details of the results from the
metabolite identification studies see Table I. N-OH-PhIP was identified by a
standard synthesized from Nitro-PhIP as reported earlier by Lin et al. (30).
BSO pretreatment
To inhibit glutathione (GSH) synthesis, we treated Wistar rats with 4 mmol
BSO/kg 3 and 1.5 h i.p. prior to beginning of the experiments. Efficiency of
BSO pretreatment was controlled by determining GSH levels [according to
the method of Tietze (24)] from liver and gut tissue pieces immediately
homogenized after killing.
UV and fluorescence spectra
UV spectra were taken online on a diode array detector (Applied Biosystems
1000S, Cheshire, UK) from 240 to 370 nm. Fluorescence spectra were
recorded on a Aminco Bowman Luminescence Spectrometer (Spectronic,
Cambridge, UK) with a fixed emission–excitation wavelength difference of
55 nm according to Crofts et al. (29).
Organs
After sample collection in each experiment, animals were killed by bleeding
from the aorta and a 4 ml blood sample was collected. Liver, whole intestine
(with feces), kidney, pancreas, heart and lungs were excised. Feces were
separated from the intestine before storage. All material was stored at –80°C.
MS identification of metabolites
Mass spectra were recorded using a Quattro 2 triple quadruple mass spectrometer from Micromass (Manchester, UK), equipped with an electrospray
source (ESP). Nitrogen was used as the spraying and drying gas while the
source temperature was kept at 80°C. Fractions from the HPLC separation,
dried under nitrogen, were reconstituted in methanol/water (1:1) and injected
directly. Positive ion scans were taken using a capillary voltage of 3500 V,
negative ion scans with one of 3000 V. The cone voltage was optimized for
each fraction, generally around 35 V. After that the collision energy, used for
the daughter-ion scans, was optimized individually, generally around 25 eV.
Argon was used as collision gas at a pressure of 2.5 µbar.
Sample treatment
Directly after sampling aliquots of urine and bile samples were injected on
HPLC for separation of metabolites and used for counting. Weighed portions
of the organs from at least two different sites were completely solubilized in
Soluene 350 (Packard, Meriden, CT, USA) (1 ml/100 mg tissue) to determine
tissue levels of radioactivity. Two additional portions were homogenized in
ethanol (1:1, w/v), centrifuged and, after another step of washing with ethanol
and centrifuging, the pellet was dissolved in Soluene as described above to
determine tissue binding of radioactivity. The supernatant and wash-out from
these steps were counted separately. All organ samples were decolorized with
10% (v/v) H2O2 and mixed with scintillation fluid (to a total volume of
10 ml) before counting. For determination of metabolites in liver tissue
homogenates, 2–3 g tissue was homogenized in ethanol, centrifuged and, after
another step of washing with ethanol and centrifuging, the supernatant was
dried under nitrogen. The sample was reconstituted in 2 ml ethanol again,
centrifuged and again dried. Then the sample was dissolved in 20 µl water
and injected directly on HPLC.
Bile and urine samples treated with enzyme for metabolite identification
(see below) were passed over a C18-OASIS 30 mg column (Waters, Milford,
MA, USA). Metabolites were eluted with methanol, dried and then reconstituted in water and injected. Total radioactivity recovery with this method
was ⬎98%.
Feces were homogenized in 50% methanol in water (1:5, w/v), the
homogenate was counted and after centrifugation the supernatant was dried,
reconstituted in water and analyzed by HPLC. Radioactivity recovery with
this method was around 95%.
For detection of metabolites in blood we used the method of Kaderlik
et al. (25).
Enzyme treatment
β-Glucuronidase (3000 U/ml, bovine as well as E.coli) and sulfatase
(20 U/ml, in presence of 1 mM saccharo-1,4-lactone to inhibit glucuronidase activity) were dissolved in 15 mM acetate buffer pH 5.0 (E.coli
glucuronidase in potassium phosphate buffer pH 6.8) and 1 ml was added to
20 µl sample. Incubation lasted 5 h at 37°C.
HPLC
Analysis was done on a system with Gynkotek-Pumps (Germering, Germany),
connected to a Rheodyne 7125 injection valve (Cotati, CA, USA) and an
Inertsil 5 ODS 3 column (Chrompack, Bergen op Zoom, The Netherlands).
Injection volume was 20 µl. We used an UV-Detector Spectroflow 757 (Kratos,
Ramsey, NJ, USA) at 315 nm, a fluorescence detector (Jasco
FP-920, Tokyo, Japan) and an on-line Scintillation Analyzer (515TR, Packard,
806
Results
Excretion of PhIP and its metabolites 2 h after intravenous
dosing of PhIP
In the 2 h following intravenous administration of PhIP, the
percentage of total radioactivity excreted into bile for Wistar
and TR– rats was 41.4 ⫾ 7.4 and 9.4 ⫾ 2.7%, respectively
(Figure 1, P ⬍ 0.001). The excretion of radioactivity in the
urine was ~2-fold higher in TR– rats (24.6 ⫾ 7.3%) than in
Wistar rats (13.8 ⫾ 4.4%; P ⬍ 0.05; Figure 1). Nevertheless,
the total body elimination of radioactivity was still significantly
lower in TR– rats compared with Wistar rats (34 ⫾ 9.7 versus
55.2 ⫾ 4.7%, P ⬍ 0.01; Figure 1).
Metabolites in bile and urine were analyzed using HPLC
(Figure 2). This revealed four major and several minor metabolites. All major metabolites and one minor metabolite could be
identified by relative retention times, UV and fluorescence
spectra, enzyme treatment with glucuronidase and sulfatase
and electrospray/tandem mass spectrometry (see Materials and
methods and Table I for details). Quantification of these
metabolites (Figure 3) showed that in the bile of TR– rats the
glucuronides of 4’-OH-PhIP and the N2-glucuronide of
N-OH-PhIP were virtually absent (P ⬍ 0.001), while the N3glucuronide of N-OH-PhIP was strongly reduced (0.88 ⫾ 0.31
versus 11.5 ⫾ 3% in the TR– and Wistar rats, respectively,
P ⫽ 0.003). The hepatobiliary secretion of the original
compound was also significantly impaired in TR– rats
(0.7 ⫾ 0.33 and 3.09 ⫾ 0.59% for TR– and Wistar rats,
respectively; P ⬍ 0.001) as was the secretion of 4’-OH-PhIP
Mrp2 deficiency and PhIP metabolism
Table I. Retention times of major metabolites on HPLC (pH 6.0) and data of enzyme treatment, UV and fluorescence spectra and MS identification (see
Materials and methods)
Nr.
Retention
time (min)
Metabolite
Enzyme treatment
UV max
(nm)
Fluoresc. max
(nm)
Mass (MS)
1/2
3
4
5
6
7
6/8
20
23
25
26
29
4’-OH-PhIP-Gluc
4’-PhIP-sulfate
Not identified
4’-OH-PhIP
N-OH-PhIP-N2-gluc.
N-OH-PhIP-N3-gluc.
Gluc.→4’-OH-PhIP
Sulf.→4’-OH-PhIP
Sulf.→?
–
B-gluc.→N-OH-PhIP
Gluc.→N-OH-PhIP
328
324
–
315
317
328
326
322
–
326
–
–
–
8
30
31
Not identified
PhIP
–
–
–
317
–
318
–
ESP- 319
–
ESP⫹ 241
–
ESP⫹ 417
ESP– 415
–
ESP⫹ 225
Peak numbers refer to Figure 3. Under the applied conditions, the indicated enzyme treatment resulted in a shift of the peak to the position of the indicated
product. Metabolites at 23, 26 and 30 min were too minor to be identified by MS. N-OH-PhIP was identified by a synthesized standard (retention time ~30.6
min).
Abbreviations: gluc., glucuronidase treatment; b-gluc., bovine glucuronidase treatment; sulf., sulfatase treatment; ESP, electrospray ionization.
Fig. 1. Excretion of radioactive material in bile (open bars) and urine
(closed bars) in the 2 h following intravenous administration of [14C]PhIP in
Wistar, TR– and BSO pretreated Wistar rats (BSO). Data represent means ⫾
SD (n ⫽ 4 in all groups). Significance compared with untreated Wistar rats:
⫹, P ⬍ 0.05; *, P ⬍ 0.01; **, P ⬍ 0.001.
Fig. 2. Representative HPLC chromatograms (pH 6) of bile samples 30 min
after intravenous [14C]PhIP administration. Above Wistar rat, below TR–
rat. For identification data see respective peak number in Table I.
(P ⫽ 0.02). The transport of unmetabolized PhIP into bile was
not significantly changed (P ⫽ 0.15; Figure 3A) in BSO pretreated Wistar rats, while glucuronides in bile were slightly
but significantly reduced.
In urine (Figure 3B), the differences were much less striking.
There was increased urinary excretion of glucuronides and
Fig. 3. Amount of identified metabolites excreted in (A) bile, (B) urine and
(C) total elimination of each metabolite in 2 h after intravenously
administered [14C]PhIP in Wistar (closed bars), TR– (open bars) and BSOpretreated Wistar rats (hatched bars). Significance levels (⫹, P ⬍ 0.05;
*, P ⬍ 0.01; **, P ⬍ 0.001) of TR– and BSO pretreated Wistar rats are
compared with untreated Wistar rats. Data are given as means ⫾ SD (n ⫽ 4
in all groups).
807
C.G.Dietrich et al.
Fig. 4. Amount of radioactivity in homogenized luminal contents of small
intestine (closed bars) and colon (without cecum, open bars) of Wistar, TR–
and BSO pretreated Wistar (BSO) rats 2 h after intravenously administered
[14C]PhIP. Significance levels (*, P ⬍ 0.01) of TR– and BSO pretreated rats
are compared with untreated Wistar rats. Data are given as means ⫾ SD
(n ⫽ 4 in all groups).
sulfate of 4⬘-OH-PhIP in TR– rats. This was not observed in
the BSO pretreated Wistar rats. Total body elimination (Figure
3C) of all glucuronides, 4⬘-OH-PhIP and the original compound
PhIP was reduced in TR– rats. There was a tendency
(P ⫽ 0.055) in TR– rats to excrete more 4⬘-PhIP-sulfate. The
BSO pretreated Wistar rats also excreted in total less amounts
of all glucuronides.
The pattern of metabolites detected in blood did not differ
from that in urine (data not shown). The activated compound
N-OH-PhIP was not detected in blood, bile or urine. Analysis
of bile and urine was done at several time points after injection,
but over time there were no significant changes in the relative
amount of metabolites in either Wistar or TR– rats except for
a minor metabolite at 30 min retention time which was not
identified further.
In the same experiment we also collected the total luminal
content of small intestine and colon (without cecum) separately.
Any radioactivity measured in this material must have been
secreted by the mucosa because PhIP was injected intravenously while bile was continuously diverted during the 2 h
experimental period. There was much less radioactivity in the
small intestinal luminal contents of the TR– rat than in that of
the Wistar rats (0.28 ⫾ 0.19% versus 0.95 ⫾ 0.22%, respectively, P ⫽ 0.007), while the difference in the colon showed
only a tendency (P ⫽ 0.074; Figure 4). In BSO pretreated
Wistar rats active secretion in the intestine was reduced to a
similar extent as in the TR– rat (0.28 ⫾ 0.26%, P ⫽ 0.008).
We analyzed the supernatant of the homogenized intestinal
contents with HPLC and found that 75% of the radioactivity
consisted of the original compound PhIP, the rest was 4⬘-OHPhIP. This was not different between Wistar, BSO pretreated
Wistar and TR– rats.
Tissue levels and tissue binding of radioactivity 2 h after
injection
To determine the distribution of radioactivity remaining in the
body at 2 h after intravenous administration, we solubilized
aliquots of several organs (liver, kidney, small and large
intestine, pancreas, heart and lung) and determined radioactivity. There was no significant difference in whole organ radioactivity between TR– and Wistar except for the liver, where a
1.7-fold increased radioactivity was measured in TR– rats
808
Fig. 5. Total (left) and covalently bound (i.e. ethanol insoluble portion,
right) levels of radioactivity in the liver of Wistar (closed bars), TR–
(open bars) and BSO pretreated Wistar rats (hatched bars) at 2 h after
intravenous administration of [14C]PhIP. Significance levels (*, P ⬍ 0.01;
**, P ⬍ 0.001) of TR– and BSO pretreated rats are compared with
untreated Wistar rats. Data are given as means ⫾ SD (n ⫽ 4 in all groups).
Fig. 6. Amounts of metabolites (percentage of injected dose) in the ethanol
soluble fraction of liver homogenate in Wistar (closed bars) and TR– rats
(open bars) at 2 h after intravenous administration of [14C]PhIP. Liver
material was homogenized in ethanol and the supernatant was analyzed by
HPLC (see Materials and methods). Significance levels (*, P ⬍ 0.01;
**, P ⬍ 0.001) of TR– rats are compared with untreated Wistar rats. Data
are given as means ⫾ SD (n ⫽ 4 in all groups).
(Figure 5). Tissue binding of radioactivity in the TR– rat, as
determined by ethanol insolubility, was more than doubled
(P ⬍ 0.001; Figure 5). Tissue levels of radioactivity in liver
of the BSO pre-treated Wistar rats were not significantly
different from untreated Wistar rats, whereas tissue binding of
radioactivity in this organ was significantly elevated (2.15 ⫾
0.42 versus 1.32 ⫾ 0.22% for BSO pre-treated and untreated
Wistar rats, respectively; P ⫽ 0.01; Figure 5).
To determine the pattern of PhIP metabolites in the liver,
we analyzed the ethanol soluble fraction of liver homogenate
by HPLC (Figure 6). The portions of the respective metabolites
in Figure 6 are given as percentage of the total dose. All
metabolites are significantly elevated in the liver of the TR–
rat, with PhIP-sulfate ⬎5-fold elevated. Surprisingly, PhIP
itself is lowered in the TR– rat. Another metabolite eluting at
~15 min which was not detected in bile, urine or blood,
Mrp2 deficiency and PhIP metabolism
accounted for ~5% of the ethanol soluble fraction of all animals
and could not be identified further (data for this metabolite
not shown in Figure 6).
Discussion
Data from the present study demonstrate that rat Mrp2 mediates
the excretion of all major phase-II metabolites of PhIP. We
could show that directly for the glucuronides, the biliary
excretion of which was dramatically reduced. Although the
absolute amount of PhIP-sulfate excretion in bile is equal in
Wistar and TR– rats, we observed 5-fold elevated levels of
this metabolite in the liver as well as higher levels in blood
and urine. Thus, in TR– rats the same sulfate excretion is only
reached at higher intracellular concentrations. Based on the
known characteristics of Mrp2, this excretory function for
glucuronidated and sulfated conjugates is not surprising; in a
large number of studies this transporter was characterized
as having affinity for a broad spectrum of organic anions,
particularly glucuronides and glutathione conjugates (21,31).
In addition, this study shows that Mrp2-mediated transport
in kidney proximal tubules (32) is relatively unimportant
compared with glomerular filtration; urinary excretion of
glucuronides was increased rather than decreased in the TR–
rat. The higher phase-II metabolite levels in urine of the
TR– rats are a consequence of higher blood levels of these
metabolites. This is most likely due to basolateral secretion of
unexcreted glucuronides and sulfates out of the liver into the
plasma. The latter process might be mediated by Mrp3 in
the hepatocyte basolateral membrane, which also transports
xenobiotics (33). Importantly, Mrp3 expression was shown to
be increased in Mrp2-deficient rats (34,35), indicating that the
actual accumulation of PhIP and/or its metabolites may partly
be counteracted.
Though compensatory excretion of glucuronides occurred
in urine of the TR– rat, excretion of all glucuronides as well
as 4’-OH-PhIP, is significantly reduced in the Mrp2-deficient
rat (Figure 3). While the higher amounts of these metabolites
in liver parenchyme only indicate a delay in excretion, the
total amounts of excreted metabolites even then are significantly lower when these amounts in liver parenchyme (and
presumably gut parenchyme since glucuronidation takes place
there, too) are taken into consideration. This additionally
indicates a reduction or at least a delay in metabolism of PhIP
in Mrp2-deficiency and confirms the recent finding (36) that
the expression level of a transport protein can also influence
metabolic reactions, probably through product inhibition of
the accumulated metabolites.
Another important finding of our study, though not unexpected, is that the gut functions as a third elimination route for
PhIP. The significant difference in intestinal excretion between
Wistar and TR– rats points at a role for Mrp2 in this function.
HPLC analysis of the radioactivity found in the intestinal
luminal contents revealed mainly unmetabolized PhIP as the
excreted compound. The role of Mrp2 in transport of unmetabolized PhIP is underscored by the data from biliary excretion.
Both the liver and the small intestine excreted unmetabolized
PhIP, and in the TR– rat transport in both organs was strongly
reduced. Walle and Walle (19) recently suggested from inhibition studies in Caco-2 monolayers that MRP2 may contribute
to transport of the parent compound. In addition, it has been
shown very recently that Mrp2 is expressed in rat intestine
(37,38). Our results support these data and show functional
involvement of Mrp2 in hepatic and intestinal defence against
PhIP. The Mrp2-expressing gut functions as an active excretory
organ. Intestinal PhIP excretion accounts for over 1% of the
dose in the first 2 h in the whole gut of Wistar rats after
intravenous administration. This amount is comparable with
the amount of unmetabolized PhIP excreted by the liver and
suggests a role for Mrp2 in reducing oral bioavailability of
this compound. The colonic excretion of PhIP was not different
between Wistar and TR– rats which is in line with the almost
absent expression of Mrp2 in this part of the gut (37). Levels
of radioactivity in the colon in our experiment were not
influenced by the changed luminal levels in the small intestine,
since the experiment lasted only 2 h and the cecum and
contents of the cecum were not used for determination of
radioactivity.
We speculate that the transport of the parent compound
PhIP is taking place together with GSH, since PhIP itself is
neither conjugated nor negatively charged. We were not able
to demonstrate a significant effect of BSO on PhIP transport
from liver into bile, but transport in the small intestine was
significantly reduced. Within the experimental period it was
difficult to reduce GSH in rat liver to an extent lower than
25% of controls, a fact also observed by others (39,40). This
level might be enough to support PhIP transport sufficiently.
While BSO treatment reduced hepatic GSH content to 29%
of control in our experiment, the intestinal content could be
reduced to 14.5% of control. The lower GSH content of the
intestine might explain why PhIP transport was reduced in
this organ and not in the liver.
In the past it was shown that glutathione depletion caused
a 5-fold increase of DNA adducts in the liver (25) or 15-fold
increase in hepatocytes (8), respectively, in the latter case
without changing the metabolic profile. It was also shown that
a concentration of 40 mM GSH in vitro inhibited DNA
binding by N-acetoxy-PhIP (8). Our data suggest that, in
addition to the reduction of DNA adduct formation, GSH
may play a role in direct extrusion of the carcinogen. It is
therefore likely that there is addition of several effects in
GSH action. Because GSH itself is also a substrate for
Mrp2, TR– rats lack biliary GSH secretion and, as a
consequence, exhibit higher hepatic GSH levels [~157% of
the Wistar rat (31)]. These high GSH levels may protect
against increased adduct formation. Consequently, one might
expect even higher tissue bound radioactivity in the liver of
the TR– rat if GSH levels were the same as in Wistar rats. In
BSO pretreated Wistar rats, liver levels of radioactivity are
equal to those found in untreated Wistar rats, compatible with
the fact that PhIP excretion into bile in these animals is not
impaired at GSH concentrations reached with our method of
BSO treatment. However, the ethanol insoluble fraction of
tissue radioactivity, representing covalently bound reactive
metabolites, is significantly higher in BSO pretreated than in
untreated Wistar rats, which possibly highlights the influence
of GSH on tissue binding. We speculate that the two described
effects of GSH on detoxification of PhIP in the liver have
different thresholds. While reduction to 29% of the normal
GSH concentration already increases tissue binding of reactive
metabolites, it still drives excretion of PhIP into bile. This
hypothesis also may explain the reduced total excretion of the
glucuronides of N-OH-PhIP in the BSO pre-treated Wistar
rats. It is not expected that GSH plays a role in transport of
glucuronides as these are charged compounds by themselves.
On the other hand, lower GSH levels will result in higher
809
C.G.Dietrich et al.
tissue binding of N-OH-PhIP derivatives, leaving less free NOH-PhIP available to be glucuronidated.
In summary, we propose a dual function of Mrp2 in the
defence against carcinogens like PhIP. It acts before phase-I
metabolism by pumping the unmetabolized xenobiotic out of
the gut mucosa and the liver. It also contributes to reduction
of the body level of toxins by rapid excretion of metabolites
which have been formed by phase I enzymes (cytochrome
P450) and phase II enzymes (glucuronyltransferases and sulfotransferases) to inhibit further damage through systemic
deconjugation. By these two functions Mrp2 reduces the
number of reactive molecules in the whole body, thereby
probably reducing adduct formation.
Complete MRP2 deficiency in humans only occurs in
rare Dubin–Johnson patients. Nevertheless it is expected that
polymorphisms in the MRP2 gene may lead to reduced
expression or activity of the protein in otherwise completely
healthy subjects. Such polymorphisms have already been
shown to occur in the MDR1 gene (41), which has a functionally
similar task in the gut epithelium. Ten percent of the normal
level of MRP2 is sufficient to prevent jaundice in the rat (42),
but at the same time such a reduction will significantly reduce
excretion of substrates such as PhIP. In so far, MRP2 could
become a marker of cancer susceptibility in the future.
Acknowledgements
The work of C.G.D. was supported by a fellowship grant from the START
fund (Nr. 92/98, Technical University of Aachen, Germany).
References
1. Skog,K.I., Johansson,M.A. and Jagerstad,M.I. (1998) Carcinogenic
heterocyclic amines in model systems and cooked foods: a review on
formation, occurrence and intake. Food Chem. Toxicol., 36, 879–896.
2. Felton,J.S., Knize,M.G., Shen,N.H., Lewis,P.R., Andresen,B.D., Happe,J.
and Hatch,F.T. (1986) The isolation and identification of a new mutagen
from fried ground beef: 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine
(PhIP). Carcinogenesis, 7, 1081–1086.
3. Ito,N., Hasegawa,R., Sano,M., Tamano,S., Esumi,H., Takayama,S. and
Sugimura,T. (1991) A new colon and mammary carcinogen in cooked
food, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP). Carcinogenesis, 12, 1503–1506.
4. Adamson,R.H. and Thorgeirsson,U.P. (1995) Carcinogens in foods:
heterocyclic amines and cancer and heart disease. Adv. Exp. Med. Biol.,
369, 211–220.
5. Alexander,J., Fossum,B.H., Reistad,R. and Holme,J.A. (1995) Metabolism
of the food carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine
(PhIP) in the rat and other rodents. Princess Takamatsu Symp., 23, 113–122.
6. Alexander,J., Wallin,H., Holme,J.A. and Becher,G. (1989) 4-(2-Amino-1methylimidazo[4,5-b]pyrid-6-yl)phenyl sulfate—a major metabolite of the
food mutagen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP)
in the rat. Carcinogenesis, 10, 1543–1547.
7. Buonarati,M.H., Roper,M., Morris,C.J., Happe,J.A., Knize,M.G. and
Felton,J.S. (1992) Metabolism of 2-amino-1-methyl-6-phenylimidazo[4,5b]pyridine (PhIP) in mice. Carcinogenesis, 13, 621–627.
8. Kaderlik,K.R., Mulder,G.J., Shaddock,J.G., Casciano,D.A., Teitel,C.H. and
Kadlubar,F.F. (1994) Effect of glutathione depletion and inhibition of
glucuronidation and sulfation on 2-amino-1-methyl-6-phenylimidazo[4,5b]pyridine (PhIP) metabolism, PhIP–DNA adduct formation and
unscheduled DNA synthesis in primary rat hepatocytes. Carcinogenesis,
15, 1711–1716.
9. Lang,N.P., Nowell,S., Malfatti,M.A., Kulp,K.S., Knize,M.G., Davis,C.,
Massengill,J., Williams,S., MacLeod,S., Dingley,K.H., Felton,J.S. and
Turteltaub,K.W. (1999) In vivo human metabolism of [2-14C]2-amino-1methyl-6-phenylimidazo[4,5-b]pyridine (PhIP). Cancer Lett., 143, 135–
138.
10. Lin,D.X., Lang,N.P. and Kadlubar,F.F. (1995) Species differences in
the biotransformation of the food-borne carcinogen 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine by hepatic microsomes and cytosols from
humans, rats and mice. Drug Metab. Dispos., 23, 518–524.
810
11. Malfatti,M.A., Kulp,K.S., Knize,M.G., Davis,C., Massengill,J.P.,
Williams,S., Nowell,S., MacLeod,S., Dingley,K.H., Turteltaub,K.W.,
Lang,N.P. and Felton,J.S. (1999) The identification of [2-(14)C]2-amino1-methyl-6-phenylimidazo[4,5-b]pyridine metabolites in humans.
Carcinogenesis, 20, 705–713.
12. McManus,M.E., Felton,J.S., Knize,M.G., Burgess,W.M., RobertsThomson,S., Pond,S.M., Stupans,I. and Veronese,M.E. (1989) Activation
of the food-derived mutagen 2-amino-1-methyl-6-phenylimidazo[4,5b]pyridine by rabbit and human liver microsomes and purified forms of
cytochrome P-450. Carcinogenesis, 10, 357–363.
13. Styczynski,P.B., Blackmon,R.C., Groopman,J.D. and Kensler,T.W. (1993)
The direct glucuronidation of 2-amino-1-methyl-6-phenylimidazo[4,5-b]
pyridine (PhIP) by human and rabbit liver microsomes. Chem. Res.
Toxicol., 6, 846–851.
14. Watkins,B.E., Esumi,H., Wakabayashi,K., Nagao,M. and Sugimura,T.
(1991) Fate and distribution of 2-amino-1-methyl-6-phenylimidazo[4,5b]pyridine (PhIP) in rats. Carcinogenesis, 12, 1073–1078.
15. Crofts,F.G., Sutter,T.R. and Strickland,P.T. (1998) Metabolism of 2-amino1-methyl-6-phenylimidazo[4,5-b]pyridine by human cytochrome P4501A1,
P4501A2 and P4501B1. Carcinogenesis, 19, 1969–1973.
16. Buonarati,M.H., Turteltaub,K.W., Shen,N.H. and Felton,J.S. (1990) Role
of sulfation and acetylation in the activation of 2-hydroxyamino-1-methyl6-phenylimidazo[4,5-b]pyridine to intermediates which bind DNA. Mutat.
Res., 245, 185–190.
17. Kadlubar,F., Kaderlik,R.K., Mulder,G.J. et al. (1995) Metabolic activation
and DNA adduct detection of PhIP in dogs, rats and humans in relation
to urinary bladder and colon carcinogenesis. Princess Takamatsu Symp.,
23, 207–213.
18. Alexander,J., Wallin,H., Rossland,O.J., Solberg,K.E., Holme,J.A.,
Becher,G., Andersson,R. and Grivas,S. (1991) Formation of a glutathione
conjugate and a semistable transportable glucuronide conjugate of N2oxidized species of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine
(PhIP) in rat liver. Carcinogenesis, 12, 2239–2245.
19. Walle,U.K. and Walle,T. (1999) Transport of the cooked-food mutagen 2amino-1-methyl-6-phenylimidazo-[4,5-b]pyridine (PhIP) across the human
intestinal Caco-2 cell monolayer: role of efflux pumps. Carcinogenesis,
20, 2153–2157.
20. Paulusma,C.C., Bosma,P.J., Zaman,G.J., Bakker,C.T., Otter,M.,
Scheffer,G.L., Scheper,R.J., Borst,P. and Oude Elferink,R.P. (1996)
Congenital jaundice in rats with a mutation in a multidrug resistanceassociated protein gene. Science, 271, 1126–1128.
21. Oude Elferink,R.P., Meijer,D.K., Kuipers,F., Jansen,P.L., Groen,A.K. and
Groothuis,G.M. (1995) Hepatobiliary secretion of organic compounds;
molecular mechanisms of membrane transport. Biochim. Biophys. Acta,
1241, 215–268.
22. Loe,D.W., Deeley,R.G. and Cole,S.P. (1998) Characterization of vincristine
transport by the M(r) 190,000 multidrug resistance protein (MRP): evidence
for cotransport with reduced glutathione. Cancer Res., 58, 5130–5136.
23. Jansen,P.L., Peters,W.H. and Lamers,W.H. (1985) Hereditary chronic
conjugated hyperbilirubinemia in mutant rats caused by defective hepatic
anion transport. Hepatology, 5, 573–579.
24. Tietze,F. (1969) Enzymic method for quantitative determination of
nanogram amounts of total and oxidized glutathione: applications to
mammalian blood and other tissues. Anal. Biochem., 27, 502–522.
25. Kaderlik,K.R., Minchin,R.F., Mulder,G.J., Ilett,K.F., Daugaard-Jenson,M.,
Teitel,C.H. and Kadlubar,F.F. (1994) Metabolic activation pathway for
the formation of DNA adducts of the carcinogen 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine (PhIP) in rat extrahepatic tissues.
Carcinogenesis, 15, 1703–1709.
26. Malfatti,M.A., Connors,M.S., Mauthe,R.J. and Felton,J.S. (1996) The
capability of rat colon tissue slices to metabolize the cooked-food
carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine. Cancer
Res., 56, 2550–2555.
27. Wallin,H.,
Mikalsen,A.,
Guengerich,F.P.,
Ingelman-Sundberg,M.,
Solberg,K.E., Rossland,O.J. and Alexander,J. (1990) Differential rates of
metabolic activation and detoxication of the food mutagen 2-amino1-methyl-6-phenylimidazo[4,5-b]pyridine by different cytochrome P450
enzymes. Carcinogenesis, 11, 489–492.
28. Kaderlik,K.R., Mulder,G.J., Turesky,R.J., Lang,N.P., Teitel,C.H.,
Chiarelli,M.P. and Kadlubar,F.F. (1994) Glucuronidation of N-hydroxy
heterocyclic amines by human and rat liver microsomes. Carcinogenesis,
15, 1695–1701.
29. Crofts,F.G., Strickland,P.T., Hayes,C.L. and Sutter,T.R. (1997) Metabolism
of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) by human
cytochrome P4501B1. Carcinogenesis, 18, 1793–1798.
30. Lin,D., Kaderlik,K.R., Turesky,R.J., Miller,D.W., Lay,J.O.Jr and
Kadlubar,F.F. (1992) Identification of N-(deoxyguanosin-8-yl)-2-amino-1-
Mrp2 deficiency and PhIP metabolism
methyl-6-phenylimidazo [4,5-b]pyridine as the major adduct formed by
the food-borne carcinogen, 2-amino-1-methyl-6-phenylimidazo[4,5b]pyridine, with DNA. Chem. Res. Toxicol., 5, 691–697.
31. Paulusma,C.C., van Geer,M.A., Evers,R., Heijn,M., Ottenhoff,R., Borst,P.
and Oude Elferink,R.P. (1999) Canalicular multispecific organic anion
transporter/multidrug resistance protein 2 mediates low-affinity transport
of reduced glutathione. Biochem. J., 338, 393–401.
32. Schaub,T.P., Kartenbeck,J., Konig,J., Vogel,O., Witzgall,R., Kriz,W. and
Keppler,D. (1997) Expression of the conjugate export pump encoded by
the mrp2 gene in the apical membrane of kidney proximal tubules. J. Am.
Soc. Nephrol., 8, 1213–12121.
33. Kool,M., van der Linden,M., de Haas,M., Scheffer,G.L., de Vree,J.M.,
Smith,A.J., Jansen,G., Peters,G.J., Ponne,N., Scheper,R.J., Elferink,R.P.,
Baas,F. and Borst,P. (1999) MRP3, an organic anion transporter able to
transport anti-cancer drugs. Proc. Natl Acad. Sci. USA, 96, 6914–6919.
34. Ortiz,D.F., Li,S., Iyer,R., Zhang,X., Novikoff,P. and Arias,I.M. (1999)
MRP3, a new ATP-binding cassette protein localized to the canalicular
domain of the hepatocyte. Am. J. Physiol., 276, G1493–1500.
35. Hirohashi,T., Suzuki,H., Ito,K., Ogawa,K., Kume,K., Shimizu,T. and
Sugiyama,Y. (1998) Hepatic expression of multidrug resistance-associated
protein-like proteins maintained in eisai hyperbilirubinemic rats. Mol.
Pharmacol., 53, 1068–1075.
36. Morrow,C.S., Smitherman,P.K. and Townsend,A.J. (2000) Role of
multidrug-resistance protein 2 in glutathione S-transferase P1-1-mediated
resistance to 4-nitroquinoline 1-oxide toxicities in HepG2 cells. Mol.
Carcinog., 29, 170–178.
37. Gotoh,Y., Suzuki,H., Kinoshita,S., Hirohashi,T., Kato,Y. and Sugiyama,Y.
(2000) Involvement of an organic anion transporter (Canalicular
multispecific organic anion Transporter/Multidrug resistance-associated
protein 2) in gastrointestinal secretion of glutathione conjugates in rats.
J. Pharmacol. Exp. Ther., 292, 433–439.
38. Mottino,A.D., Hoffman,T., Jennes,L. and Vore,M. (2000) Expression and
localization of multidrug resistant protein mrp2 in rat small intestine.
J. Pharmacol. Exp. Ther., 293, 717–723.
39. Minchinton,A.I., Rojas,A., Smith,K.A., Soranson,J.A., Shrieve,D.C.,
Jones,N.R. and Bremner,J.C. (1984) Glutathione depletion in tissues after
administration of buthionine sulphoximine. Int. J. Radiat. Oncol. Biol.
Phys., 10, 1261–1264.
40. Drew,R. and Miners,J.O. (1984) The effects of buthionine sulphoximine
(BSO) on glutathione depletion and xenobiotic biotransformation. Biochem.
Pharmacol., 33, 2989–2994.
41. Hoffmeyer,S., Burk,O., von Richter,O., Arnold,H.P., Brockmoller,J.,
Johne,A., Cascorbi,I., Gerloff,T., Roots,I., Eichelbaum,M. and
Brinkmann,U. (2000) Functional polymorphisms of the human multidrugresistance gene: multiple sequence variations and correlation of one allele
with P-glycoprotein expression and activity in vivo. Proc. Natl Acad. Sci.
USA, 97, 3473–3478.
42. Paulusma,C.C., Kothe,M.J., Bakker,C.T., Bosma,P.J., van Bokhoven,I., van
Marle,J., Bolder,U., Tytgat,G.N. and Oude Elferink,R.P. (2000) Zonal
down-regulation and redistribution of the multidrug resistance protein 2
during bile duct ligation in rat liver. Hepatology, 31, 684–693.
Received October 19, 2000; revised December 29, 2000;
accepted February 2, 2001
811