Abstract
The induction of hepatic and intestinal cytochrome P450, NAD(P)H:quinone oxidoreductase (QOR), glutathioneS-transferase (GST), and UDP-glucuronosyltransferase (UGT) activities by intragastric administration of 1,7-phenanthroline, 2,2′-dipyridyl, and oltipraz has been investigated in rats. In the liver, all three compounds induced phase II drug-metabolizing enzymes without inducing overall cytochrome P450 concentrations and, in a direct comparison, all agents induced the enzymes to a greater extent than did the same dose of tert-butyl-4-hydroxyanisole. With a 75 mg/kg daily, 3-day regimen, UGT, GST, and QOR activities were induced by all compounds. The changes in hepatic GST, QOR, and UGT activities induced by N-heterocyclic compounds were accompanied by increases in the amounts of mRNA for GST Ya (2–2.4-fold), QOR (1.6–2.8-fold), and the UGTs UGT2B1 (4–6-fold) and UGT1A6 (4–10-fold). Changes in the amounts of UGT2B1 mRNA and UGT1A6 mRNA were highly correlated (r = 0.9), but there was no correlation between changes in either UGT2B1 or UGT1A6 mRNA and GST Ya mRNA. No significant mRNA changes were elicited bytert-butyl-4-hydroxyanisole. Neither GST nor UGT activities were induced in the small intestinal mucosa by any agent. QOR activity was slightly induced by oltipraz. The data suggest that requirements for induction of phase II enzymes in the intestine are markedly different from requirements in the liver.
The metabolism of xenobiotics often involves a sequential oxidation-conjugation pathway. The oxidations, which are largely catalyzed by cytochrome P450s, can lead to bioactivation of the molecule, which can result in cell toxicity. Conjugation reactions are the major detoxification pathways in drug metabolism. The formation of conjugates improves the polarity and water solubility of a xenobiotic or oxidized metabolite, facilitating its excretion in urine or bile. Prominent conjugation reactions are glucuronidation, catalyzed by UGT1 enzymes located in the endoplasmic reticulum, and glutathione and sulfate conjugation, catalyzed by GST and sulfotransferase enzymes, respectively, located in the cytosol.
All drug-metabolizing reactions, with the possible exception of sulfate conjugation, are subject to induction by a wide range of chemically unrelated xenobiotics. The induction phenomenon most often involves increases in cytochrome P450-dependent phase I oxidation/activation reactions and one or more phase II conjugation/inactivation reactions. However, some xenobiotics elicit induction of only phase II enzyme activities. Such compounds can have utility as chemoprotectants because they enhance detoxification without increasing cytochrome P450-dependent bioactivation. Best known among compounds that preferentially, or selectively, induce phase II enzymes are antioxidants. Antioxidants coordinately elevate GST and QOR enzymes through a common transcriptional activation pathway controlled by an antioxidant response element. Because they are often concomitantly induced with QOR and GST, UGTs and microsomal epoxide hydrolase are believed to be induced through a similar mechanism. The induction process has been extensively investigated in the liver. Although the liver is the most important site of conjugation in the body, extrahepatic tissues also possess conjugation capabilities. Conjugation in the gastrointestinal tract is of importance because this tissue is often the first site of xenobiotic exposure and, therefore, is the first organ where metabolism can occur. Compared with the number of studies devoted to changes in the liver, induction of conjugation enzymes in the intestine has not been as exhaustively studied.
Induction studies with the antioxidant BHA have shown that drug-metabolizing enzymes in both the liver and intestine can be induced. In mice, induction in both organs has been demonstrated for QOR activity (Benson et al., 1980; Sparnins et al., 1982; De Long et al., 1985; Prochaska et al., 1985) and GST activity (De Long et al., 1985;Prochaska et al., 1985; Benson et al., 1979;Jaeschke and Wendel, 1985). Changes in mouse intestinal UGT and cytochrome P450 after BHA administration have not been reported. BHA elicits induction of UGT activity without affecting cytochrome P450 in murine liver (Cha et al., 1978, 1982; Hazelton et al., 1985; Cha and Heine, 1982). In rats, GST (Benson et al., 1979; Nijhoff and Peters, 1992) and UGT (Goon and Klaassen, 1992; Kashfi et al., 1994) activities are induced in both liver and intestine by BHA. QOR is induced by BHA in liver (Cha and Heine, 1982; Buetler et al., 1995), but there are conflicting reports on whether cytochrome P450 is (Cha and Heine, 1982;Buetler et al., 1995) or is not (Thompson et al., 1982) increased. Neither QOR nor cytochrome P450 has been examined in rat intestine after BHA administration.
Oltipraz, although thought to operate through the same antioxidant response element mechanism as BHA for the induction of hepatic enzymes (Prochaska and Talalay, 1988; Talalay et al., 1995), has shown some species differences in the ability to induce drug-metabolizing enzymes in the intestine. Hepatic and intestinal GST and hepatic and intestinal QOR activities were concomitantly induced by oltipraz in mice, but induction was confined to the liver in rats (Ansher et al., 1983). Rat liver UGT activity is also induced by oltipraz (Kensler et al., 1987) but, except for two recent reports of two UGT mRNA changes (Kessler and Ritter, 1997;Grove et al., 1997), induction of this enzyme activity has not been monitored in the intestine.
In this study, we have examined the relative effects of threeN-heterocycle-containing compounds (2,2′-dipyridyl, 1,7-phenanthroline, and oltipraz; formulae shown in fig.1), capable of preferentially inducing hepatic phase II drug-metabolizing enzymes (Franklin, 1991; Franklinet al., 1993), on the drug-metabolizing enzymes of the liver and intestine of rats. The inducing abilities of these compounds were directly compared with that of BHA by using the same dosing regimen.
Materials and Methods
Chemicals.
Oltipraz was supplied by Dr. C. Grubbs, University of Alabama (Birmingham, AL). 1,7-Phenanthroline was purchased from Aldrich Chemical Co. BHA, 2,6- dichlorophenolindophenol, 2,2′-dipyridyl, estrone, 7-ethoxyresorufin, FAD, glutathione, 4-hydroxybiphenyl, 2-hydroxybiphenyl, 7-methoxyresorufin, 3-methylcholanthrene, 4-methylumbelliferone, NADH, NADPH, 2-naphthol, 4-nitrophenol, 7-pentoxyresorufin, phenolphthalein, testosterone, and UDP-glucuronic acid were purchased from Sigma Chemical Co. 1-Naphthol was obtained from J. T. Baker Inc. [14C]UDP-glucuronic acid (293.6 mCi/mmol) and [α-32P]dCTP were purchased from DuPont-NEN. Morphine was purchased from Merck and Co., phenobarbital was purchased from Ganes Chemical Works, and 1-chloro-2,4-dinitrobenzene was purchased from Eastman Kodak Co. Trizol solution for RNA isolation was purchased from Gibco-BRL, agarose was from Fischer Scientific, Nytran membranes were from Schleicher and Schuell, and the Multiprime DNA labeling kit was from Amersham.
Animals, Treatment, and Microsomal Preparation.
Adult, male, Sprague-Dawley rats (125–275 g) were maintained under a 12-hr light/dark cycle in a temperature- and humidity-controlled environment, with free access to food and water. Rat treatments were all at 75 mg/kg daily for 3 days (with the exception of 3-methylcholanthrene, which was administered at 20 mg/kg), in aqueous or corn oil (3-methylcholanthrene or BHA) vehicle, by the intragastric (1,7-phenanthroline, 2,2′-dipyridyl, oltipraz, and BHA) or ip (phenobarbital and 3-methylcholanthrene) route. Untreated rats served as experimental controls for rats receiving compounds administered in aqueous vehicle, and animals receiving corn oil vehicle served as controls for BHA- and 3-methylcholanthrene-treated animals. All animals were euthanized 24 hr after the last dose. Microsomal and cytosolic fractions were prepared from intestine and liver as described previously (Harmsworth and Franklin, 1990; Franklin and Estabrook, 1971). Protein concentrations of the tissue fractions were determined by the procedure of Lowry et al. (1951). Tissue fractions were stored at −70°C until assays were performed.
Enzyme Assays.
Hepatic microsomal cytochrome P450 concentrations were determined from dithionite-reduced CO-difference spectra, using an extinction coefficient of 91 mM−1cm−1 (Omura and Sato, 1964). Cytochrome P450 concentrations in intestinal microsomes were determined from the dithionite-difference spectra of microsomes gassed with CO for 5 min, using an extinction coefficient of 100 mM−1cm−1 (Estabrook et al., 1972). 7-Methoxy-, 7-ethoxy-, and 7-pentoxyresorufin dealkylase activities were determined from the rates of resorufin formation detected by the absorbance increase at 572 nm (Klotz et al., 1984) and were used to indicate CYP1A2 (Nerurkar et al., 1993), CYP1A1/2 (Burke and Mayer, 1975), and CYP2B1 (Lubet et al., 1985), respectively.
Hepatic UGT activities toward 1-naphthol, morphine, testosterone, and estrone were determined by direct quantitation of glucuronide formation by reverse-phase HPLC (Liu and Franklin, 1984), and UGT activity toward 4-nitrophenol was determined from the disappearance of 4-nitrophenol (Franklin and Finkle, 1986). Intestinal UGT activities toward 1-naphthol and 4-nitrophenol and both hepatic and intestinal activities toward 2-naphthol, 4-hydroxybiphenyl, 2-hydroxybiphenyl, 4-methylumbelliferone, and phenolphthalein were determined by a TLC assay based on that described by Bansal and Gessner (1980) and modified as described by Vargas and Franklin (1997). Cytosolic GST activity toward 1-chloro-2,4-dinitrobenzene was determined by spectrophotometry, using the method of Habig and Jakoby (1981). Determination of QOR activity was by the dicoumarol-inhibited reduction of 2,6-dichlorophenolindophenol by NADH at pH 7.4, as described by Bensonet al. (1980).
Northern Blot Analysis.
mRNA detection and quantitation were performed by Northern blotting using high-stringency hybridization conditions, as described by Leet al. (1996); hybridized blots (16 hr at 42°C) were washed for 30 min at 42°C in 2× saline sodium citrate/0.1% sodium dodecyl sulfate and then twice for 30 min at 42°C and once for 45 min at 54°C in 0.1× standard saline citrate/0.1% sodium dodecyl sulfate. UGT mRNA and cyclophilin mRNA levels were determined simultaneously with combined cDNA probes on the same membrane (example shown in fig. 2). GST Ya and QOR mRNA levels were determined from duplicate membranes. The cDNA probe used for UGT1A6 encompassed positions +28 to + 810 (Iyanagi et al., 1986), that for UGT2B1 positions −18 to +728 (Mackenzie, 1986), that for QOR positions +32 to +1420 (Robertson et al., 1986), and that for GST Ya positions −36 to +791 (Telakowski-Hopkins et al., 1986). Plasmids containing cDNA probes were kindly provided by Dr. J. Ritter, Medical College of Virginia, Virginia Commonwealth University (Richmond, VA).
Statistics.
Statistical analyses were performed using analysis of variance, followed by Fisher’s partial least-squares difference, multiple-range test. Differences were considered significant at p < 0.05.
Results
The induction of intestinal and hepatic drug-metabolizing enzymes by BHA, oltipraz, and two other N-heterocyclic compounds known to selectively induce phase II enzymes (1,7-phenanthroline and 2,2′-dipyridyl) has been examined for the tissue and enzyme selectivity of the inductive effect in male rats. Hepatic GST activity was significantly increased by 1,7-phenanthroline and 2,2′-dipyridyl, by 60% and 50%, respectively (table 1). The mean 36% increase in GST activity after oltipraz treatment was also statistically significant, as was the 33% increase produced by BHA. QOR activity was induced by all four agents, more so by 1,7-phenanthroline (4-fold) than by 2,2′-dipyridyl and oltipraz (∼2.5-fold); all three N-heterocyclic compounds induced QOR activity to a greater extent than did BHA (1.3-fold). In contrast to the effects on GST and QOR activities in the liver, none of the compounds produced significant changes in the small intestinal mucosa except for oltipraz, which caused a small but significant increase (20%) in QOR activity.
None of the compounds induced microsomal cytochrome P450, either in the liver or in the intestine (table 1), although BHA effects on intestinal cytochrome P450 were not determined. With the present treatment regimen, a minor (compared with treatment with a polycyclic aromatic hydrocarbon, 3-methylcholanthrene) but significant increase in ethoxyresorufin O-deethylase activity was seen with oltipraz (0.062 ± 0.016 vs. 0.014 ± 0.006 nmol/mg/min), but without any hypsochromic shift in the absorbance maximum of the ferrous cytochrome P450-carbon monoxide complex. 3-Methylcholanthrene treatment increased ethoxyresorufin O-deethylase activity from 0.048 ± 0.033 to 1.927 ± 0.142 nmol/mg/min (N = 3) and caused a 2-nm blue shift in the cytochrome P450-carbon monoxide spectrum. There was no change in methoxyresorufin (0.067 ± 0.011 vs. 0.054 ± 0.011 nmol/mg/min) and pentoxyresorufin (0.042 ± 0.015 vs. 0.019 ± 0.007 nmol/mg/min) O-dealkylase activities with oltipraz treatment. However, there were major changes in the UGT activities of the microsomes with all three N-heterocyclic compounds (table 2). Only a minor change (26% increase) in a single UGT activity (4-nitrophenol) was seen with BHA (table 3). For theN-heterocyclic compounds, the increases in UGT activities for the different aglycone substrates varied in magnitude from 1.5- to 3.5-fold (table 2). Both 2,2′-dipyridyl and 1,7-phenanthroline induced UGT activities toward 4-nitrophenol, 1-naphthol, 2-hydroxybiphenyl, morphine, and testosterone. 1,7-Phenanthroline also induced hepatic UGT activities toward 4-hydroxybiphenyl, 2-naphthol, and 4-methylumbelliferone. Neither compound induced phenolphthalein or estrone glucuronidation. The pattern of induction by oltipraz was different from the patterns for both 2,2′-dipyridyl and 1,7-phenanthroline; there was no significant induction of 2-hydroxybiphenyl and testosterone glucuronidation. Inductions by oltipraz and 2,2′-dipyridyl were similar in the magnitude of induction of 4-nitrophenol, 1-naphthol, and morphine glucuronidation and in the lack of induction of 4-hydroxybiphenyl glucuronidation. Oltipraz and 1,7-phenanthroline inductions were similar with respect to the magnitude of induction of 2-naphthol and 4-methylumbelliferone glucuronidation. UGT activity toward estrone was significantly decreased after treatment with oltipraz (table 2). In contrast to the effects on UGT activities in rat liver, neither 1,7-phenanthroline, 2,2′-dipyridyl, nor oltipraz produced significant changes in any of the six UGT activities examined in rat small intestine (table4). The factors controlling constitutive expression of the UGT activities also differed between the two organs, inasmuch as intestinal activities varied between approximately 100% (4-nitrophenol) and 10% (1-naphthol) of the hepatic values.
After 3 days of treatment, levels of mRNAs for the phenobarbital-inducible UGT2B1 and 3-methylcholanthrene-inducible UGT1A6 were significantly elevated by all three nitrogen-containing heterocyclic compounds in the liver (table5). BHA was without effect. When the changes in the mRNAs for the two UGT isozymes in individual animals were compared (fig. 3), they were found to be highly correlated (r = 0.9, p < 0.05). GST and QOR mRNAs appeared to be increased by all threeN-heterocyclic agents, reflecting the increases in enzyme activity, but the elevation was significant only with oltipraz treatment. Correlations between GST mRNA and UGT2B1 and UGT1A6 mRNAs were only 0.15 and 0, respectively, and were not significant.
Discussion
The present study demonstrates that 1,7-phenanthroline, 2,2′-dipyridyl, and oltipraz are inducers of phase II drug-metabolizing enzymes in the liver and that little or no concomitant effect is seen in the small intestine. Similar organ selectivity was observed for UGT activities after phenobarbital, 2,3,7,8-tetrachlorodibenzo-p-dioxin, and pregnenolone-16α-carbonitrile administration when harmol and acetaminophen glucuronidation were under investigation (Goon and Klaassen, 1992) and after 3-methylcholanthrene and phenobarbital administration when o-aminophenol UGT activity was evaluated (Hanninen and Aitio, 1968). The observation of induction in the liver and no effect in the gastrointestinal tract is not limited to UGT activities but has also been made for GST activities. Induction of GST activity toward 3,4-dichloronitrobenzene by 3-methylcholanthrene and 3,4-benzo(a)pyrene was observed in the liver of rats but not in the small intestine (Clifton and Kaplowitz, 1978).
In contrast to the liver-selective inductions mentioned above, concomitant induction of phase II enzymes in liver and small intestine has been observed for several agents, including BHA, as described in the introduction. UGT activity toward 3-hydroxybenzo(a)pyrene was induced by omeprazole in rat liver and small intestine (Kashfi et al., 1995), as was 1-naphthol glucuronidation by 3-methylcholanthrene and benzo(a)pyrene (Goon and Klaassen, 1992). Both hepatic and intestinal GST activity toward 1-chloro-2,4-dinitrobenzene were induced after treatment with omeprazole (Kashfi et al., 1995) and phenobarbital (Clifton and Kaplowitz, 1978). In the present study, concomitant induction of hepatic and intestinal enzymes by BHA was not observed. This may be a dose-related effect, because the 75 mg/kg dose used here for direct intercompound comparisons is 1 order of magnitude less than that used in reported studies. It appears that intestinal conjugation enzymes show more limited induction by a variety of known inducers, compared with hepatic enzymes. With intragastric administration, the gastrointestinal tract is exposed to the full doses of the compounds investigated; however, in the present study, there was no induction of either UGT or GST activities with any agent. The lack of induction of intestinal GST by oltipraz in the present study agrees with the observations of Ansher et al. (1986). The only significant change in intestinal enzyme activities observed was an increase in QOR activity after oltipraz administration.
1,7-Phenanthroline and 2,2′-dipyridyl were previously found to produce large increases in hepatic phase II enzymes without inducing cytochrome P450 (Franklin, 1991; Franklin et al., 1993; Franklin and Moody, 1992). This effect has been confirmed in the present study, where significant increases in QOR activity, GST activity toward 1-chloro-2,4-dinitrobenzene, and UGT activity toward many aglycones in rat liver after treatment with 1,7-phenanthroline and 2,2′-dipyridyl were observed, without effects on cytochrome P450 content. The changes were of similar magnitudes as those produced by the same dose of oltipraz, a dithiolthione-containing N-heterocycle thought to protect animals against various carcinogens largely as a result of its enzyme-inducing activity. Although oltipraz has been considered a selective inducer of phase II enzymes, studies by Kensler et al. (1987) and Buetler et al. (1995) found significant increases in hepatic microsomal cytochrome P450 and oxidase activities toward aniline, aminopyrine, methoxyresorufin, and benzyloxyresorufin, as well as increased mRNA levels for several cytochrome P450 isoenzymes. With the present treatment regimen, we found a minor (compared with treatment with a polycyclic aromatic hydrocarbon, 3-methylcholanthrene) but significant increase in 7-ethoxyresorufinO-deethylase activity after our oltipraz treatment regimen but no change in methoxyresorufin and pentoxyresorufinO-dealkylase activities.
Induction of hepatic phase II enzymes such as GST and UGT by oltipraz has also been observed in rats and mice (Buetler et al., 1995; Ansher et al., 1983, 1986; Kensler et al., 1987; Davidson et al., 1990). Our results are in general agreement with these findings. We have demonstrated induction of UGT activities toward both morphine and more planar molecules,i.e. 4-nitrophenol, 1-naphthol, 2-naphthol, and 4-methylumbelliferone. The present study also showed significant changes in QOR activity in rat liver with oltipraz treatment, which is at variance with the findings of both Ansher et al. (1986)and Buetler et al. (1995). The latter study found no increased QOR activity after oltipraz treatment, even though increased levels of QOR mRNA were observed.
From published information on UGT activities (table6), it was evident that multiple isozymes of UGT were induced in the liver by N-heterocycles and oltipraz. UGT activity for morphine was increased, and this compound is the major substrate for only one known isozyme; therefore, UGT2B1 was the form induced. This result derived from substrate activity changes was confirmed by changes in the mRNA, which was increased by all three agents. Because 1-naphthol glucuronidation was increased and UGT2B1 does not catalyze this reaction, other isozymes (UGT1A6 or UGT2B12, of the characterized forms) must also have been induced. Because UGT1A1 catalyzes phenolphthalein glucuronidation and this activity was not increased, UGT1A1 was not induced. The increases in 2-hydroxybiphenyl glucuronidation must, therefore, have arisen from an increase in UGT2B12. Among the substrates evaluated so far, there are none that differentiate UGT1A6 and UGT2B12. Thus, from changes in enzyme activity, induction of UGT2B12, UGT2B1, and possibly UGT1A6 was indicated. Resolution of whether UGT1A6 induction occurred was obtained from changes in the mRNA, which showed that the mRNA for this isozyme was increased after 3 days of treatment. Oltipraz induction of UGT1A6 was previously shown by Buetler et al. (1995), Kessler and Ritter (1997), and Grove et al. (1997). That the induction of UGT1A6 by 1,7-phenanthroline and 2,2′-dipyridyl occurred through a mechanism other than an aryl hydrocarbon/xenobiotic response element mechanism is strongly suggested by the absence of CYP1A induction (7-ethoxyresorufin O-deethylase activity) by either of these two agents (Franklin, 1991; Franklin et al., 1993). The mechanism for the induction of the phenobarbital-inducible isozyme UGT2B1, indicated in the present study by changes in mRNA levels and morphine conjugation activity and indicated by recent findings on mRNA levels reported by Kessler and Ritter (1997), is unknown, especially because in the present study there was no significant concomitant induction of 7-pentoxyresorufin O-deethylase activity. The high degree of correlation between elevations in UGT2B1 mRNA and the mRNA for the aryl hydrocarbon/xenobiotic response element-responsive isozyme UGT1A6 suggests that regulation of enzymes by compounds containing N-heterocycle moieties may be viaunusual mechanisms that require additional studies for full elucidation.
This study has demonstrated that three N-heterocyclic compounds containing two nitrogen atoms, i.e.1,7-phenanthroline, 2,2′-dipyridyl, and oltipraz, induce phase II drug-metabolizing enzymes in the liver but, with the exception of a minor change in a single activity, do not induce drug-metabolizing enzymes in the small intestine. The general absence of significant induction in the intestine with dosing regimens ofN-heterocyclic compounds that increase hepatic enzymes contrasts with the reported effects of antioxidants such as BHA, although BHA administration has generally been in the diet and at much higher doses. The hepatic induction of the UGTs and GST was demonstrated at two biochemical levels, i.e. increases in mRNA concentrations and enzyme activities. Two UGT isozymes that are individually induced by polycyclic aromatic hydrocarbons and phenobarbital were both induced by each of theN-heterocycle-containing agents studied.
Footnotes
-
Send reprint requests to: Dr. Michael R. Franklin, Room 112, Skaggs Hall, College of Pharmacy, University of Utah, Salt Lake City, UT 84112.
-
This project was made possible in part by Grant ES05687 from the National Institute of Environmental Health Sciences, awarded to M.V.
- Abbreviations used are::
- UGT
- UDP-glucuronosyltransferase
- GST
- glutathioneS-transferase
- QOR
- NAD(P)H:quinone oxidoreductase
- BHA
- tert-butyl-4-hydroxyanisole
- Received April 7, 1997.
- Accepted October 6, 1997.
- The American Society for Pharmacology and Experimental Therapeutics