Effects of 3-Methylcholanthrene, P-Naphthoflavone, and Phenobarbital on the 3-Methylcholanthrene-inducible Isozyme of Cytochrome P-450 within Centrilobular, Midzonal, and Periportal Hepatocytes*

Rabbit antiserum produced against cytochrome P- 450 MC-B, the major isozyme of cytochrome P-450 isolated from hepatic microsomes of 3-methylcholan- threne-pretreated rats, was utilized in unlabeled antibody peroxidase-antiperoxidase and indirect fluores- cent antibody staining techniques to investigate the effects of 3-methylcholanthrene, P-naphthoflavone, and phenobarbital on the hemeprotein within centrilobular, midzonal, and periportal hepatocytes in livers of male rats. In livers of untreated and vehicle-pretreated rats, midzonal and periportal hepatocytes were found to bind the anti-cytochrome P-450 MC-B to equal extents while centrilobular hepatocytes bound approximately 25% more antibody. Pretreatment of rats with pheno- barbital did not result in alterations in either the intensity or pattern of immunohistochemical staining for cytochrome P-450 MC-B within the liver lobule. In liv- ers of rats which had been pretreated with 3-methyl-cholanthrene, microfluorometric analyses revealed that the extent of anti-cytochrome P-450 MC-B

Rabbit antiserum produced against cytochrome P-450 MC-B, the major isozyme of cytochrome P-450 isolated from hepatic microsomes of 3-methylcholanthrene-pretreated rats, was utilized in unlabeled antibody peroxidase-antiperoxidase and indirect fluorescent antibody staining techniques to investigate the effects of 3-methylcholanthrene, P-naphthoflavone, and phenobarbital on the hemeprotein within centrilobular, midzonal, and periportal hepatocytes in livers of male rats. In livers of untreated and vehicle-pretreated rats, midzonal and periportal hepatocytes were found to bind the anti-cytochrome P-450 MC-B t o equal extents while centrilobular hepatocytes bound approximately 25% more antibody. Pretreatment of rats with phenobarbital did not result in alterations in either the intensity or pattern of immunohistochemical staining for cytochrome P-450 MC-B within the liver lobule. In livers of rats which had been pretreated with 3-methylcholanthrene, microfluorometric analyses revealed that the extent of anti-cytochrome P-450 MC-B binding increased by approximately 215% within both the midzonal and periportal regions of the lobule and by approximately 150% within centrilobular regions. Pretreatment of rats with p-naphthoflavone resulted in alterations in the hemeprotein which were similar to those produced by 3-methylcholanthrene. Thus, after rats had been pretreated with either 3-methylcholanthrene or 8-naphthoflavone, the pattern of intralobular distibution of cytochrome P-450 MC-B was altered so that the antibody bound to approximately equal extents to hepatocytes within the three regions of the lobule. These results demonstrate that 3-methylcholanthrene and P-naphthoflavone exert similar inductive effects on cytochrome P-450 MC-B within the liver lobule, with the least degree of induction of the hemeprotein being produced within centrilobular hepatocytes.
In earlier studies (24-26), we employed qualitative and semiquantitative immunohistochemical techniques to demonstrate that two different forms of hepatic cytochrome P-450, the major isozymes induced by phenobarbital and 3methylcholanthrene, exhibit different nonuniform patterns of intralobular distribution in livers of untreated rats. In the present study, these techniques were used to investigate the effects of 3-methylcholanthrene, j3-naphthoflavone, and phenobartial on cytochrome P-450 MC-B, the major cytochrome P-450 isozyme isolated from livers of 3-methylcholanthrenepretreated rats, within hepatocytes in the three regions of the liver lobule.

EXPERIMENTAL PROCEDURES
Male albino Holtman rats weighing 180-210 g were used in this study and were fasted for 24 h prior to killing. Groups of rats were pretreated either for 3 days with 3-methylcholanthrene (25 mg/kg/ day, intraperitoneally, in corn oil) or P-naphthoflavone (40 mg/kg/ day, intraperitoneally, in corn oil) or for 4 days with ph.nobarbital (40 mg/kg/day, intraperitoneally, in 0.9% NaCl solution). Control rats were either untreated or administered appropriate volumes of vehicle. Rats were killed by decapitation 24 h after the last treatment, and the median lobe from each liver was removed and fixed by immersion in parabenzoquinone (26).
Details of the procedures for the immunohistochemical localization of cytochrome P-450 MC-B in liver sections using the unlabeled antibody peroxidase-antiperoxidase and the indirect fluorescent antibody staining techniques have been presented elsewhere (24, 26). Sections stained using the unlabeled antibody peroxidase-antiperoxidase method were examined by transmitted light microscopy and were photographed using Kodak Panatomic-X film (ASA 32) at an ASA setting of 32. Sections stained using the indirect fluorescent antibody technique were examined by incident-light fluorescence Kodak Tri-X pan film (ASA 400) at an ASA setting of 6400. microscopy as described previously (26) and were photographed using Measurements of the intensity of fluorescence emitted at 525 nm Induction of Cytochrome P-450 MC-B within the Liver Lobule were obtained as described previously (26) following indirect fluorescent antibody staining from circular areas (diameter = 6 pm) on serial tissue sections exposed to rabbit antiserum to rat hepatic microsomal cytochrome P-450 MC-B, normal (nonimmune) rabbit serum, and rabbit anti-cytochrome P-450 MC-B serum which had been adsorbed with cytochrome P-450 MC-B. For the microfluorometric analyses, a t least 4 sections from each liver were exposed to each serum, and 10 microfluorometric measurements were taken from within centrilobu-. . . lar, midzonal, and periportal regions in each section. The extents of anti-cytochrome P-450 MC-B binding to centrilobular, midzonal, and periportal hepatocytes were determined by subtracting the mean fluorescence intensity emitted from regions in sections exposed to normal rabbit serum from each individual microfluorometric measurement obtained from corresponding regions in serial sections exposed to rabbit antiserum to cytochrome P

Induction of Cytochrome P-450MC-B within the Liver Lobule
determined using livers obtained from 16-24 rats. The microfluorometric determinations were analyzed statistically using the group Student's t test.

Localization a n d Distribution of Cytochrome P-450 MC-B
within Livers of Untreated and Vehicle-pretreated Rats-When fixed paraffin-embedded sections prepared from livers of untreated male rats were exposed t,o rabbit antiserum raised against rat hepatic microsomal cytochrome P-450 MC-B in both the unlabeled antibody peroxidase-antiperoxidase and the indirect fluorescent antibody staining procedures, hepatocytes throughout the liver were found to be stained for this cytochrome P-450 isozyme (Figs. IA and 2 A ) . Immunohistochemical staining for cytochrome P-450 MC-B was not apparent within either cells associated with the hepatic vasculature, Kupffer cells, or sinusoidal cells. Although antibodies produced against hepatic microsomal cytochrome P-450 MC-B cross-react with hepatic nuclear cytochrome P-450 (8), hepatocyte nuclei did not appear to be appreciably stained for this isozyme of the hemeprotein. However, as reported in a previous communication (26), hepatocyte nuclei were frequently observed to be surrounded by rings of stain after unlabeled antibody peroxidase-antiperoxidase staining (Fig. 1, E, Z, and M ) . In addition, intracellular staining for cytochrome P-450 MC-B was frequently observed to occur as large granular deposits scattered throughout the cytoplasm of hepatocytes, especially those within the midzonal and periportal regions of the lobule (Fig. 1, Z and M ) . Although the data are not presented, identical observations were made after rats had been pretreated with the two vehicles used in this study, corn oil and saline.
Microfluorometric measurements of the intensity of fluorescence emitted from 28.3-pm2 circular areas within the three regions of the liver lobule in sections after indirect fluorescent antibody staining revealed that, while the anti-cytochrome P-450 MC-B bound to the same extent to hepatocytes within the midzonal and periportal regions, between 20 and 25% more antibody bound to centrilobular hepatocytes in livers of untreated and vehicle-pretreated rats (Table I and Fig. 3). The data presented in Table I also demonstrate that binding of the anti-cytochrome P-450 MC-B to hepatocytes throughout the liver was abolished after the antibody had been removed from the antiserum by adsorption with the purified rat hepatic microsomal hemeprotein.
Effects of 3-Methylcholanthrene, P-Naphthoflavone, a n d Phenobarbital on Cytochrome P-450 MC-B within the Liver Lobule-After rats had been pretreated with 3-methylcholanthrene, immunohistochemical staining for cytochrome P-450 MC-B was markedly enhanced within all regions of the liver lobule (Fig. 1,B, F, J, and N , and Fig. 2B). Although the intensity of staining for the hemeprotein was dramatically increased within the cytoplasm and around nuclei, the administration of the polycylic aromatic hydrocarbon did not appear to result in an appreciable degree of staining of hepatocyte nuclei for cytochrome P-450 MC-B. Visually, both unlabeled antibody peroxidase-antiperoxidase (Fig. 1B) and indirect fluorescent antibody (Fig. 2B) staining for cytochrome P-450 MC-B appeared to be uniform throughout the lobule in livers of 3-methylcholanthrene-pretreated rats. This observation was verified by microfluorometric analyses of liver sections after indirect fluorescent antibody staining (Figs. 3 and 4). The data presented in Fig. 4 show that while 3-methylcholanthrene pretreatment resulted in an increase of approximately 150% in the extent of anti-cytochrome P-450 MC-B binding to centrilobular hepatocytes, antibody binding to midzonal and periportal hepatocytes increased between 210 and 225%. Thus, after pretreatment of rats with 3-methylcholanthrene, the intralobular pattern of distribution of cytochrome P-450 MC-B is altered so that hepatocytes within the three regions of the liver lobule bind the anti-cytochrome P-450 MC-B to similar extents (Fig. 3). methylcholanthrene (B), B-naphthoflavone (C), and phenobarbital (0) pretreated rats. The photomicrographs show areas in 7-pm thick sections which had been exposed to rabbit antiserum to rat hepatic microsomal cytochrome P-450 MC-B in the indirect fluorescent antibody staining protocol. V, central vein; P, portal triad.  The administration of P-naphthoflavone to male rats resulted in alterations in cytochrome P-450 MC-B which were indistinguishable from those seen after rats had been pretreated with 3-methylcholanthrene. Thus, unlabeled antibody peroxidase-antiperoxidase (Fig. 1, C, G, K, and 0 ) and indirect fluorescent antibody (Fig. 2C) staining for cytochrome P-450 MC-B were both markedly enhanced in hepatocytes throughout the livers of P-naphthoflavone-pretreated rats. Microfluorometric analyses of liver sections obtained from Pnaphthoflavone-and corn oil-pretreated rats showed that pnaphthoflavone pretreatment resulted in 130% more anti-cytochrome P-450 MC-B binding to centrilobular hepatocytes and 210 to 220% more antibody binding to cells within the midzonal and periportal regions of the lobule (Fig. 4). Furthermore, the noncarcinogenic polycyclic aromatic hydrocarbon affected this isozyme of cytochrome P-450 in such a manner that it was now uniformly distributed within the liver lobule (Fig. 3).

Induction of Cytochrome P-450MC-B within the Liver Lobule
In contrast to the effects of 3-methylcholanthrene and pnaphthoflavone, the administration of phenobarbital did not

DISCUSSION
Antiserum produced against cytochrome P-450 MC-B, the major isozyme of cytochrome P-450 isolated from livers of rats pretreated with 3-methylcholanthrene, has been utilized in unlabeled antibody peroxidase-antiperoxidase and indirect fluorescent antibody staining techniques to demonstrate that the hemeprotein is present within hepatocytes throughout the livers of untreated and xenobiotic-pretreated rats. Consistent with earlier observations (24-26), cytochrome P-450 MC-B was shown to be distributed nonuniformly within the lobule in livers of untreated and vehicle-pretreated rats; while hepatocytes within the midzonal and periportal regions of the lobule contain equal amounts of the hemeprotein, centrilobular hepatocytes contain significantly more (approximately 25% more) enzyme. This finding is also consonant with reports that centrilobular hepatocytes in livers of untreated rats possess the greatest content of total cytochrome P-450 (27), the greatest amount of smooth endoplasmic reticulum (28), and the greatest aryl hydrocarbon hydroxylase (29) and 7-ethoxycoumarin 0-de-ethylase (30) activities. In addition, this observation may aid in explaining the centrilobular location of hepatotoxicities produced by many xenobiotics which are oxidatively metabolized by the hepatic monooxygenase enzyme systems to highly reactive and toxic electrophiles (31)(32)(33)(34).
The results of the present study also demonstrate that cytochrome P-450 MC-B is induced within hepatocytes throughout the liver following the pretreatment of rats with 3-methylcholanthrene and P-naphthoflavone. Moreover, the two polycyclic aromatic hydrocarbons were found to exert similar, if not identical, inductive effects on the hemeprotein. This observation is in agreement with reports (4, 5,35,36) that the major isozyme of hepatic cytochrome P-450 induced by 3-methylcholanthrene is also induced by P-naphthoflavone. However, while both 3-methylcholanthrene and pnaphthoflavone induce cytochrome P-450 MC-B within all hepatocytes, the induction does not occur uniformly through-Induction of Cytochrome P-450 MC-B within the Liver Lobule out the liver lobule; although the enzyme is induced by similar extents within midzonal and periportal hepatocytes, sign& cantly less induction of this isozyme of cytochrome P-450 is seen within centrilobular hepatocytes. Furthermore, after rats have been pretreated with either 3-methylcholanthrene or /3naphthoflavone, the intralobular pattern of distribution of the hemeprotein is altered so that it is now distributed uniformly within the lobule.
The observation that cytochrome P-450 MC-B is induced to the least extent within centrilobular hepatocytes following the administration of 3-methylcholanthrene to rats appears to be at variance with the histochemical observations of Wattenberg and Leong (29) who reported that aryl hydrocarbon hydroxylase activity was enhanced by 3-methylcholanthrene to the greatest degree within the centrilobular regions of rat liver. Although cytochrome P-450 MC-B appears to be the form of rat hepatic cytochrome P-450 which is most active toward benzo(a)pyrene (18,35), and although anti-cytochrome P-450 MC-B preparations are capable of inhibiting more than 90% of the aryl hydrocarbon hydroxylase activity catalyzed by hepatic microsomes of 3-methylcholanthrenepretreated rats (38), the reasons for this discrepancy are unknown at the present time. However, our immunohistochemical finding that the polycylic aromatic hydrocarbon induces cytochrome P-450 MC-B to significantly greater extents within the midzonal and periportal regions of the liver lobule may aid, at least in part, in explaining why 3-methylcholanthrene affords protection against the centrilobular hepatotoxicities produced by certain xenobiotics (32)(33)(34)(39)(40)(41).
The findings that cytochrome P-450 MC-B is not induced uniformly throughout the liver lobule and that its pattern of intralobular distribution is altered as a consequence of the administration of certain xenobiotics are similar to those reported previously by this laboratory (42) for the induction of hepatic NADPH-cytochrome c (P-450) reductase by the steroid pregnenolone-l6a-carbonitrile. Moreover, similar observations have also been made for the induction of epoxide hydrolase within hepatocytes following the pretreatment of rats with either phenobarbital or trans-stilbene oxide.' Thus, not only can significant enzymatic differences exist among centrilobular, midzonal, and periportal hepatocytes, but there may also be significant differences in the inductions of enzymes such as cytochrome P-450 isozymes within hepatocytes in the different regions of the liver lobule.