Organ and enzyme selectivity and toxicological significance of cytochrome P450 induction and inhibition elicited by nitrogen heterocycles

The metabolism of xenobiotics by extrahepatic cytochrome P450 (P450) may be integral in the development of organ-selective toxicities. Some xenobiotics also induce extrahepatic P450, and the polycyclic aromatic hydrocarbon (PAH)-type agents are generally considered the most effective inducers in the rat. N-benzylimidazole (NBI), induced rat hepatic P450 in a PAH-type manner; therefore, the effects of NBI and 15 other nitrogen heterocycles on P450 in rat kidney, lung and intestine were investigated. Those compounds that were high magnitude inducers of hepatic P450 also induced in some extrahepatic tissues, but, only two, NBI and N-methylnaphthylimidazole (NMN), induced in all four tissues. Induction was greater in kidney and intestine than in lung, but, except for induction by NBI and NMN, the degree did not approach that seen in liver. Several compounds induced extrahepatic P450 isozymes (determined by monooxygenase activities) that are not induced by PAHs, but interorgan inconsistencies in the extent and the pattern of isozyme induction were observed. Further investigations revealed that these interorgan inconsistencies in induction were probably not caused by the persistence of residual inhibitory nitrogen heterocycle, since evidence for the persistence of NBI in liver (inhibition of monooxygenase activities relative to P450, slowed development of CO-ferrous P450 spectra, and alteration of type II binding characteristics) and in extrahepatic tissues (alteration of type II binding characteristics) was not consistent with the degree of P450 induction by NBI in the tissues. The yield of hepatic, but not extrahepatic, microsomes was decreased by 30-40% by three compounds, including NBI. Further studies with NBI indicated that the decreased yield was due to an alteration of the sedimentation properties of hepatic subcellular organelles, which resulted in an increased loss of microsomes and mitochondria from liver homogenates during low-speed centrifugation. The ramifications of renal P450 induction by nitrogen heterocycles to the development of chloroform nephrotoxicity were investigated using a rat kidney slice model. The results of these and additional studies with other inducers and with P450 inhibitors were inconclusive, suggesting that the rat kidney slice model may not be valid when nephrotoxicity is thought to arise from reactive metabolites generated by P450.

ORGAN AND ENZYME SELECTIVITY AND TOXICOLOGICAL SIGNIFICANCE OF CYTOCHROME P450 INDUCTION AND INHIBITION ELICITED BY NITROGEN HETEROCYCLES by Wendy Lynn Harmsworth A dissertation submitted to the faculty of The University of Utah in partial fuHiliment of the requirements for the degree of' Doctor of Philosophy Department of Pharmacology and Toxicology The University of Utah March 1991 Copyright © Wendy Lynn Harmsworth 1991 All Rights Reserved THE UNIVERSITY OF UTAH GRADUATE SCHOOL SUPERVISORY COMMITTEE APPROVAL of a dissertation submitted by Wendy Lynn Harmsworth This dissertation has been read by each member of the following supervisory committee and by majority vote has been found to be satisfactory. ' Chair: i chae 1 R. Frank 1 in j 1/ ,/ AJcLldLa ( Q~I~r.?~ Douglass E. Rollins Garold s. ydit THE UNIVERSITY OF UTAH GRADUATE SCHOOL FIN AL READING APPROV AL To the Graduate Council of the University of Utah: I have read the dissertation of Hendy I yon Harmsworth in its final fonn and have found that (1) its format, citations and bibliographic style are consistent and acceptable; (2) its illustrative materials including figures, tables, and charts are in place; and (3) the final manuscript is satisfactory to the supervisory conunittee and is ready for submission to The Graduate School. Dale Michael R. Franklin Chair, Supervisory Committee Approved for the Major Department ChairjDean Approved for the Graduate Council B. Gale Dick Dean of The Graduate School ABSTRACT The metabolism of xenobiotics by extrahepatic cytochrome P450 (P450) may be integral in the development of organ-selective toxicities. Some xenobiotics also induce extrahepatic P450, and the polycyclic aromatic hydrocarbon (PAH)- type agents are generally considered the most effective inducers in the rat. N-benzylimidazole (NBI), induced rat hepatic P450 in a PAH-type manner; therefore, the effects of NBI and 15 other nitrogen heterocycles on P450 in rat kidney, lung and intestine were investigated. Those compounds that were high magnitude inducers of hepatic P450 also induced in some extrahepatic tissues, but, only two, NBI and N-methylnaphthylimidazole (NMN), induced in all four tissues. Induction was greater in kidney and intestine than in lung, but, except for induction by NBI and NMN, the degree did not approach that seen in liver. Several compounds induced extrahepatic P450 isozymes (determined by monooxygenase activities) that are not induced by PAHs, but interorgan inconsistencies in the extent and the pattern of isozyme induction were observed. Further investigations revealed that these interorgan inconsistencies in induction were probably not caused by the persistence of residual inhibitory nitrogen heterocycle, since evidence for the persistence of NBI in liver (inhibition of monooxygenase activities relative to P450, slowed development of CO-ferrous P450 spectra, and alteration of type II binding characteristics) and in extrahepatic tissues (alteration of type II binding characteristics) was not consistent with the degree of P450 induction by NBI in the tissues. The yield of hepatic, but not extrahepatic, microsomes was decreased by 30-400/0 by three compounds, including NBI. Further studies with NBI indicated that the decreased yield was due to an alteration of the sedimentation properties of hepatic subcellular organelles, which resulted in an increased loss of microsomes and mitochondria from liver homogenates during low-speed centrifugation. The ramifications of renal P450 induction by nitrogen heterocycles to the development of chloroform nephrotoxicity were investigated using a rat kidney slice model. The results of these and additional studies with other inducers and with P450 inhibitors were inconclusive, suggesting that the rat kidney slice model may not be valid when nephrotoxicity is thought to arise from reactive metabolites generated by P450. v TABLE OF CONTENTS ABSTRACT. . . . .iv LIST OF FIGURES. . viii LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . x Chapter 1. INDUCTION OF HEPATIC AND EXTRAHEPATIC P450 AND MONOOXYGENASE ACTIVITIES BY NITROGEN HETEROCYCLES . 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . .1 Methods .......................... 3 Resutts . . . . . . . . . . . . . . . . . . ......... 7 Discussion . . . . . . . . . . . . . . . . . . . . . . . . 15 2. ALTERATION OF SEDIMENTATION PROPERTIES OF SUBCELLULAR ORGANELLES BY NBI . . . . . . . . . . . . . . . . . . . . 22 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 22 Methods .......................... 23 Resutts . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Discussion . . . . . . . . . . . . . . . . . . . . . . . . 29 3. INTERORGAN DIFFERENCES IN THE IN VITRO BINDING OF NBI TO P450 AND POSSIBLE MODIFICATION BY RESIDUAL COMPOUND FOLLOWING IN VIVO TREATMENT ....... 35 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 35 Methods .......................... 37 Resutts . . . . . . . . . . . . . . . . . . ........ 40 Discussion . . . . . . . . . . . . . . . . . . . . . . . . 52 4. EFFECTS OF ALTERED KIDNEY P450 ON IN VITRO CHLOROFORM NEPHROTOXICITY IN A KIDNEY SLICE MODEL . 64 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 64 Methods .......................... 67 Resutts . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Discussion . . . . . . . . . . . . . . . . . . . . . . . . 92 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . .102 vii LIST OF FIGURES Figure 1.1. Chemical structures of the 16 nitrogen heterocycles orally administered to rats. . . . . . . . . " .....4 2.1. Alteration of liver weight! body weight ratio and hepatic microsomal yield by inducing agents. . . . . . .. ... 25 2.2. Effect of treatment of rats with NBI on the sedimentation of total hepatic protein and microsomal and mitochondrial enzymes in liver homogenates during centrifugation. .,. 27 3.1. Time course of NBI induction of P450 concentration, p-nitroanisole demethylase and EROO activities in microsomes from rat liver, kidney, lung and intestine. . . . . . 41 3.2. SOS-Polyacrylamide gel of kidney cortical microsomes from untreated rats or rats treated with BNF or NBI. . . . .53 4.1. Slicing apparatus for preparation of renal cortical slices. . . . . . . . . . . . . . . . .. .... 69 4.2. Chloroform concentration-response curves for in vitro parameters of nephrotoxicity. . . . . . . . . . . . . . . 73 4.3. Effect of rat pretreatment with N-substituted imidazoles on renal P450 and chloroform nephrotoxicity determined in vitro. . . . . . . . . . . . . . . . . . .. .., 76 4.4. Effect of incubation of kidney slices with P450 inhibitors on chloroform nephrotoxicity. . . . . . . . .. ... .79 4.5. Effect of P450 inhibitors on constitutive aminopyrine demethylase activity of rat kidney microsomes. . . . . . . . .82 4.6. Destruction of P450 by ABT in untreated rat kidney microsomes. . . . . . . . . . . . . . . . . . . . . . 83 4.7. Effect of rat pretreatment with BNF on chloroform nephrotoxicity determined in vitro. . . . . . . . . . .85 4.8. Effect of slice incubation with supplementary hepatic or renal microsomes on in vitro chloroform nephrotoxicity. ... .88 4.9. Effect of rat pretreatment with BSO on chloroform nephrotoxicity determined in vitro. . . . . . . . . . . . . .90 ix LIST OF TABLES Table 1 .1 . Microsomal yield from liver, kidney cortex, lung and small intestine following treatment with N-heterocycles. . _ . . . . .8 1 .2. Effect of N-heterocycle administration on P450 administration on P450 concentration and p-nitroanisole and EROD activities in rat tissues. . . . . . . " ... 9 1 .3. Effect of N-heterocycle administration on demethylase activities in liver and selected extrahepatic tissues in the rat. . . . . . . . . . . . . . . . . . .. . _ . 13 3.1. Microsomal protein yield from liver, kidney, intestine and lung following treatment with NBI and BNF. . . . . . . . _ .44 3.2. NBI inhibition of EROD in BNF- and NBI-induced rat microsomes from liver, kidney, lung and intestine. . 3.3. Development of CO-ferrous P450 spectra in the presence and absence of exogenous NBI in hepatic and extrahepatic .. 46 microsomes from untreated, BNF- and NBI-treated rats. . .47 3.4. NBI P450 type II binding spectra in hepatic and extrahepatic microsomes from untreated, BNF- and NBI-treated rats. . . . . . . . . . . . . . . 4.1. Possible relationship between metabolism of chloroform . . . 49 by kidney P450 and nephrotoxicity. . . . . . . . . . . . .66 CHAPTER 1 INDUCTION OF HEPATIC AND EXTRAHEPATIC P450 AND MONOOXYGENASE ACTIVITIES BY NITROGEN HETEROCYCLES Introduction Most P450-catalyzed xenobiotic metabolism in mammals occurs in liver, largely due to the high concentrations of P450 isozymes present and the large size of the organ. However, P450 enzymes are ubiquitous, as they have been been found in virtually every mammalian tissue examined, albeit in much smaller quantities and concentrations than in liver. P450 in extrahepatic tissues often selectively catalyzes the oxidation of endogenous substrates; however, extrahepatic metabolism of xenobiotics by P450 also occurs. Extrahepatic P450 usually plays an insignificant role in the overall metabolism of axe no biotic, due to its low concentration; however, extrahepatic P450 may be an important contributor to the generation of any organ-selective toxicities from such compounds. The concentration of hepatic xenobiotic metabolizing isozymes is altered following exposure to foreign compounds. The ability of prototypical hepatic P450 inducing agents, such as 3-methylcholanthrene (MC) and phenobarbital (PB), to induce P450 in extrahepatic tissues has been extensively investigated. The polycyclic aromatic hydrocarbon (PAH)- type inducing agents, such as MC, are generally considered the most effective inducers of extrahepatic P450 in the rat. A large number of compounds containing a nitrogen heterocyclic ring moiety, such as imidazoles {Bossche, 1987; Rush, Smith, Mulvey, Graham and 2 Chaplin, 1987; Morita, Ono, and Shimakawa, 1988; Hitchcock, Dickinson, Brown, Evans and Adams, 1990), triazoles (Middleton, Milne, Moreland and Hasmall, 1986; Hitchcock, et al., 1990; Shaw, Tarbit and Troke, 1987; Rahier and Taton, 1990), pyridines (Anagnostopulos, Bartlett, Eiben and Stoll, 1989), pyrimidines (Lindstrom and Whitaker, 1987) and imidazopyridines (Bernstein and Franklin, 1986), are either in use or are being tested as therapeutic and agricultural agents. Investigations with several N-substituted imidazole compounds, which are potent inhibitors of P450, have found that some of these compounds can also induce hepatic drug metabolizing enzymes (Kahl, Friederici, Kahi, Ritter and Krebs, 1989; Ritter and Franklin, 1987a; Rodrigues, Waddell, Ah-Sing, Morris, Wolf and loannides, 1988; Papac and Franklin, 1988; Magdalou, Totis, Boiteux-Antoine, Fournel-Gigleux, Siest, Schladt and Oesch, 1988; Hostetler, Wrighton, Molowa, Levin and Guzelian, 1989; Khan, Kuhn, Merk, Park, Gelboin, Bickers and Mukhtar, 1989). Two N-substituted imidazoles, clotrimazole (CLTZ) and N-benzylimidazole (NBI), are high magnitude (3- to 4- fold) inducers of hepatic P450. Both compounds are mixed­type inducing agents: CL TZ induced in a manner like both PB and glucocorticoids, while NBI was both a PB-type and a PAH-type inducer. CL TZ increased selective monooxygenase parameters and intensified the same molecular weight protein bands on sodium dodecylsulfate polyacrylamide gel electrophoresis that were induced by PB and glucocorticoids (Ritter and Franklin, 1987). These results were supported by other reports (Rodrigues, et al., 1988; Hostetler, et al., 1989; Khan, et aI., 1989), in which immunoquantification of P450 isozymes and quantification of mRNA using cDNA probes revealed that CL TZ induced the hepatic levels of both PB (IIB1/2)- and glucocorticoid (1IIA1)- inducible isozymes, as well as the levels of mRNA for these isozymes. The confirmatory studies also validated the use of enzyme activities as monitors of P450 isozymes. In similar enzyme activity studies, NBt selectively induced hepatic monooxygenase activities characteristic of both PB-type induction and PAH-type induction (Papac and Franklin, 1988), and immunoblots with microsomes from NBI- treated rats revealed induction of isozymes of P450 inducible by both PB (IIB1/2) and MC (IA 1) (Magdalou, et al., 1988). Since NBI induced in a PAH- like manner in liver, and because PAH- like compounds have been found to be the most effective inducers of extrahepatic P450, the present study was undertaken to determine the extent to which NBI, and possibly other nitrogen heterocycles, are able to induce P450 in extrahepatic tissues. An additional aim was to determine the range of isozymes induced, using monooxygenase activities as functional indicators of isozyme activity. Methods Chemicals All biochemicals were obtained from Sigma Chemical Co. (St Louis, MO). 3 Erythromycin was a gift from Abbott Laboratories (North Chicago, IL). 7-Ethoxyresorufin and ethylmorphine were obtained from Pierce Chemical Co. (Rockford, IL) and BOH (Canada), respectively. The structures of the nitrogen­heterocycle compounds investigated in this study (seven N-substituted imidazoles, two triazoles, three pyridines, three imidazopyridines and one pyrimidine) are shown in Figure 1.1. NBI (99% pure) and diphenyl-4-pyridyl methane (DPP) were obtained from Aldrich Chemical Co. CL TZ was obtained from Sigma Chemical Co. N-[2-Naphthylmethyl]imidazole (NMN) and nafimidone (NAF) were gifts from Syntex Research (Palo Alto, CA). 1-[2-Chloro-2-(2,4-dichlorophenyl)vin-1-yl] imidazole (CECP) was provided by Janssen Research Foundation (Belgium). The triazoles, fluconazole [FLU] and IC1151885, were gifts from Pfizer (U.K.) and Imperial Chemical Industries (U.K.), respectively. The imidazopyridines, sulmazole (SUM), LY163252 and L Y175326, were gifts from Eli Lilly (Indianapolis, IN). All compounds were used as supplied. IMIDAZOLES Q# CI ~ N?'\. -0 ~N-C \ J 6 Clotrimazole (CL TZ] '" N ... O CN-CH20 CI 5-Chloro-3-[ l-imidazolmethyO benzo-Idl-isoxazole [MCB1 N~N'" -C<~?~ oN~"' O ~~ CI 5-Chloro-3-[( l-imidazol-l-yl)ethyQ benzo-{d]-isoxazole (ECB) N-Benzylimidazole [NBI] C>-CH2-OO N-(2-Naphlhylmethyllimidazole (NMN] N~ II 0-00 ~N-CH,-C '% I ., Nafimidone (NAF] CI N~ J=\ . < ~N-Q-1=~~Cl CI 1_[2_Chloro-2-(2.4-dichlorophenyl)vin-l-y1l imidazole [CECP] LY56110 4 PYRIDINES TRIAZOLES Q N H OH~ r~N-zO-C~F < t-CH 6 ?"\ ~ F IC1151885 Diphenyl-4-pyridylmethane ~N\. ~ 9H ~ ,N~I [DPP) I N-C-C-C-N II N::::J ~I H '\.-N C2HS F ?" ~HOH~ I N~N_LN~ ~ 86034 C2Hs Fluconazole (FLU) IMIDAZOPYRIDINES CXN~ M. 0 ~ I Tv-S-CH:t N • H Sulmazole (SUM] Figure 1.1. Chemical structures of the 16 nitrogen heterocycles orally administered to rats. 5 Animals. Ireatments. and Mjcrosome Preparation Adult, male Sprague-Dawley rats (Charles River Labs, Portage, MI) were housed under a 12-hour light-dark cycle in an environment with a constant temperature and humidity. All drugs were administered by gastric lavage (in volumes of 0.2% of the body weight) as suspensions, except for NBI (which was solubilized in dilute HCI). Ihe suspending agent, 30% polyethylene glycol 400, had no effect on hepatic parameters when administered alone. Ihe dosing regimen was 75 mglkg, daily, for 3 days, followed by a 48 hour period before killing by C02 asphyxiation. The 48 hour wait was included to allow sufficient time for clearance of the drugs from the organs, since N-substituted imidazoles are potent inhibitors of P450-catalyzed monooxygenase reactions. The livers were perfused with ice-cold saline, homogenized in 0.25M sucrose and microsomal fractions were prepared as described previously (Franklin and Estabrook, 1971). The following modifications of this procedure were made to prepare microsomal fractions from intestine, lung and kidney. Tissues from three to five animals were pooled for each preparation. An intestinal segment consisting of the 25 cm length proximal to the pylorus was removed, flushed thoroughly with ice-cold Iris-KCI buffer (150n1M KCI, 50mM Iris-HCI, pH 7.4), slit open and the lumen scraped to remove mucosal tissue. Ihe tissue was homogenized (1:10 weight Ivolume) in buffer to which trypsin inhibitor (chicken egg white, partially purified ovomucoid, containing ovoinhibitor) had been added (5 mglg wet weight) to reduce proteolytic degradation of P450 during microsome preparation (Stohs, Graftstrom, Burke, Moldeus, and Orrenius, 1976a). Ihe lungs were perfused with Iris-KCI buffer in situ through the right ventricle of the heart, removed, and thoroughly flushed through the trachea to reduce surfactant contamination. A tissue press was used to remove coarse connective tissue, and the exudate was homogenized in buffer (1 :4, weight Ivolume). The kidneys were removed, perfused through the renal pelvis with Iris-KCI buffer, the cortices were separated from the medullary tissue, and the cortical tissue was homogenized in buffer (1 :5, weight Ivolume). 10 increase the microsomal yield (Fouts and Devereux, 1973; Lindeskog, Haaparanta, Norgard, Glaumann, Hanson and Gustafsson, 1986), the homogenates of the lung and intestinal tissues were subjected to 15 seconds of sonication at an output of 4.5 on a Branson sonicator, model W185, equipped with a microtip. Determjnation of Protein and P450 Concentration Protein concentrations were determined using Folin phenol reagent (Lowry, Rosebrough, Farr and Randall, 1951). Hepatic microsomal P450 concentrations were determined from dithionite-reduced CO-difference spectra, using an extinction coefficient of 91 mM-1 cm-1 (Omura and Sato, 1964). To decrease the spectral interference from haemoglobin and other 6 haemoproteins, P450 concentrations in extrahepatic tissues were determined from the dithionite-difference spectrum of microsomes gassed with CO for 5 minutes, using an exti nction coefficient of 100 ITIM-1 cm-1 (Estabrook, Peterson, Baron and Hildebrandt, 1972). Enzyme Activities p-Nitroanisole O-demethylase and 7-ethoxyresorufin O-deethylase activities were determined from the linear portion of kinetic assays which were followed for 15 minutes. The chromophores monitored were p-nitrophenol (Netter and Seidel, 1964) and resorufin (Klotz, Stegeman and Walsh, 1984), respectively. Rates of N-demethylation of erythromycin, ethylmorphine, aminopyrine and benzphetamine were determined from the formaldehyde produced (Nash, 1953) at 5, 10, and 15 minutes, a period over which the reaction proceeded linearly. Substrate concentrations for these demethylation reactions were 1,8,8, and 1 mM, respectively. Statistics Statistical determinations were performed using Student's two-tailed t-test on unpaired sample means. Differences between the groups were considered statistically significant for p :5 0.05. Resutts Sixteen nitrogen heterocycle compounds (7 imidazoles, 2 triazoles, 3 pyridines, 3 imidazopyridines and 1 pyrimidine; Figure 1.1) were administered to male rats for three days and microsomes were prepared from liver, kidney cortex, lung and small intestinal mucosa. The effect of these treatments on microsomal yield is shown in Table 1 .1. Most of the compounds did not significantly alter the yield of microsomal protein from any tissue. However, the hepatic microsomal yield was decreased by 18% following treatment with SUM, and by 30-40% after treatment with NAF, NBI and DPP. 7 The inductive effects of the compounds on rat hepatic and extrahepatic P450 concentrations and selected monooxygenase activities were examined. The effect of the treatments on P450 concentrations in liver, kidney cortex, small intestinal mucosa, and lung microsomes are shown in Table 1.2. All seven N-substituted imidazoles induced hepatic P450 concentration, but to varying extents: CL TZ and NBI were "high magnitude," inducing over 3-fold; NMN and CECP induced to a lesser extent (approximately 2.5-fold), and 5-chloro-3-(1-imidazolmethyl]benzo-[d]-isoxazole (MCB), 5-chloro-3-[(1-imidazol-1-yl)ethyl]benzo-[d]-isoxazole (ECB) and NAF were lower magnitude inducers. Induction of hepatic P450 by the non-imidazole nitrogen heterocycles was inconsistent, with most compounds inducing minimally, or not at all. However, the triazole, ICI151885, was a high-magnitude inducer, inducing the hepatic concentration 4-fold, and two other compounds, the pyridine, DPP, and the pyrimidine, LY5611 0, each induced to an extent comparable to that of NMN and CECP. In contrast, only a few of the 16 compounds induced P450 in selected extrahepatic tissues, and only two, NBI and NMN, induced to an extent above controls which approximated "high magnitude." Six compounds, ECB, NAF, NBI, NMN, ICI151885 and DPP induced kidney cortical P450 concentration to extents ranging from 1.5- to 4-fold. Two of these (the imidazoles, NBI and NMN) induced intestinal P450 5-fold, and two others (the triazole, IC1151885, and the pyridine, DPP) induced the P450 concentration two- to three-fold. NBI and Table 1.1. Microsomal yield from liver, kidney cortex, lung and small intestine following treatment with N-heterocycles. 8 Microsomes were prepared from the liver, kidney cortex, lung and small intestinal mucosa of untreated rats and rats treated with one of the nitrogen heterocycle compounds shown in Figure 1.1, as described in Methods. Data are expressed as a percent of the mean control value, calculated from concurrently determined controls. Values are expressed as the mean ± the standard error of the mean for sample numbers (n) = 3-11. If no standard error is given, n=1-2, and values are given. Average control values in mg microsomal protein per gram tissue (except for the lung yield, which was determined as the total yield of microsomal protein in mg per sample) in liver, kidney, lung and intestine, respectively, are: 6.2±0.4 (n=53); 3.0±0.2 (n=24); 11.1±0.6 (n=21); 4.7±0.3 (n=18). Statistical significance was determined using Student's two-tailed t-test; samples were considered statistically significant (*) for p ~ 0.05. EERCENT QE QONTROL YIEI..O Trealrneot Li~e[ ~idoe:i Luog loleslioe Imidazoles: MCB 99±7 104±14 98±8 104±2 ECB 111±17 121,92 89,104 88,82 CECP 99±11 95,116 79,81 86,86 CLTZ 112±7 106±10 82±12 92,93 NAF 62±11* 91,74 N[)a ND NBI 65±8* 123±6 85±3 108±5 NMN 97±23 90,110 73, 70 91,103 Triazoles: ICI151885 123±12 163,142 99,88 111, 125 FLU 89±6 97±10 95±19 91±4 P:iridjoes: 26020 95±17 105±7 102±8 99±7 86034 76±4 87±14 143±30 103±5 DPP 68±4* 82,133 74,96 96,86 I rnidazQC:i[idioes' LY163252 98±7 96 125 73 LY175326 95±7 102 110 76 SUM 82±3* 107 88 95 P:irj midioe; LY56110 93±7 107 103 92 aN D = not determined Table 1.2. Effect of N-heterocycle administration on P450 concentration and p-nitroanisole and EROO activities in rat tissues. Microsomes were prepared from the liver, kidney cortex, lung and intestinal mucosa of untreated (CON) rats and rats treated with one of the nitrogen heterocycle compounds shown in Figure 1.1. The microsomal concentration of P450 (nmoles/mg) and the rates of the demethylation of p-nitroanisole and the deethylation of 7-ethoxyresorufin (nmoles/mg/min) were determined as described in Methods. Values are expressed as the mean ± the standard error of the mean for sample numbers (n) > 2 (for n =1-2, values are given). Statistical analysis was performed using Student's two-tailed t-test. Values of p s; 0.05 were considered statistically significant and are indicated by an asterisk (*). Eararneter; P-450 p:r:JiUoaoisole Demetb~lase EBOO Ii~~ue; Liver Kidne~ Intestine Lung Liver Kidne~ Intestine Lung Liver Kidney Intestine Luna Irealrneot; CON 0.844± 0.117± 0.048± 0.089± 0.584± 0.013± 0.001± 0.042± 0.034± 0.001± 0.004± 0.003± 0.026 0.008 0.003 0.004 0.031 0.003 0.001 0.011 0.005 0.001 0.001 0.001 MCB 1A10± 0.136± 0.042± 0.116± 2.302± 0.029± 0.013± 0.065± 0.508± 0.007± 0.008± 0.014± 0.071 • 0.021 0.007 0.012 0.136 • 0.016 • 0.007 0.024 0.075 • 0.003 0.002 0.003 • ECB 1.700± 0.173, 0.049, 0.096, 2.267± 0.012, 0.011, 0.127, 0.503± 0.009, 0.007, 0.014, 0.115 • 0.171 • 0.074 0.098 0.226 • 0.016 0.027 0.135 • 0.105 • 0.007 0.001 0.002 CECP 1.960± 0.219, 0.072, 0.058, 2.960± 0.024, 0.010, 0.049, 0.890± 0.001 0.028, 0.008, 0.127 • 0.150 0.069 0.084 00430 • 0.021 0.048 0.073 0.096 • 0.005· 0.010 CLTZ 3.360± 0.089± 0.064± 0.134± 2.750± 0.025± 0.024± 0.105± 0.128± O.OOO± 0.002± 0.004± 0.160 • 0.015 0.007 0.012 • 0.070 • 0.009 0.010 • 0.015 • 0.009 • 0.000 0.001 0.002 NAF 1.647± 0.234, NOa NO 2.970± 0.066, ND NO 1A60± 0.053, NO ND 0.047 • 0.198· 0.050 • 0.064 • 0.320 • 0.043· NBI 2.820± 0.217± 0.248± 0.143± 3.380± o A03± 0.357± 0.126± 2.240± 0.348± 0.164± 0.036± 0.140 • 0.020 • 0.054 • 0.026 0.360 • 0.106 • 0.057 • 0.022 • 0.170 • 0.077 • 0.037 • 0.005 • NMN 2.058± 0.506, 0.280, 0.160, 4.120± 0.237, 0.083, 0.120, 4.1211 0.322, 0.143, 0.045, ~ 0.129 * 0.438· 0.193 " 0.141* 0.240 * 0.569 * 0.096" 0.159 " 0.245 * 0.602" 0.125 " 0.068* Table 1.2. Continued --- Ea[arn~l~[; P450 p:~~roanisole Demet~lase ERGD Ii~~u~; Liver Kidne~ Intestine Lung Liver Kidne~ Intestine Lung Liv~r Kidn~~ Intestine Lung I[~atrn~nl; I[ia~QI~s' ICI151885 3.362± 0.235± 0.153, 0.097, 2.333± 0.023± 0.021, 0.057, 0.049± 0.010± 0.000, 0.011, 0.090* 0.055* 0.167* 0.107 0.247* 0.005 0.000 0.020 0.015 0.006 0.002 0.021* FLU 1.255± 0.106± 0.046± 0.126± 1.366± 0.008± 0.027± 0.134± 0.046± 0.001± 0.006± 0.001± 0.096* 0.005 0.008 0.008* 0.105* 0.003 0.026 0.011 * 0.024 0.002 0.001 0.004 E~[idiD~~' 26020 0.936± 0.186± 0.077± 0.123± 0.920± 0.034± O.OOO± 0.138± 0.062± 0.008± 0.010± 0.016± 0.053 0.039 0.005* 0.034 0.065* 0.013 0.000 0.078 0.011 0.003 0.003 0.003* 86034 1.164± 0.166± 0.075± 0.111± 1.644± 0.007± 0.276± 0.116± 0.044± 0.008± 0.027± 0.016 0.033* 0.015 0.013 0.019 0.254* 0.006 0.136 0.063 0.025 0.003 0.004* 0.003* DPP 1.906± 0.207, 0.119, 0.108, 2.035± 0.556 0.038, 0.000. 0.082± 0.000. 0.007, 0.001, 0.210* 0.243* 0.089* 0.094 0.279* 0.000 0.001 0.018 0.006 0.007* 0.001 IrnidazQg~ddi[l~~ ; LY163252 0.976± 0.123 0.038 0.077 0.940± 0.003 0.004 0.075 0.090± 0.003 0.000 0.016 0.052 0.096* 0.017 LY175326 1.009± 0.165 0.043 0.089 1.000± 0.031 0.038 0.094 0.274± 0.003 0.001 0.012 0.035* 0.038* 0.036* SUM 0.961± 0.165 0.038 0.095 0.706± 0.022 ND 0.092 0.122± 0.001 0.000 0.007 0.046 0.044 0.024* ~[irnidi[l~; LY56110 1.750± 0.153 0.081 0.085 3.951± 0.007 0.002 0.088 0.068± 0.000 0.004 0.006 0.223* 0.387* 0.026 aND = not determined 0 NMN were the best inducers in all tissues, as they also induced pulmonary P450 to the greatest extent (1.5-fold). However, two other compounds, the imidazole, CLTZ, and the triazole, FLU, also increased P450 concentration in lung to approximately the same extent as NSI and NMN, although they did not induce in kidney and intestine. p-Nitroanisole demethylation, a monooxygenase activity catalyzed extensively by several isozymes of cytochrome P-450, was induced to varying extents in liver by all compounds examined with the exception of the imidazopyridine, SUM (Table 1.2). In kidney cortex, only NSI and NMN and possibly OPP induced this activity to the largest extent. The induction of p-nitroanisole de methylase activity in the small intestinal mucosa did not 11 parallel the induction of P450 concentration in that the extent of demethylase induction after treatment with NSI was greater than after NMN treatment, despite the equivalent increase in P450 concentration by the two imidazoles. In lung, an organ in which constitutive p-nitroanisole demethylase activity was much greater than that in the other two extrahepatic tissues examined, five of the compounds were good inducers (ECS, CL TZ, NSI, NMN and FLU) although the maximum degree of induction achieved in lung was less than that seen in the kidney and intestine. 7-Ethoxyresorufin deethylation (EROO) is an activity which is considered selective for P450c (IA1), an isozyme induced by PAHs and a limited number of similar compounds. All imidazoles as well as the imidazopyridines and one pyridine, OPP, induced EROO activity in liver to varying extents, although the induction following treatment with CL TZ and the nonimidazoles was relatively minor (Table 1.2). In the three extrahepatic tissues, only NSI and NMN induced this activity to a great degree, with the induction being greatest in kidney and least in lung. The other compounds either did not induce EROO activity or only induced the activity to a small extent compared to that seen following treatment with NSI or NMN. Erythromycin demethylation activity is generally associated with hepatic P450 (lilA 1) induction by glucocorticoids. Among the nitrogen heterocycles 12 examined, the imidazole, CLTZ, and the pyridine, DPP, induced this activity to a large extent (12-fold and 7.5-fold, respectively) in liver, while the other agents only induced it up to 4.5-fold (Table 1.3). In the small intestine, NBI and NMN, which were the best inducers of P450 concentration, also induced erythromycin demethylase activity, although compounds which were not inducers of P450 concentration in this tissue (the pyridine, 86034, the pyrimidine, L Y5611 0, and possibly the imidazopyridine, L Y175326) induced this activity to the sanle extent (approximately two-fold). In kidney cortex, the induction of erythromycin demethylase activity by two compounds (the imidazole, ECB, and the pyridine, 86034) and possibly two others (the imidazopyridine, L Y175326, and the pyrimidine, L Y56110) was not accompanied by induction of P450 concentration in this tissue. Induction of ethylmorphine demethylation is associated with hepatic P450 induction by either glucocoriticoids or PB. Two N-substituted imdazoles, CL TZ and CECP, and the pyridine, DPP, induced this activity more than three-fold in liver (Table 1.3). Both triazoles and the pyrimidine also induced this activity in the liver, but to a lesser extent (approximately two-fold). Changes in ethylmorphine demethylase activity were minimal for any N-substituted imidazole in the two extrahepatic tissues which were studied, with the exception of NBI, which induced this activity two-fold in the kidney cortex. The demethylations of aminopyrine and benzphetamine were examined in kidney cortex and lung, respectively, due to precedents in the literature, since these substrates are preferential, but not specific, for PB-inducible P450 isozymes. Increases in P450 concentration in the kidney by the N-substituted imidazoles, NAF, NBI and NMN and the triazole, ICI151885, were paralleled by increases in aminopyrine demethylase activity. The increase in pulmonary P450 concentration by CL TZ was accompanied by a small increase in benzphetamine demethylase activity, which was not seen with the other inducers of P450 concentration in this organ. However, two compounds that did not induce pulmonary P450 concentration (the pyridines 26020 and 86034) did induce benzphetamine demethylase in the lung to an extent at least as great as Table 1.3. Effect of N-heterocycle administration on demethylase activities in liver and selected extrahepatic tissues in the rat. Data are expressed as a per cent of the mean control value (calculated from concurrently determined controls) ± the standard error of the mean for sample numbers (n) = 3-11. If no standard error is given, n =1- 2, and values are given. Statistical significance was determined using Student's two-tailed t-test; samples were considered statistically significant (*) if P ~ 0.05. Average control values in nmollmg/min are: for erythromycin demethylase in liver, intestine and kidney, 0.47±0.04 (n=25), 0.12±O.02 (n=11), 0.06±O.01 (n=18), respectively; for ethylmorphine demethylase in liver, intestine and kidney, 7.77±O.48 (n=27), 0.13±0.03 (n=11), 0.17±0.02 (n=18), respectively; for aminopyrine de methylase in the kidney, 0.23±O.02 (n=18); and for benzphetamine demethylase in the lung, 1.24±0.22 (n=12). Oemethylation of: Erythromycin Ethylmorphine Aninopyrine Benzphetamine Tissue: Liver Intestine Kidney Liver Intestine Kidney Kidney Lung I(~atm~[]l; ImidazQI~ MCB 273±21* 127±23 130±19 183±11* 158±61 150±29 66±17 116±10 ECB 364±27* 65,235 222,448 185± 9* NO NO 97, 108 NO CECP Noa 0,63 47,57 329±13* 7,40 50,58 106,120 133, 173 CLTZ 1202±59* 86,104 98, 153 353±27* 61,67 71± 8 106±18 147±20 NAF 268±29* NO 0,0 145±12* NO 100,16 130, 186 NO NBI 200±18* 189±87 99±39 194±18* 141±45 199±40 144±10* 129±21 NMN 118, 192 175,185 0, 133 NO 133,182 42,59 196,295 98, 123 w Table 1.3. Continued. Demethylation of: Erythromycin Ethylmorphine Aminopyrine Benzphetamine Tissue: Liver Intestine Kidney Liver Intestine Kidney Kidney Lung I[~atm~Dt; I[ial;QI~5; ICI151885 203±18* 128,86 199,0 212±40· ND 93, 78 135±6 88, 140 FLU 324±26* 146±19 95±18 246±5* 83±16 148±17 95±24 94±9 ~[idiD~5' 26020 105±28 221,42 155,44 140±4 150,63 98 116,67 163,125 86034 159±14* 231±119 282±11* 160±13* 98±129 99±3 128±16 220±32 DPP 752±162* 83,86 0,8 389±55* 85,63 62, 121 69,118 80,52 Imidal;QC~[idiD~5 ; LY163252 93±4 100 101 98±16 62 75 114 93 LY175326 115±18 177 174 121±9 24 73 100 119 SUM 80±8 16 59 75±7 7 43 79 108 ~[imidiD~; LY56110 452±37* 219 366 180±16* 150 42 88 76 aND = not determined ..p,. that seen with CL TZ. Discussion This investigation has shown that compounds containing a nitrogen heterocyclic moiety, including imidazoles, triazoles and pyridines, can induce P450 in extrahepatic organs. Prior investigations have demonstrated that, with a few exceptions, PAHs or closely related compounds were the major inducers of extrahepatic P450 (Nebert, Gielen and Goujon, 1972; Stohs, et al., 1976a; Stohs, et al., 1976b; Anders, 1980; Endou, Koseki, Hasamura, Kakuno, Hojo and Sakai, 1982; Funae, Seo, and Imaoka, 1985; Keith,Olson, Wilson and Jefcoate, 1987). Two of the sixteen N-substituted imidazoles examined consistently induced in a PAH- type manner (EROD induction) in extrahepatic tissues. Monooxygenase activities which are not extensively induced by PAHs (benzphetamine, aminopyrine, and erythromycin demethylases) were also induced in extrahepatic tissues by some of the heterocyclic compounds. 15 The substrate specificity of constitutive renal P450 is largely directed toward the hydroxylation of endogenous prostaglandins and fatty acids, such as lauric acid, while the extent that testosterone and xenobiotics are metabolized by constitutive renal P450 is much less. However, low levels of aminopyrine demethylase activity, an activity catalyzed by many isozymes, but most efficiently by isozymes inducible by PB (IIB1/2), have been reported in uninduced rat kidney microsomes (Orrenius, Ellin, Jakobsson, Thor, Cinti, Schenkman and Estabrook, 1973; Utterst, Mimnaugh and Gram, 1977). More recently, characterization of three P450s purified from rat kidney cortex revealed two forms which had detectable levels of catalytic activity toward two xenobiotics, aminopyrine and 7-ethoxycoumarin, as well as lauric acid (Imaoka, Nagashima and Funae, 1990). Immunocytochemical studies have found very low levels of in1munoreactivity in uninduced rat kidney microsomes to antibodies to hepatic PAH-inducible P450 isozyme c (IA 1) ( Foster, Elcombe, Boobis, Davies, Sesardic, McQuade, Robson, Hayward and Lock, 1986; 16 Christou, Wilson and Jefcoate, 1987), although a subsequent study did not find detectable levels of staining for this isozyme (Sesardic, Cole, Edwards, Davies, Thomas, Levin and Boobis, 1990). The immunoreactive constitutive isozyme was deemed functional in one of the former studies (Christou, et aI., 1987), based upon antibody inhibition of 7,12-dimethylbenz[a]anthracene metabolism. Another group of investigators has found low levels of transcription in rat kidney of the genes for two PAH-inducible isozymes, c (IA 1) and d (IA2), although mature mRNA was only detected for isozyme c (IA 1) (Pasco, Boyum, Merchant, Chalberg and Fagan, 1988). Compounds which induce hepatic P450 in a PAH- type manner have also been reported to induce P450 in kidney cortex of several species (Nebert, et aI., 1972; Atlas, Thorgeirsson, Boobis, Kumaki and Nebert, 1975; Litterst, et al., 1977; Zenser, Mattammal and Davis, 1978; Anders, 1980; Hook, Elcombe, Rose and Lock, 1982; Rush, Pratt, Lock and Hook, 1986). Early investigations in rats showed that MC or benzo(a)pyrene induced renal P450 concentration 2-fold, and induced renal aryl hydrocarbon hydroxylase (AHH) activity (Weibel, Leutz, Diamond and Gelboin, 1971; Nebert, et al., 1972; Grundin, Jakobsson and Cinti, 1973; Ciaccio and De Vera, 1976). MC treatment of rats, in more recent studies, dramatically increased the level of immunoquantitatable isozyme c (IA 1) in kidney microsomes (Christou, et al., 1987; Sesardic, et aI., 1990). Treatment of rats with the PAH-type inducer, B-naphthoflavone (BNF), was found to increase renal P450 concentration approximately 2.5-fold, and ERGO (Rush, et aI., 1986) and 7-ethoxycoumarin deethylase (Wolf, Hook and Lock, 1982) activities were concurrently induced. BNF also dramatically induced a P450 that was reactive to isozyme c (IA 1) antibody (Foster, et aI., 1986) as well as the level of mRNA for isozyme c (IA 1 ) in rat kidney, although this increase in message was not associated with an increased rate of transcription (Pasco, et al., 1988). Few studies have examined the ability of compounds other than PAHs to induce renal P450. Cortical P450 concentration, monooxygenase activities, and immunoreactive P450 b/e (IIB1/2), were not induced in rat kidney by either 17 PB or 2, 4,2', 4'-tetrachlorobiphenyl (Uehleke and Greim, 1968; Anders, 1980; Hook, et al., 1982; Rich, Sesardic, Foster, Davies and Boobis, 1989), although Ct"lristou, et al. (1987) reported immunoquantifiable, albeit very low, levels of functional P-450b (liB 1) in three out of six sets of rat kidney microsomes following PB treatment. Other induction studies found that both polybrominated biphenyls and isosafrole induced rat renal P450 concentration, and EROD and 7-ethoxycoumarin deethylase activities, although the induction of these activities was less than that seen after treatment with BNF (Hook, et al., 1982; Rush, et al. 1986). All seven N-substituted imidazoles and several of the nonimidazole nitrogen heterocycles examined in the present study induced hepatic P450 concentrations to varying extents: however, only five of these, three imidazoles, NAF, NBI and NMN, the triazole IC1151885, and the pyridine, DPP, were large inducers of P450 in the kidney. These five each induced cortical P450 concentration to approximately 200% of control, with the exception of NMN, which increased the concentration to 400%, an extent similar to that seen after MC treatment. Induction of renal P450-dependent activities by the three N-substituted imidazoles, NAF, NBI and NMN, and the triazole, IC1151885, included activities associated with non-PAH- inducible P450 isozymes. In addition, two compounds that did not induce cortical P450 concentration (ECB, 86034) did induce erythromycin demethylase activity, an activity inducible by dexamethasone. Constitutive concentrations of P450 are considerably less in rat small intestinal mucosa than in liver. However, low constitutive levels of drug metabolizing activities such as ethylmorphine demethylase (Thomas, Baba, Greenberger and Salsburey, 1972) and EROD (Burke, Prough and Mayer, 1977; Hassing, AI-Turk and Stohs, 1989) are detectable in the mucosal tissue. Recent immunocytochemical investigations have yielded conflicting evidence with regard to the constitutive levels of xenobiotic metabolizing isozymes of P450 in the rat intestine. Isozymes inducible by PB in the liver (b/e; IIB1/2) were detected in uninduced male Wistar rat small intestinal enterocytes by Rich, 18 et al. (1989), but not in microsomes isolated from the small intestine of untreated male Sprague-Dawley rats (Christou, et al., 1987). However, a cDNA probe recognizing mANA of both isozymes b (IIB1) and e (IIB2), revealed detectable levels of message in untreated Sprague-Dawley rat intestine (Traber, Chianale, Florence, Kim, Wojcik and Gumucio, 1988). Findings from three studies of isozyme c (IA 1) have been consistent (Foster, et aI., 1986; Christou, et al., 1987; Sesardic, et a/., 1990) in showing low levels of EAOD activity in untreated rat intestinal mucosa, but no immunoquantitatable isozyme c (IA 1 ). Treatment with PAHs induced immunoreactive isozyme c (IA 1) in rat intestine, despite undectectable constitutive levels of this isozyme, and the concentration of mucosal P450 was induced to a degree similar to that seen in the liver and kidney (Stohs, et a/., 1976a, b; Dubey and Singh, 1988; Christou, et al., 1987; Sesardic, et a/., 1990). However, while treatment with the PAH-like inducer, BNF, increased P450 isozyme(s) in intestine which are immunochemically similar to those induced by BNF in the liver (Lindeskog, et al., 1986), the pattern of intestinal isozymes induced following MC treatment differed from the pattern induced in liver, based upon qualitative differences in benzo(a)pyrene metabolite formation (Stohs, et a/., 1976a). Consistent with the results of the latter investigation, Sesardic, et a/. (1990) found that MC treatment of rats induced both isozymes c (IA 1) and d (IA2) in the liver, but only isozyme c (IA 1) in the small intestine (Sesardic, et a/., 1990). PB did induce P450 concentration in rat intestinal mucosa (Dubey and Singh, 1988) and epithelial cells (Bonkovsky, Hauri, Marti, Gasser and Meyer, 1985), in contrast to the lack of inductive response observed in rat kidney. However, monooxygenase activities in the intestine were induced to a lesser degree by PB than by PAHs (Stohs, et al., 1976a, b). Christou, et al. (1987) reported slight induction in the small intestine of immunoreactive isozyme b (IIB1), but not e (IIB2), following treatment of rats with PB, and, in support of this finding, Traber, et al. (1988) found that both PB and Aroclor 1254 induced the amount of mANA for isozyme b (IIB1) but not e (IIB2). In addition to the induction of PB-inducible isozymes, evidence has also been presented for the 19 induction of glucocorticoid-inducible isozymes in intestine following treatment with dexamethasone (Watkins, Wrighton, Schuetz, Molowa and Guzelian, 1987) or pregnenolone-16alpha-carbonitrile (Stohs, et al., 1976b). In the former study (Watkins, et al., 1987), a two-fold increase in erythromycin demethylase activity in rat enterocytes was accompanied by a three-fold increase in immunoreactive steroid-inducible P450 (isozyme p; IliA 1). Intestinal P450 concentration was induced 5-fold by the imidazoles, NBI and NMN, and two- and three-fold by the triazole, IC1151885, and the pyridine, OPP, respectively. The magnitude of induction seen following NBI and NMN treatment was similar to that seen in this tissue after induction with either PB (Bonkovsky, et al., 1985) or MC (Stohs, et al., 1976a). The two imidazoles induced the PAH-inducible EROO activity to a great extent, but induction of erythromycin demethylase activity, which is catalyzed by a steroid-inducible isozyme, was also observed. Neither the triazole nor the pyridine induced these activities; however, some induction of erythromycin de methylase was observed following treatment with two compounds that did not induce P450 concentration in this tissue (FLU, 86034). These results support those of previous investigations that found that isozymes other than those induced by PAHs can be induced by xenobiotics in the intestinal mucosa. The cells of untreated rat lung contain detectable amounts of xenobiotic metabolizing P450 isozymes. Recently, Voight, Kawabata, Burke, Martin and Guengerich (1990) found immunohistochemical staining for the major forms of P450 induced by PB (liB 1 ), BNF (lA 1) and pregnenolone-16alpha-carbonitrile (lilA 1) in several cell types of rat lung, including epithelial, Type II and Clara cells. These results agree with those of Rampersaud and Walz (1986), Rich, et al. (1989) and Christou, et al. (1987), who all found staining for isozyme b (1IB1), but not e (1182), in untreated rat lung. However, results from other investigations of the localization of isozyme c (IA 1) are inconsistent. Both Christou, et al. (1987) and Foster, et al. (1986) observed low levels of staining to antibodies for the PAH-inducible isozyme in rat lung, but a later investigation (Sesardic, et al., 1990) was unable to duplicate these results. Consistent with the presence of xenobiotic-metabolizing isozymes in untreated rat lung cells, constitutive levels of activities which are preferentially catalyzed by these isozymes have been found, notably benzphetamine demethylase (Matsubara, Prough, Burke and Estabrook, 1974), AHH (Voight, et al., 1990) and EROO (Burke, et al. 1977). The hepatic P450 inducers, PB, polychlorinated biphenyls, 2,3,7,8-tetrachlorodibenzo-p-dioxin, and PAHs have all been found to selectively induce certain P450 isozymes in rabbit lung (Serabjit-Singh, Albro, Robertson and Philpot, 1983), but induction of lung P450 in the rat has generally been limited to that arising from treatment with PAHs (Matsubara, et al., 1974; Sagami and Watanabe, 1983; Funae, et al., 1985; Foster, et al., 1986; Christou, et al., 1987; Keith, et al., 1987; Sesardic,et al., 1990). Investigations with PB, in striking contrast to PAH induction, have generally not shown induction of rat lung P450 concentration (Matsubara, et al., 1974; 20 Funae, et al., 1985; Keith, et al., 1987) or selective isozymes, as determined by monooxygenase activities (Funae, et al., 1985). Immunocytochemical studies with antibodies to PB-inducible isozymes, rat isozyme b (1IB1) (Keith, et al., 1987), rabbit isozyme 5 (Vanderslice, Domin, Carver and Philpot, 1987), and a monoclonal antibody to isozymes band e (IIB1/2) (Rich, et al., 1989), did not show significant enhancment of the staining in rat lungs by phenobarbital pretreatment. Another study supported and extended these results by observing that isozyme b (IIB1), but not e (1IB2), was present in untreated rat lung, but neither form was inducible by PB (Rampersaud and Walz, 1986). However, Christou, et al. (1987) reported slight induction in rat lung of both isozymes b (IIB1 ) and e (IIB2) by PB. The two N-substituted imidazoles, NBI and NMN, induced P450 concentration in lung, as in other tissues. The imidazole, CL TZ, and the triazole, FLU, which did not induce in any other extrahepatic tissue examined, induced lung P450 to about the same extent as that seen after NBI and NMN treatment. NMN and NBI both induced EROD in the lung, as well as p-nitroanisole demethylation. In contrast, CL TZ and FLU induced p-nitroanisole and benzphetamine demethylases, but not EROO. The induction by CL TZ and FLU of benzphetamine demethylase (an activity selectively catalyzed by PB·inducible isozymes) is good evidence that treatment with selected agents can induce non-PAH-inducible isozymes in rat lung. 21 Some comparisons in the inductive response of the organs to the nitrogen heterocycles may be made, although pharmacokinetic distribution studies are needed to determine the concentrations that the compounds attain in each each organ. Those compounds that were high magnitude inducers of hepatic P450 were also inducers in extrahepatic tissues, but some organ selectivity was observed: the imidazole, CLTZ, only induced in the lung; the triazole, ICI151885, induced in kidney cortex and small intestine; and NBI induced in all three extrahepatic tissues examined. NMN, which did not induce hepatic P450 to the extent that NBI, CLTZ and ICI151885 did, was a good inducer of P450 in all three extrahepatic tissues examined. The pattern of P450 isozyme induction in extrahepatic tissues by these compounds sometimes differed from the pattern induced in the liver. For example, NSI, which induced in the liver in a manner similar to both PAHs and PB (EROO and ethylmorphine demethylase), did not induce in a PB-type manner (benzphetamine demethylase) in the lung. This selectivity of induction could arise from differing populations of constitutive isozymes or 'from variations in the regulation of induction in the different organs. However, consistency of response in all organs was observed for some compounds. Both NSI and NMN were good inducers of EROO in the liver, and they were also good inducers of this activity in all three extrahepatic tissues. Many compounds containing the imidazole moiety have been found to be inducers of hepatic P450. In the present study, some of these compounds and other nitrogen heterocycles have also proven to be good inducers of extrahepatic drug metabolizing P450 isozymes. Thus, compounds other than PAHs have been found to selectively induce extrahepatic P450 in the rat, and, as in the liver, it is evident from activity determinations that isozymes other than those induced by PAHs are increased by nitrogen heterocycles. CHAPTER 2 ALTERATION OF SEDIMENTATION PROPERTIES OF HEPATIC SUBCELLULAR ORGANELLES BY NBI Introduction Induction of hepatic and extrahepatic P450 by nitrogen heterocycles, reported in Chapter 1 , revealed that two N-substituted imidazoles, NBf and NMN, were consistently good inducers in aU tissues examined. However, 14 other nitrogen heterocycles studied showed little interorgan consistency in inductive response. Structure activity relationships could not be utilized to seek an explanation for this inconsistency, because an insufficient number of closely related compounds were examined in the induction studies. One potential causal factor that could be examined was the possibility that, with induction, nonuniform changes in the structure of the endoplasmic reticulum (ER) might occur, and this would result in the production of a heterogeneous population of microsomes during homogenization. Changes in the physical nature of the membrane could alter the sedimentation of microsomes during preparative centrifugation, thus altering the microsomal yield. Induced P450 might not appear in the microsomal pellet, if it were differentially located in either more or less dense microsomes. P450 contained in more dense microsomal fractions could sediment along with the mitochondrial fraction during low- speed centrifugation, or P450 contained in less dense microsomes could remain in the cytosolic fraction during high- speed centrifugation. However, a comparative examination of the hepatic and extrahepatic microsomal yields (see Table 1.1 ) found that most of the compounds did not siQnificantly affect this parameter. Thus, this is not likely to be a causal factor for the inconsistent interorgan 23 induction observed with most of the compounds. However, treatment with some of the compounds did decrease microsomal yields in selected tissues. Two compounds that were consistently good inducers of P450 concentration in all organs, NSI and NMN, markedly decreased microsomal yields in liver and lung, respectively. Three other compounds, NAF, OPP and SUM also significantly decreased hepatic microsomal yield. The effects that this occasional phenomenon might have on both the quantitative and qualitative analysis of data from induction studies suggested that the phenomenon warranted investigation. The nature of this effect in the liver following treatment with NSI was investigated by examining the sedimentation of subcellular organelles from liver homogenates by differential centrifugation. Assays for three microsomal parameters, P450 concentration and two monooxygenase activities, p-nitroanisole de methylase and EROO, and for a marker of the outer mitochondrial membrane, monoamine oxidase, were conducted using the supernatants of centrifuged liver homogenates. The results showed that NSI treatment alters the sedimentation characteristics of both E R and mitochondria. Methods Kynuramine, the substrate for monoamine oxidase activity determinations, was obtained from Sigma Chemical Co. (St. Louis, MO). The sources for all other chemicals and biochemicals were the same as described in Chapter 1 . The maintenance and NSI treatment regimen for male, Sprague-Dawley rats was the same as that used in previous induction experiments, with the exception of one experiment in which only a single 75 mg/kg dose of NSI was employed and only a 3 hour period elapsed before sacrifice. Animals were sacrificed, and the livers were perfused, removed and homogenized, as described in Chapter 1. In these sedimentation experiments, care was taken to use the same Potter-Elvehjem teflon pestle and glass homogenizer tube throughout to reduce variability in the homogenization arising from differences 24 in the sheer force resulting from variations in clearance. Equal volume aliquots of the homogenates were centrifuged in 10 ml tubes at different g-forces (500, 1,000,3,000,5,000, 18,000 g) for 20 minute intervals using a JA-17 rotor in a Beckman J2-21 centrifuge. The supernatants from these centrifugations were removed and assayed spectrophotometrically for microsomal parameters (P450 concentration, p-nitroanisole demethylase and EROO) and protein content. In addition, a spectrophotometric assay for a mitochondrial marker, monoamine oxidase (Weissbach, Smith, Daly, Witkop and Udenfriend, 1960), was also conducted. The P450 concentration and all enzyme activities were determined using a constant volume of supernatant from each centrifugation. Since the supernatants contained proteins other than those associated with the subcellular organelles of interest, P450 concentration and the enzyme activities were not calculated per mg protein, but as a percent of that present in the 500 g supernatant. The 500 g supernatant was used as the baseline control value because attempts to determine enzyme activities in uncentrifuged, whole homogenates found the absorbance to be excessively variable. When microsomal fractions were prepared, the supernatants from the 18,000 g centrifugation were centrifuged at 105,000 g for 50 minutes. The microsomal pellet obtained was resuspended in 0.15M KCI and recentrifuged at 105,000 g for 40 minutes. Resutts The effects of treatment of rats with NBI on the liver weight! body weight ratio and the hepatiC microsomal yield are shown in Figure 2.1. The effects on these parameters following treatment with another N-substituted imidazole, clotrimazole, and with the classical inducer, phenobarbital, are also depicted, for comparative purposes. Like phenobarbital, both N-substituted imidazoles significantly increased liver weight, resulting in significant increases in the liver weight! body weight ratio. Clotrimazole treatment was similar to phenobarbital treatment in that the yield of microsomal protein obtained was significantly D "0 (i) 's;... 12 "ffi E g 10 e u 'E 8 ~ o r.:«1 ~ 6 4 2 o 25 CONTROL PB CLTZ NBI TREATMENT Figure 2.1. Alteration of liver weight! body weight ratio and hepatic microsomal yield by inducing agents. Male Sprague-Dawley rats (150-250g), either untreated (CONTROL; n=16) or treated with PB (n=5) (80mg/kg, i.p., daily for 4 days, followed by a 24 hour period after the last dose), or one of two N-substituted imidazoles, CL TZ (n=6), or NBI (n=22), were sacrificed, the livers perfused, removed and weighed, and microsomes prepared by differential centrifugation (see Chapter 1 for imidazole dosing regimen and microsomal preparation protocol). The effect of inducers on the mean liver weight, calculated as a percent of the body weight, is depicted by the stippled bars with the standard error of the mean. The mean yield of microsomal protein, in mg per g liver, is represented by the unshaded bars with the standard error of the mean. 26 increased. In marked contrast, NBI treatment resulted in a 55% decrease in liver microsomal yield compared to untreated controls. To investigate this anomalous low microsomal yield accompanying induction, the effect of NBI treatment on the sedimentation of hepatocyte ER and mitochondria, as determined by marker enzyme activities, was examined. The sedimentation of total protein and of ER and mitochondrial marker enzymes in liver homogenates from untreated and NBI-treated rats is shown in Figure 2.2. NBI treatment altered the sedimentation of total liver protein (Figure 2.2a) slJch that, at each of the g-forces examined, a greater percentage of protein was removed from the supernatants of NBI- treated rat liver homogenates than from those of untreated controls. For example, approximately 25% of the protein in NBI homogenates had been sedimented after centrifugation at 5,000 g, compared to 15% for controls. The protein removed from the supernatant after centrifugation at 18,000 g (the g-force usually used during microsomal preparation to clear the homogenate of unwanted subcellular organelles prior to separation of microsomal membranes from the cytosol at 105,000 g) was approximately 350/0 and 200/0 for NBI-treated and control rat liver homogenates, respectively. The sedimentation patterns of microsomal parameters are shown in Figure 2.2b-d. NBI treatment altered the sedimentation of P450 and p-nitroanisole demethylase activity in a manner similar to that seen with total hepatic protein. At any g-force, the cytochrome P-450 and p-nitroanisole demethylase removed from the supernatant was greater for the NBI-treated than for the untreated rat liver homogenates. The effect of NBI treatment on the sedimentation of EROO activity could not be compared with control sedimentation patterns, since this activity is selectively catalyzed by nonconstitutive, PAH-inducible isozymes. However, the pattern of EROO sedimentation indicated that these selective isozymes were sedimented to a greater extent at low g-forces than total P450 and p-nitroanisole demethylase (Le., 80% vs. 30% and 40%, respectively, at 5,000 g). The amount of EROD remaining in the supernatant after the 18,000 9 27 Figure 2.2. Effect of treatment of rats with NBI on the sedimentation of total hepatic protein and microsomal and mitochondrial enzymes in liver homogenates during centrifugation. Livers from untreated, control (shaded circles) and NBI-treated (unshaded squares) rats were homogenized and centrifuged at g-forces of 500, 1 ,000, 3,000, 5,000, 18,000 g for 20 minutes. Supernatants from these centrifugations were assayed for a) protein, b) P450, c) p-nitroanisole demethylase activity, d) EROD activity, and e) monoamine oxidase activity. The data for supernatants of increasing g-force (abscissa, g-force x 0.001) were calculated as a percent of that of the 500 g supernatants (ordinate). Each data point is the mean of 4-8 samples and standard errors were determined. The average control and NBI-treated 500 g values for a) protein, b) P450, c) p-nitroanisole demethylase activity, d) EROD activity, and e) monoamine oxidase activity were: a) 26.70 ± 0.47 and 26.22 ± 1.09; b) 1.59 ± 0.09 and 5.56 ± 0.48; c) 1.26 ± 0.08 and 6.44 ± 0.95; d) 0.66 ±O.1 0 (NBI-treated only); and e) 1.31 ± 0.10 and 1.27 ± 0.39. 28 a) Total Hepatic Protein b) P450 100 100 90 90 80 80 O'l 70 70 8 60 60 LO '0 50 50 ?!! 40 40 30 30 20 20 10 10 0 0 0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 c) p-Nitroanisole Oemethylase d) EROO 100 100 90 90 80 80 O'l 70 70 00 60 60 LO '0 50 50 ?!! 40 40 30 30 20 20 10 10 0 0 0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 g-force X 0.001 e) Monoamine Oxidase 100 90 80 01 70 8 60 LO '0 50 ?!! 40 30 20 10 0 0 2 4 6 8 10 12 14 16 18 g-force X 0.001 29 centrifugation was not detectable by the spectrophotometric assay under the conditions used, although activity could subsequently be detected in the more concentrated microsomal fraction prepared from this supernatant (data not shown). The sedimentation pattern of the mitochondrial marker enzyme, monoamine oxidase, differed from those of total protein, P450 and p-nitroanisole demethylase (Figure 2.2e), in that a larger percentage of the activity was removed at the lower g-force centrifugations (Le., at 1,000,3,000 and 5,000 g). This pattern was seen in homogenates from both control and NSI-treated rats, but NSI pretreatment altered the sedimentation such that a greater amount of monoamine oxidase was removed from the homogenates at lower g-forces, as observed with the microsomal markers. Discussion The technique of differential centrifugation has been used extensively to fractionate subcellular organelles ever since the pioneering studies of Albert Claude (1946a,b). The technique is based upon the separation of particles with differing sedimentation coefficients (as defined for spherical particles of a specific density and radius by the Svedberg equation; Svedberg and Pedersen, 1940) by suspending them in a medium and subjecting them to centrifugal force. The ability to obtain pure populations of subcellular organelles with the technique is limited by both breakage and adhesion of particles, and by the fact that populations of subcellular organelles are often heterogeneous and contain particles with a range of sedimentation coefficients (de Duve, 1975). However, a relatively pure microsomal membrane fraction can be separated from mitochondria and other subcellular organelles using the technique, despite these limitations. The ease and success with which the technique is routinely employed may result in oversights, unless an alteration of the magnitude found in the present study is encountered. A 55% decrease in 30 hepatic microsomal protein yield was observed following treatment with the N-substituted imidazole, NBI. This decreased yield was particularly surprising because NBI treatment increased liver weight to an extent comparable to that observed with both the classical inducer, phenobarbital, and another N-substituted imidazole, clotrimazole (Figure 2.1). All three compounds induced hepatic P450 concentration [three- to four-fold for the two imidazoles (data in Chapter 1) and two-fold for phenobarbital (data not shown)], but NBI distinguished itself from the other two by producing a dramatic decrease in microsomal yield. The nature of this effect was an alteration in the sedimentation properties of the microsomal membranes such that NBI pretreatment increased the amount of microsomal marker enzymes (P450 and p-nitroanisole demethylase activity, Figure2.2b,c) and total protein (Figure 2.2a) removed from liver homogenates after centrifugation at low g-forces. The pattern of sedimentation in NBI homogenates of the nonconstitutive monooxygenase activity, EROO, distinguished itself from the patterns for P450 and p-nitroanisole demethylase, even though the sedimentation could not be measured in untreated rat liver homogenates. The percent of EROD removed at any g-force was greater than that removed for the other microsomal parameters, indicating that the PAH-inducible isozymes induced by NBI were preferentially sedimented at the lower g-forces. An explanation for this could be that the selectively induced isozymes might be located within ER with different biochemical and/or physical properties from the remaining membrane. This could result in the production of microsomal vesicles of greater density, and in the subsequent loss of the induced isozyme in these vesicles during centrifugation at lower g-forces. Membranes, in general, are considered to be dynamic, heterogeneously constructed entities whose individual constituents are continually being replaced, reorganized and modified (Parry, 1978), and, in this regard, ER is no exception. Furthermore, the organization of ER protein constituents is not random. For example, Winqvist and Dallner (1976), using sucrose density gradients, were able to separate a population of microsomes 31 containing high concentrations of phosphatases from other populations containing NADH- and NADPH-cytochrome c reductases. Fouts and Gram (1969) found evidence that drug metabolizing enzymes were also differentially distributed or concentrated throughout the microsomal fraction. It is apparent from earlier studies with the classical P450 inducing agents, phenobarbital and PAHs, that xenobiotics can modify the relative distribution of P450 between rough and smooth microsomes, as well as alter the sedimentation characteristics of microsomal vesicles. Gram, Rogers and Fouts (1967) found that treatment of rabbits with the PAH, Me, increased the P450 concentration only in a rough microsomal fraction, and phenobarbital induced the concentration in both rough and smooth fractions. However, examination of the so-called rough microsomal fraction from Me-treated rabbits found that it was predominantly composed of large smooth surfaced vesicles with a mean dian1eter greater than those seen in control or phenobarbital-treated animals, indicating that Me treatment was associated with an apparent change in the sedimentation characteristics of smooth vesicles. Thus, the authors concluded that the increased P450 concentration by Me was most likely localized in these large, smooth vesicles. This evidence was supported in part in a later study by Murphy, Van Frank and Williams (1969), which found that treatment of rats with Me induced P450 concentration in more dense (higher sedimentation coefficient) microsomal vesicles while phenobarbital primarily increased the concentration in lighter vesicle fractions, which was presumed to correspond to the induction of smooth ER known to occur with phenobarbital treatment. NBI could be differentially inducing selective P450 isozymes in ER which yield microsomal vesicles of greater size and/or density upon homogenization, similar to the effects of Me noted in the above two studies. The loss of induced isozymes contained in microsomal vesicles of higher sedimentation coefficient could account for the greater percent loss of P450 and p-nitroanisole demethylase activity from liver homogenates of NBI- treated rats, as compared to the loss of these parameters from untreated rat homogenates. In addition, 32 the greater sedimentation of nonconstitutive EROO activity, as compared to the sedimentation of P450 and p-nitroanisole demethylase activity in NBI-induced rat liver homogenates, suggests the possibility that the greater percent loss of microsomal parameters may be due to the specific loss of PAH-inducible isozymes in vesicles of higher sedimentation coefficient. While both the above-mentioned cell fractionation studies (Gram et al., 1967; Murphy et aI., 1969) agreed that Me induced P450 in microsomal fractions with higher sedimentation coefficients, the two studies were not in complete agreement with regard to phenobarbital induction of P450 in hepatic microsomal fractions. A later study by Massey and Butler (1979), which incorporated both cell fractionation and morphological techniques, examined the proliferation of the ER n1embranes in rat hepatocytes following phenobarbital treatment through morphometric analysis. A substantially greater increase was found in the total area of hepatic ER (both rough and smooth membranes) after phenobarbital treatment (329%) than in the amount of microsomes measured after subcellular fractionation (92%). This discrepancy was due in large part to the loss of large rough and smooth ER vesicles in the 9500 g pellet, which were of much greater size than those recovered in the final microsomal pellet. However, the measurement of P450 as a microsomal marker in liver homogenates and microsomal pellets showed that phenobarbital administration had no significant effect on the recovery of P450. This suggested to the authors that the large microsomal vesicles in the 9500 g pellet had a relatively low content or specific activity of P450 compared to the vesicles recovered in the 105,000 g pellet. To account for this, they proposed that phenobarbital may have disproportionately increased the phospholipid content of the ER such that the lost larger vesicles might have had a higher phospholipid/P450 ratio, resulting in a lower specific activity of P450 for these vesicles. This could result in an unaltered recovery of P450 in the liver microsomal fraction of phenobarbital-treated rats. Such a proposal was not unreasonable, given the results of an earlier study by Ishidate and Nakazawa 33 (1976), which found that both phenobarbital and polychlorinated biphenyls altered the lipid content of hepatic ER, phenobarbital by increasing the synthesis of phosphatidylcholine, and polychlorinated biphenyls by inhibiting the catabolism of menlbrane phospholipids. The altered sedimentation of microsomal vesicles following NBI treatment in the present study may also reflect a change in the lipid content of the newly synthesized ER, due either to increased synthesis or decreased catabolism of phospholipids. These mechanisms have not been explored as possible causes of the effects of NBI treatment on microsomal sedimentation, but the possibility that NBI might be exerting a more direct effect, by interacting with and perturbing the structure of membrane components, has been examined. This possibility was suggested by evidence 'from Sgaragli, Della Corte, Rizzotti-Conti and Giotti (1977), who found that monocyclic compounds having an aliphatic chain containing at least two carbon atoms interact with biomembranes producing solubilization and release of proteins. Their study found that lipophilicity was not the only property necessary for solubilization, and that the more electron-dense aromatic ring structures, such as 2,tert-butyl,4 methoxyphenol (BHA) and 3,5,di-tert-butyl, 4, hydroxytoluene (BHT), exerted the most remarkable effects. A preliminary experiment in which NBI was administered to rats 3 hours prior to sacrifice (a time assumed to be sufficient for NBI to exert any solubilization effect, but not adequate for induction or alterations in lipid synthesis or catabolism) found that, using P450 as a marker, the sedimentation of microsomes in liver homogenates from these rats was not altered from that in untreated rat homogenates (data not shown). In addition to its effects on microsomal parameters, NBI also altered the sedimentation of the outer mitochondrial membrane enzyme, monoamine oxidase, although the extent of the alteration was not as great as that seen with the microsomal parameters. The pattern of sedimentation of this enzyme closely approximated the sedimentation of EROD activity. This observation suggests that NBI is inducing selective isozymes which are localized within ER 34 membrane that is physically associated with the outer mitochondrial membrane. Close approximation and physical connection between the ER and outer mitochondrial membrane has been reported, and from electron micrographs of freeze-etched tissue, which visualized the membrane interiors, the apposed membranes were found to structurally resemble each other (Tewari and Malhotra, 1973). Furthermore, exchange of phospholipids between microsomes and inner and outer mitochondrial membranes has been reported (Slok, Wirtz and Scherphof, 1971). Additionally, rough ER and mitochondria may comprise a structural unit necessary for enzyme induction and membrane biogenesis, a suggestion that was based upon an increased frequency of juxtaposition of these organelles following phenobarbital treatment. This juxtaposition may be formed due to the requirement for heme, synthesized from the a-aminolevulinic synthetase and ferrochelatase in the mitochondria, that is necessary for complete P450 synthesis in the ER (Mills and Jones, 1974). However, the effects of NBI on the sedimentation of microsomal parameters are not likely due solely to a physical connection with mitochondrial membranes, if this increased juxtaposition is assumed to occur with a/l inducing agents, since a/l inducers do not decrease the microsomal yield as NBI does. In conclusion, treatment of rats with the N-substituted imidazole, NSI, markedly decreased microsomal yield, although the liver weight was significantly increased. This effect was due to an alteration in the sedimentation properties of subcellular organelles which led to increased removal of microsomes and mitochondria from liver homogenates during low speed centrifugation. Further experiments are necessary to ascertain the mechanism of this alteration. CHAPTER 3 INTERORGAN DIFFERENCES IN THE IN VITRO BINDING OF NBI TO P450 AND POSSIBLE MODIFICATION BY RESIDUAL COMPOUND FOLLOWING IN VIVO TREATMENT I ntraduction The inductive effects of 16 nitrogen heterocycle compounds on rat hepatic and extrahepatic P450, investigated in Chapter 1 , were not consistent between organs. Induction of P450 concentration in one organ by a compound was not necessarily paralleled by induction of this enzyme in other tissues. Similarly, consistent induction of a P450 isozyme-selective activity by a given compound was rarely observed in all organs examined. Two compounds, NBI and NMN, distinguished themselves from the olajority in that they consistently induced P450 in all organs. While these two compounds were those that induced P450 concentration to the greatest extent in each tissue, quantitative differences in the degree of induction between tissues were seen. The concentration of P450 was induced to the greatest extent in the small intestinal mucosa (five-fold), to an intermediate extent (three­to four-fold) in the liver and kidney cortex, and to the least extent in the lung (1.5-fold). Similarly, while p-nitroanisole de methylase and EROD activities were induced in all organs by the two compounds, interorgan variation in the extent of induction of these activities also occurred. Quantitative differences in the induction of additional monooxygenase activities by the two compounds were also observed in several organs. For example, following NBI treatment, 36 erythromycin demethylase activity was induced approximately two-fold in both liver and intestine, but was not altered in the kidney cortex. Changes in P450 concentration and selected monooxygenase activities in hepatic and extrahepatic tissues following treatment with one of these compounds, NBI, were examined further to explore the possible causes of these interorgan inconsistencies. Part of this investigation examined whether the dosing regimen that was utilized throughout the induction studies (which included a 48 hour period following the final dose to clear residual compound from the tissues) might have precluded the observation of short-lived inductive responses. Such an inductive response may have been missed at the single time point examined if interorgan differences in P450 lability existed. These differences could allow disproportionate losses in P450 to occur in certain tissues and might account for the apparent inconsistencies in inductive effects. Nitrogen heterocycles can reversibly inhibit P450 by binding to the ferric heme at the sixth axial coordination position. This ligation results in a shift in the distribution of ferric d-orbital electrons to a low-spin state, and can be observed in vitro as a type II binding spectrum (Schenkman, Remmer and Estabrook, 1967; Jefcoate, 1978; Schenkman, Sligar and Cinti, 1981). Compounds containing an N-substituted imidazole moiety are among the most potent of this class of inhibitors. Numerous structure activity relationship studies have determined that both lipophilic and steric factors are important determinants of imidazole inhibitory potency, with a lipophilic substituent in the 1- or 4(5)-position and a sterically accessible lone electron pair provided by an imidazole nitrogen being requirements for optimal potency (Wilkinson and Hetnarski, 1974; Rogerson, Wilkinson and Hetnarski, 1977; Kapetanovic and Kupferberg, 1985; Murray, 1987a). Some imidazoles are widely used as therapeutic agents, and their inhibition of P450 in vivo can result in clinically important drug interactions. For example, the orally active antimycotic, ketoconazole, has been shown to inhibit the human metabolism of cyclosporin (Anderson and Blaschke, 1986), mephenytoin and debrisoquine (Atiba, 37 Blaschke and Wilkinson, 1989). Differences in P450 susceptibility to N-substituted imidazole inhibition could be responsible for apparent inconsistencies in inductive response observed between organs. P450 isozymes in one tissue may have a greater affinity for a particular imidazole inhibitor than isozymes in other tissues or than other isozymes in the same tissue, resulting in differential effects on expressed monooxygenase activities. Alternatively, interorgan variations in the distribution or metabolism of the imidazole could lead to its persistence in certain tissues, and its inhibition of all P450 isozymes in these tissues could result in variations in the observed induction. Therefore, the persistence of residual imidazole inhibitor was examined using microsomes prepared from liver and extrahepatic organs of rats treated with NBI followed by "wash-out" periods of various durations after the last dose administered. Evidence for imidazole persistence was sought through inhibited monooxygenase activities, slowed development of CO- ferrous P450 binding spectra, and alterations in type II binding characteristics. Additionally, microsomes were examined to ascertain any differences between tissues in the susceptibility of P450 to in vitro inhibition by NBI. Methods Male Sprague-Dawley rats were treated with NBI as described in Chapter 1 , except that the standard "wash-out" period of 48 hours following the last dose administered was modified to include times of 36 and 72 hours. Another group of animals was administered high doses of BNF (80mg/kg, daily) for 3 days and sacrificed 24 hours after the last dose. Microsomes were prepared from liver, kidney cortex, lung and small intestinal mucosa, and were assayed for protein, P450 concentration and p-nitroanisole demethylase and EROD activities, as described in Chapter 1. Inhibition of EROD activity by NBf in vitro was assessed in NBI- and BNF-induced hepatic and extrahepatic microsomes (constitutive EROD activity in microsomes from all organs of 38 untreated rats is insufficient to allow any assessment of NBI inhibition). Microsomes (0.10 and 0.42 mg protein/ml for hepatic and extrahepatic microsomes, respectively) were diluted in buffer (1 OOmM potassium phosphate, pH 7.8) containing substrate (6 JlM), and placed in two cuvettes. Metabolism was initiated by adding NADPH (0.2 and 0.5mM for hepatic and extrahepatic microsomes, respectively) to the sample cuvette, and the uninhibited EROO activity (rate of increase in absorbance at 572nm) was monitored for 2.5 minutes. A small volume (less than 0.5% dilution of cuvette contents) of NBI dissolved in buffer was then added to the sample cuvette and the rate monitored for an additional 2.5 minutes. The assay was repeated with several different concentrations of NBI, and the concentration of NBI which inhibited the basal rate by 50% (Le., IC50) was determined. The development of the ferrous P450 CO-binding spectrum, which is normally used to measure the concentration of P450 in microsomes, was monitored over time. The protocol used in Chapter 1 for extrahepatic microsomes was modified so that the spectral development could be followed. Hepatic and extrahepatic microsomes were diluted (2mg protein/ml for liver and 1.7mg protein/ml for extrahepatic microsomes) in buffer (50mM Tris HCI, 150mM KCI, 10mM MgCI2, pH 7.4), placed in a cuvette and bubbled with CO for 3-5 minutes. The cuvette was then placed in the light path of a spectrophotometer in dual beam mode, and light beams of 450 and 490 nm were balanced. The difference in the absorbance of the CO-saturated microsomes at 450 nm and at 490 nm was monitored for 30 seconds to establish stability. A small amount (a few Jlg) of the chemical reductant, sodium dithionite, was then added to the cuvette, the contents quickly stirred, and the development of the difference spectrum was monitored for 3-10 minutes. The development of the CO spectrum in a second equivalent aliquot of microsomes in the presence of 100JlM NBI was also monitored. For these determinations, a small volume (0.25% dilution of cuvette contents) of NBI in buffer was added to the diluted microsomal suspension prior to bubbling with CO. The time required 39 for the development of 670/0 of the n1aximal absorbance in uninhibited (no added NBI) microsomes was determined (this time was arbitrarily chosen for ease of measurement). The time required to develop this amount of absorbance (67% of maximal in uninhibited microsomes) in the same microsomes in the presence of 100llM NBI was also determined (with NBI added, the maximal absorbance rarely reached that observed in uninhibited microsomes within 10 minutes). The type II binding spectrum of NBI in microsomal samples was determined by difference spectroscopy of the absorbance changes in the Soret region, and quantitated as the absorbance difference between 431 and 390nm. Small volumes of NBI in buffer (50mM Tris HCI, 150mM KCI, 10mM MgCI2, pH 7.4) were successively added to microsomes diluted in the same buffer (dilution of microsomes with NBI solution additions never exceeded 0.7 and 1.7% in hepatic and extrahepatic suspensions, respectively), and the change in absorbance was monitored with each addition. Double reciprocal plots of the absorbance changes versus the NBI concentration were constructed from the data. The plots for the hepatic microsomes were linear, with correlation coefficients greater than 0.9. An apparent dissociation constant (Ksapp [JlM]), and the maximal absorbance (absorbance units, A.U.) were determined from the x- and y-intercepts, respectively. The plots of extrahepatic tissue data were biphasic, and two distinct lines with correlation coefficients greater than 0.9 could be derived if the data were divided above and below an NBI concentration of 8JlM. The two lines were considered to reflect NBI type II binding of high and low affinity, and apparent dissociation constants were determined for each binding affinity. Maximal absorbances were also determined for each affinity and were expressed as a percent of the overall maximal absorbance. For all hepatiC and extrahepatic microsomal samples, the overall maximal type II absorbance change and the P450 concentration in the assay were used to calculate an extinction coefficient for NBI binding (A.U.j.1.M-1cm-1 ). 40 Polyacrylamide gel electrophoresis was performed according to the method of Laemmli (1970). Kidney cortical microsome samples (0.3 mg) were added to 200 JlI of a solution containing 1S0 mg sodium dodecyl sulfate (SOS), 110 mg Tris, 2.S ml glycerol, and 2.S ml O.s% bromophenol blue. To denature microsomal protein, 20 JlI of B-mercaptoethanol was added, samples were sealed and placed in a boiling water bath for S minutes. Following electrophoresis through a 3% polyacrylamide stacking gel (4 cm, 40 rnA) and a 7.So/0 polyacrylamide running gel (8 cm, 70 rnA), the gel was removed and placed in a solution composed of acetic acid: methanol: water (1 :2:7) for 10 minutes. To visualize protein bands, the gel was then stained overnight with Coomassie Brilliant Blue (O.Os% in the solution of acetic acid, methanol and water), as described by Fairbanks, Steck and Wallach (1971). Oestaining was achieved by soaking the stained gel in the solution of acetic acid, methanol and water for several days, with the solution changed daily. Resutts The changes in P4S0 concentration in liver, kidney cortex, lung and intestinal mucosa following treatment of rats with NBI are shown in Figure 3.1 (a-d). The concentrations in tissues from untreated control rats and from rats treated with the PAH-like inducing agent, BNF, are also shown in this figure for comparison. P-4S0 concentration was induced approximately 3.S-, 2.S-, 2-, and 7-fold in liver, kidney, lung, and intestine, respectively, when the final dose of NBI was followed by a 48 hour clearance time. Induction of hepatic P4S0 by BNF (two-fold) was significantly less than that seen with NBI, but the extent of induction by BNF in the extrahepatic tissues more closely paralleled that of NBI. Examination of the P4S0 concentration present at earlier and later time points (36 and 72 hours) after the final NBI dose revealed that, in general, induction was maximal around the 36 and 48 hour time pOints in all tissues, with the exception of liver, where the concentration was slightly higher at 72 hours (up from 2.60 nmollmg at 48 hours to 3.00 nmol/mg at 72 hours). 41 Figure 3.1. Time course of NBf induction of P450 concentration, p-nitroanisole de methylase and EROO activities in microsomes from rat liver, kidney, lung and intestine. Untreated (CON) rats, or rats treated with BNF or NBI ,were sacrificed (for NBI-treated rats, at 36,48, and 72 hours after the 'final dose), and microsomes from liver, kidney cortex, small intestinal mucosa and lung were prepared. P450 concentration (solid bars, nmol/mg), and p-nitroanisole demethyJase and EROO activities (hatched and stippled bars, respectively, nmol/mg/min) were assayed in microsomes. Results are expressed as the mean of 3 or more samples ± the standard error of the mean. Significant differences from control were determined using Student's two-tailed t-test, and are indicated by an *. An ANOVA analysis of variance was also performed on the data for each parameter, and the Tukey multiple comparisons test was used to determine differences between the values for the different treatment groups. A significant difference from the values obtained with microsomes prepared from NBI-treated rats sacrificed at 72 hours after the last dose is indicated by the symbol @. a) Liver b) Kidney 5r 0.5 ~ * 4i 0.4 * * * :f * 0.3 I' ~~~ . .. 0.2 •• 0.1 o 1 __ 0.0 CON BNF NBI36 NBI48 NBI72 CON BNF NBI36 NBI48 NBI72 c) Lung d) Intestine @ 0.25 r * * 0.5 0.20 * * 0.4 * 0.15 * I 0.3 0.10 0.2 0.05 T * 0.1 0.001- 1111 -.. i i - Rm2' i __ cmw; • 0.0 CON BNF NBI36 NBI48 NBI72 CON BNF NBI36 NBI48 NBI72 ..j:::::. N 43 The changes in two monooxygenase activities, p-nitroanisole demethylase and the PAH-inducible activity, EROO, following treatment with NSI are also shown in Figure 3.1. In the liver, both activities were lower at 36 and 48 hours than at the 72 hour time point, relative to P450 concentration. This pattern of lower relative activity at the earlier time points was not repeated in extrahepatic tissues, with the exception of the lung, where EROO activity was lower relative to pulmonary P450 concentration at the 36 hour time point. The decline of P450 and the two monooxygenases differed between the tissues. All three parameters were at maximal levels in the liver at 72 hours after the final NSI dose, while, at this time point in the kidney, both p-nitroanisole de methylase and EROO, but not P450 concentration, had decreased to less than 50% of the values observed at 48 hours. Smaller decreases in both monooxygenase activities were seen in the lung by 72 hours, but, in this tissue, these changes were accompanied by a decline in P450 concentration. The decay of parameters in the intestine was quicker than in the other tissues, as marked decreases to control or nearly control levels in P450 concentration and both monooxygenase activities were seen at 72 hours after the final NSI dose. The yield of microsomal protein from hepatic and extrahepatic tissues following three-day treatment of rats with NSI is shown in Table 3.1. The only significant alterations in this parameter in the extrahepatic tissues were increases in the yield of kidney cortical microsomes at 36 and 48 hours after the final dose. In marked contrast, the yield of hepatic microsomes at these two time points was markedly decreased to 23 and 65% of control, respectively. However, the yield had returned to control values in this organ by 72 hours after the last NSI dose. Interestingly, treatment with SNF also decreased hepatic microsomal yield (to 77% of control), although not to the extent seen with NSI treatment. SNF also appeared to decrease intestinal microsomal yield to an extent comparable to the decrease seen with hepatic microsomes, although the data for this observation is limited. The in vitro inhibition of EROO activity by NSI was examined in hepatic and Table 3.1. Microsomal protein yield from liver, kidney, intestine and lung following treatment with NBt and BNF. 44 Untreated (CONTROL) rats or rats treated with BNF or NBt were sacrificed (for NBI-treated rats, at 36,48 and 72 hours after the final dose), and microsomes from liver, kidney cortex, small intestinal mucosa and lung were prepared. The yield of total microsomal protein (mg) was determined for all tissues. For liver, kidney and intestine the yield was calculated as a fraction of the original tissue weight (mg/g), but lung yield was recorded as the total mg obtained per sample (pooled tissue from four to five rats), since the weight of the starting tissue could not be measured due to the large proportion of connective tissue and fluid removed by the tissue press prior to homogenization. The percent of the average control yield was calculated for microsomal samples from treated rats, and the mean of 3 or more samples and the standard error are reported (except for BNF-induced extrahepatic tissues, where the sample number was 2 and both values are reported). Significant difference from control is indicated by an *, and was assessed by Student's two-tailed t-test. Average yields for contro/liver, kidney, intestine and lung were 6.2±0.4 mg/g (n=53), 3.0±0.2 mg/g (n=28), 4.9±0.4 mg/g (n=20), and 12.2±0.8 mg (n=21), respectively. PERCENT OF CONTROL YIELD Treatment Liver Kidney Intestine Lung BNF 77±5* 170,117 83,66 124,88 NB13x+36 23±3* 186±17* 104±6 106±12 NB13x+48 47±5* 157±8* 113±4 100±2 NB/3x+72 113±36 120±14 89±1 123±27 45 extrahepatic microsomes from both BNF- and NBI-induced rats (Table 3.2). NBI was an equally potent inhibitor of this activity in BNF- and NBI-induced microsomes from all organs with ICSOs in the range of 1-3 J.1M, and, in all extrahepatic microsomes, the ICSO was independent of both P4S0 concentration and EROO activity. The ICSO in hepatic microsomes prepared from NBI-treated rats was less at the 72 hour clearance time (0.7J.1M) than that at the 36 hour time (1.8~M). A 70% increase in EROO activity and a slight increase (1So/0) in P4S0 concentration occurred between these two time pOints. However, an increase in ICSO 'from that seen at 36 hours after the last NBI dose was seen in hepatic microsomes prepared from NBI-treated rats at 48 hours after the last dose (from 1.8 to 3.2J.1M), a time interval in which no change in P4S0 concentration or EROO activity was evident. Therefore, changes in hepatic ICSO also appear unrelated to either microsomal P4S0 concentration or expressed EROO activity. The characteristics of the development of the ferrous P4S0 CO-binding spectrum in hepatic and extrahepatic microsomes is shown in Table 3.3. The time required to reach 670/0 of the maximal absorbance obtained was 2,8, 14 and 17 seconds in microsomes from untreated rat liver, kidney, lung, and intestine, respectively. These results suggested a possible inverse relationship between P4S0 concentration (0. 71S, 0.093, 0.084, 0.041 nmol/mg in liver, kidney, lung and intestine, respectively) and CO-binding spectral development time. The spectral development time was not altered by rat treatment with either BNF or NBI in the kidney or lung, except for a marked slowing in lung microsomes prepared from NBI-treated rats 72 hours after the last dose. The development time was slightly accelerated in intestinal microsomes prepared from rats treated with either BNF or NBI, with the exception of microsomes prepared 72 hours after the last NBI dose, in which the development time was the same as that in uninduced microsomes. In contrast to the alterations observed in extrahepatic tissues, the development of CO-binding spectra in hepatic microsomes was conSistently slower when microsomes were prepared Table 3.2. NBI inhibition of EROO in BNF- and NBI-induced rat microsomes 'from liver, kidney, lung and intestine. 46 BNF-induced rats or rats treated with NBI were sacrificed (for NBI-treated rats, at 36,48, and 72 hours after the final dose), and microsomes from liver, kidney cortex, lung and intestinal mucosa were prepared. Some microsomal samples were also used to obtain the data presented in Figure 3.1, but additional samples were prepared for these determinations. Microsomal P450 concentration (nmol/mg), the uninhibited EROO (nmollmg/min) activity and the NBI IC50 (JlM) were determined, as described in Methods. All data are represented as the mean of 3 or more samples ± the standard error of the mean, with the exception of BNF-induced and NBI (3x+72)-induced intestinal microsomes (n=2, both values given). An ANOVA analysis of variance was performed on the data for each parameter, and the Tukey multiple comparisons test was used to determine differences between the values for the different treatment groups. A significant difference from the values obtained with microsomes prepared from NBI-treated rats sacrificed at 72 hours after the last dose is indicated by the symbol @. Organ Treatment [P450] Uninhibited EROO IC50 BNF 1.123±0.129@ 1.461 ±O .164@ 2.1±O.7 LIVER NB13x+36 2.676±0.124 1.659±0.384 1.8±O.3 NB13x+48 2.600±0.037 1. 709±0.1 09 3.2±0.2 NB13x+72 3.001±0.183 2.770±0.315 0.7±O.1 BNF 0.220±0.026 0.275±0.061 2.1±O.4 KIDNEY NB13x+36 0.240±0.052 0.280±0.042 2.5±0.5 NBf 3x+48 0.222±0.023 0.292±0.030 1.8±O.6 NB13x+72 0.212±0.013 0.107±0.026 2.3±O.3 BNF 0.146±0.007 o .033±0. 005 2.2±O.2 LUNG NB13x+36 0.182±0.013 O. 029±0 .003 2.6±O.8 NB13x+48 0.187±0.033 O. 043±0. 006 2.3±O.8 NB13x+72 0.117±0.009 0.019±0.001 2.9±O.9 BNF 0.199,0.222 0.126,0.175 1.9,1.9 INTESTINE NB13x+36 0.318±0.012 0.133±0.027 0.9±O.1 NB13x+48 0.305±0.071 0.145±0.032 1.1±0.5 NB13x+72 0.056, 0.061 0.015, 0.009 0.3,1.3 Table 3.3. Development of CO-ferrous P450 spectra in the presence and absence of exogenous NBI in hepatic and extrahepatic microsomes from untreated, BNF- and NBI-treated rats. 47 Untreated (CONTROL) rats or rats treated with BNF or NBI were sacrificed (for NBI-treated rats, at 36,48 and 72 hours after the final dose), and microsomes ·from liver, kidney cortex, small intestinal mucosa and lung were prepared. Microsomes were suspended in buffer (2 and 1 .67 mg/ml for hepatic and extrahepatic microsomes, respectively) and bubbled with carbon monoxide (CO) for 3-5 minutes. The suspension was divided between two cuvettes, dithionite was added to the sample cuvette and the development of the CO-binding spectrum was recorded over time as the difference in absorbance between 450 and 490 nm. The development of the spectrum was also recorded in the presence of 100fJ.M NBI, which was added to the suspension prior to bubbling with CO. The sample number, n, for each entry was 3-4 for hepatic microsomes and 2 for all extrahepatic microsomal determinations, with two exceptions, for which n=1 (*). All entries are reported as the mean, with the standard error of the mean added for hepatic determinations. Time to Deyelop 67% of the Maximal P450 (sec) Treatment in vitro NBI LIVER CONTROL + BNF + NB13x+36 + NB13x+48 + NB13x+72 + aND= not determined 2±1 112±12 5±4 413±168 16±3 198±22 20±3 198±7 9±4 182±32 KIDNEY 8 19 6 122 7 26 8 116 8 93 LUNG 14 190* 15 NDa 15 ND 12 ND 33 ND INTESTINE 17 52 9 150* 10 195 12 304 20 163 48 from rats treated with NBf. The slowing was most marked in microsomes prepared 36 and 48 hours after the last NBI dose, but the development time in microsomes prepared at 72 hours after the last dose, although still slowed, was much closer to control values. Treatment with BNF did not appreciably alter the development of the CO-binding spectrum in hepatic microsomes. The development of the CO-binding spectrum was slowed when NBI (100JlM) was added in vitro to both hepatic and extrahepatic microsomes (Table 3.3). In untreated rat microsomes, the development was slowed by in vitro NBI to a much greater extent in hepatic (and possibly lung) microsomes than in those from kidney and intestine. The spectral development times in the presence of in vitro NBI were even slower in hepatic and intestinal microsomes from rats pretreated with either BNF or NBI. This was also true of kidney microsomes from BNF- or NBI-pretreated rats, with the exception of the seemingly spurious short development time in NBI-supplemented microsomes prepared from NBI-induced rats 36 hours after the last dose. The in vitro ligand binding of NBI to P450 heme was assessed in hepatic and extrahepatic microsomes through the examination of type II binding spectra (Table 3.4). In hepatic microsomes from untreated (control), BNF- or NBI-induced rats, NBI bound to P450 heme with Ksapp varying from 2-5JlM (the variation was correlated [R=O.83] with changes in P450 concentration in the assay, giving a mean KSapP/[P-450] ratio of 1.00±O.13). Induction of hepatic P450 with either BNF or NBI increased the extent of type II binding (maximal absorbance, A.U.) from that seen in control hepatic microsomes. However, calculation of extinction coefficients revealed differences in NBI type II binding between BNF- and NBI-induced hepatic microsomes. Pretreatment with BNF did not change the extinction coefficient for NBI type II binding from that of untreated hepatic microsomes. In contrast, NBI treatment elevated the extinction coefficient in hepatic microsomes prepared at 48 and 72 hours after the last dose. However, some of the induced P450 in hepatic microsomes prepared from NBI-treated rats at 36 hours after the last dose appeared to be Table 3.4. NBI P450 type II binding spectra in hepatic and extrahepatic microsomes from untreated, BNF- and N BI-treated rats. Untreated (CONTROL) rats or rats treated with BNF or NBI were sacrificed (at 36,48 or 72 hours after the last dose of NBI), and microsomes were prepared from liver, kidney cortex, lung and small intestinal mucosa (n = 3-6, except for two groups (*), where n = 2). Microsomal P450 concentration (nmol/mg) was determined. The type II binding spectrum of NBI in diluted microsomal samples (1.67 and 2 mg protein/ ml for extrahepatic and hepatic microsomes, respectively, exceptfor hepatic microsomes from NBI-induced rats at 36 hrs, which, due to low microsomal yield, were diluted to an average concentration of 0.90 mg protein/ ml) was measured as the difference in absorbance between 431 and 390 nm. Double reciprocal plots of the absorbance versus the NBI concentration were constructed and apparent dissociation constants (Ksapp, ~M) and maximal type II binding (A.U.) were determined from the x- and y-intercepts, respectively, from lines with correlation coefficients greater than 0.9. Two separate lines were formed from the extrahepatic data on either side of an 8~M dividing point, and high and low affinity dissociation constants and extent of binding for each affinity (expressed as a per cent of the total binding) were determined from these lines. An apparent extinction coefficient (E.C.app) was calculated from the overall maximal type" absorbance and the cytochrome P-450 concentration in the assay . .Qrg.ao Treatment If.45Ql assay (P4S0] hW ~ ~ ~ ~~l.tli.gl]l ~!lc.w.l. ~ __ ----=c-:...--::-7=-:::-;--.-l(:..;;;,nm~o:,:-:Vm..::.....git.L) (pM) (A.U.) (A.U.~-1cm·1) (%) (%) ~) ~ CONTROL 0.708 1.416 0.0778 0.0555 100 2 BNF 1.064 2.127 0.1190 0.0578 100 2 LIVER N BI 3x+36 2.676 2.406 0.1024 0.0423 100 2 NBI3x+48 2.188 4.376 0.2705 0.0651 100 5 NB13x+72 3.004 6.008 0.4047 0.0678 100 - 4 _~ ______ ~ _____ _ CONTROL 0.107 0.178 0.0154 0.0895 34 --'---66---1-.0- 22 BNF 0.243 0.405 0.0286 0.0702 53 47 0.7 14 KIDNEY NB13x+36 0.240 0.400 0.0235 0.0619 33 67 0.6 23 NBI 3x+48 0.243 0.405 0.0273 0.0676 32 68 0.5 26 NBI3x+72 0.212 0.353 0.0253 0.0732 26 74 CONTROL 0.084 0.140 0.0087 0.0608 --~---- ~- -2'6'- BNF 0.146 0.243 0.0129 0.0538 80 20 0.5 4 LUNG NBI 3x+36 0.182 0.303 0.0110 0.0365 74 26 0.2 4 NBI3x+48- 0.204 0.340 0.0136 0.0421 67 33 0.3 5 NB13x+72 0.117 0.195 0.0089 0.0480 71 CONTROL 0.042 0.070 0.0059 0.0885 5i"'---'-- BNF- 0.210 0.351 0.0266 0.0758 60 40 0.2 5 INTESTINE NB13x+36 0.318 0.530 0.0228 0.0426 65 35 0.4 8 NB13x+48 0.305 0.508 0.0230 0.0456 59 41 0.5 11 NB13x+72 0.060 0.100 0.0080 0,0821 64 36 0.4 9 50 unable to bind NBI in vitro, as evidenced by a decrease in extinction coefficient from that calculated for control microsomes. Characteristics of NBI type" binding in extrahepatic microsomes differed from those in hepatic microsomes in several respects. In addition, differences in ligand binding between extrahepatic tissues were revealed. The extinction coefficient for NBI ligand binding to control lung P450 was comparable to that determined for control hepatic microsomes, but the extinction coefficients for both control kidney and intestinal P450 were considerably higher. The induction of P450 by either BNF or NBI in the extrahepatic tissues was accompanied by an increase in the extent of in vitro NBI type II binding, as was seen with hepatic P450. However, the effects of induction on the extinction coefficient of type" binding differed between extrahepatic tissues and the liver. BNF induction did not alter and NBI induction increased the extinction coefficient in hepatic microsomes, but both inducers lowered the extinction coefficient in extrahepatic microsomes. NBI treatment decreased the extrahepatic extinction coefficient to a greater extent than did BNF treatment, particularly when microsomes were prepared at 36 and 48 hours after the last dose of NBI. The extent that the extinction coefficients were lowered by BNF and NBI in extrahepatic microsomes from a particular tissue did not correlate with either the magnitude of the extinction coefficient or the P450 concentration in control (uninduced) microsomes from the same tissue. The extent to which the extinction coefficient was decreased by NBI and BNF could be correlated with the amount of induced P450 (R=O.90) in the intestine, but not in lung or kidney; thus, in the intestine, NBI and BNF may induce common isozymes which are less able to bind NBI. However, in all extrahepatic microsomes prepared from NBI-induced rats at 72 hours, the extinction coefficient had increased toward control values; therefore, part of the decrease in extinction coefficients seen in NBI-induced extrahepatic microsomes at 36 and 48 hours may have been due to induced P450 being unable to bind NBI at these time points. 51 The most striking contrast between the characteristics of NBI type II binding to P450 in the extrahepatic tissues and those in the liver was seen in the biphasic nature of ligand binding in extrahepatic tissues. NBI bound to extrahepatic P450 heme with both a high and a low affinity, and a Ksapp was calculated for each affinity. The high affinity binding had a Ksapp of 0.1-1.0~M in control and induced extrahepatic microsomes, but the Ksapp for low affinity binding varied between tissues (approximately 20J..LM for kidney, 5J..LM for lung and 1 OJ..LM for intestine). Additionally, the division of NBI type II binding between high and low affinity varied between tissues, with lung and intestine having a greater percent of high affinity binding (approximately 70 and 60%, respectively), while in kidney microsomes, the greatest percent of NBI type II binding was of low affinity (approximately 700/0). The dissociation constants in extrahepatic microsomes did not appear to increase with increasing assay P450 concentration, as had been observed for the monophasic Ksapp in hepatic microsomes; however, variations in the low affinity Ksapp were proportional to the percent of low affinity binding in the extrahepatic tissues (R=0.97). Overall, the induction of extrahepatic P450 by NBI or BNF did not alter the proportions of high and low affinity P450 in microsomes; therefore, NBI and BNF each induced both high and low affinity extrahepatic P450. [e.g., The proportions of kidney P450 (0.243 nmollmg microsomal protein) in microsomes prepared from NBI-treated rats sacrificed at 48 hours after the last dose were 320/0 high affinity (0.078 nmol/mg) and 68% low affinity (0.165 nmol/mg). Similar calculations with control kidney microsomes determined that the proportions of high and low affinity P450 were 0.036 and 0.071 nmol/mg, respectively. The values of control microsomes were subtracted from those of induced microsomes to determine the amount of renal P450 of high (0.042 nmol/mg) and low (0.094 nmol/mg) affinity induced by NBI in microsomes prepared at 48 hours after the last dose.] An exception to this generality was seen with BNF induction in the kidney, where there was an increase of high 52 affinity binding to 50%, 'from the 30% found in control and NBI-induced kidney microsomes. However, calculations indicate that, in addition to the preferential induction of high affinity renal P450 (0.093 nmol/mg), some induction of low affinity P450 also occurred (0.043 nmol/mg). Further investigation of this shift in the proportions of high and low affinity binding to renal P450 in BNF- but not NBI-induced microsomes was undertaken through analysis of kidney microsomal proteins by SDS polyacrylamide gel electrophoresis (Figure 3.2). This analysis showed that protein bands with minimum MWs of 50,000 and 48,000 daltons were differentially intensified by BNF and NBI treatment, and that a band with a minimum MW of 52,000 daltons was intensified by NBI, but not BNF treatment. These protein bands are wit~lin the MW range of most known P450s; thus, these findings indicate that BNF and NBI may differentially induce renal P450 isozymes, an observation which may account for the differences in the amount of high and low affinity NBI type II binding in microsomes prepared from BNF- and NBI-treated rats. Changes in the intensity of other bands (44,000, 47,000 and 200,000 MW) outside the range of P450s were also noted, indicating that induction of renal proteins by BNF and NBI is not limited to P450s. Discussjon This investigation examined possible causal factors responsible for the interorgan inconsistencies observed in the induction of P450 by nitrogen heterocycles. A study of the changes in P450 concentration and two monooxygenase activities in microsomes from rats pretreated with a single nitrogen heterocycle, NBI, followed by varying IIwash-outll periods (Figure 3.1) addressed the possibility that interorgan variations in P450 (protein) turnover could be responsible for the observed differences in induction between tissues. The concentration of P450 in microsomes from NBI-induced rats at 48 hours was never significantly less than that in microsomes prepared at 36 hours. Thus, there is no short-lived inductive response to account for the interorgan Figure 3.2. SOS-polyacrylamide gel of kidney cortical microsomes from untreated rats or rats treated with BNF or NBI. 53 Kidney cortical microsomes from untreated (CON) or BNF-induced rats or rats treated with NBI followed by 36, 48, or 72 hrs after the last dose were electrophoresed (each lane contained 300J.1g of microsomal protein). Protein standards (Std) with their MW were myosin (200,000), B-galactosidase (116,300), phosphorylase B (92,500), bovine serum albumin (66,000) and ovalbumin (45,000). 'I. f , t J '1\ ~~ I I ~ ~ 0 CO 0 ~ C\I ~ I , I I ~ ~ C\I CO 0) CO t t I ~ LO ~ u.. Z a:l -a:l Z ~- a:l Z CO ct) -al Z z o () 55 differences in induction, and the generalized use of a 48 hour "wash out" period is supported. Interestingly, the decay of induced enzyme differed between the tissues, with hepatic and kidney cortical P450 concentrations remaining elevated for 72 hours after the last dose, while concentrations in the lung and intestine had declined by this time. The variation in induced enzyme decay appeared to be unrelated to the degree of induction. The possibility that selective inhibition of P450 isozymes by residual NBI in different organs could be responsible for the apparent differences in isozyme induction between tissues was investigated. The organ-selective inhibition could arise from differences in imidazole distribution, from quantitative or qualitative differences in the affinities of the P450 isozymes for the imidazole ligand, and/or from inhibition by locally-produced metabolites. Previous studies by Ritter and Franklin (1987a) and other investigators (Rodrigues, Waddell and lonnides, 1988; Kojo, Honkakoski, Jarvinen, Pelkonen and Lang, 1989) found evidence of residual imidazole inhibition of P450-dependent monooxygenase activities in hepatic microsomes prepared from animals pretreated with N-substituted imidazoles. Selective inhibition of specific P450 isozymes by nitrogen heterocycles has been demonstrated for benzimidazoles (Murray, 1987b), imidazoles (Murray and Wilkinson, 1984; Sheets, Mason, Wise and Estabrook, 1986; Rodrigues et al., 1988; Kojo et al., 1989), pyridines (Murray and Wilkinson, 1984) and phenothiazines (Murray, 1989). Indeed, isozyme selective inhibitors have been actively sought in the development of antifungal agents (specific inhibition of 'fungal lanosterol 14-demethylation) and agents for the control of estrogen-dependent mammary tumors (selective inhibition of P450-dependent aromatase) (Ortiz de Montellano and Reich, 1986; Poulos, 1988). Selective isozyme inhibition is not limited to liver P450, as a study by Kan, Hirst and Feldman (1985) found that ketoconazole and related imidazole antifungals inhibited C-17 ,20 lyase (microsomal) and cholesterol-side-chain-cleavage (mitochondrial) activities in rat testis, although another testicular P450-dependent microsomal activity, progesterone 56 17-hydroxylase, was unaffected. Since the heme iron ligation with imidazole nitrogen is the same for all P450 isozymes, the selectivity of inhibition is dependent upon interactions of the inhibitor with the protein around the heme; thus, the active site topology of individual isozymes must differ. In particular, studies by Murray and Wilkinson (1984), and Murray (1987b, 1989) with phenobarbital- and SNF-induced microsomes have suggested that the PAH-inducible isozymes possess a larger and more accessible active site relative to the phenobarbital-inducible isozymes. Their studies indicate that steric determinants are most important in the inhibition of EROD activity, as opposed to the inhibition of phenobarbital-inducible pentoxyresorufin depentylase activity, in which lipophilic determinants are more important. In microsomes from both liver and lung of NSI-treated rats, EROD activity was lower relative to P450 concentration at earlier time points after the last dose, and in the liver, but not in the lung, relative p-nitroanisole demethylase activity paralleled that of EROD (Figure 3.1). These differences in specific activities relative to P450 concentration in lung and liver disappeared with longer "wash-out" periods, leading to speculation that they could be due to the presence of residual inhibitory NSI in these organs, which did not occur in either the kidney or intestine. This suggests that interorgan differences in clearance of residual inhibitor may be responsible for the observed differences in the induction of monooxygenase activities between tissues at earlier time points after the last NSI dose. These studies might also suggest that residual NSI may preferentially inhibit PAH-inducible P450 isozymes, since only EROD (selectively catalyzed by PAH-inducible isozymes), and not p-nitroanisole de methylase was decreased in NSI-induced lung microsomes at earlier time points. Selective inhibition of PAH-inducible isozymes by NSI in the liver might also account for the decreases in both EROD and p-nitroanisole demethylase activities in this tissue, since the latter activity is catalyzed by multiple isozymes, including those induced by PAHs. An alternative explanation for the selective decreases in monooxygenase 57 activity relative to P450 concentration in lung and liver might employ the observed alterations in sedimentation of microsomal membranes by NBI treatment. NBI-induced alterations in membrane sedimentation, as discussed in Chapter 2, could result in the selective loss of microsomal membrane containing PAH-inducible isozymes. The decreased yield of hepatic microsomes prepared from NBI-induced rats at 36 and 48 hours after the last dose supports this possibility (Table 3.1). However, alterations in sedimentation cannot explain the relative differences in EROD activity versus P450 concentration in the lung, since lung microsomal yield was not decreased. Additional evidence that residual NBI may be inhibiting monooxygenases in the liver can be found in the results of in vitro NBI inhibition of EROD in NBI-induced hepatic microsomes. The increase in EROD activity in hepatic microsomes prepared from NBI-induced rats at 72 hours after the last dose, over that in NBI-induced microsomes prepared at 36 and 48 hours, was accompanied by a decrease in NBI IC50 for this activity. Residual NBI may be preferentially inhibiting PAH .. inducible cytochrome P-450 isozymes in hepatic microsomes prepared at 36 and 48 hours after the last NBI dose, as suggested above, and this could account for the decreased EROD activity relative to P450 concentration at these time points. PAH-inducible isozymes would be released from inhibition if NBI were cleared from the liver by the time microsomes were prepared 72 hours after the last dose, and, as a result, an increase in EROD activity relative to P450 concentration would be observed in these microsomes. These uninhibited isozymes with the highest affinity for NBI would be the first to be inhibited in vitro by low concentrations of NBI added to microsomes prepared at 72 hours after the last dose; therefore, 50% inhibition would come at a lower concentration (Le., a decrease in IC50 would be observed). The mechanism of imidazole inhibition of P450 is, at least in part, through ligand binding to the heme. CO also binds as a heme ligand to produce the characteristic P450 spectrum with an absorption maximum at 450 nm; therefore, any residual NBI from in vivo treatment retained as a heme ligand 58 would be expected to slow or prevent the development of the CO-binding spectrum. This expectation is supported by previous investigations with hepatic microsomes from rats pretreated with high doses of the antimycotic agents, ketoconazole (Meredith, Maldonado and Speeg, 1985), and clotrimazole (Ritter and Franklin, 1987b). Results from the latter investigation found that pretreatment only slowed the rate of formation of the CO-binding spectra in hepatic microsomes, while the former investigation found that ketoconazole pretreatment decreased the amount of microsomal P450 detectable by its CO-binding spectra. In vitro studies with clotrimazole and ketoconazole supported the observations of altered CO-binding following in vivo treatment. A high concentration of ketoconazole (50JlM) added in vitro reduced the absorbance difference between 450 and 490nm of the CO-binding spectrum by 37% (Meredith et al., 1985). In contrast, only the rate of formation of the CO-binding spectrum in hepatic microsomes was slowed by in vitro clotrimazole, and, if adequate time was allowed (> 5 min), all the P450 was detectable as a CO-adduct, regardless of the clotrimazole concentration (Ritter and Franklin, 1987b). Rat pretreatment with NBI, but not with BNF, which is not a heme ligand, markedly slowed the formation of CO-binding spectra in hepatic microsomes (Table 3.3). This difference might be attributed to NBI induction of hepatiC isozymes with less affi nity for CO or with greater resistance to dithionite reduction than isozymes induced by BNF. However, these possibilities are unlikely, since the CO-binding spectrum development time in NBI-induced microsomes prepared at 72 hours had returned toward the development time in untreated (control) microsomes (Table 3.3), although P450 concentration had not decreased from that present at 36 and 48 hours after the last dose (Figure 3.1). No evidence of slowed formation of CO-binding spectra in NBI-induced extrahepatic microsomes prepared at 36 and 48 hours after the last dose was found. In vitro studies with 100JlM NBI demonstrated that NBI can slow and possibly prevent (with NBI added, the n1aximal absorbance rarely reached that 59 observed in uninhibited microsomes within 10 minutes) the development of CO-binding in microsomes from all tissues. Therefore, the selective slowing of CO-binding spectrum development in hepatic microsomes from NBI-pretreated rats suggests that residual NBI was only present in the liver. Alternatively, NBI pretreatment could be producing selective changes in hepatic, and not extrahepatic, microsomal membranes, which, in addition to selectively altering microsomal yield (Table 3.1), could also be responsible for both the selective slowing of CO-binding spectrum development and the lower monooxygenase activities relative to P450 concentration seen in the liver (Figure 3.1 ). The differences noted between tissues in the time required for the development of the CO-binding spectra of P450 and the different degrees to which 100ilM NBI added to control microsomes slowed CO-binding suggest that the affinity of P450 heme for ligands may vary between tissues. This possibility is strongly supported by evidence from binding studies with NBI (Table 3.4), which revealed the biphasic nature of NBI type II binding in the extrahepatic microsomes. The percentage of high and low affinity type II binding of NBI varied between extrahepatic microsomes, with lung and intestinal microsomes each having more high affinity binding than kidney microsomes (Le., lung> intestine> kidney). The percentage of high affinity binding in extrahepatic microsomes corresponded with the extent that the development of CO-binding spectra was slowed by 100ilM NBI (13.6-fold in lung, 3.1-fold in intestine, and 2.4-fold in kidney, Table 3.3), suggesting a possible relationship. A biphasic nature has also been reported for the type II binding of n-alkylamines (Jefcoate, Gaylor and Calabrese, 1969), ketoconazole (Meredith, et al., 1985) and pyridylalkanamides (Repond, Bulgheroni, Meyer, Mayer and Testa, 1986) to hepatic P450, but not for phenylpyridines, phenylimidazoles and pyridylimidazoles (Murray and Wilkinson, 1984), and for NBI in the present investigation. Thus, the structural characteristics of the nitrogen heterocycles determine whether type II binding in the liver is biphasic. 60 A plausible explanation for biphasic binding involves different spin states of the ferric iron (Repond et aI., 1986). It was postulated that, at low ligand concentrations, binding is to the sixth axial position of the portion of the P450 population that exists in the low-spin state due to the presence of endogenous ligand. Displacement of endogenous ligand by exogenous ligand results in a type II binding spectrum known as lib. Binding at higher ligand concentrations is to the high-spin (pentacoordinated) iron atom, producing a shift to the low-spin state (hexacoordinated) that results in a type Ita spectrum. The recorded binding spectrum is observed as the sum of the lib and lIa spectra. The biphasic nature of NBI type II binding in microsomes from extrahepatic tissues could be due to the presence of P450 isozymes which preferentially exist in different spin states. Electron paramagnetic resonance spectroscopy has shown that rabbit P450 isozymes LM2 and LM4 (induced by phenobarbital and MC, respectively) are predominantly low and high spin in the ferric state (Haugen and Coon, 1976). White and Coon (1982) found that several imidazoles, including NBI, have a higher affinity for low spin rabbit P450 LM2 (apparent dissociation constant for NBI binding=1 flM) than for high spin LM4 (apparent dissociation constant for NBI binding=50flM). However, individual isozymes are not completely low or high spin. The relatively low free energy difference between the two spin states allows for the existence of mixed spin states (Sligar, 1976; Ristau, Rein, Janig and Ruckpaul, 1978). Thus, the biphasic nature of NBI type II binding might also result from the existence of both high and low spin states in an individual isozyme. This possibility is supported by Hajek, Cook and Novak (1982), who demonstrated biphasic type II binding of imidazole to purified rabbit liver P450 isozymes, with dissociation constants of 6flM and 61 flM for LM2 and 6flM and 160flM for LM4. The hypothesis that differences exist in the proportions of extrahepatic P450 ferric heme existing in high and low spin states (whether due to the presence of different proportions of P450 isozymes existing preferentially in different spin states, or to the presence of individual isozymes existing in both 61 high and low spin states, or a combination thereof) can account for the observed biphasicity in NBI type II binding affinity. Why this does not occur in the liver is unknown. If the relationship between spin state and dissociation constant for NBI by White and Coon is applied universally, then the greater percent of low affinity kidney P450 represents a predominance of ferric heme in the high spin state, while the greater percent of high affinity P450 in lung and intestine represents a predominance of ferric heme in the low spin state. Overall, the induction of extrahepatic P450 by NBI or BNF did not alter the proportions of high and low affinity P450 in extrahepatic microsomes. Thus, if these proportions represent the amount of extrahepatic ferric heme existing in high and low spin states, NBI and BNF each induced extrahepatic P450 existing in both high and low spin states. An exception to this generality was seen with BNF-induction in the kidney, where there was an increase of high affinity binding to 50%, from the 30% found in control and NBI-induced kidney microsomes. However, calculations indicate that, in addition to the preferential induction of high affinity (low spin) renal P450 (0.093 nmollmg), some induction of low affinity (high spin) renal P450 (0.043 nmoles/mg) also occurred. Analysis of kidney microsomal proteins on an SDS polyacrylamide gel (Figure 3.2) showed that protein bands with minimum MWs of 50,000 and 48,000 daltons were differentially intensified by BNF and NBI treatment, and that a band with a minimum MW of 52,000 daltons was intensified by NBI, but not BNF, treatment. These findings suggest that BNF and NBI differentially induced renal P450 isozymes, and, if these isozymes preferentially exist in different spin states, this differential isozyme induction may have contributed to the observed differences in the percent of high and low affinity binding by NBI to renal P450 in kidney microsomes prepared from BNF- and NBI-induced rats. The extinction coefficients for NBI type II binding to low and high spin P450 may be similar, if high and low affinity NBI type II binding represents the existence of ferric heme in low and high spin states, respectively. Renal and intestinal control microsomes had similar extinction coefficients, although they had differing percentages of high and low affinity binding. Therefore, differences in spin state would not account for the differences in extinction coefficient observed between control microsomes from the various organs. 62 Treatment with SNF and NSt produced differing effects on NSI type II binding extinction coefficients in the liver. NSI treatment increased the coefficient in microsomes prepared at 48 and 72 hours after the last dose, although the extinction coefficient in liver was unchanged by SNF treatment. These differential effects on the hepatic extinction coefficient may result from the different pattern of P450 isozymes induced in the liver by SNF and NSI. While both agents induce in a PAH-type manner in the liver, NSI also induces hepatic isozymes that are inducible by phenobarbital, and the phenobarbital-inducible isozymes may have a different molar absorptivity than either constitutive or PAH-i nducible hepatic isozymes. Soth NSI and SNF induction decreased the NSI type II binding extinction coefficients in extrahepatic tissues from those determined in control microsomes, in contrast to the differential effects of the two inducers on the hepatic extinction coefficient. The extent to which the extinction coefficient in microsomes from intestine, but not from lung or kidney, was decreased following NSI or SNF treatment was correlated with the amount of induced P450 (R=O.90). However, the extent to which the extinction coefficient was decreased was much greater in NSI-induced microsomes prepared 36 and 48 hours after the last dose. Additionally, the extinction coefficient was decreased in NSI-induced hepatic microsomes prepared at 36 hours after the last dose, although the extinction coefficient was increased in microsomes prepared at 48 and 72 hours. These alterations in extinction coefficients in the various tissues following NSI treatment may indicate the presence of residual NSI inhibitor, which prevents some of the induced enzyme from binding NSI added in vitro. In conclusion, the results of this investigation of the possible causal factors of the interorgan variations in P450 induction found that the lability of induced cytochrome was not responsible for the differences in induction. The possibility 63 that residual inhibitory imidazole is responsible for some of the apparent differences in induction was supported by several lines of evidence in the liver (changes in NSf type II binding extinction coefficient; slowed CO-binding spectrum development; changes in EROD and p-nitroanisole demethylase activities relative to P450 concentration following NSf treatment; and changes in EROD inhibition in vitro). However, observed decreases in the extinction coefficient of NSI type II binding in extrahepatic microsomes prepared from NSI-induced rats at