Mapping of phenytoin-inducible cytochrome P450 immunoreactivity in the mouse central nervous system.

The distribution of phenytoin-inducible cytochrome P450 in non-treated mouse brain and spinal cord was analysed immunohistochemically using polyclonal antibodies against phenytoin-induced mouse cerebral microsomal P450. This P450 protein was proved in Ouchterlony [Volk B. et al. (1988) Neurosci. Lett. 84, 219-224], Western blot, and immunohistochemical analyses to be reactive to the specific antibodies and an IgG fraction raised against phenobarbital-induced rat liver microsomal P450IIB1. The phenytoin-induced P450 is designated P450IIB1* because immunologically it is comparable with P450IIB1; however, it has not yet been analysed for other characteristics of this enzyme. Immunocytochemistry was performed on acetone-fixed serial cryosections of the whole brain using the avidin-biotin-peroxidase detection system. Negative controls included incubations with preimmune serum of the immunized animal instead of the primary antibody and preabsorption of the antibody with the corresponding immunogen. The pattern of immunoreactive sites indicates that P450IIB1* is not distributed evenly throughout the CNS. It was found to be restricted to only some cellular populations. The most striking aspect of immunostaining was a predominant reactivity in the evolutionary old brain parts. Neuropil and neuronal staining was found in the spinal cord (motor neurons of the ventral horn), medulla oblongata (hypoglossal nuclei, magnocellular part of the lateral reticular nuclei), pons (trigeminal, facial, cochlear and pontine nuclei), cerebellum (granule cells), midbrain (dorsal raphe nucleus) and limbic lobe (hippocampal pyramidal cells). Neuropil reactivity alone appeared in cerebellar nuclei, midbrain, thalamus, basal ganglia, neopallium and olfactory brain. Generally, pia mater/arachnoid, ependyma, choroid plexus, vascular system and some astrocytic populations were found to be strongly P450IIB1* immunoreactive. In comparison with astroglia, which is characterized by glial fibrillary acidic protein-positiveness, the astrocytes, which are also P450IIB1* reactive, occurred only in subpial and subependymal layers, and in large fiber tracts of the spinal cord and brainstem, where they were attached to the vascular system. Otherwise, the glial fibrillary acidic protein-positive astrocytes were not P450IIB1* immunoreactive in the cerebellar molecular layer (fibers of Bergmann glia), in remaining neuropils and in white matter areas.

Comparison of enzyme phenotypes in human bladder tumours and experimentally induced hyperplastic and neoplastic lesions of the rat urinary bladder. A combined histochemical and immunohistochemical approach.

The expression of a number of enzymes involved in drug metabolism, membrane function etc. was compared in hyperplastic and neoplastic lesions of the rat bladder and in human bladder tumours. Transitional cell carcinomas (TCC) in both rat and Man were characterized by decreased alkaline phosphatase (ALP) and increased gamma-glutamyl transpeptidase (GGT), beta-glucuronidase (beta-G1), succinate dehydrogenase (SD) and glucose-6-phosphate dehydrogenase (G6PD) activities. In addition, binding for antibodies specific for different cytochrome P-450 species (UT50, PB3a, MC1, MC2) and microsomal epoxide hydrolase (mEHb) was elevated in both murine and human tumours. Comparison of the enzyme phenotype in hyperplastic lesions induced by freeze ulceration or uracil administration with that in preneoplastic papillary or nodular hyperplasia (PNH) and TCC suggested, however, that most of the alteration in enzyme content or activity was non-specific and related to requirements for epithelial cell proliferation. On the other hand, the decreased ALP, and increased GGT and beta-G1 activity appeared more directly related to neoplastic transformation. The results suggested that qualitative differences exist between reactive hyperplasia and preneoplastic or neoplastic lesions in the urinary bladder. The finding of increased cytochrome P-450, in clear contrast to the reduction characteristic of preneoplastic hepatic lesions, may be important with regard to the observed difference in neoplastic transformation between the bladder and liver in response to drug metabolising enzyme inducers.

Effect of nutritional imbalances on cytochrome P-450 isozymes in rat liver.

Male Sprague-Dawley rats were fed for six weeks either a control diet containing 22% casein (C) and 5% fat (F) or a low-protein diet (6% C, 5% F) or high-lipid diet (30% C, 30% F). A group of rats received a control diet containing 50 ppm of Phenoclor DP6. Three major forms of cytochrome P-450, UT 50, BP 3a and MC 2 were purified from livers of DP6-fed rats and only two forms, UT 50 and PB 3a, were purified from control and dietary groups. The amino acid composition and the catalytic activities towards all substrates tested were only significantly modified in the purified UT 50 P-450 isozyme from rats fed the low-protein diet. The N-terminal sequence analysis shows that cytochrome P-450 UT 50 (from control group) and UT 501 (from low-protein group) are two distinct proteins.

Effect of modifying agents on the phenotypic expression of cytochrome P-450, glutathione S-transferase molecular forms, microsomal epoxide hydrolase, glucose-6-phosphate dehydrogenase and gamma-glutamyltranspeptidase in rat liver preneoplastic lesions.

The expression of A and P forms of glutathione S-transferase (GST-A and P), two cytochrome P-450 isoenzymes (P-450 PB3a and P-450 MC2), microsomal epoxide hydrolase (mEHb), glucose-6-phosphate dehydrogenase (G6PD) and gamma-glutamyltranspeptidase (gamma-GT) was compared in preneoplastic liver lesions and background parenchyma of F344 rats post-treated with butylated hydroxyanisole (BHA), ethoxyquin (EQ) or acetaminophen (AAP). These latter three compounds have been shown to inhibit hepatocarcinogenesis after initial treatment with N-ethyl-N-hydroxyethylnitrosamine (EHEN) and a significant decrease in the number of enzyme-altered foci and nodules positive for GST-P, GST-A, G6PD and gamma-GT and negative for P-450 PB3a, P-450 MC2 was associated with their administration. Whereas in the foci case the decrease was most prominent for non-discrete (heterogeneous) type lesions, the results of quantitation of nodules revealed a most significant alteration in the discrete homogeneously staining population. This indicates that BHA, EQ and AAP have the potential to inhibit the growth of the phenotypically stable lesions thought most likely to be the immediate precursors of hepatocellular carcinomas. The two anti-oxidants were associated with periportal increase of all enzymes investigated, whereas AAP induced GST species and mEHb in the perivenular zone. Irrespective of slightly elevated enzyme levels in surrounding parenchyma, mEHb antibody binding levels within lesions showed a reciprocal shift from positive to negative in rats treated with BHA, EQ and AAP.

First evidence of cytochrome P-450 induction in the mouse brain by phenytoin.

Long-term administration of phenytoin (60 mg/kg) to male C57Bl/6J mice causes an average drug serum level of 16-19 micrograms/ml. Cytological changes consisting of focal axonal swellings in the deep cerebellar nuclei are characterized by an increasing accumulation of vesiculotubular membranes in the dense, but ribosome-deficient axoplasm. Microsomal fractions of brain and liver were prepared at intervals until the 80th day of treatment to determine their cytochrome P-450 activities and isoenzyme characteristics. The cerebellar tissue was shown to be distinguished by a 15-fold enhancement of cytochrome P-450 PB3a activity, that means twice the extent of that found in liver and cerebrum.

Effect of antibodies against cytochrome P-450 on demethylation and denitrosation of N-nitrosodimethylamine and N-nitrosomethylaniline.

Rat liver microsomes which were induced either with ethanol or PB were incubated with NDMA or NMA. Formaldehyde generation and nitrite formation were measured as metabolic parameters for oxidative bioactivation and denitrosation, respectively. The influence of antiserum PB3a1 and PB22 containing antibodies against the corresponding cytochrome P-450 species on both metabolic functions was investigated. The results showed that the influence on formaldehyde production and denitrosation varied independently in that both parameters were either not affected, or influenced in an opposite way, or inhibited to a different degree. Especially remarkable was the 80% inhibition of formaldehyde generation accompanied by no change in the nitrite formation by antiserum PB3a in ethanol-induced microsomes when NDMA was used as substrate. The 80% inhibition of formaldehyde production by antiserum PB3a was in contrast to no inhibition by antiserum PB2 under the same conditions. Denitrosation of NMA by PB-induced microsomes was inhibited by antiserum PB3a, but activated by antiserum PB2. It is concluded that oxidative demethylation and reductive denitrosation can be mediated by different cytochrome P-450 forms.

Effect of unbalanced diets on incorporation of delta-aminolevulinic acid into cytochrome P-450.

The in vivo syntheses of two liver microsomal cytochromes P-450 PB3a, P-450 UT50 [(1987) Eur. J. Biochem., submitted] (Mr 50,000, 52,000) have been estimated by measuring the specific activity 2 h after i.p. administration of delta-[3H]aminolevulinic acid to male Sprague Dawley rats. The animals were fed either a standard rat chow (5% lard, 22% casein) or unbalanced diets (high lipid, 30% lard or low protein, 6% casein) with or without 50 ppm Phenoclor DP6. The high-lipid diet supported a more rapid body weight gain but had little impact on cytochrome P-450 content, expressed either per whole liver or per mg microsomal protein, and on the incorporation of the precursor into cytochrome P-450. The latter was determined by measuring the radioactivity incorporated into the cytochrome P-450 fraction, partially purified by affinity chromatography, as well as into two cytochrome P-450 isozymes (Mr 50,000 or 52,000) purified by DEAE-52 cellulose ion-exchange chromatography. The low-protein diet, on the other hand, severely depressed body weight gain and cytochrome P-450 content as well as incorporation of radioactivity, the lower-Mr cytochrome (Mr 50,000) being particularly affected. However, when a potent inducer, Phenoclor DP6, was added to the low-protein diet, cytochrome synthesis was restored indicating that the effect was reversible.

Altered drug metabolizing potential of acinar cell lesions induced in rat pancreas by hydroxyaminoquinoline 1-oxide.

Foci of atypical acinar cells observed in male rats 1 year after a single injection of hydroxyaminoquinoline 1-oxide (HAQO) were assessed immunohistochemically for altered expression of a number of enzyme forms considered to play important roles in drug metabolism. The pancreatic lesions, classified as of either basophilic or eosinophilic type on histological appearance, demonstrated distinctive patterns of altered enzyme phenotype. On the one hand, the basophilic foci composed of enlarged cells/nuclei with very prominent nucleoli were characterized by increase in GST-P, G6PD and P450 PB1 and MC2 forms. The eosinophilic type, in contrast, comprised smaller cells demonstrating elevated P450 MC1 and PB1 but not MC2, normal G6PD and strong GST-P binding limited only to a proportion of the nuclei. Both shared decreased GGT and almost total lack of GST-B positive connective tissue and ductular elements. Apparent islet cell lesions and normal islet tissue were characterized by a distinct enzyme phenotype strongly positive for all P450 species investigated. The results indicate that HAQO-induced putative preneoplastic pancreatic lesions, like equivalent carcinogen associated with focal populations in liver, kidney and ductular pancreas, demonstrate a non-random altered expression of specific drug metabolizing enzyme species.

The effect of dietary imbalances on the activation of benzo[a]pyrene by the metabolizing enzymes from rat liver.

Male Sprague-Dawley rats (70-80 g) were fed ad libitum a standard control diet (22% casein, 5% lard), or a high lipid diet (30% lard) or a low protein diet (6% casein) or a standard diet containing 50 ppm phenoclor DP6. After 6 weeks on these diets, the cytochrome P-450 microsomal content, the benzo[a]pyrene monooxygenase (BaP-MO) and the epoxide hydrolase (EH) were assayed. The formation of mutagenic B(a)P metabolites which covalently bind with DNA was compared. The activity of BaP-MO and of EH were increased by the high lipid diet (+27% and 106% respectively) and by the phenoclor DP6 treatment (+63% and 400% respectively), compared to the standard diet. In animals fed a low protein diet the BaP-MO was decreased (-34%) and the EH activity was strongly increased (+262%) compared to those fed a standard diet. All experimental diets increased both the activation of BaP to metabolites able to bind DNA and the mutagenicity of BaP versus TA98 Salmonella typhimurium strain. It was concluded that dietary imbalances can be considered as a factor in chemical carcinogenesis.

Immunohistochemically demonstrated altered expression of cytochrome P-450 molecular forms and epoxide hydrolase in N-ethyl-N-hydroxyethylnitrosamine-induced rat kidney and liver lesions.

A comparison of N-ethyl-N-hydroxyethylnistrosamine (EHEN)-induced preneoplastic and neoplastic lesions in the rat liver and kidney was made with respect to the expression of different drug metabolizing enzymes. Four cytochrome P-450 species (cyt. P-450 UT50, PB3a, MC1 and MC2) and microsomal epoxide hydrolase (mEHb) were investigated along with two glutathione S-transferase species (GST-P and A forms) earlier shown to be elevated in putative preneoplastic lesions in the liver and kidney, respectively. In contrast to the liver lesions, which showed clear decrease in all forms of cyt. P-450s and increase of mEHb, elevated levels of cyt. P-450 PB3a and, to a lesser extent, the other P-450 forms and early elevation to late decrease in mEHb characterized the renal tubular lesions. Thus opposite shift in enzyme phenotype was observed in carcinogen-induced focal lesions of the two organs. Variation in binding levels in the different nephron segments and zones of the liver acinus indicated physiological specialization with regard to the enzymes investigated and suggested that the altered phenotype of preneoplastic populations might be of adaptive significance.

Isoenzyme-specific phosphorylation of cytochromes P-450 and other drug metabolizing enzymes.

A series of fourteen cytochrome P-450 isoenzymes was treated with three different protein kinases and found to divide into isoenzymes phosphorylated by both the cyclic AMP-dependent kinase and the calcium-phospholipid-dependent kinase (P-450 PB 3a and PB 2e), by none of these kinases (P-450 PB 1b, MC 1b, UT 1, and thromboxane synthase), and by either the cyclic AMP-dependent kinase (P-450 LM 2, PB 2d, and PB 3b) or the calcium-phospholipid-dependent kinase (P-450 PB 1a, PB 2a, MC 1a, LM 3c, and LM 4). Other components of the monooxygenase system, cytochrome P-450 reductase, cytochrome b5, cytochrome b5 reductase as well as microsomal epoxide hydrolase, were poor substrates for the kinases employed. On the other hand, glutathione transferases 1-2 and 4-4, but not 3-3, were relatively good substrates for the calcium-phospholipid-dependent kinase.

Identification of human cytochromes P-450 analogous to forms induced by phenobarbital and 3-methylcholanthrene in the rat.

Antibodies to four rat liver forms of cytochrome P-450, two phenobarbital-inducible (PB1 and PB2) and two 3-methylcholanthrene-inducible (MC1 and MC2) proteins, have been used to make a structural and functional comparison of rat and human cytochromes P-450. Proteins from both species were identified on Western blots by their reaction with these antibodies. In the human liver preparations, structurally related proteins to PB1 and to PB2 were identified in all the samples tested with apparent Mr values of 51 800 and 54 800 for PB1 and 53 600 and 57 200 for PB2. Considerable variation in the content of the lower-Mr proteins was measured between samples and, as with the rat enzymes, samples which reacted well with anti-PB1 also reacted with anti-PB2, indicating that these proteins are regulated at least to some degree, co-ordinately. The apparent Mr values of the major human proteins identified with anti-MC1 and anti-MC2 were 54 400 and 57 000 respectively. Only six (of 31) human samples contained significant amounts of these proteins. The same six samples which reacted with anti-MC1 also reacted with anti-MC2, again indicating co-ordinate regulation of these two proteins. Antibody inhibition of microsomal 7-ethoxycoumarin and 7-ethoxyresorufin metabolism demonstrated a degree of conservation of substrate specificity related to specific P-450 isoenzymes between the species. However, the contributions of the different P-450 isoenzymes to the human microsomal activity were not always related to the rat enzyme with the highest activity towards these substrates.