Novel involvement of a mitochondrial steroid hydroxylase (P450c11) in xenobiotic metabolism.

Adrenocortical mitochondrial cytochrome P450 isozymes of the Cyp11 family normally synthesize steroids with a very strict substrate specificity. However, for the first time, P450c11 was additionally shown to metabolize and bioactivate the adrenotoxic environmental pollutant 3-methylsulfonyl-2,2-bis(4-chlorophenyl)-1,1-dichloroethene (MeSO2-DDE). This conclusion is based on a striking correlation between inductions of MeSO2-DDE and deoxycorticosterone metabolism by forskolin in the adrenocortical cell lines Y1 and Kin-8, inhibition of P450c11-dependent activities in Y1 cells by MeSO2-DDE, and metabolism of MeSO2-DDE by non-steroidogenic COS cells after transfection with a cDNA encoding P450c11. The interaction between xenobiotics and glucocorticoid synthesis should focus more attention to xenobiotic-induced hormonal disturbances.

MeSO 2 -DDE (3-methylsulfonyl-2,2-bis(4-chlorophenyl)-1,1dichloroethene) 1 is a metabolite of the still extensively used insecticide DDT and is therefore a global environmental pollutant. Because of its lipophilicity and persistence, MeSO 2 -DDE is bioaccumulated in the food chain, even leading to pollution of human breast milk (1,2). Animal experiments have shown MeSO 2 -DDE to be adrenotoxic to several species after bioactivation in adrenal cortex mitochondria (3)(4)(5)(6)(7). Furthermore, human adrenal mitochondria have also been demonstrated to bioactivate MeSO 2 -DDE in vitro (8). It has been suggested, but never demonstrated, that the bioactivation is mediated by mitochondrial cytochrome P450 (7). The adrenal gland contains several P450 isozymes, but in contrast to hepatic P450s, adrenal P450s are generally not active in xenobi-otic metabolism. The genes encoding the mitochondrial P450 11␤-hydroxylase (P450c11), cholesterol side-chain cleavage (P450scc), and aldosterone synthase (P450aldo) belong to the Cyp11 subfamily of the P450 superfamily of genes (9). These enzymes show a strict substrate specificity and are therefore not believed to metabolize xenobiotics. P450scc is the first and rate-limiting enzyme in steroid hormone synthesis, localized in all steroidogenic tissues, whereas P450c11 and P450aldo are specifically localized in the adrenal cortex. The latter two participate as the last enzymes in the synthesis of glucocorticoids and mineralocorticoids, respectively. Given the specific mitochondrial damage in the adrenal cortex after MeSO 2 -DDE exposure, we decided to investigate the possible involvement of steroidogenic mitochondrial P450s in the metabolism of MeSO 2 -DDE. For this purpose we studied the bioactivation of MeSO 2 -DDE, i.e. the metabolic formation of reactive MeSO 2 -DDE intermediates capable of binding irreversibly to macromolecules or cellular antioxidants, in the mouse adrenocortical tumor cell line Y1. We also studied the bioactivation of MeSO 2 -DDE in a mutant of the Y1 cell line, Kin-8, that harbors a defect in the cAMP-dependent protein kinase (10), which normally participates in the hormonal regulation of the steroid-metabolizing enzymes. In Kin-8 cells, as opposed to Y1 cells, ACTH, forskolin, or cAMP are inefficient inducers of the P450 isozymes expressed in these cells, i.e. mitochondrial P450scc, P450c11, and P450aldo. To unambiguously demonstrate the relative importance of the adrenal mitochondrial P450s in the bioactivation of MeSO 2 -DDE, experiments were finally conducted in monkey kidney COS cells. These cells do not normally metabolize steroids or MeSO 2 -DDE, so by investigating the MeSO 2 -DDE-metabolizing capacity in COS cells transfected with expression plasmids encoding the three mitochondrial P450s we tried to identify the enzyme responsible for bioactivation, and presumably toxicity, of MeSO 2 -DDE. Enzyme assays were performed after growing cells in the presence of steroid substrate (deoxycorticosterone, corticosterone, or 22-hydroxycholesterol). A 500-l aliquot of the medium was removed and extracted with chloroform:methanol (2:1). After one additional extraction of the medium with chloroform, the chloroform phases were pooled, evaporated, and dissolved in 50% acetonitrile for HPLC analyses of corticosterone and aldosterone as described previously (11). Pregnenolone in the medium was quantitatively converted to progesterone with 3␤-hydroxysteroid dehydrogenase (Sigma) according to Talalay and Dobson (12) before extraction of the medium. Formed progesterone was then quantified by HPLC as above.
Metabolism of MeSO 2 -DDE was examined after growing cells in the presence of [ 14 C]MeSO 2 -DDE (13.4 mCi/mmol). The medium was removed, and the cells washed with phosphate-buffered saline and scraped into 500 l of cold phosphate-buffered saline. The culture medium and the cell suspensions were immediately extracted with three volumes of chloroform:methanol (2:1), followed by centrifugation (300 ϫ g, 15 min). The medium was extracted once more with chloroform, and the concentration of hydrophilic [ 14 C]MeSO 2 -DDE metabolites in the hydrophilic phase was measured by liquid scintillation. The protein pellet was exhaustively extracted by repeatedly solubilizing the protein in 1% sodium dodecyl sulfate and precipitating with acetone as described by Baker and Van Dyke (13). The protein was finally dissolved in NaOH, and aliquots were assayed for protein (14) and irreversibly bound radioactivity.
Transfection studies were performed on COS cells grown on 30-mm plates until almost confluent, when they received new medium contain-ing 3 g of DNA (2 g of P450 ϩ1 g of adrenodoxin (ADX)) in the transfection reagent DOTAP (Boehringer Mannheim). After a 20-h transfection period, the cells were grown for 24 h in fresh medium and then new medium containing 20 M steroid substrate or 4 M [ 14 C]MeSO 2 -DDE was added. Medium and cells were harvested after 24 h of incubation, and enzyme activities as well as [ 14 C]MeSO 2 -DDE metabolism assayed as described above.

RESULTS AND DISCUSSION
Upon treatment of adrenocortical Y1 cells with forskolin, which stimulates adenylate cyclase, these cells metabolized MeSO 2 -DDE to intermediates that reacted with cellular protein (Fig. 1a) and gave rise to hydrophilic MeSO 2 -DDE metabolites (Fig. 1b). By thin layer chromatography analyses of the hydrophilic MeSO 2 -DDE metabolites, the presence of ninhydrin-positive MeSO 2 -DDE metabolites, presumably glutathione conjugates, was indicated (data not shown). The lack of MeSO 2 -DDE metabolism in unstimulated cells and the minimal MeSO 2 -DDE metabolism in forskolin-stimulated Kin-8 cells (Fig. 1)  We next tested the ability of the three mitochondrial P450s present in adrenocortical Y1 cells to metabolize MeSO 2 -DDE. This metabolism was assessed in COS cells transfected with expression plasmids encoding mouse P450c11, P450scc, and P450aldo, respectively, either alone or in combination with a plasmid encoding bovine adrenocortical ADX (15,16). Adrenodoxin is an iron-sulfur protein that participates in the transfer of electrons from NADPH to the mitochondrial P450s, a process previously shown to be rate-limiting for steroid hydroxylations in transfected COS cells (16). As can be seen in Table I, the transfections were successful, although cotransfection with the ADX plasmid was necessary to obtain high P450 activities with the respective substrates. Only cells transfected with P450c11 metabolized MeSO 2 -DDE to any significant extent, as determined by the formation of hydrophilic MeSO 2 -DDE metabolites in the medium (Table I), and irreversible binding of MeSO 2 -DDE to protein (Fig. 3). Thus, both formation of hydrophilic metabolites and irreversible protein binding of MeSO 2 -DDE is mediated by the same isozyme, i.e. P450c11.
In vivo exposure of mice to MeSO 2 -DDE and in vitro experiments with subcellular adrenal homogenates have shown a selective irreversible binding of MeSO 2 -DDE in the adrenal cortex (4,5,7). The present studies provide direct evidence for an involvement of adrenocortical mitochondrial P450s in the bioactivation of MeSO 2 -DDE, by showing that minimal MeSO 2 -DDE or deoxycorticosterone metabolism occurred in uninduced adrenocortical cells and that forskolin induction increased the metabolism of both substrates to the same extent. Furthermore, we unambiguously ascribe the metabolism of MeSO 2 -  In addition, at concentrations only slightly affecting the cell number (assessed as amount of protein per plate), MeSO 2 -DDE was a potent inhibitor of the P450c11-activity in the Y1 cells (Fig. 4). In contrast, equal concentrations of the non-adrenotoxic DDE did not affect P450c11 or cell number (data not shown). This finding may provide an explanation to the previously observed reduced capacity of mice exposed to MeSO 2 -DDE to synthesize glucocorticoids (6).
This study thus illustrates two mechanisms, bioactivation of xenobiotics and inhibition of steroid hydroxylases, whereby xenobiotics may interfere with glucocorticoid synthesis. Because the adrenal gland generally accumulates high levels of lipophilic xenobiotics (17,18), and since a normal capacity to produce glucocorticoids is vital for organisms during stressful periods, the ability of certain xenobiotics to interfere with steroid hormone synthesis may have great significance and deserves further attention.