Self-augmentation Effect of Male-specific Products on Sexually Differentiated Progesterone Metabolism in Adult Male Rat Liver Microsomes*

It is well known that several 3-keto-4-ene steroids such as progesterone and testosterone are metabolized in a gender-specific or -predominant manner by adult rat liver microsomes. In the male, these steroids are primarily metabolized into two oxidized (16α-hydroxyl and 6β-hydroxyl) products mainly by the respective, male-specific cytochrome P450 subforms, CYP2C11 and CYP3A2, while they are primarily metabolized into the 5α-reduced products by female-predominant 5α-reductase in the female. These sexually differentiated enzyme activities are largely regulated at the transcription level under endocrine control. In the present study, we show that unlabeled 16α-hydroxyprogesterone and 6β-hydroxyprogesterone inhibited the 5α-reductive [3H]progesterone metabolism by adult male rat liver microsomes without significantly inhibiting the CYP2C11 and CYP3A2 activities producing themselves, whereas 3α-hydroxy-5α-pregnan-20-one and 5α-pregnane-3,20-dione not only stimulated the 5α-reductive metabolism producing themselves but also inhibited the male-specific oxidative metabolism. This finding compels us to propose a novel hypothesis that adult male rat liver microsomes may possess a self-augmentation system regulated by the male-specific products on sexually differentiated steroid metabolism, besides regulation by gene expressions of the related enzymes.

In the course of our investigation on structural requirements of substrates and/or inhibitors for active sites of CYP2C11 and CYP3A2 in male rat liver microsomes (to be published elsewhere), we unexpectedly found that male-specific products, 16␣-OH-P and 6␤-OH-P, inhibited female-predominant [ 3 H]PROG 5␣-reductase activity without significantly inhibiting the CYP2C11 and CYP3A2 activities producing themselves, while 3␣-OH-5␣-P and 5␣-P, female-predominant products by the 5␣-reductase, not only stimulated this enzyme activity but also inhibited the male-specific oxidative [ 3 H]PROG metabolism.
In the present paper, we extend these findings and suggest a novel self-augmentation effect of the male-specific products on sexually differentiated steroid metabolism in adult male rat liver microsomes, not involving gene expressions of the related enzymes.

EXPERIMENTAL PROCEDURES
Materials- [1,[2][3] H]PROG (specific activity, 49.2 Ci/mmol) was obtained from PerkinElmer Life Sciences and purified by a paper chromatographic system of hexane, saturated with formamide. Unlabeled steroids were purchased from Sigma and Steraloids Inc. (Wilton, NH). Goat anti-rat NADPH P450 reductase antiserum and rat CYP3A2 supersomes were purchased from Daiichi Pure Chemicals Co., Ltd. (Tokyo, Japan), and Whatman No. 1 filter papers used for paper chromatographies were from Whatman Ltd. Other reagents were of analytical grade.
Preparation of Adult Male Rat Liver Microsomes-Male Wistar rats, originally provided by Japan Charles River K. K., were bred in our colony. They were castrated on the 70th day after birth and used 3-4 weeks later. The liver microsomes were prepared as described previously (11). The experiments were performed according to institutional guidelines for the care and use of laboratory animals.
[ 3 H]PROG Metabolism by Rat Liver Microsomes-Effects of various unlabeled steroids on [ 3 H]PROG metabolism by liver microsomes were examined, according to our previously described procedure (12). Briefly, the microsomal suspension (400 -600 g of protein/2.2 ml, total volume of the reaction mixture) was preincubated with [ 3 H]PROG (20 nM) in the absence or presence of an unlabeled steroid (0.0316 -10 M) at 36°C for 30 min. Then NADPH (3.16 M) was added, and the reaction mixture was incubated for a further 5 min. After the incubation, two * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
identical samples were mixed and extracted with toluene. In some cases, before the above described incubation procedure, microsomal suspension (250 g of protein/1.1 ml, total volume of the reaction mixture) was preincubated with goat anti-rat NADPH P450 reductase antiserum (50 l) at 25°C for 30 min in order to inhibit P450-dependent oxidative [ 3 H]PROG metabolism. The toluene-extractable [ 3 H]PROG metabolites (more than 98%) were isolated by various paper chromatographic systems and then identified by the recrystallization method (13) [ 3 H]PROG Metabolism by Rat CYP3A2 Supersomes-In order to examine the direct effects of some unlabeled steroids on the oxidative [ 3 H]PROG metabolism, we used rat CYP3A2 supersomes, microsomes (82.5 g of protein/1.1 ml, total volume of the reaction mixture) of insect cells (BTI-TN-5B1-4) containing the cDNA-expressed rat CYP3A2, rat NADPH P450 reductase, and human cytochrome b 5 . Other experimental conditions were the same as those using the rat liver microsomes. The purified efficiency of [ 3 H]6␤-OH-P, exclusively formed by the supersomes, was 95.31 Ϯ 3.81%.

Evaluation of the Present Assay System for [ 3 H]PROG Metabolism by Rat Liver
Microsomes-In the present study, the respective final concentrations of [ 3 H]PROG and NADPH were adjusted to be 20 nM and 3.16 M, although these were approximately 2-4 orders of magnitude lower than those of customary enzyme assay systems (4,5,7,9). The reasons are as follows. 1) When the final concentration of ethanol (used for solubilizing [ 3 H]PROG and an unlabeled steroid) exceeded 2% (v/v), this induced aggregation of the microsomes, 2 and Wiebel et al. (14) have shown that some P450-dependent enzyme activities could be affected by more than 1% (v/v) of ethanol. Therefore, ethanol concentration was fixed to be 0.68% (v/v) in the present study, by which some unlabeled steroids became insoluble in the reaction mixture at their final concentrations over 1.0 M. 2) The [ 3 H]PROG concentration of 20 nM used seems physiological rather than those of the customary systems, since the plasma PROG concentration is estimated to be about 10 nM in adult male rats (15,16).
3) The yields of unidentifiable [ 3 H]PROG metabolites, included in both the water-soluble and tolueneextractable fractions, increased in a dose-dependent manner when either lower concentrations of [ 3 H]PROG or higher concentrations of NADPH were used. 2 The [ 3 H]PROG metabolism of the representative result for the 37 experimental batches performed in the present study is shown in Table I (1,9). However, the ratio of the sum of oxidized to 5␣-reduced products seemed to be severalfold to 10-fold lower than those of other investigators' data (1,7,8,10). This discrepancy may be partly related to the fact that we used adult male rats castrated for 3-4 weeks (in order to decrease endogenous steroids and increase [ 3 H]5␣-reduced metabolites), because such a postpubertal castration is known to induce a partial feminization of liver microsomal steroid metabolisms by repressing the CYP2C11 and CYP3A2 gene expressions and conversely stimulating the 5␣-reductase gene expression (2,17). It should, however, be noted that there were several reports showing similar results to ours, using intact male rats (2,18).  dative metabolism, but the latter did not inhibit it. Most interestingly, 3␣-OH-5␣-P and 3␣,11␤-(OH) 2 -5␣-P, compared with PROG, not only showed stronger inhibitory effects on the oxidative metabolism but also conversely stimulated the 5␣-reductive metabolism.

Classification of Various Steroids Based on Their Respective Effects on the Oxidative and 5␣-Reductive [ 3 H]PROG Metabolisms by Rat Liver
Microsomes-We found that various unlabeled steroids used could be divided into six groups, A, B, C, D, E, and F, based on their respective effects on the oxidative and 5␣-reductive [ 3 H]PROG metabolisms ( Fig. 2 and Table I). The group A steroids such as 3␣-OH-5␣-P and 5␣-P showed inhibitory effects on the oxidative metabolism, while having stimulatory effects on the 5␣-reductive metabolism producing themselves. The group B steroids, PROG and TEST, inhibited both metabolisms as probably alternative substrates. The group C steroids, 5␤-A-17␤-ol and 3␤-OH-P, showed inhibitory effects on the oxidative metabolism with no effect on the 5␣-reductive metabolism, and conversely, the group D steroids, COR and 11␤-OH-P, showed stimulatory effects on the 5␣-reductive metabolism with no effect on the oxidative metabolism, despite possessing a 3-keto-4-ene structure that might be catalyzed by the 5␣-reductase. Other 3-keto-4-ene steroids (group E), 16␣-OH-P and 6␤-OH-P, inhibited only the 5␣-reductive metabolism without the product inhibition effects on the oxidative metabolism producing themselves. It is noteworthy that 16␣-OH-P, as well as 20␣-OH-P and 4-AN-CA already reported by other investigators (19,20), were of the 3-keto-4-ene steroids showing the strongest inhibitory effect on the 5␣-reductase activity. Finally, the group F steroids, 11␣-OH-P and cholesterol, showed a slight effect or no effect on both of the metabolisms.
By the way, one may envisage a possibility that such a stimulatory effect of group A steroids on the 5␣-reductive metabolism may result from the increasing utilizations of free [ 3 H]PROG and NADPH, left over by their inhibitory effects on the oxidative [ 3 H]PROG metabolism and vice versa. However, this possibility may be largely refuted by the results of the following two experiments using the anti-rat NADPH P450 reductase antiserum and rat CYP3A2 supersomes.

Inhibition of P450-dependent [ 3 H]PROG Metabolism by Antirat NADPH P450
Reductase Antiserum-We examined the direct effects of representative steroids on the 5␣-reductive [ 3 H]PROG metabolism, using the rat liver microsomes pretreated with goat anti-rat NADPH P450 reductase antiserum (Fig. 3). By this means, more than 85% of the P450-dependent, oxidative [ 3 H]PROG metabolism was inhibited, irrespective of the absence or presence of an unlabeled steroid. Under such an experimental condition, PROG and 16␣-OH-P inhibited the 5␣-reductive metabolism, while 11␤-OH-P, 3␣-OH-5␣-P, and 3␣,11␤-(OH) 2 -5␣-P stimulated it, as the intact microsomes did (see Fig. 2). This result clearly shows that the effects of these steroids on the 5␣-reductive metabolism could be brought about by their intrinsic properties, not affected by the co-existence of P450-dependent metabolism in intact rat liver microsomes.
For additional information, an addition of normal goat serum, as compared with the 130 mM KCl-based buffer (12), induced a tendency to decrease the oxidative metabolism and increase the 5␣-reductive metabolism. Although the mechanism inducing such a tendency is wholly unclear at present, this may have been associated with lower stimulatory effects of 11␤-OH-P, 3␣-OH-5␣-P, and 3␣,11␤-(OH) 2 -5␣-P on the 5␣-reductive metabolism by the antiserum-treated microsomes, compared with the intact microsomes.
[ 3 H]PROG Metabolism by Rat CYP3A2 Supersomes-In order to examine also the direct effects of representative steroids on the male-specific P450-dependent [ 3 H]PROG metabolism, we used rat CYP3A2, but not CYP2C11, supersomes, which were composed of the microsomes of insect cells containing the cDNA-expressed rat CYP3A2, rat NADPH P450 reductase, and human cytochrome b 5 , since a recombinant CYP2C11 expression system has not come into the market, and we found that various unlabeled steroids showed a similar inhibitory pattern on rat liver microsomal [ 3 H]PROG 6␤-oxidation and 16␣-oxidation, mainly catalyzed by CYP3A2 and CYP2C11, respec-

FIG. 3. Effects of representative unlabeled steroids on [ 3 H]progesterone metabolism by adult male rat liver microsomes, pretreated with goat anti-rat NADPH P450 reductase antiserum.
The microsomal suspension (250 g of protein/1.1 ml, total volume of the reaction mixture) was preincubated with or without normal goat serum (50 l) or goat anti-rat NADPH P450 reductase antiserum (50 l) at 25°C for 30 min, and then they were incubated with [ 3 H]PROG (20 nM), an unlabeled steroid (1.0 M), and NADPH (3.16 M), according to the procedure shown in Fig. 1. a, 130 mM KClbased buffer without goat serum (12). b, the antiserum-pretreated microsomal suspension, incubated without an unlabeled steroid, was used as control. The data are means Ϯ S.D. of at least three experiments for separate rats. The sum of oxidized (f) or 5␣-reduced (Ⅺ) products is the same as shown in Fig. 1. Fig. 1 except for the concentration of unlabeled steroids (1.0 M in this experiment). The data are means Ϯ S.D. of at least three experiments for separate rats. tively (Fig. 4). 2 When the CYP3A2 supersomes were incubated with [ 3 H]PROG, an exclusively formed product was [ 3 H]6␤-OH-P (data not shown), and the inhibitory pattern of unlabeled steroids on the [ 3 H]6␤-OH-P formation resembled that obtained from the intact rat liver microsomes (Fig. 5).

FIG. 4. Comparison of the effects of representative unlabeled steroids on the [ 3 H]progesterone 6␤-or 16␣-oxidizing activity by adult male rat liver microsomes. The formation of [ 3 H]6␤-OH-P (f) or [ 3 H]16␣-OH-P (Ⅺ) was determined according to the procedure shown in
Furthermore, we examined the types of inhibition and the inhibitor constant (K i ) values of unlabeled PROG and 3␣-OH-5␣-P against [ 3 H]6␤-OH-P formation by rat CYP3A2 supersomes, according to a simple graphic method using two [ 3 H]PROG concentrations (21). From this graphic presentation (so-called Dixon's plot) shown in Fig. 6, it turned out that both of the unlabeled steroids behaved like a competitive inhibitor, the K i value of 3␣-OH-5␣-P was about 10-fold lower than that of PROG, and this K i value ratio agreed well with the IC 27.5 (against [ 3 H]6␤-OH-P formation) and IC 40 (against 16␣-OH-P formation) value ratios obtained using rat liver microsomes (Table II). Since not only unlabeled PROG but also 3␣-OH-5␣-P (22) must be metabolized into 6␤-OH-and/or 16␣-OH-products, it is most likely that these unlabeled steroids compete with [ 3 H]PROG as alternative substrates, but not as true competitive inhibitors, for rat CYP3A2 and/or CYP2C11 with 3␣-OH-5␣-P possessing about 10-fold higher affinity for the substratebinding pockets of these enzymes and that the effects of these unlabeled steroids on the CYP3A2 (and probably CYP2C11) activity also can be brought about by their intrinsic properties, independent of a difference of the microsomal structures between the rat liver and insect cells.
In conclusion, the present study clearly shows that the malespecific products, 16␣-OH-P and 6␤-OH-P, inhibited the female-predominant 5␣-reductase activity without significantly inhibiting the male-specific CYP2C11 and CYP3A2 activities producing themselves. On the other hand, the female-predominant products, 5␣-P and 3␣-OH-5␣-P, not only inhibited the male-specific P450 activities but also stimulated the 5␣-reductase activity producing themselves, and such effects can be brought about by the intrinsic properties of these steroids. Thus, we can propose a novel hypothesis, as described under "Discussion," on the regulation system of sexually differentiated steroid metabolisms in adult male rat liver. DISCUSSION It is well known that various 3-keto-4-ene steroids such as PROG, TEST, and 4-AN are primarily metabolized into 16␣-(in some cases, 2␣-also) and 6␤-oxidized products mainly by the respective, male-specific P450 subforms, CYP2C11 and CYP3A2, in male rat liver microsomes, whereas they are primarily metabolized into the 5␣-reduced products by femalepredominant 5␣-reductase in the female (1-10), and it is known that expressions of these sexually differentiated enzyme activities are largely regulated in transcription level under endocrine control of which GH plays a major role (6, 8 -10).
In the present in vitro study using adult male rat liver microsomes (Tables I and II; Figs. 1-4) and rat CYP3A2 supersomes (Figs. 5 and 6; Table II), we showed for the first time that two major male-specific oxidized PROG metabolites, 6␤-OH-P and especially 16␣-OH-P, strongly inhibited the femalepredominant 5␣-reductase activity without significantly showing the inhibitory effects on the CYP3A2 and CYP2C11 activities producing themselves, and these events may be fur- ther enhanced by high levels of CYP2C11 and CYP3A2 gene expressions in the male (6, 8 -10, 17). On the other hand, 5␣-P and especially 3␣-OH-5␣-P not only inhibited both the CYP2C11 and CYP3A2 activities but also stimulated the 5␣reductase activity producing themselves. However, such adverse effects of the 5␣-reduced products on the male pattern metabolism may be attenuated by a scanty expression of the 5␣-reductase gene in the male (8,10). Thus, our results compel us to propose a very interesting hypothesis, summarized in Fig.  7, that adult male rat liver microsomes may possess a selfaugmentation system by the male-specific products on sexually differentiated steroid-metabolizing activities, coupled with the regulation system by gene expressions of the related enzymes under endocrine control. In other words, the results may also explain the reason why adult male rat liver should preserve not only much higher levels of CYP2C11 and CYP3A2 gene expressions but also lower 5␣-reductase gene expression, as compared with the female.
Furthermore, it is of great interest and importance to investigate whether the female rat liver also possesses such a selfaugmentation system, although the present results strongly suggest that at least female-predominant 5␣-reductase activity (1,3,7,8,10,18) may be further enhanced by its products, 5␣-P and especially 3␣-OH-5␣-P. As regards these, an important question for future study is to elucidate the reason why adult male rat liver microsomes must metabolize PROG first into more hydrophilic products, 16␣-OH-P and 6␤-OH-P, while the female must metabolize it into more hydrophobic products, 5␣-P, under the strictly regulated systems described above.
By the way, a similar scenario may occur on the androgen metabolism, since 3-keto-4-ene androgens such as TEST and 4-AN are also known to be catalyzed sex-dependently by the same enzyme systems (1-3, 5, 7-10, 18), and the effects of various 3-keto-4-ene and 5␣-reduced androgens, especially TEST and 5␣-A-3␣,17␤, on the [ 3 H]PROG metabolism showed a similar pattern to those of various 4-pregnene and 5␣-pregnane steroids described here (Fig. 2). 2 As regards another interesting finding obtained from the present study, it has been reported that endogenous COR production in rat adrenal cortex is suppressed by exogenously administrated COR or cortisol in in vivo and in cell culture systems and that this inhibition probably results from the various effects of these steroids, namely inhibiting ACTH secretion from the pituitary, decreasing ACTH sensitivity of adrenal cortex (23), and stimulating the adrenal 5␣-reductase activity metabolizing COR into its 5␣-reduced products (24). However, several recent studies have clearly shown that the two 11␤-OH corticosteroids, COR and cortisol, are of the poorest substrate group for 5␣-reductases of various organs probably including the adrenal cortex itself (19,20,25), and we showed in the present study that COR and 11␤-OH-P, but not 11␣-OH-P, rather stimulated [ 3 H]PROG 5␣-reductase activity of rat liver microsomes (Fig. 2). These results suggest that the C-11␤-OH group of a steroid molecule may strongly disturb access of the steroid to the active site of the 5␣-reductase, and we can propose another possibility that adrenal cortex may FIG. 7. Supposed self-augmentation system on sexually differentiated progesterone metabolism in adult male rat liver. 16␣-OH-P and 6␤-OH-P are produced by male-specific microsomal P450s, CYP2C11 and CYP3A2, respectively. Both products inhibit PROG 5␣-reduction without significant product-inhibition effects on the above P450 activities, and these events may be further enhanced by high levels of CYP2C11 and CYP3A2 gene expressions in the male. Actually, 16␣-OH-P may make a higher contribution to the inhibitory effect on the 5␣-reductase activity than 6␤-OH-P, judging from the result shown in Fig. 2 and a higher expression of CYP2C11 gene than CYP3A2 gene (1,9). On the other hand, 5␣-P and 3␣-OH-5␣-P are produced by the 5␣-reductase and subsequently cytosolic (and, to a lesser degree, microsomal) 3␣-hydroxysteroid dehydrogenase (26 -28), respectively. Although these products, especially 3␣-OH-5␣-P, not only inhibit the CYP2C11 and CYP3A2 activities but also stimulate the 5␣-reductase activity, such adverse effects of these 5␣-reduced products on the male pattern metabolism may be actually attenuated by lower expressions of both the 5␣-reductase (1,3,7,8,10,18) and 3␣-hydroxysteroid dehydrogenase (27,28) genes in the male, compared with the female. 16␣-OH-P, 6␤-OH-P, or 3␣-OH-5␣-P can be further metabolized into its glucuronide or sulfate and eventually excreted into urine and/or bile. It is most likely that the same self-augmentation system operates on the androgen metabolism. m, microsomal enzyme. c, cytosolic enzyme.
In conclusion, we can propose two novel hypotheses on 1) the self-augmentation system on sexually differentiated steroid metabolism in adult male rat liver and 2) a short negative feedback system of COR production in adrenal glands. Although the action mechanisms operating these regulatory systems are largely unclear at present, an attempt to clarify them is currently under investigation in our laboratory.