Correlation of hepatic cytosolic androgen binding proteins with androgen induction of hepatic microsomal ethylmorphine N-demethylase in the rat

Vol. 21, No. 3, pp. 243S2.52, 1984 Printed in Great Britain. All rights reserved

.I. sreroid Biochem.

0022-4731/84 $3.00+ 0.00

Copyright Q 1984Pergamon PressLtd

CORRELATION OF HEPATIC CYTOSOLIC ANDROGEN BINDING PROTEINS WITH ANDROGEN INDUCTION OF HEPATIC MICROSOMAL ETHYLMORPHINE N-DEMETHYLASE IN THE RAT RICHARD C. RUMBAUGH, ZADDOCK MCCOY and GEORGEW. LUCIER* Laboratory of Pharmacology, National Institute of Environmental Health Sciences, National Institutes of Health, P.O. Box 12233, Research Triangle Park, NC 27709, U.S.A. (Receiued

19 Sepie~~e~

1983)

Sammary-Previous reports have demonstrated the presence of moderate to high affinity binding for androgens in the cytosol of livers from male rats. This binding was significantly lower in female rats or in immature rats of either sex. The hepatic androgen binding protein, which sedimented at approx. 4 S on sucrose density gradients, has been called a receptor which mediates the actions of androgens in the liver. The experiments in the present study were designed to evaluate the hepatic androgen binding protein for characteristics which have been attributed to receptors in other tissues and to correlate the presence of androgen binding with androgen induction of hepatic drug metabolism. In the current studies, we have shown that cytosol from the livers of male rats bound [3H]dihydrotestosterone ([3H]DHT) and translocated this steroid ligand to the nucleus in a time and temperature dependent manner. Cytosol prelabeled with [3H]DHT, when passed over a column of denatured DNA cellulose, eluted in three radioactive peaks. Two of these peaks were absent when cytosol from livers of female or hy~phy~ctomized males was used. In addition, the presence of high concentrations of hepatic androgen binding correlated well with the ability of androgen to induce

ethylmorphine ~~~ethylase, a marker of microsomal cytochrome P-45~-de~ndent drug metabolism. Values for both parameters were higher in males than in either females or hypophysectomized males. Testosterone treatment induced both parameters in ovariectomized females and lir&estradiol repressed both in males. However, testosterone treatment failed to induce hepatic androgen binding in hypophysectomized males and immature males, both of which are also unresponsive to androgen induction of drug metabolism. The results suggest that one or more hepatic cytosolic androgen binding proteins possess several characteristics associated with steroid receptors in reproductive tract tissue. Furthermore, this binding may be implicated as a mediator for the androgen induction of at least one component of hepatic drug metabolism.

INTRODUCTION

It is now widely recognized that a principal requirement for the action of steroid hormones is the interaction of the hormone with a highly specific intracellular protein receptor (1,2]. There is a large body of literature describing receptors for gonadal steroids in reproductive tract tissuesp-51. There is, however, relatively little information available on the mechanism of action of sex steroids in extragenital organs, despite the fact that a number of tissues including skin [6], skeletal muscle [7], kidney [8], brain [9] and liver [IO-121 have been shown to contain steroid receptors. Previous studies have demonstrated that the livers of both male and female rats are responsive to estrogen admi~stration and also contain an estrogen receptor with characteristics similar to that of the uterus f 121.Hepatic tissue also responds to androgens by the induction of a low molecular

*To whom correspondence should be addressed. A preliminary report of this investigation was presented at the 64th Annual Meeting of the Endocrine Society, San Francisco, June 16-18, 1982.

weight protein called cr,,-globulin [13] and by the induction of a number of cytochrome P45~dependent drug and steroid metabolizing enzymes [14]. However, the mechanism of action of androgen in the liver is not yet clear. As early as 1974, Roy et af.[iO] described an androgen binding component in the cytosol of male rats which bound dihydrotestosterone (DHT) with high affinity and 17fl-estradiol with somewhat lower affinity. These authors [lo] suggested that this binding component was a receptor for DHT and was obligatory for androgen induction of a,,,-globulin. The androgen binding component of hepatic cytosol from the male rat has recently been further characterized by gel ~ltration, Scatchard analysis and specificity studies [ 1.5,161. Although it has been known for many years that a number of cytochrome P-450-dependent hepatic oxidative drug and steroid metabolizing enzymes are androgen responsive [14, 171, the mechanism of this androgen induction is unknown although it has been reported that androgen effects on liver cytochrome P-450 are pituitary dependent [14]. However, it seems likely that if the androgen binding component of

243

244

RICHARD

C.

RUMBAUGH et

hepatic cytosol is a true receptor, then the androgen responsiveness of hepatic monooxygenases may be mediated through this mechanism. Therefore, we have investigated several properties of the hepatic androgen binding component which are also characteristic of steroid receptors in reproductive tract tissue. Furthermore, in intact and hypophysectomized male and female rats, we have correlated binding of DHT to androgen responsiveness and induction of hepatic drug metabolism. MATERIALS

AND METHODS

Animals and treatment

Male and female Sprague-Dawley rats were used for all experiments. The animals were housed 5 per cage in a room with a 12 h light-dark cycle and constant temperature of 25°C. Rat chow (NIH31 mouse and rat ration, Ziegler Bio.) and tap water were available ad libitum. In some experiments, rats were hypophysectomized by the breeder (Charles River Laboratories, Wilmington, MA) using the parapharyngeal route under light ether anesthesia at 60 days of age. Control animals received sham operations. Animals were received at 80-days of age and allowed to recover for 2 weeks prior to experimentation. Hypophysectomized animals were given a solution of 0.9% NaCl and 5% glucose in tap water ad libitum to enhance survival. In one experiment, female rats were ovariectomized under anesthesia with ketamine (Bristol Laboratories, Syracuse, NY) and xylazine (Cutter Laboratories, Shawnee, KS). These animals were allowed to recover for 2 weeks prior to experimentation. Some animals were treated with a single injection of either Depo-Testosterone (Upjohn Co., Kalamazoo, MI), 100 mg/kg, or DepoEstradiol (Upjohn), 25Opg/kg 2 weeks prior to sacrifice. These doses of long-acting preparations of steroid have been shown to maintain sex accessory tissue weights in castrated animals at normal values for at least 4 weeks after injection. All control animals received an equal volume of vehicle only. In one experiment, male rats of 21 days of age were treated with Depo-Testosterone, 100 mg/kg and were returned to the dams for 7 days prior to sacrifice. Determination of speciJic cytosolic binding of DHT

Rats were sacrificed by decapitation and the liver was quickly exised and placed on ice. One gram portions of minced liver were homogenized in 10 ml of a buffer containing 10 mM Tris, pH 7.4, 1 mM dithiothreitol (DTT) and 1 mM EDTA (Buffer A). The homogenates were centrifuged at 105,OOOg for 60min and the supernatants carefully removed to avoid the lipid layer. Cytosolic binding of [3H]DHT was measured in duplicate by adding 0.2 ml cytosol (l&12mg protein/ml) to an equal volume of [1,2,4,5,6,7-3H(N)]dihydrotestosterone (New England Nuclear, 11&150 Ci/mmol) for 2 h at 4°C. Protein binding of [3H]DHT reached equilibrium

al.

after I h at 4°C and was stable for up to 7 h. The final concentration of t3H]DHT was 40nM [IO]. After incubation, the samples were added to an equal volume of dextran-coated charcoal (DCC) (O.OS”,; Dextran, O.S’;/, activated charcoal) in Buffer A for 15 min and the bound radioactivity was measured as described previously for estradiol binding [18, 191. The data in Tables l-4 are uncorrected for the efficiency of DCC which was greater than 9976. Specific binding of [‘H]DHT was calculated as the difference in bound radioactivity in the absence and presence of lOO-fold excess unlabeled DHT. In some experiments, aliquots of DCC-treated samples were pooled and layered on linear 5520”/6 sucrose density gradients. Following centrifugation at 200,OOOg for 18 h, each gradient was fractionated and the fractions were assessed for radioactivity by liquid scintillation spectrometry. The sedimentation marker used was bovine serum albumin (4.6s). Cell ,free nuclear translocation Cytosol was prepared by homogenizing a 2g sample of minced liver in 10 ml of a buffer containing 50 mM Tris, pH 7.4, 25 mM KCl, 10 mM MgCl,, 0.25 M sucrose, and 1 mM DTT (Buffer C). This homogenate was centrifuged at 105,OOOgfor 60 min and the supernatant was carefully removed. A 2.5 ml aliquot of cytosol was incubated with an equal volume of [3H]DHT in buffer C (40 nM final concentration) for 1.5 h at 4’C. In one experiment, a lOO-fold excess unlabeled DHT was also included in the incubation. The incubates were then added to a pellet prepared from an equal volume of DCC suspension in Buffer C, held on ice for 20min and centrifuged at 1000 g for 5 min to remove DCC containing unbound steroid. The DSS-treated cytosol mixtures were then incubated with nuclei. Nuclei were prepared by homogenizing a 4 g portion of minced liver in 35 ml of a buffer containing 10 mM Tris, pH 7.5, 0.5 M sucrose and 5 mM MgCI, (Buffer B). The homogenates were filtered through cheesecloth and centrifuged at 1OOOgfor 15 min at 4°C. The resultant pellet was resuspended in Buffer B and centrifuged again. Nuclei were then purified by centrifugation through a cushion of 2.2 M sucrose containing 1 mM MgCl, as previously described [ 121. Nuclei were checked for purity by electron and light microscopy; nuclei were intact and free from extranuclear contamination as shown previously [ 121. Moreover, nuclear preparations did not contain detectable alkaline phosphatase activity. Purified nuclei were resuspended in 2 ml Buffer C and combined with pre-labeled DCC-treated cytosol. The mixtures were incubated at different temperatures as indicated and aliquots were removed at various times. Following incubations, nuclei were isolated by centrifugation for 1 min in a Beckman Microfuge. The nuclear pellet was washed three times; once with Buffer C, once with buffer C containing 0.27: (w/v) Triton X-100 to remove the outer nuclear membrane and finally with

Correlation of hepatic androgen binding and drug metabolism Buffer C. The bottom of the tube containing the washed nuclear pellet was cut and the pellet was measured for nuclear bound radioactivity. Scufchard unalysis

Livers were obtained from male and female rats and a 20”/, homogenate was prepared in Buffer AG (Buffer A containing 10mM Na’ molybdate and 10% glycerol). In order to generate consistent Scatchard data, it was necessary to remove large quantities of non-specific binding. To accomplish this, cytosol was prepared and partially purified by incubating aliquots with 35% (w/v) (NH,&S04 for 30 min at 4°C with constant shaking. Following centrifugation at 12,000g for 30 min the resultant pellets were resuspended in Buffer AG. Aliquots of resuspended protein were incubated with various concentrations of I3H]DHT in the absence and presence of lOO-fold excess unlabeled DHT. After 2 h at 4°C the unbound steroid was precipitated with DCC as described above. Aliquots of the supernatant were assessed for bound radioactivity. The data were calcuiated and plotted according to Scatchard and the line fitting the points was determined. Similar data were obtained when Scatchard analysis was performed on liver samples, prepared in the absence of Na+ molydbate. Data on Scatchard analysis (Fig. 4) using molybdate was selected in order to demonstrate that the tow amount of specific binding could not be increased by molybdate ions. Binding to DNA -cellulose

The livers of 2 rats were pooled and minced together. A 6 g portion of minced liver was homogenized in 30 ml of buffer containing 20mm Tris, pH 7.4, 1 mM EDTA, I mM DTT and 10% (w/v) glycerol (TEDG Buffer). The homogenate was centrifuged at 105,000 g for 60 min. An aliquot of 2.5 ml of cytosol was incubated with an equal volume of [‘H]DHT in TEDG (40 nM final concentration) for 2 h at 4°C followed by 0.5 h at 29°C. All subsequent procedures were performed at 4°C. The incubates were added to a pellet of DCC prepared from an equal volume of DCC in TEDG, incubated for 20min and centrifuged at 1OOOg for 5 min. The supernatant containing only bound radioactivity was applied to a column (1.0 x 15 cm) of denatured DNA-cellulose (P-L Biochemicals, Milwaukee, WI) which had been equilibrated with TEDG buffer, washed with TEDG containing 2 M KC1 and reequilibrated with TEDG. Following a period of equilibration of 90 min the column was washed with TEDG at a flow rate of Sml/h for at least 16 h. Fractions of 4ml were collected and the wash was continued until radioactivity in the fractions reached a low level. The column was eluted with a 200 ml linear gradient of O--l.0 M KC1 in TEDG at 8 ml/h. When the gradient was complete, 1 ml aliquots of the fractions were assessed for radioactivity by liquid scintillation spectrometry.

245

Measurement of drug metabolism

One gram aliquots of minced liver were homogenized in lOm1 of buffer containing 0.05 M Tris, pH 7.4 and 1.15% (w/v) RCl. The homogenates were centrifuged at 900 g for IO min followed by 9500 g for 20 min to remove nuclei and mitochondria. Aliquots of post-mitochondrial supernatant were removed for measurement of ethylmorphine N-demethylase activity and the remainder was centrifuged at 105,OOOgfor 60min to prepare the microsomes. The microsomal pellet was resuspended in KCI-Tris buffer and centrifuged again. The washed microsomal pellet was resuspended in KCl-Tris buffer and cytochrome P450 concentrations were determined as the dithionite reduced CO bound difference spectrum using an extinction coefficient of 91 mM-’ cm’ according to the method of Omura and Sato 1201.The activity of ethylmorphine N-demethylase was assayed by the spectrophotometric determination of formaldehyde according to the method of Nash [21] as described previously [22]. Aliquots of post-mitochondrial supernatant equivalent to 0.1 g liver were used and the c reaction was linearly dependent on both protein concentration and time under the incubation conditions employed. Other methods

Protein in microsomes and cytosol was measured by the method of Lowry(231 using bovine serum albumin as the standard. Statistical differences were determined by Student’s r-test and differences were significant when P < 0.05.

RESULTS Sucrose gradients

Figure 1 illustrates sucrose gradient analysis of male rat liver cytosol pre-labeled with [3H]DHT. There was one large peak of radioactivity sedimenting at approximately 4s. When IOO-fold excess unlabeled DHT was included in the incubation mixture, the peak at 4s was reduced but not completely eliminated, indicating the presence of both specific and nonspecific binding in the 4s sedimenting peak. When cytosol from livers of female rats was used in an identical experiment, the gradients in the presence and absence of IOO-fold excess unlabeled DHT appeared identical, suggesting that all the androgen binding in liver of female rats was nonspecific (data not shown). Cell free nuclear transiocation When hepatic nuclei from male rats were incubated with cytosol which had been pre-labeled with 13H]DHT, nuclear translocation of [3H]DHT occurred. The rate of translocation was approx. three times greater when the cytosol-nuclei incubation was conducted at 22°C then at 4°C (Fig. 2). Similarly, when the cytosol was pre-labeled with [3H]DHT

246

RICHARD

C. RUMBAUCH

et al.

IO

20

i 30

40

50

60

Time hn)

J

Fig. 1. Sucrose density gradient analysis of 13H]DHT binding in liver cytosol from male rats. Aliquots of hepatic cytosol from male rats incubated with 4OnM 13H]DHT (+---0) or 40 nM [3H]DHT plus lOO-fold excess unlabeled DHT (0-O) were analyzed on linear S-20”/, sucrose gradients as described in the text. Migration from left to right is top to bottom. The arrow indicates the position of BSA in parallel gradients.

IOO-fold excess unlabeled DHT, there was very little subsequent nuclear translocation, even when the cytosol-nuclei incubation was performed at 22°C (Fig. 3). When buffer was substituted for the cytosol in a control experiment, there was no si~ificant a~umuiation of radioactivity in the nucleus (not shown).

Fig. 3. Cell free nuclear translocation rH]DHT from liver cytosol of male rat at 22°C. O-0, cytosol incubated with 40 nM rH]DHT. This curve is the same as shown in Fig. 2. C+---0, cytosoi incubated with 40nM 13H]DHT + 4 FM unlabeied DHT.

livers of male and female rats using partially purified cytosol (Fig. 4). Because we used partially-purified preparations, these Scatchard analyses provide qualitative (&) data but not good quantitative data. It was impossible to generate reliable Scatchard plots

containing

r

180

0

160

~

Scatchard analysis

Binding

affinities for DHT were determined

in

3000

~

20

4.0

60

80

100

120

L 140

Bound (dpm/mllxlO-’

IO

20

30

40

50

60

Time (mm)

Fig, 2. Cell free nuclear translocation of [3HfDHT from iiver cytosol of male rats: ~m~rature dependence. The radioactivity is nuclear bound [-‘H]DHT resulting from translocation of cytosolic bound material as described in the text.

Fig. 4. Scatchard analysis of [3H]DHT binding in hepatic cytosol from male and female rats. Cytosol was prepared from the pooled livers of 3 individual animals and partially purified as described in “Methods” and incubated with various concentrations of t3H]DHT for 2 h at 4°C in the presence and absence of 1Wfoid excess unlabeled DHT. After the incubation, samples were treated with DCC as described in “Methods” and the supernatant was assessed for bound radioactivity. The data points for specific binding were plotted and the line fitting the points was dete~ined by linear regression. (Q-----O), male; (O-O), female.

Correlation of hepatic androgen binding and drug metabolism

00

I; I /

,,’

75

I

I

50

E Y

25

241

solid circles). In a parallel experiment using cytosol of male rat livers but a column of cellulose only, no significant radioactivity eluted from the column in the salt gradient (Fig. 5, open circles). Cytosol from livers of female rats (Fig. 6a) and hypophysectomized male rats (Fig. 6b) was also passed over columns of denatured DNA-cellulose. In both cases, only one peak eluted early in the salt gradient. However, 2 peaks were absent indicating that two components of androgen binding normally found in male rats were absent in females and hypophysectomized males. Correlation of binding activity with response

IO

20

30

40 Fraction

50

60

70

60

no

Fig. 5. Binding of [)H]DHT labeled liver cytosol from male rats to denatured DNA cellulose (e-0) or cellulose (0-O). Cytosol from male rats (7-9 mg protein/ml final concentration) was incubated with [)H]DHT (40 nM final) for 2 h at 4°C and 0.5 h at 29°C. The incubations were treated with DCC to remove free ligand and applied to a column of denatured DNA-cellulose. After equilibration, the column was washed with TEDG and eluted with TEDG containing a linear gradient of O.&I.OM KC1 (-----). Fractions were collected and aliquots of the fraction were assessed for radioactivity. Details are described in “Materials and Methods.”

using whole cytosol due to huge amounts

of nonspecific binding. Hepatic cytosol from male rats exhibited a high affinity, low capacity binding component with a Kd = 3.1 nM and concentration of 4.7 fmol/mg cytosolic protein. In the female, this component was apparently absent, although binding was detected with Kd = 43.4 mM and concentration of 8.3 fmol/mg cytosolic protein. Binding to DNA In vitro experiments were performed in which ahquots of cytosol pre-labeled with [‘H]DHT were passed over columns of denatured DNA-cellulose. After equilibration, the column was washed extensively, then eluted with a gradient of KCI. When cytosol from livers of male rats was used in the experiment, bound radioactivity eluted from the DNA-cellulose column in three distinct peaks (Fig. 5,

We next performed a series of experiments for the purpose of correlating the presence of high concentrations of androgen binding activity with hepatic androgen responsiveness. The enzymatic marker of responsiveness was ethylmorphine n-demethylase P-450-dependent mono(EMD), a cytochrome oxygenase activity known to be androgen sensitive [24]. A comparison of male and female rats (Table 1) indicated that the normal adult male had high concentrations of androgen binding activity, while the adult female had a significantly lower amount. Likewise, the activity of EMD was high in males and low in females. However, the total concentration of cytochrome P-450 was not significantly different in the two sexes. Similarly, when the hypophysectomized male was compared to the intact male, a parallel change in androgen binding and EMD activity occurred, both being decreased by hypophysectomy. Cytochrome P-450 concentrations were not significantly different in hypophysectomized and intact rats. It has been demonstrated that oxidative drug metabolism in the hypox rat is unresponsive to androgen treatment, however, the reason for the lack of effect is unknown [24]. Therefore, we examined androgen binding in the hypophysectomized animals treated with testosterone. In the hypophysectomized males, androgen binding was low and was unchanged by testosterone treatment (Table 2). The activity of EMD was also low in the animals, however, with testosterone treatment, there was a very small, but statistically significant increase in EMD activity. This increase was far smaller than that produced by testosterone in the intact animal [24]. Therefore, hypophysectomy markedly diminished responsiveness of hepatic EMD to androgens. The hypophy-

Table 1. Correlation of hepatic cytosolic androgen bonding and microsomal ethylmorphine

”

DHT binding (fmol/mg protein)

A’-demethylase

Ethylmorphine N-demethylase (nmol/min/g liver)

Male Female

455.6 + 96.7 66.9 5 23.9’

281.0 f 28.4 62.6 f 4.9*

HYPES male

161.6 f 27.9’

28.9 ?r:4.8’

-~

Cytochrome P-450 (nmol/mg protein) 0.72 f 0.08 0.61 f 0.02 0.57 + 0.03

Values are expressed as the mean f SEM of groups of 5-6 individual animals. Asterisk indicates a significant difference (P -z 0.05) from the appropriate male control group.

248

RICHARD C. RUMBAUGH

et al.

Table 2. Effects of testosterone administration on hepatic cytosolic androgen binding and microsomal ethylmorphine N-demethylase in hypophysectomized (hypox) male and female rats DHT binding (fmol/mg protein)

Ethylmorphine N-demethylase (&ol/min/g liver)

Cytochrome P-450 (nmol/mg protein)

108.9 * 9.7

71.6 + 4.3

ND

102.4 i_ 20.1

86.4 f 1.1;

ND

HYPES female

73.2 f 17.9

57.9 * 4.8

0.84 + 0.02

HYPES female + testosterone

76.9 + 12.5

65.2 f 3.9

0.74 f 0.05

Hypox male Hypox male + testosterone

Values are expressed as the mean i SEM of groups of 5-6 individual animals. Asterisk indicates a significant difference (P < 0.05) from the vehicle-treated control group. ND = Not determined.

sectomized female was completely unresponsive to androgen treatment (Table 2). Both androgen binding and EMD activity were present at low levels and were unchanged by administration of testosterone. Likewise, the concentration of cytochrome P-450 was not affected by hormone treatment. Estrogen administration to male rats effected hepatic androgen binding activity and EMD activity similarly (Table 3). Males which received estrogen treatment exhibited a decrease in hepatic androgen binding activity and rate of EMD activity, and a significant decrease in cytochrome P-450 concentration. Conversely, although the normal female did not exhibit high concentrations of hepatic androgen binding or rates of EMD, it was possible to induce both by administration of testosterone to the ovariectomized adult female (Table 3). There was a significant increase in both [3H]DHT binding and EMD activity following 2 weeks of testosterone

treatment in these females although the amount of binding was not equal to that found in intact male rats. Total cytochrome P-450 concentration was not affected by this hormone regimen. Table 4 presents the results of an experiment in which the effect of testosterone adminstration on hepatic function in prepubertal rats were examined. Rats of 21 days of age received testosterone continuously for 7 days and were sacrificed at 28 days of age. The dose of testosterone used was equivalent to the dose which, in adult animals, is sufficient to induce both androgen binding and drug metabolism. However, this dose in prepubertal animals was ineffective at inducing either parameter. There was no detectable specific androgen binding in either control or treated animals. EMD activity was present at a low level but was not affected by testosterone treatment, nor were cytochrome P-450 concentrations.

Table 3. Effects of administration of estradiol to the male and testosterone to the ovariectomized female on heoatic cvtosolic androeen bindine and microsomal ethvlmoruhine N-demethvlase

Male

DHT binding (fmol/mg protein)

Ethylmorphine N-demethylase (nmol/min/g liver)

Cytochrome P-450 (nmol/mg protein)

636.2 k 59.2

216.6 + 29.0

0.68 + 0.02

Male + estradiol

85.7 + 13.48

43.9 + 1.18

0.44 + 0.05*

Ovariectomized female

13.2 f 8.3

44.1 + 2.7

0.50 * 0.05

Ovariectomized female + testosterone

212.7 + 31.6’;

79.0 i 9.0**

0.48 & 0.04

Values are expressed as the mean f SEM of groups of M individual animals. Single asterisk indicates a significant difference (P < 0.05) from the male control group. Double asterisk indicates a significant difference (P i 0.05)from the ovariectomized female control group. Table 4. Effects of testosterone administration to immature male rats on hepatic cytosolic androgen binding and microsomal ethylmorphine N-demethylase DHT binding (fmol/mg protein)

Ethylmorphine N-demethylase (nmol/min/g liver)

Cytochrome P-450 (nmol/mg protein)

Immature male

ND

71.5 * 8.9

0.52 + 0.02

Immature male + testosterone

ND

83.1 + 7.8

0.51 + 0.01

Details on the ages of the animals and dosages used are given in the text. Values are expressed as the mean + SEM of 6 determinations done on samples prepared from a pool of livers from 2 rats each. ND = Not detectable.

Correlation of hepatic androgen binding and drug metabolism DISCUSSION

The mechanisms by which male sex steroids influence hepatic biochemistry remain unclear despite the fact that at least two actions of androgen on liver have been well described. The androgenic induction of the secretory protein cr,,-globulin was first described by Roy and coworkers [13]. Androgens are also known to affect the rates of hepatic metabolism of certain drug and steroids through actions on cytochrome P-450-dependent monooxygenases [24,25]. In 1974, Roy rt al. characterized a hepatic protein in cytosol which bound androgens and also had a low affinity for estrogens [1O].This protein was described as a receptor although it was characterized only in terms of its specificity and sedimentation pattern [lo]. It is likely that the bound [-‘HJDHT which we detect on sucrose gradients (Fig. 1) and with the one-point androgen binding assay (Tables 14) contains a mixture of several proteins which bind [3H]DHT with moderate to high affinity. Therefore, it is probable that only a fraction of the total bound [-‘H]DHT exhibits characteristics of steroid receptors. The remainder of the bound steroid is probably associated with enzymes or other proteins of moderate binding affinity and unknown function. At least one moderate affinity binding component is likely to be the higher capacity, lower affinity estradiol binding sites, which have been described elsewhere [26,27]. These proteins are known to bind both androgens and estrogens with moderate affinity. However, it is not yet known whether these steroids compete for the same site or bind to different sites on the same protein. Other attributes of the hepatic androgen binding protein have been described elsewhere. Eisenfeld and coworkers have described high affinity androgen binding in rabbit liver which translocates to the nucleus [28]. We have used their technique of partial purification to determine that male rat liver possesses a high affinity, low capacity binding component for DHT while female rat liver has only a lower affinity binding component (Fig. 4). Ota and coworkers [16] describe a two-component model of androgen binding in which one component has high affinity and one low affinity for androgen. In addition to cytosolic binding, high affinity, saturable binding of DHT has been demonstrated in the nucleus of liver cells [29]. It is now generally accepted that cytosolic steroid receptors must also meet other criteria. Specifically, they must translocate steroids to the nucleus, interact with nuclear material and initiate a measurable biological response. The data in the present report suggests that the hepatic androgen binding activity possesses some of these characteristics. The concentration of t3H]DHT used in our studies to label hepatic cytosol (40 nM) is identical to that used by Roy in an earlier report [lo]. In addition, we have found 40 nM [3H]DHT to saturate displace-

249

able hepatic binding of DHT. The sedimentation coefficient of hepatic androgen binding activity is approximately 4s in confirmation of previous observations [lo]. Although many steroid receptors from reproductive tract tissue sediment in similar gradients at 8-9S, there is no indication of any binding in this region of our sucrose gradients (Fig. 1). Although the specific binding of DHT occurred in the 4s region of the gradients in the same area as binding by serum components, it is unlikely that the binding we observed was due to the contamination from plasma or serum. The total binding of [‘H]DHT by hepatic cytosol from male rats is much greater than that seen when pure serum is used. Furthermore, serum from male and female rats show nearly identical amounts of binding, whereas androgen binding by hepatic cytosol of male rats is nearly ‘I-fold greater than that of female rats (Table 1). In addition, immature rats, which do not reveal any specific cytosolic binding of [3H]DHT (Table 4), exhibit [3H]DHT serum binding capacity similar to adult males. Translocation of ligand from cytosol to the nucleus is a characteristic generally attributed to steroid receptors. Our results (Figs 2 and 3) suggest that the hepatic androgen binding protein possesses this characteristic. The activity was temperature dependent (Fig. 2) and was reduced to an apparent baseline level when the cytosol was prelabeled in the presence of lOO-fold excess unlabeled ligand (Fig. 3). The baseline activity may represent the interaction of [3H]DHT with nonspecific binding sites. The nature of the interaction of [3H]DHT with the nucleus is unknown, although it has been shown that there are sites in the nucleus of the liver cell which bind 13H]DHT with high affinity [29]. The interaction of [‘H]DHT labeled cytosol with denatured DNA-cellulose (Fig. 5) suggests that these protein-steroid complexes may have a functional role in c&o. It is interesting to note that the DNAcellulose elution pattern of labeled cytosol from female rat liver lacks two major peaks which are found with cytosol from males (Fig. 6a). Furthermore, liver cytosol from hypophysectomized male rats, which are marginally responsive to testosterone treatment, also lacks two peaks (Fig. 6b). These observations indicate that in animals which possess high levels of androgen binding protein and are responsive to androgen treatment, the androgen binding component also has the ability to interact with DNA. By far the most important criterion of receptor action is the ability of the putative receptor to initiate and to correlate with a measurable biological response. As an indicator of androgen action in the liver, we have chosen the activity of the cytochrome P-450-dependent enzyme, ethylmorphine Ndemethylase. This enzymatic activity is reproducible and markedly sensitive to androgen stimulation. Although androgen stimulation is the predominant

250

RICHARD C. RUMBAUGH

Fraction

no

Fig. 6a. Binding of [‘H]DHT labeled liver cytosol from female rats to denatured DNA-cellulose. Cytosol from livers of female rats (7-9 mg protein/ml final concentration) was incubated with [‘H]DHT (40nM final) and passed over a column of denatured DNA-cellulose as described in the legend to Fig. 4.

/

/

/

500

\-c E

I

-

300

-

200

-

100

-

ct

L 10

20

30

I

I

40

Fracl~on

I

025

0

50

60

70

with concentrations of hepatic cytosolic androgen binding which are already elevated or which are also induced by testosterone treatment. Table 1 indicates that males have greater androgen binding and EMD activity than either females or hypophysectomized males. The data in Table 2 demonstrate that hypophysectomized animals, which do not respond to testosterone treatment, also lack high concentrations of androgen binding protein. Table 3 indicates that sex steroids which affect rates of oxidative metabolism also affect the amount of specific androgen binding by the cytosol with estradiol reducing and testosterone increasing both parameters identically. Although it is not yet known how the sex steroids affect drug metabolism, the data in Tables l-3 suggest that their effects may be due, in part, to their actions on cytosolic binding of androgens. Neither is it known why hypophysectomized animals are unresponsive to androgen treatment. However, the data contained in Table 2 suggests the possibility that the lack of response might be due to the lack of a functional receptor for androgen and/or the ability of androgen to induce synthesis of its own receptor. The data in Tables l-3 also imply that the level of androgen binding protein in cytosol may be regulated by its own ligands. Sato and coworkers have previously noted that testosterone administration increases [30] and estradiol treatment decreases [31] the concentration of hepatic androgen binding protein. Our results (Table 3) confirm this observation. Roy has proposed a model[32] of hepatic cytosolic androgen binding in which the same binding protein contains different binding sites for androgen and estrogen. This model postulates that the androgen binding site is a stimulatory one, causing translocation to the nucleus and androgenic effects on protein synthesis including stimulation of the production of more androgen binding protein. The estradiol site in this scheme is inhibitory, prevents androgen action, and represses synthesis of androgen binding proteins. Our observations thus far are consistent with such a hypothesis. It is significant to note that in the experiment described in Table 4, the rats were prepubertal during the entire course of androgen treatment. These rats were found to have no cytosolic androgen binding activity and were also unresponsive to testosterone treatment. However, if these rats had been allowed to mature normally, they would have developed both DHT binding and responsiveness. This observation is strongly suggestive of a functional correlation between hepatic binding of DHT and the in vim responsiveness of the organ to treatment with androgen. Although it is unclear why immature animals did not respond to androgen treatment, our results indicate that there is a developmental component to the regulation of both androgen binding and responsiveness. We have not proven that testosterone acts on the liver through a classical steroid-receptor mechanism,

;f

40505

/’

I

80

no

Fig. 6b. Binding of [‘HIDHT labeled liver cytosol from hypophysectomized male rats to denatured DNA-cellulose. Cytosol from livers of hypox male rats (7-9 mg protein/ml final concentration) was incubated with [‘H]DHT (40 nM final) and passed over a column of denatured DNAcellulose as described in the legend to Fig. 4.

hormone effect on hepatic monooxygenases in male rats, these enzyme activities are subject to regulation by multiple hormones in addition to the gonadal steroids. Among the hormonal influences on cytochrome P-450-dependent enzymes are thyroid hormones glucocorticoids, and growth hormone [14]. Thus, observed differences with activities of EMD may also be due, in part, to alterations in the extra-genital hormonal environment. However, in the experiments which we have performed, the presence of androgen induction of drug metabolism correlates

et al.

Correlation of hepatic androgen binding and drug metabolism however, some evidence supports this hypothesis. We have shown that cytosolic proteins which bind androgen in the liver translocate the steroid to the nucleus. Protein bound steroid also has the ability to bind reversibly with DNA. In addition, androgenic induction of drug metabolism correlates with cytosolic androgen binding. Furthermore, androgen induction is blocked by either anti-androgens [33] or by the syndrome of testicular feminization in which androgen binding protein is absent [34]. In nearly all other steroid responsive tissues, receptor mechanisms have also been identified. High affinity androgen binding had been characterized in a number of extragenital tissues including skeletal muscle [7], skin [6] and brain [9]. Our finding of displaceable binding for DHT in liver coupled with evidence for nuclear translocation of steroid and subsequent interaction with DNA and the correlation between androgen binding androgen responsiveness is strongly suggestive that such a mechanism may be operative in liver also. Further studies must now be performed to partially purify the androgen binding component and to analyze the androgen-mediated effects on hepatic gene expression. REFERENCES

1. Jensen E. V. and Jacobson J. I.: Basic guides to the

mechanism of estrogen action. Recent Prog. Hormone Res. 18 (1962) 387-41 I. 2. Bruchovsky N. and Wilson J. D.: The conversion of testosterone to 5~-androstan-l7~~l-one by rat urostate in viva and in vitro. J. biol. Chem. 243 (1968) 20 (2-2021.

3. Jensen E. V.. Suvuki T.. Kawashima T.. Stumuf W. E.. Jungblet P. W.- and DeSombre E. R.: A *two-step 4.

5.

6.

7. 8.

9.

10.

11. 12.

mechanism for this interaction of estradiol with rat uterus. Proc. nam. Acad. Sci., U.S.A. 59 (1968) 632-638. O’Malley B. W., Toft D. 0. and Sherman M. R.: Progesterone-binding com~nents of chick oviduct. II. Nuclear components. J. biof. Chem. 246 (1971) 1117-1122. Jensen E. V., Numata M., Brecher P. I. and DeSombre E. R.: Hormone-receptor interaction as a guide to biochemical mechanism. In The Biochemistry of Steroid Hormone Acfion (Edited bv R. M. S. Smell& Academic Press, New York (1971) pp. 133-159. Eppenberger V. and Hsia S. L.: Binding of steroid hormones by the 105,000 xg supernatant fraction from homogenates of rat skin and variations during the hair cycle. J. biol. Chem. 274 (1972) 546335469. Kochakian C. D.: Definition of androgens and protein anabolic steroids. Pharm. Ther. 1 (1975) 149-177. Gehring U., Tomkins G. M. and Ohno S.: Effect of the androgen-insensitivity mutation on a cytoplasmic receptor for dihydrotestosterone. Nature 232 (1971) 106107. Sar M. and Stumpf W. E.: Cellular and subcellular localization of radioactivity in the rat pituitary after injection of l,2-3H-testosterone using dryautoradiography. Endocrinology 92, (1973) 631-635. Roy A. K., Milin B. S. and McMinn D. M.: Androgen receptor in rat liver: Hormonal and developmental regulation of the cytoplasmic receptor and its correlation with the androgen-dependent synthesis of a,,-globulin. Biochim. biophys. Acta 354 ( 1974) 2 13-232. Eisenfeld A. J., Aten R., Weinberger M., Hasalbacher G., Halpern K. and Krahoff L.: Estrogen receptor in the mammalian liver. Science 191 (1976) 862-865. Powell-Jones W., Nayfeh S. N. and Lucier G. W.:

251

Translocation of 4s and 5s forms of estrogen-receptors into rat liver nuclei in vitro. Biochem. biophys. Res.

13 Commun. 85 (1978) 167-173. Roy A. K.: Hormonal regulation of a,,-globulin. In ’ Biochemical Actions of Hormones (Edited by G. Litwack). Academic Press, New York (1979)pp. 48 1-517. 14. Colby H. D.: Regulation of hepatic drug and steroid metabolism by androgens and estrogens. In Advances in Sex Hormones (Edited by J. A. Thomas and R. Singhal). Urban and Schwarzenberg, Baltimore, Vol. 4 (1980) pp. 27-73. 1.5.Sato N., Ota M. and Obara K.: Presence of binding component~s) for testosterone in rat liver cytosol.

16, &doer. Jup. 27 (1980) 315-319.

Ota M., Sato N., Takahashi S. and Ono S.: Characteristics of the binding component for testosterone and androstenedione in rat liver cytosol. Endow. Jap. 27 (1980) 321-328.

f7. Yates F. E., Herbst A. L. and Urquhart

J.: Sex difference in rate of ring A reduction of delta-4-3-ketosteroids in vitro by rat liver. ~ndocrjno~ogy 63 (1980) 321-328.

18. Rumbaugh R. C., Clark J. C., McDaniel 0. S. and Lucier G. W.: Feminization of the concentrations of hepatic estrogen-binding proteins by ectopic pituitary. Endocrinology 112 (1985).1363-1369.

J9 Powell-Jones W.. Thomnson C., Raeford S. and Lucier ’ G. W.: Effect of gonahectomy on the ontogeny of estrogen-binding components in rat liver cytosol. Endocrinology 109 (1981) 628-636. 20. Omura T. and Sato R.: The carbon monoxide-binding pigment of liver microsomes. J. biol. Chem. 239 (1964) 237&2378. 21. Nash T.: The calorimetric estimation of formaldehyde by means of the Hantzch reaction. Bdochem. J. 55 (1953) 416421. 22. Rumbaugh R. C. and Colby H. D.: is growth hormone the pituitary feminizing factor mediating the actions of estradiol on hepatic drug and steroid metabolism. Endicrinology 17 ( 1980) 7 19-724.

23. Lowry 0. H., Rosebrough N. J., Farr A. L. and Randall R. J.: Protein measurement with the Folin phenol reagent. J. biol. Chem. 193 (1951) 265-275. 24. Kramer R. E., Greiner J. W., Rumbaugh R. C.. Sweeney T. D. and Colby H. D.: Requirement of the pituitary gland for gonadal hormone effects on hepatic drug metabolism in rats. J. pharmac. exp. Ther. 208 (1978) 19-23. 25. Testa B. and Jenner P.: Drug Metabolism-Chemicul and Biochemical Aspects. Marcel Dekker. New York

(1976) pp. 389-397. 26. Dickson R. B., Aten R. F. and Eisenfeld A. J.: An unusual sex steroid-binding protein in mature male rat liver cytosol. Endocrinology 103 (1978) 16361646.

27. Thompson C. L., Powell-Jones W. and Lucier G. W.: Sex differences in hepatic estrogen-binding proteins. Biochem. J. 194 (1981) l-8.

28. Eisenfeld A. J., Sheets C. S. and Aten R. F.: Androgen Receptor in Mammalian Liver. Abstract qf’ the 65th Annual Meetina of the Endocrine Societv. San Antionio, Texas (1983). kbstr. 137. 29. Lefebvre Y. A. and Morante S. J.: Binding of dihvdrotestosterone to a nuclear-envelope fraction from the male rat liver. Biochem. J. 202 (1982) 225-230. 30. Sato N., Takahashi S. and Ota M.: Effect of testosterone on the androgen receptor in the liver cytosol and nuclei of castrated male rats. Endocr. Res. Commun. 7 (1980) 241-252. 31. Sato N.. Ota M. and Takahashi S.: Estradiol treatment reduces a cytosol androgen binding protein in male rat liver. Experientia 36 (1980) 8777878. 32. Roy A. K., Chatterjee B. and Deshpande A. K.:

252

RICHARDC. RUMBAUGHet al.

Hormone-dependent expression of a,,-globulin gene in rat liver. In Gene Regulation by Steroid Hormones (Edited by A. K. Roy and J. H. Clard). Springer, New York (1978)pp. 230-246. 33. Brown T. R., Green F. E. and Bardin C. W.: Androgen receptor dependent and independent activities of testos-

terone on hepatic microsomal drug metabolism. Endocrinology 99 (1976) 1352-1362.

34. Brown T. R., Green F. E., Bullock L. P. and Bardin C. W.: Effect of the 7”m locus on the hepatic ethylmorphine N-demethylase system in mice. Endocrinolog> 103 (1978) 13741382.