Progestins can mimic, inhibit and potentiate the actions of androgens

Pharmac. Ther. Vol. 23, pp. 443 to 459, 1984 Printed in Great Britain. All rights reserved

0163-7258/84 $0.00 +0.50 Copyright © 1984 Pergamon Press Ltd

Specialist Subject Editors: J. CLARKand B. M. MARKAVER1CH

PROGESTINS

CAN MIMIC, INHIBIT AND POTENTIATE THE ACTIONS OF ANDROGENS

C. WAYNE BARDIN, TERRY BROWN, VELI V. ISO~IAA and OLLI A. J,~NNE The Population Council, 1230 York Avenue, New York, New York 10021, U.S.A.

1. INTRODUCTION Progesterone and synthetic progestins, like other classes of steroids which stimulate the reproductive tract, are defined according to their biological activities. Among the most important of these are effects on the endometrium which permit implantation and subsequent maintenance of pregnancy. In addition to their effects on the uterus, some progestational steroids are known to stimulate the growth of the male reproductive tract and to virilize the external genitalia of the female fetus (Bullock and Bardin, 1977). In this respect, progestins behave like androgens, and compounds structurally related to 19-nortestosterone are, in general, more androgenic than those that are derivatives of progesterone. In addition to their inherent androgenic potential, progestins may also potentiate and inhibit the actions of testosterone (Bullock et al., 1975; Gupta et al., 1978; Mowszowicz et al., 1974). These modifying effects of progestins have been termed their synandrogenic and antiandrogenic actions which, like the androgenic effects of this class of steroids, depend on steroid structure and the responsiveness of individual tissues. Present evidence suggests that progestins may require a functional androgen receptor to initiate their androgenic, synandrogenic, and antiandrogenic actions (Bullock et al., 1978; Mowszowicz et al., 1974). This report reviews some of the experiments that have defined how progestins mimic or modify androgen action through involvement of steroid metabolic and receptor mechanisms. The mouse kidney, submaxillary gland, and prostateseminal vesicle have been used in many of these studies. The mouse is of particular interest since there are a number of well-defined genetic mutants in this species in which the actions of androgens and progestins are modified. 2. THE EFFECT OF ANDROGENS AND PROGESTINS ON SPECIFIC KIDNEY PROTEINS The androgen-induced hypertrophy of the cells of the Bowman's capsule and of the proximal convoluted tubule of the mouse is associated with an increase in many kidney constituents and has been summarized in recent reviews (Bardin et al., 1978; Bardin and Catterall, 1981). Of the androgen-responsive renal proteins, fl-glucuronidase and ornithine decarboxylase (ODC) are perhaps the best studied (Bardin et al., 1973; Bardin et al., 1978; Pajunen et al., 1982; Isomaa et al., 1983). All androgens tested to date stimulate the activities of these enzymes, and the androgen receptor is necessary for these effects (Bardin et al., 1973; Pajunen et al., 1982). These increases in enzyme activities are preceded by rises in mRNA and enzyme synthesis (J/inne et al., 1984; Paigen et al., 1979). Although these enzymes comprise much less than 1% of total kidney protein when fully induced, they are useful for the study of hormone action inasmuch as androgens produce 20- to 600-fold increases in a specific activity. In addition, several genetic loci have been identified that modify the androgen-induced response of fl-glucuronidase and ODC genes (Bardin et al., 1978; Swank et al., 1978; Bullock, 1983). Recent studies have indicated that ODC exhibits a rapid response to androgen treatment. For example, this enzyme increases rapidly within 443

BOMT Flutamide

Others 6ct-bromo- 17fl-hydroxy- 17~t-methyl-4-oxo-5~t-androstan-3-one 3-trifluoromethyl-4-nitro-isobutyranilide

0 0

375 155 75

i'50 2500 1 2500 100 0

very potent 0 2000

Progestational

0 0

+ + + + +

0 0 + + + 0 + + + + 0 + 0 0

+ + + + +

0 0 0

+ + + + + 0 0 + + + 0 0 + 0 0

Antiandrogenic

Relative biological activities* Androgenic

0 0

+ + +

+ ++ 0 + + + + ++ 0 0 + + 0

Synandrogenic

*Relative progestational activities are from Boris and DeMartino (1971), Smith et al. (1974) and Kontula et al. (1975), progesterone = 100. The relative androgenic, antiandrogenic and synandrogenic activities are from Mowszowicz et al. (1974) and Gupta et al. (1978). A + + + response was the maximal seen for any progestin listed; + was a response that could just be distinguished fron~ 0 usually with 1 dose of steroid; 0 indicates no response at the doses tested. (From Bardin et al., 1978.)

d-norgestrel Lynestrenol Norethynodrel

Cyproterone acetate Cyproterone Medroxyprogesterone acetate (MPA) 17ct-hydroxy-6a-methylprogesterone 6~t-methylprogesterone 17~t-acetoxyprogesterone 17ct-hydroxyprogesterone Megestrol acetate Progesterone caproate Hydrocortisone

Trivial name

C-18 Progestins d- 13fl-ethyl- 17~t-ethynyl, 17fl-hydroxy-4-gonen-3-one 17ct-ethynyl-17fl-hydroxy-4-estren-3-one 17a-ethynyl-17fl-hydroxy-5(10)-estren-3-one

17a-hydroxy-4-pregnene-3,20-dione 17a-acetoxy-6~t-methyl-6-dehydro-4-pregnene-3,20-dione 17~t-caproxy-4-pregnene-3,20-dione 11fl, 17~t,21-trihydroxy-4-pregnene-3,20-dione

17a-acetoxy-4-pregnene-3,20-dione

C-21 Steroids 17~t-acetoxy-6-chloro-l,2ct-methylene-4,6-pregnadiene-2,30-dione 17ct-hydroxy-6-chloro- 1,2ct-methylene-4,6-pregnadiene-3,20-dione 17ct-acetoxy-6ct-methyl-4-pregnene-3,20-dione 17~t-hydroxy-6a-methyl-4-pregnene-3,20-dione 6a -methyl-4-pregnene-3,20-dione

Structural name

Steroid

TABLE 1. The Relative Androgenic, Antiandrogenic and Synandrogenic Activities o f Progestational and Nonprogestational Agents

z e~

Progestins and the action of androgens 35 4 ~-

I

i

I

I

71.

//

3O

I

o

/

x E

445

/

60

/

~x

2...

5

70

50

2o

0

0

0

30 ~ ' ~

20

5

.~

10

._J

I

0.1

I

0.3

I

0.5

I

10

II

I

2.0

STEROID DOSE (mg/day)

F1G. 1. The effect of testosterone or medroxyprogesterone acetate (MPA) on fl-glucuronidase activity per gram wet weight and on the relative rate of fl-glucuronidase synthesis in mouse kidney. The five animals in each group were treated for six days. Mean 4- SEM (from G u p t a et al., 1981).

hours after testosterone administration, whereas days are required for fl-glucuronidase (Pajunen et al., 1982). A series of experiments by Mowszowicz (Mowszowicz and Bardin, 1974; Mowszowicz et al., 1974; Bullock, 1983) indicated that certain progestins, such as medroxyprogesterone acetate (MPA), also stimulate fl-glucuronidase, ODC, and other renal proteins. These observations prompted an examination of a series of steroids to determine the relationship between steroid structure and the ability to stimulate these proteins. The studies on fl-glucuronidase (Table 1) may be summarized as follows: progesterone itself was inactive, but was agonistic (androgenic) following the addition of a 6~-methyl group. The activity of this latter steroid (6~-methylprogesterone) was further modified by the addition of functional groups at the 17a position; a 17a-hydroxyl group inhibited, whereas a 17a-acetoxy group increased, the androgenic activity of the 6or-substituted progestin. The introduction of a double bond at the C-6 position of MPA also reduced the androgenicity of this progestin. In addition to the C21 steroids, C18 progestins were also androgenic. Besides fl-glucuronidase and ODC, the androgenic activities of progestins have been shown on other kidney proteins, including alcohol dehydrogenase, 3-keto reductase, and the T proteins found in kidney membrane fractions (Bardin et al., 1978; Gupta et al., 1978; Mowszowicz and Bardin, 1974; L i n e t al., 1978). In order to investigate how progestins increase renal proteins, a procedure was used which measures in vivo incorporation of [3H]amino acid into immunoprecipitable fl-glucuronidase (Gupta et al., 1981). This assay was then used to demonstrate that MPA increased fl-glucuronidase activity by increasing the rate of synthesis of new enzyme molecules. The results of these studies are summarized in Fig. 1. These observations are consistent with the hypothesis that the androgenic action of progestins like that of testosterone requires gene activation and protein synthesis. 3. THE SYNANDROGENIC AND ANTIANDROGENIC ACTIONS OF PROGESTINS ON MOUSE KIDNEY Cyproterone acetate (CypAc) is known to be a potent progestin in the female reproductive tract and an effective antiandrogen in the male (Neumann et al., 1970). The effect of simultaneous administration of testosterone and CypAc on fl-glucuronidase activity in kidney of female mice is shown in Fig. 2. When the dose of CypAc was the same or up to 2-5 times higher than the dose of testosterone, androgen action was potentiated J.P.T. 23/2--1

446

C.W. BARDINet al.

~ 2000

7,

d.,

o12~

o'.s

0%

,Io

//

,~,

CYPROTERONE ACETATE ( r a g / d ) + TESTOSTERONE 0.1 mg/d

FIG. 2. The syn- and antiandrogenic actions of cyproterone acetate on renal fl-glucuronidase activity of female mice. Testosterone treatment (0.1 mg/day for 6 days) increased fl-glucuronidase activity 2-fold in this study (Horizontal shaded bar shows the mean _+SEM). When this treatment was combined with various doses of cyproterone acetate, the response was first potentiated and then inhibited (solid line). Cyproterone acetate alone produced no effect on fl-glucuronidase activity. The arrow indicates the enzyme activity in untreated animals (from Mowszowicz et al., 1974). (Mowszowicz et al., 1974). However, when increasing amounts of CypAc were given with a constant dose of testosterone, an antiandrogenic effect was observed (Fig. 2). A similar pattern of response was observed when 5~-dihydrotestosterone was used instead of testosterone (Gupta et al., 1978). It is of note that CypAc alone had no androgenic activity on either the kidney or the other androgen-responsive tissues of the mouse such as the submaxillary gland or male reproductive tract. In an experiment similar to that shown in Fig. 2, M P A was given to female mice in place of CypAc. That is, a fixed dose of testosterone was administered alone and with various doses of M P A from 1 to 10 times that of the androgen; other groups of mice were treated with M P A alone. A low dose of M P A was inactive when administered alone but was synandrogenic when given with an equal mass of testosterone. Higher doses of this progestin mimicked testosterone when given alone (as noted above), and potentiated the androgen action when administered with testosterone. These observations indicated that both androgenic (MPA) and antiandrogenic (CypAc) progestins can potentiate the actions of testosterone on mouse kidney when the progestin:androgen ratio is appropriate (Bardin et al., 1978). A series of steroids was investigated to determine whether the synandrogenic property of progestins relates to steroid structure. Several different analogs of M P A were tested (Table 1). M a n y compounds with the 6or-methyl substitution showed synandrogenic activity. Although these observations suggested the importance of 6~-substitution, progesterone caproate and several other progestins without the 6~-methyl group were also synandrogenic. It may be significant, however, that the synandrogenic response of these latter steroids was produced only with very large doses. As a consequence, the mechanism by which they produce synergism may be different from those of the 6~t-methyl series which synergise at lower progestin:androgen ratios (Gupta et al., 1978; Bardin et al., 1978). In view of the fact that antiandrogenic progestational steroids such as CypAc were found to have synandrogenic activity, it was pertinent to investigate the actions of nonprogestational antiandrogens. Two steroidal (cyproterone and BOMT) and one nonsteroidal (flutamide) compounds were studied (Table 1). Flutamide and BOMT had

Progestins and the action of androgens

447

no effect on kidney fl-glucuronidase when administered alone, whereas cyproterone reduced the activity of this enzyme by 25~o. None of these agents demonstrated synandrogenic activity when administered with testosterone, although all the compounds exhibited antiandrogenic activities on the male reproductive tract and the mouse kidney. This indicated that the synandrogenic and antiandrogenic activities are not strictly related. Although all but one of the androgenic progestins tested enhanced testosterone action to some extent, three synandrogens were not androgenic. Furthermore, all synandrogenic agents tested were potent progestins, but the reverse was not true. The results of the studies summarized in Table 1 indicate that the synandrogenic activity was best correlated with progestins which were substituted in the 6e-position (Bardin et al., 1978). The results of these and other studies suggest that the standard requirements may be quite limited while those for an antiandrogen are quite diverse (Starka et al., 1982; Yates and King, 1981; Brown et al., 1981c). The potentiating effects of progestins, like their agonistic actions, were related to an increase in new enzyme molecules. Further studies did, in fact, indicate that MPA potentiated the action of testosterone at all doses tested when the rate of fl-glucuronidase synthesis was used as an end point. Similarly, low doses of CypAc potentiated and high doses inhibited the testosterone stimulated rate of fl-glucuronidase synthesis (Gupta et al., 1981). This later experiment confirmed the conclusions from the study shown in Fig. 1 based upon the specific activity of the enzyme. When the synandrogenic effect of progestins was first reported, it was proposed that this activity might be exerted on several androgen-responsive proteins in kidney. To date, however, the only ones which show this response are fl-glucuronidase and 3-keto reductase. This is in marked contrast to the androgenic and antiandrogenic actions of progestins which are manifest on most androgen-responsive proteins (Bardin et al., 1978). 4. INTERACTION OF A N D R O G E N S AND PROGESTINS WITH THE fl-GLUCURONIDASE LOCUS As noted above, one of the best studied and most important androgen-induced proteins in the kidney is fl-glucuronidase. A genetic analysis of fl-glucuronidase activity in several strains of mice led to the identification of the complex locus, Gus. The structural gene, Gus-s, is the source of both the lysosomal and microsomal glucuronidases. The gene is located near the distal end of chromosome 5, and the mutations at this locus code for a thermolabile enzyme (Gus-sh) and two electrophoretic variants of fl-glucuronidase (Gus-s ~ and Gus-s b) (Swank et al., 1978). Mouse strains generally fall into two groups according to their temporal and quantitative responses to androgens. In some, fl-glucuronidase activity increases rapidly to a high plateau, and in others the response is less rapid and of lesser magnitude. These different responses are defined by the regulatory locus, Gus-r, within the Gus complex. The A/J and the C57BL/6J mice are typical of high and low inducing strains, respectively (Paigen et al., 1979). Strain A/J is homozygous for the allele, Gus-r a, and its glucuronidase can be induced to a level several times higher than in C57BL/6J, which is homozygous for the Gus-r b allele as shown in Fig. 3. The regulatory locus acts in a cis manner on the structural gene (Lusis et al., 1980) and, on the basis of breeding studies, is near or within the Gus-s structural gene (Swank et al., 1978). Previous studies have shown that Gus-r determines the rate of enzyme synthesis from Gus-s during androgen treatment (Swank et al., 1978). In view of the regulatory role of Gus-r on the androgen induction of fl-glucuronidase, it was pertinent to determine if other classes of fl-glucuronidase inducers would be recognized by the Gus-r a and Gus-r ° alleles. Figure 3 shows the response of A/J and C57BL/6J mice treated for varying periods of time with MPA (Bardin et al., 1978). As was the case with testosterone, this progestin produced a greater stimulation of fl-glucuronidase activity in animals carrying the Gus-r a allele. It is significant, however, that MPA was not as potent as testosterone in either of the strains (Fig. 3). These studies demonstrate that both the steroid and the Gus-r locus determine

448

C.W. BARDIN et al. i

40,000 z~ E

i

i

q

TESTOSTERONE ----- MEDROXYPROGESTERONE ACETATE

30,000

~.~

/

A/J

~ 20,000

z g

O =0 lO,O00

~/

/

, , ~ "

IC57 BLACK/6J

q,..c~

~~___~

,,,-"-~, - f - - -

,

,

,

,

0

7

I0

15

21

2 4

DAYSOF TREATMENT FlG. 3. The effect of testosterone and medroxyprogesterone acetate on fl-glucuronidase activity in the kidneys of female mice. A/J mice (O, ©) are homozygous for the Gus-r a allele and C57BL/6J animals ( & , / k ) are homozygous for the Gus-# allele (from Bardin et al., 1978).

the rate and magnitude of the steroid-induced response from the fl-glucuronidase gene. By analogy with other steroid responsive systems, it seems likely that Gus-r regulates glucuronidase synthesis in part by affecting the synthesis of its mRNA. Experiments using frog oocytes to measure m R N A levels indicate that an increase in the rate of enzyme synthesis is preceded by an increase in fl-glucuronidase m R N A activity (Paigen et al., 1979). However, the extent to which the appearance of new enzyme molecules in response to testosterone and progestin treatment reflects transcriptional, as opposed to posttranscriptional, control remains to be established. The influence of the Gus-r locus on the synandrogenic and antiandrogenic actions of progestins was also investigated. When administered with testosterone, low doses of CypAc produced a synandrogenic response in animals carrying the Gus-r" allele. By contrast, little or no synandrogenic response was noted in animals carrying the Gus-r b allele that were treated with a comparable dose. However, when testosterone and CypAc were administered in a 1 : 10 ratio, androgen action was inhibited in animals carrying both Gus-r alleles. It is of note that all the studies on the synandrogenic effects described above (Table 1) were performed in mice with the Gus-r" allele. In summary, progestins may mimic or modify the action of androgens on the mouse kidney. The synandrogenic effects are most easily demonstrated using fl-glucuronidase as an end point in animals with the Gus-r a, but not the Gus-r b allele; this effect is not readily observed on most other androgen responsive renal proteins. By contrast, the antiandrogenic and androgenic actions of progestins have been observed on almost all kidney proteins examined to date (Bardin and Catterall, 1981). In view of this, we have reviewed how some of these varied effects of progestins are manifest in other tissues. 5. T H E E F F E C T S OF P R O G E S T I N S ON O T H E R A N D R O G E N RESPONSIVE ORGANS The mouse submaxillary glands, like the kidneys, are dimorphic structures in male and female animals. In females and castrate males, testosterone increases the activity of several enzymes, as well as the concentration of epidermal and nerve growth factors ( E G F and N G F ) in this organ (Bullock et al., 1975). Of the progestins studied, MPA caused the greatest androgenic response when given alone, with large doses producing a 40-fold increase in E G F concentration. Progesterone and megestrol acetate also had weak androgenic activities. CypAc had no activity when administered alone and acted as a potent antiandrogen when given with testosterone. The only synandrogenic response

Progestins and the action of androgens

449

elicited was with a high dose of progesterone caproate. This is in striking contrast to the mouse kidney in which 100-fold lower doses of a variety of progestins, such as MPA, CypAc and megestrol acetate, were synergistic with testosterone (Bullock et al., 1975). In the rat, progesterone stimulates preputial gland growth by an androgen receptordependent mechanism (Bardin et al., 1973). When the action of several progestins was tested using mouse preputial gland weight as the end point (Mowszowicz et al., 1974), the response was similar to that in the mouse submaxillary gland (Bullock et al., 1975). Of the progestational agents tested, only MPA had a significant effect on preputial gland weight in the absence of testosterone. When progestins were administered with testosterone, only high doses of progesterone caproate were synandrogenic. High doses of CypAc and megestrol acetate were antiandrogenic but the synandrogenic activity produced by low doses of these steroids on kidney was not observed in the preputial gland. Thus, the action of progestins on the preputial gland is strikingly different from that on the kidney (Bardin et al., 1973; Mowszowicz et al., 1974). Progestins are known to be potent inducers of hepatic microsomal enzymes. The effect of several of these steroids on mouse liver was studied to determine their inherent potency and ability to modify the biological activity of testosterone. When increased hepatic ethylmorphine demethylase activity and cytochrome P450content were used as end points, the biological potencies of progestins per se on the liver were ranked as follows: CypAc > progesterone > MPA > progesterone caproate --- controls (Brown et al., 1977b). The induced alterations of these parameters were, therefore, unrelated to the reported progestational or androgenic activities of these steroids on the uterus and male reproductive tract. A similar conclusion was reached when hepatic weight and microsomal protein contents were used as end points. When various doses of these progestins were administered with testosterone, the effects were additive. No synergistic or antagonistic actions of progestins were noted. This is in contrast to their actions on the kidney, submaxillary, and preputial glands (Brown et al., 1977a). It is of note that the nonprogestational antiandrogen, flutamide, was able to inhibit some of the actions of testosterone on the liver (Brown et al., 1976). Since progestins alone are better inducers of hepatic proteins than testosterone, the ability to demonstrate their modifying effect on the action of androgens will be difficult (Brown et al., 1978). The effects of progestins on the male reproductive tract have been studied by many investigators (for review see Bardin et al., 1978; Bardin and Catterall, 1981; Bullock and Bardin, 1977). As a consequence, the ability of these steroids to mimic the action of androgens is widely recognized. When progestins are assayed in the adult castrate rat, progesterone and its derivatives have only minimal androgenic effects, whereas compounds structurally related to 19-nortestosterone stimulate the male reproductive tract. By contrast, both progesterone and 19-nortestosterone derivatives masculinize the external genitalia of female fetuses. Some agents, such as MPA, are as potent as testosterone propionate on the fetus, whereas megestrol acetate, which is structurally related to MPA, has much less androgenic activity and then only on selected parts of the reproductive tract (Salander and Tisell, 1976; Tisell and Salander, 1975). In addition to these structure-function differences, there are species dependent variations in responsiveness. This point is particularly well illustrated by studies indicating that CypAc virilizes the external genitalia of guinea-pigs but not the rat. The antiandrogenic actions of progestins such as CypAc are evident to a variable extent on the reproductive tract and external genitalia of many species (Neumann et al., 1970). The synandrogenic action of this class of steroids is much more difficult to demonstrate, however. When an experiment similar to that shown in Fig. 2 was performed in castrated male mice, an antiandrogenic without a synandrogenic effect was observed on the weight of prostate-seminal vesicles (Mowszowicz et al., 1974). By contrast, Coffey and associates noted a marked increase in prostatic DNA synthesis over that expected for testosterone alone when CypAc was administered to rats several hours before testosterone; when CypAc was given simultaneously with testosterone it was inhibitory (Coffey, 1974). Similarly, CypAc also potentiated the androgen-induced carnitine accumulation by

450

C . W . BARD1Net al.

epididymis (Bohmer and Hansson, 1975). In addition, low doses of megestrol acetate, which have no effect on their own, acted synergistically with estradiol in stimulating epithelial growth in the prostate (Salander and Tisell, 1976). In organ culture of rat prostate, CypAc appeared to have a different effect: when used alone in low doses this steroid stimulated growth, whereas higher doses were inhibitory; when CypAc was administered with testosterone, marked hyperplasia occurred (Lasnitzki and Robel, 1969). The reason for the discrepancy between the in vivo and in vitro results is not known. It is clear, however, that progestins can have similar effects on the male reproductive tract as on mouse kidney, depending on the conditions used. 6. EVIDENCE THAT THE ANDROGEN RECEPTOR IS INVOLVED IN THE ACTION OF PROGESTINS 6.1. THE ACTION OF PROGESTINSIN THE ANDROGEY-INsENSlTIVE MOUSE (TFM/Y) Tfm/Y mice have inherited end organ insensitivity to androgens. Several studies indicate that mice with this gene defect have defective androgen receptors (Bullock and Bardin, 1972, 1974). As a consequence, testosterone does not stimulate androgen responsive tissues as it does in normal littermates. Since MPA and other androgenic progestins stimulate many of the same proteins as testosterone, it was pertinent to examine the effect of this class of steroids in Tfm/Y mice. Several experiments demonstrated that progestins do not increase renal or submaxillary gland proteins in the absence of a functional androgen receptor (Bullock et al., 1975; Mowszowicz et al., 1974). A similar conclusion was made from studies on androgen-insensitive rats (Bardin et al., 1973) and in humans with testicular feminization (Perez-Palacios et al., 1981). These observations suggested that some of the androgenic actions of progestins can be mediated by the androgen receptor. As noted above, progestins are stimulators of many hepatic proteins, and their potencies were not correlated with their androgenic or progestational activities. In Tfm/Y mice, the dose-dependent increase in some proteins was not distinguishable from that of normal littermates with a full complement of androgen receptor (Brown et al., 1976, 1977b). These observations suggest that many progestin-mediated effects on liver are not mediated via the androgen receptor as is the case in other organs. 6.2. IN V t v o AND I N V I T R O BINDING OF PROGESTINS TO ANDROGEN RECEPTORS The inference that the androgen rather than the progesterone receptor mediates the responses of progestins on some androgen sensitive tissues was tested in a series of in vivo experiments in rats and mice. The specific nuclear uptake of [3H]MPA or [3H]androgens in prostate-seminal vesicle, kidney, and submaxillary glands was more readily blocked by the simultaneous administration of androgens than by MPA or other progestins (Brown et al., 1979a; Bullock et al., 1978). In addition, there was no evidence of nuclear uptake of either [3H]MPA or [3H]androgens in Tfm/Y mice (Bullock et al., 1978). These observations imply that the androgen (rather than progestin) receptors are involved in the nuclear retention of [3H]MPA. The hypothesis that androgens and progestins may share the same receptor was also supported by ultracentrifugal analysis of progestin and androgen binding to mouse kidney cytosol (Bullock et al., 1978). [3H]MPA and [3H]testosterone were both bound with high affinity to macromolecules with sedimentation coefficients of 8S. This binding was displaced by nonradioactive MPA or T in a dose-dependent fashion and by a 100-fold excess of other progestins and estradiol, but not dexamethasone. There was no detectable binding of [3H]MPA or [3H]testosterone to 8S components in kidney cytosol from Tfm/Y mice. These observations, along with the biological studies mentioned above, strongly support the conclusion that the androgenic, like the antiandrogenic, effects of progestins are mediated via the androgen receptor. The significance of these experiments was increased by the inability to demonstrate a binding protein in the mouse kidney with the same steroid specificity as the progesterone receptor from the uterus (unpublished, O. Jfinne). The ability of progesterone, CypAc, and other progestins to bind to the androgen

Progestins and the action of androgens

451

receptor has been demonstrated by many investigators. In fact, progestin binding is now accepted as one of the features that distinguishes androgen receptor from extracellular androgen binding proteins such as ABP and TeBG. The relative binding affinities of various progestins for [3H]testosterone binding sites on the androgen receptor in mouse kidney cytosol was studied using equilibrium binding analysis. The apparent affinities of several progestins for the androgen receptor were not directly correlated with the magnitude or direction (agonist or antiagonist) of their effects on renal fl-glucuronidase (Bullock et al., 1978; Brown et al., 1979a). In addition, when kidney slices were incubated with [3H]testosterone in the presence or absence of unlabeled progestins, the relative binding affinities to the nuclear androgen receptor differed from those obtained with the cytosol assay mentioned above (Brown et al., 1979a). The results from the in vitro nuclear uptake assay were similar to in vivo studies of the relative potency of progestins to inhibit nuclear testosterone uptake by mouse kidney. Results from these latter experiments suggest that the nuclear uptake of progestins relative to testosterone correlates with their androgenic but not their antiandrogenic and synandrogenic actions. From these considerations we conclude that the multiple actions of progestins in mouse kidney cannot be accurately predicted from a simple analysis of their affinities to androgen receptors. As a consequence, the binding curves were examined with a more complex model. 6.3. A STERIC-ALLOSTERIC MODEL FOR ANDROGEN-PROGESTIN INTERACTION Unlike the classic allosteric model for interaction between identical or dissimilar ligands, the steric-allosteric model includes the competitive binding of similar ligands to the same site, as well as their effects on the putative equilibrium between active and inactive states of the receptor (Bullock et al., 1978). The observation that the relative affinities for the receptor are unrelated to the direction of their physiological effects was compared with the predictions of the model for combined steric and allosteric interactions. According to this model, the response elicited by a mixture of steroids is determined by the combined effects on an equilibrium between an active and inactive state of the receptor. Each steroid influences this allosteric equilibrium according to the ratio of its dissociation constants for the two states. The observation that androgen binding to mouse kidney cytosol was similarly inhibited by progestins with disparate physiological effects (androgenic, antiandrogenic, and synandrogenic or undetectable) was consistent with the prediction of the steric-allosteric model. It should be noted that the computations from the model were based on the assumption that steroid receptors have two binding sites. The steric-allosteric model predicted that a monomeric receptor could also mediate a range of responses produced by progestins. In contrast, a cooperative response to one steroid or a synergistic response to a mixture of steroids is consistent with the model only if two or more binding sites are involved in the allosteric transition. Thus, the synandrogenic action of certain progestins may involve either a coordinated transition between states of a steroid receptor with two or more subunits, or a cooperative interation at a later stage of the hormone response such as a concerted effect of several steroid receptor complexes on the genome. Although the individual effect of progestins on androgen responsive tissue (androgenic, antiandrogenic, and synandrogenic activity) could be predicted for the steric-allosteric model, it was difficult for the model to predict how a synandrogenic effect could be produced by one dose of a given progestin, and an antiandrogenic effect by a higher dose. These considerations led to the conclusion that the interaction of the steroid receptor complex with the nucleus, as well as steroid binding to the receptor, must be important in determining the biological response observed (Bullock et al., 1978). 7. THE RELATIONSHIP OF ANDROGEN RECEPTOR KINETICS TO BIOLOGICAL RESPONSE In order to examine the change in the concentrations of androgen receptor in the cytoplasm and nucleus following testosterone treatment, it was necessary to develop assays

C.W. BARDIN et al.

452

A

E] 0 . 3 m g T [ ] 1.0 m g T 3.0 mg T

o

oj

1

2

6

1B

24

TIME AFTER TESTOSTERONE

Z

o_

i'1 IIrl ~1 •

B

I-E t-Z Z 0 0

48

72

120

216

A D M I N I S T R A T I O N (H)

o.3 mg T 1 mg T 3 rng T lOmgT

200 Z

1oo 0 Z

0.5

1

2

4

6

TIME AFTER TESTOSTERONE

12

18

24

48

I N J E C T I O N (h)

FIG. 4. (A) Long term effects of a single dose of testosterone on cytosol androgen receptor (AR) concentrations in the kidney of castrated male mice. The bars show mean + S.E. for at least 6 animals per group. The shaded area represents the receptor concentrations (mean + S.E. range) in vehicle-treated control animals. The values significantly different from the control level are indicated by asterisks. T, testosterone. (B) Nuclear androgen receptor concentrations in the mouse kidney after administration of testosterone. Two days following castration male mice of the NCS strain were given single intraperitoneal doses of testosterone (T) and killed at timed intervals. Nuclear androgen receptors were measured from kidneys pooled from 4-6 animals. The crosshatched area shows the mean + S.E. range for receptor values in vehicle-treated control animals, while each bar represents mean values for 2-3 separate experiments (a receptor concentration of 100 fmol/mg of DNA equals 360 receptor/nucleus) (from Pajunen et al., 1982).

capable of quantifying receptor in the presence of endogenous ligand. Attempts to do this in the past have failed due to receptor loss during nuclear isolation, and the inability to quantitatively extract the androgen receptor from the nucleus. Recent studies have shown that sodium molybdate can stabilize the receptor in cytosol (Wright et al., 1981) and that the recovery of nuclear receptor can be enhanced using hexylene glycol for nuclear isolation and pyridoxal phosphate to extract the receptor (Isomaa et al., 1982). With these innovations, it was possible to study the quantitative loss and replenishment of androgen receptor in various cellular compartments, an approach which has been used to study estrogen receptors for years (Anderson et al., 1972; Clark and Markaverich, 1982). 7.1. ANDROGEN RECEPTOR KINETICS IN MOUSE KIDNEY

Following a single dose of testosterone, there was a dose-dependent decrease in cytosol androgen receptor concentrations with a nadir at 30-60min (Pajunen et al., 1982). Thereafter, cytosolic receptor content rose and returned to control values between 4-8 hr after low and intermediate doses of testosterone, whereas a high dose brought about prolonged decline in the cytosol androgen receptor content, which returned to the pretreatment level between 72-120 hr after steroid injection (Fig. 4a). In the same animals,

Progestins and the action of androgens

453

>,. I,-

'~150

°

~

~'

•

~

•

.

o

;,

26 's

TIME AFTER TESTOSTERONE (H)

< E ul

Z nO I

I

I

i

I

I

i

I

I

I

0

1

2

3

4

5

6

7

8

9

TIME AFTER TESTOSTERONE INJECTION (DAYS)

FIG. 5. Mouse renal ornithine decarboxylase activity after testosterone administration. Groups of male NCS mice (castrated 2 days) were treated with single doses of testosterone (A, 0.3 mg; I-q, 1 mg; II, 3 mg; O, 10mg) and killed at indicated time intervals. Each symbol with a vertical bar shows the mean + S.E. for a given experimental group comprising at least 6 animals. Asterisks indicate significantly different values from the control ornithine decarboxylase (ODC) level which was 0.83_ 0.21 nmol of CO 2 formed/mg of protein in 30 min. Inset, changes in renal ornithine decarboxylase activity during the first 48 hr after testosterone administration. Note the 5-fold expanded scale on the ordinate (from Pajunen et al., 1982).

the concentration of renal nuclear androgen receptors increased after testosterone administration and was 5 times higher than in the vehicle-treated control animals at 30 min (Fig. 4b). Interestingly, at this time nuclear receptor accumulation was not dose-dependent. The androgen receptor concentrations in renal nuclei reached their maxima within one hour, at which time there were about 600 receptors/cell in animals treated with the smaller doses and about 800-850 receptors/cell with the larger doses. After the first hour, the duration of nuclear androgen receptor residence was related to the dose of testosterone in that the smaller the steroid dose, the faster receptor concentration decline to the level of vehicle-treated animals (Fig. 4b). In mice that received low and intermediate doses of testosterone the nuclear androgen receptors returned to control levels within 6-12 hr. However, in animals that were treated with a high dose, a 3-fold increased receptor concentration was still present at 120 hr posttreatment (Fig. 4b). The prolonged residence of androgen receptors in nuclei was thus associated with slow replenishment of cytosol receptor after a single dose of testosterone (Fig. 4). 7.2. THE RELATION OF NUCLEAR ANDROGEN RECEPTORS TO BIOLOGICAL RESPONSE ODC activity increased rapidly in mouse kidney after a single dose of testosterone (Fig. 5). The stimulation occurred in a dose-dependent manner until 8 hr of steroid administration (Fig. 5, inset); thereafter, a high dose of testosterone produced an increase in ODC activity that was proportionally much greater than those achieved with the lower doses. It is of note, however, that the smaller doses of testosterone increased ODC activity in a dose-dependent manner during all the time intervals with the peak activities at 12 hr after treatment. The high dose of testosterone brought about a striking and long-lasting stimulation of ODC activity which increased 150-fold at 18 hr and only returned to normal after 5-9 days (Fig. 5). In addition, the lag period required to detect increased ODC activity was shorter after the large dose of testosterone (Pajunen et al., 1982). The effects of testosterone on #-glucuronidase activity were similar to those on ODC, but the time required to achieve the response was markedly different (Pajunen et al., 1982). The small doses of testosterone brought about a slight but dose-dependent stimulation of renal #-glucuronidase activity which was not observed until the 4th posttreatment day (Fig. 6). Interestingly, this dose-related effect was observed even though the receptor-

454

C . W . BARDIN et al.

>~-

50 ~r

o< ~40 z .-~ 20 =

,o

i

0

~ltlllr~~r~llPill

~r

IllilIIII~IUlli

0.25 1 2 3 4 5 9 TIME AFTERTESTOSTERONEINJECTION(DAYS)

FIG. 6. The effect of testosterone administration on renal fl-glucuronidase activity in castrated male mice. Groups of male mice of NCS strain (castrated 2 days) were administered a single intraperitoneal dose of testosterone (13, 0.3 mg; IN, 1 mg; N, 3mg; II, 10mg) and killed at indicated time intervals. Each bar represents the mean ___S.E. for at least 6 animals/experimental group. The cross-hatched area shows the mean + S.E. range for the vehicle-treated control animals. Asterisks indicate significant differences between the experimental and the corresponding control groups (from Pajunen et al., 1982).

initiated events occurred 4 days earlier. By contrast, a large dose elicited an earlier and larger increase in enzyme activity. These observations suggest that the high and sustained concentration of nuclear androgen receptor was associated not only with greater stimulation of ODC and fl-glucuronidase activities but also a shorter lag in the time required for these increases to be observed. These observations on both enzymes suggest that the magnitude of the response and the lag period are related to the nuclear residence time of the androgen receptor. 8. THE BINDING OF ANTIANDROGENS TO THE NUCLEAR ANDROGEN RECEPTOR As discussed previously, in vivo and in vitro studies with [3H]MPA demonstrated that this steroid binds to the cytoplasmic and nuclear forms of the androgen receptors in multiple organs of mouse and rat (Bullock et al., 1978; Brown et al., 1979b). These experiments also showed that MPA rather than a metabolite was bound to the androgen receptor; that binding was competitive with androgens; and that the nuclear uptake of [3H]MPA in vivo was low compared to that of [3H]testosterone, as might be expected from its low androgenic activity. These observations were confirmed with the nuclear exchange assay mentioned above (Isomaa et al., 1982). Therefore, MPA, a steroid with androgenic and synandrogenic activities behaved like a weak androgen in this receptor binding system. These observations raised the question as to how agonists were distinguished by the receptor as being different from antagonists. One postulated way of distinguishing an androgen from an antiandrogen is the relatively short biological half-life (rapid rate of dissociation) of the cytosol receptor-antagonist complex (Raynaud et al., 1979; 1980). It was originally proposed that steroid agonists and antagonists could be discriminated by using two different conditions for measuring the relative binding of compounds for androgen receptor: a short-term incubation to achieve estimates for the relative rates of association and a long-term incubation to determine the rate of dissociation of the complex. Steroid antagonists typically associate with kinetics similar to those of the agonists, but dissociate at faster rates. As a consequence, the antiandrogens should have a higher relative binding affinity during the short-term than the long-term incubation, while androgens should exhibit roughly identical binding affinities under the two conditions (Raynaud et al., 1979; 1980). By using the above approach, Raynaud et al. (1979) concluded that both steroidal and non-steroidal antiandrogens elicit their action through forming a rapidly-dissociating complex with the cytosol androgen

Progestins and the action of androgens

455

receptor. By contrast, Wakeling et al. (1981) using a similar assay principle found that, although the relative binding affinities of various antiandrogens were, as expected, markedly lower in long-term than short-term incubations, there was no obvious correlation between antiandrogenic potency in vivo and relative receptor binding activity in vitro. Furthermore, these and other investigators were also unable to distinguish unequivocally between androgen agonists and antagonists on the basis of the above studies (Bullock et al., 1978; Brown et al., 1979a). The findings that many of the antiandrogens do not show any agonistic activity in vivo suggested (i), that the compounds are not translocated at all to target cell nuclei or (ii), that the nuclear androgen receptor-antiandrogen complexes are not biologically active, i.e. the antagonists hold the receptors in a state which is different from that of the receptor-androgen complex as indicated in the steric-allosteric model discussed above. Some recent findings are pertinent in this respect. Firstly, the experiments with steroidal and non-steroidal antiandrogens, some of which were devoid of agonistic activity, have indicated that these compounds are capable of translocating low levels of cytosol androgen receptors to nuclei in vivo (Isomaa et al., 1982, unpublished observations). Studies on androgen receptor dynamics with exchange assays have also shown that, when given concomitantly with, or prior to, testosterone or 5~-dihydrotestosterone, non-steroidal antiandrogens prevent androgen-elicited translocation of the expected quantity of cytosol androgen receptors to nuclei in mouse kidney (J~inne et al., unpublished observations), as well as rat prostate and epididymis (Callaway et al., 1982; Tezon et al., 1982) Even more interestingly, a single dose of a non-steroidal antiandrogen, flutamide, is capable of depleting nuclear androgen receptors in mouse kidney in animals implanted with testosterone-releasing rods, despite the continuous presence of the implants during the experiment. The depletion was not, however, complete; rather, a level of nuclear receptor was reached which was comparable to that achieved when the antiandrogen was given alone. Secondly, some very potent new non-steroidal antiandrogens have also been shown to possess partial agonistic activity (Wakeling et al., 1981) similar to that of non-steroidal antiestrogens (Katzenellenbogen et al., 1979; Clark and Markaverich, 1982). It thus appears that some antiandrogens seem to be capable of initiating androgen action in a manner similar to that of the androgens; the mechanisms leading to a complete or partial abortion of the action in the subsequent steps remain to be elucidated. 9. INTERACTION OF PROGESTINS WITH OTHER BINDING PROTEINS Part of the action of antiandrogens may be mediated via binding proteins other than the androgen receptor. This was suggested by the studies on 6ct-methylprogesterone, a progestin with weak androgenic activity that potentiates and inhibits the action of testosterone under appropriate conditions (Brown et al., 1981a). 6~-Methylprogesterone given alone is able to translocate androgen receptors to renal nuclei in vivo (Gupta et al., 1978; Isomaa et al., 1982) thus explaining its androgenic properties (Brown et al., 1981a). In addition to the androgen receptor, nuclear uptake of 6~-methylprogesterone and its 20~-hydroxylated metabolite seemed to occur by way of another receptor or steroidbinding protein in the mouse kidney (Brown et al., 1981b). On the basis of these findings, it is tempting to suggest that in some tissues nuclear accumulation of this progestin through an additional binding-protein could be responsible for the inhibition of the action of concurrently administered testosterone (Brown et al., 1979b). As noted above, 6~-methylprogesterone also has a synandrogenic effect on the mouse kidney, but not on the prostate-seminal vesicle. Interestingly, androgens, progestins, and glucocorticoids will potentiate the non-androgen receptor mediated uptake of 6~-methylprogesterone and its metabolite in the former but not the latter organ (Fig. 7). It is possible that the facilitated uptake of this progestin could be related to its ability to potentiate the action of androgens. It is of note that the androgen receptor is required for testosterone to stimulate increased nuclear uptake of 6~-methylprogesterone even though the increase in progestin binding is not to the receptor (Brown et al., 1981a,b). The ability

456

C.W. BARDINet

al.

3 H-6ot-M ETHYLPROGESTERONE (iv) KIDNEY 14

+ T

(sc)

+ 6 M P (sc)

£3 ~12 E ro I 10

o

E

D

4

u; _J

2¸

<= z

0

-90

-60

-30

-15

0

-90

-60

-30

-15

Control

0 Control

PRETREA]~MENT TIME (m,n)

FIG. 7. In vivo uptake of [3H]6~-methylprogesterone(6MP) by mouse kidney nuclei following testosterone (T) or 6MP pretreatment. Each bar indicates the total steroid uptake of nuclei from the kidneyspooled from three animals which were killed 30 min after an i.v. injectionof[3H]steroid (time 0). Animals designated by the open bars receivedthe indicated non-radioactivesteroids (T 'or 6MP) at the various times before receiving the [3H]steroid.Animals designated by the closed bar received only [3H]steroid(from Brown et al., 1981b).

of progestins to modify androgen action was previously believed to be strictly a function of androgen receptor. The present observations suggest that the situation is much more complex since in v i v o components, including steroid transport, metabolism, and multiple binding sites, may be required to describe the complex androgen-progestin interaction.

10. S U M M A R Y A N D C O N C L U S I O N S There is an extensive background on the androgen responsiveness of the mouse kidney which can be demonstrated histologically by hypertrophy of the Bowman's capsule and the proximal convoluted tubule. Although androgens increase many renal proteins, /3-glucuronidase and ODC are distinguished by exquisite genetic regulation of the magnitude of the response induced by testosterone. Both the qualitative and quantitative expression of the genes for these enzymes are strain specific, and are dependent upon regulatory alleles. Ornithine decarboxylase is of particular interest since the response of this enzyme is rapid compared to that of #-glucuronidase. Recent studies using a newly developed androgen receptor assay have demonstrated that the duration of retention of the androgen receptor complex in the nucleus correlates with the magnitude of the androgenic response. Progestins can mimic, inhibit, or potentiate the action of androgens. These responses have been termed the androgenic, antiandrogenic and synandrogenic actions of progestins, respectively. The androgenic and antiandrogenic action of this class of steroids are manifest on many tissues and on many endpoints within a given organ. These effects are believed to involve an early step(s) of androgen action which is common to all sensitive tissues. Results to date suggests that this early step involves the androgen receptor. By contrast, the synandrogenic action of progestins is limited in that it is not observed on all tissues, and not even on all endpoints within a single organ. In the mouse kidney, the synandrogenic actions of progestins have been most extensively studied on /~-glucuronidase. With this enzyme this unusual response to progestins can be demonstrated only in mice which carry the Gus-r a allele. This observation suggests that the potentiating action of progestins on/3-glucuronidase is manifest directly on the Gus gene complex. It is not certain at this time whether a similar mechanism is involved in the potentiation of androgen action on other organs such as the prostate.

Progestins and the action of androgens

457

The androgenic action of progestins is believed to be similar to that of other androgens. Androgenic progestins such as MPA bind to the androgen receptors and translocate them to nuclei. This is followed by a dose dependent increase of proteins similar to what is observed after testosterone administration. In addition, the regulatory genes which modulate androgen action have the same effect on the androgenic effect of progestins. The fact that the potency of progestins such as MPA is less than that of testosterone is believed to relate in part to their lower affinity for the androgen receptors. To date, the mechanism of action of antiandrogens is less certain. These compounds bind androgen receptor with lesser affinity than testosterone, and in competitive binding assays cannot be distinguished from androgenic progestins. It is assumed therefore, that they bind to the androgen receptor and hold it in a state whose conformation is different from that of the testosterone receptor complex. The main reason for this assumption is that antiandrogens usually have little if any androgenic activity in relation to the amount of receptor that they bind and transfer to the nucleus. Preliminary data suggests that the antiandrogen-receptor complex can be retained in the nucleus but never to the level attained by the androgen-receptor complex. At present it is not possible to postulate a single mechanism that will explain the synergistic action of progestins on all tissues. This is, in part, due to the fact that this action is limited to a few endpoints. Another reason is because the synergistic action of progestins is highly dose- and time-dependent. In addition the progestins which produce this effect may also be androgenic, antiandrogenic, or both. REFERENCES ANDERSON, J. N., CLARK,J. H. and PECK, E. J. (1972) Oestrogen and nuclear binding sites: Determination of specific sites by [3H]oestradiol exchange. Biochem. J. 126: 561-567. BARDIN, C. W., BROWN, T. R., MILLS, N. C., GUPTA, C. and BULLOCK, L. P. (1978) The regulation of the fl-glucuronidase gene by androgens and progestins. Biol. Reprod. l& 74-83. BARDIN, C. W., BULLOCK, L. P., SHERINS, R., MowszowIcz, I. and BLACKBURN,W. R. (1973) Part II. Androgen metabolism and mechanism of action in male pseudohermaphroditism: A study of testicular feminization. Recent Prog. Horm. Res. 29: 65-109. BARDIN, C. W. and CATTERALL, .,I F. (1981) Testosterone: A major determinant of extragenital sexual dimorphism. Science 211: 1285-1294. BOHMER, T. and HANSSON, V. (1975) Androgen-dependent accumulation of carnitine by rat epididymis after injection of [3Hlbutyrobetaine in vivo. Molec. cell. Endocr. 3:103-115. BORlS, A. and DEMARTINO, L. (197 l) The utilization of uterine weight as an adjunct to histology in the evaluation of progestational steroids. Steroidologica 2: 57-64. BROWN, T. R., BARD1N, C. W. and GREENE, F. E. (1977a) Hormonal control of cytochrome P-450-dependent ethylmorphine N-demethylase activity of the mouse. In: Pharmacology o f Steroid Contraceptive Drugs, pp. 327-343, GARATTINI, S. and BERENDES, H. W. (eds) Raven Press, New York. BROWN, T. R., BULLOCK, L. and BARDIN, C. W. (1979a) In vitro and in vivo binding of progestins to the androgen receptor of mouse kidney. Correlation with biological activities. Endocrinology 105: 1281-1287. BROWN, T. R., BULLOCK, L. P. and BARD1N, C. W. (1979b) In vivo metabolism and binding of 6ct-methylprogesterone: A progestin with anti-androgenic and synandrogenic activities. In: Steroid Hormone Receptor Systems, pp. 269-280, LEAVITT, W. W. and CLARK,J. H. (eds) Plenum, New York. BROWN, T. R., BULLOCK, L. P. and BARDIN, C. W. (1981a) The biological actions and metabolism of 6ct-methylprogesterone: A progestin which mimics and modifies the effects of testosterone. Endocrinology 109: 1814-1820. BROWN, T. R., BULLOCK, L. P. and BARDIN, C. W. (1981b) The nuclear uptake of [3H]6ct-methylprogesterone and its 20ct-hydroxy metabolite: The requirement for multiple receptors. Endocrinology 109: 1821-1829. BROWN, T. R., GREENE, F. E. and BARDIN, C. W. (1976) Androgen receptor dependent and independent activities of testosterone on hepatic microsomal drug metabolism. Endocrinology 99: 1353-1362. BROWN, T. R., GREENE, F'. E. and BARDIN, C. W. (1977b) The additive effects of progestins on testosteronestimulated hepatic ethylmorphine metabolism and cytochrome P-450 content. Steroids 30: 805-814. BROWN, T. R., G'REENE, F. E., BULLOCK, L. P. and BARDIN, C. W. (1978) Effect of the Tfm locus on the hepatic ethylmorphine N-demethylase system in mice. Endocrinology 103:1374-1382. BROWN, Z. R., ROTHWELL, S. W., SULTAN,C. and MIGEON, C. J. (1981c) Inhibition of androgen binding in human foreskin fibroblasts by antiandrogens. Steroids 37: 635-648. BULLOCK, L. P. (1983) Androgen and progestin stimulation of ornithine decarboxylase activity in the mouse kidney. Endocrinology 112: 1903-1909. BULLOCK, L. P. and BARDIN, C. W. (1972) Androgen receptors in testicular feminization. J. clin. Endocr. Metab. 35: 935-937. BULLOCK, L. P. and BARDIN, C. W. 0974) Androgen receptors in mouse kidney: A study of male, female and androgen-insensitive (Tfm/Y) mice. Endocrinology 94: 746-756. BULLOCK, L. P. and BARDIN, C. W. (1977) Androgenic, synandrogenic, and antiandrogenic actions of progestins. Ann. N.Y. Acad. Sci. 286: 321-330.

458

C . W . BARDIN et al.

BULLOCK, L. P., BARDIN,C. W. and SHERMAN,M. R. (1978) Androgenic, antiandrogenic, and synandrogenic actions of progestins: Role of steric and aUosteric interactions with androgen receptors. Endocrinology 103: 1768-1782. BULLOCK,L. P., BARTHE,P. L., MOWSZOWICZ,l., ORTH, D. N. and BARDIN,C. W. (1975) The effect of progestins on submaxillary gland epidermal growth factor: Demonstration of androgenic, synandrogenic and antiandrogenic actions. Endocrinology 97: 189-195. CALLAWAY,T. W., BRUCHOVSKY,N., RENNIE, P. S. and COMEAU,T. (1982) Mechanism of action of androgens and antiandrogens: Effects of antiandrogens on translocation of cytoplasmic androgen receptor and nuclear abundance of dihydrotestosterone. The Prostate 3: 599-610, CLARK, J. H. and MARKAVERICH,B. M. (1982) Agonistic-antagonistic properties of clomiphene: a review. Pharmac. Ther. 15: 467-519. COFFEY, D. S. (1974) The effects of androgens on DNA and RNA synthesis in sex accessory tissue. In: Male Accessory Sex Organs. Structure and Function in Mammals, pp. 307-328, BRANDES,D. (eds) Academic Press, New York. GUPTA, C., BULLOCK, L. P. and BARDIN, C. W. (1978) Further studies on the androgenic, anti-androgenic, and synandrogenic actions of progestins. Endocrinology 102: 736-744. GUPTA, C., MILLS, N. C., BULLOCK,L. P. and BARD1N,C. W. (1981) The effects of testosterone and progestins on the rate of fl-glucuronidase synthesis in mouse kidney as studied with an immunoprecipitation assay. Int. J. Androl. 4: 342-354. ISOMAA,V., PAJUNEN,A. E. I., BARD1N,C. W. and J,~NNE,O. A. (1982) Nuclear androgen receptors in the mouse kidney: validation of a new assay. Endocrinology 111: 833-843. ISOMAA, V. V., PAJUNEN, A. E. I., BARDIN, C. W. and J~,NNE, O. A. (1983) Ornithine decarboxylase in mouse kidney: Purification, characterization, and radioimmunological determination of the enzyme protein. J. biol. Chem. 258: 6735-6740. J~.NNE, O. A., KONTULA, K. K., ISOMAA, V. V., TORKKELI, T. K. and BARDIN, C. W. (1984) Androgen receptor-dependent regulation of ornithine decarboxylase gene expression in mouse kidney. In: Steroid Hormone Receptors: Structure and Function, GUSTAESSON, J.-A. and ERIKSSON, H. (eds) Elsevier/North Holland Biomedical Press, Amsterdam (in press). KATZENELLENBOGEN, B. S., BHAKOO, H. S., FERGUSON, E. R., LAN, N. C., TATEE, T., TSAI, T. L. and KATZENELLENBOGEN, J. A. (1979) Estrogen and antiestrogen action in reproductive tissues and tumors. Recent Prog. Horm. Res. 35: 259-300. KONTULA, K., JXNNE, O., VIHKO, R., DEJAGER, E., DEVlSSER, J. and ZEELEN, F. (1975) Progesterone binding proteins: In vitro binding and biological activity of different steroidal ligands. Acta endocr. 78: 574--592. LASNITZKI, I. and ROBEL, P. (1969) Effects of cyproterone on the rat prostate gland grown in organ culture. Adv. Biosci. 3: 175-184. LIN, Y. C., BULLOCK, L. P., BARDIN, C. W. and JACOa, S. T. (1978) Effect of medroxyprogesterone acetate and testosterone on solubilized RNA polymerases and chromatin template activity in kidney from normal and androgen-insensitive (Tfm/Y) mice. Biochemistry 17: 4833-4838. LUSIS, A. J., CHAPMAN,V. M,, HERBSTMAN,C. and PAIGEN, K. 0980) Quantitation of cis versus trans regulation of mouse fl-glucuronidase. J. biol. Chem. 255: 8959-8962. MowszowIez, I. and BARDIN, C. W. (1974) In vitro androgen metabolism in mouse kidney: high 3-keto-reductase (3~t-hydroxysteroid dehydrogenase) activity relative to 5ct-reductase. Steroids 23: 793-807. MowszowIez, I., BIEBER,D. E., CHUNG, K. W., BULLOCK, L. P. and BARDIN, C. W. (1974) Synandrogenic and antiandrogenic effect of progestins: Comparison with nonprogestational antiandrogens. Endocrinology 95: 1589-1599. NEUMANN, F., BERSWORDT-WALLRABE, R., ELGER, W., STEINBECK, H., HAHN, J. D. and KRAMER, M. (1970) Aspects of androgen-dependent events as studied by antiandrogens. Recent Prog. Horm. Res. 26: 337-410. PAIGEN, K., LAaARCA, C. and WATSON, G. (1979) A regulatory locus for mouse fl-glucuronidase induction, gur, controls messenger RNA activity. Science 203: 554-556. PAJUNEN, A. E. I., ISOMAA,V. V., J~.NNE, O. A. and BARDIN, C. W. (1982) Androgenic regulation of ornithine decarboxylase activity in mouse kidney and its relationship to changes in cytosol and nuclear androgen receptor concentrations. J. biol. Chem. 257: 8190-8198. PEREZ-PALACIOS,G., CHAVEZ, B. ESCOBAR,N., VILCHIS, F., LARREA,F., LINCE, M. and PEREZ, A. E. (1981) Mechanism of action of contraceptive synthetic progestins. J. Steroid Biochem. 15: 125-130. RAVNAUD, J.-P., BONNE, C., BOUTON, M.-M., LAGACE, L. and LABmE, E. (1979) Action of a non-steroidal antiandrogen, RU 23908, in peripheral and central tissues. J. Steroid Biochem. 11: 93-99. RAYNAUD, J.-P., BOUTON, M. M., MOGUILEWSKY, M., OJASOO, T., PH1LIBERT, D., BECK, G., LABRIE, F. and MORNON, J. (1980) Steroid hormone receptors and pharmacology. J. Steroid Biochem. 12: 143-157. SALANDER,H. and TISELL, L.-E. (1976) Effects of megestrol on oestradiol induced growth of the prostatic lobes and the seminal vesicles in castrated rats. Acta endocr. 82: 213-224. SMITH, H. E,, SMITH, R. G., TUFT, D. O., NEERGAARD,J. R., BURROWS,E. P. and O'MALLEY, B. W. (1974) Binding of steroids to progesterone receptor proteins in chick oviduct and human uterus. J. biol. Chem. 249: 5924-5932. STARKA, L.~ HAMPL, R., KASAL, A. and KOHOUT, L. 0982) Androgen receptor binding and antiandrogenic activity of some 4,5-secoandrostanes and ring B cyclopropanoandrostanes. J. Steroid Biochem. 17: 331-334. SWANK, R. T., PAIGEN, K., DAVEY, R., CHAPMAN,V., LABARCA,C., WATSON,G., GANSCHOW,R., BRANDT,E. J. and NOVAK, E. (1978) Genetic regulation of mammalian glucuronidase. Recent Prog. Horm. Res. 34: 401-436. TEZON, J. G., VAZQUEZ, M. H. and BLAQUIER,J. (1982) Androgen-controlled subcellular distribution of its receptor in the rat epididymis: 5ct-Dihydrotestosterone-induced translocation is blocked by antiandrogens. Endocrinology 111: 2039-2045. TISELL, L.-E. and SALANDER, H. (1975) Androgenic properties and adrenal depressant activity of megestrol acetate observed in castrated male rats. Acta endocr. 78: 316-324.

Progestins and the action of androgens

459

WAKELING,A. E., FURR, B. J. A., GLEN, A. T. and HUCHES,L. R. (1981) Receptor binding and biological activity of steroidal and nonsteroidal antiandrogens. J. Steroid Biochem. 15: 355-359. WRIGHT, W. W., CHAN, K. C. and BARD1N,C. W. (1981) Characterization of the stabilizing effect of sodium molybdate on the androgen receptor present in mouse kidney. Endocrinology 108: 2210-2216. YATES. J. and KXNG, R. J. B. (1981) Antiandrogen effects on androgen-responsive mammary turnout cells in culture. J. Steroid Biochem. 14: 819-822.