Molecular and Cellular Endocrinology

3 (1975) 83-101. Q North-Holland

Publ. Comp.

ANDROGEN-BINDING PROTEINS IN RAT EPIDIDYMIS: PROPERTIES OF A CYTOPLASMIC RECEPTOR FOR ANDROGEN SIMILAR TO THE ANDROGEN RECEPTOR IN VENTRAL PROSTATE AND DIFFERENT FROM ANDROGEN-BINDING PROTEIN (ABP) Donald

E. Martin Departments University

J. TINDALL, Vidar HANSSON*, RITZEN**, Shihadeh N. NAYFEH

of Pediatrics of North

Received 11 November

and Biochemistry Carolina

1974

School

William S. MCLEAN, and Frank S. FRENCH

and The Laboratories of Medicine,

Chapel

for

Reproductive

Hill, N.C.

27514,

Biology, U.S.A.

Accepted 24 December 1974

The cytoplasmic receptor (CR) in rat epididymal 105,000 g supernatant was separated from the androgen-binding protein (ABP) by gel electrophoresis following labeling with [l ,2,6,7-3H]testosterone in vivo. ABP disappeared from epididymal supernatants after castration ot hypophysectomy, while CR remained unchanged. CR was evenly distributed between caput and cauda, while much more ABP was present in caput. Properties of CR in epididymis and prostate were similar and distinctly different from ABP. Binding to CR was destroyed by charcoal treatment (1 mg/mg protein) of supernatant at 0 “C for 6 h, heating at 50 “C for 30 min, or exposure to the sulthydryl blocking reagent, p-chloromercuriphenylsulfonate (1 mM) at 25 “C for 30 min, while binding to ABP was unaffected. The isoelectric pH of CR (5.8) was higher than that of ABP (4.6). Dissociation of radioactive Sa-dihydrotestosterone (DHT) from CR and nuclear receptors was extremely slow (half-time at 0 “C ~2 days), while dissociation from ABP was rapid (half-time at 0 “C- 6 min). Cyproterone acetate (250 mg/lOO g body weight) inhibited binding to CR both in epididymis and ventral prostate but did not affect binding to ABP. Nuclear uptake was inhibited by cyproterone to the same extent as

*V.H. was a USPHS Fogarty International Postdoctoral Fellow, 1973-74, Department of Pediatrics and The Laboratories for Reproductive Biology, University of North Carolina School of Medicine, Chapel Hill, N.C. 27514, U.S.A. Present address: Institute of Pathology, Rikshospitalet, Oslo 1, Norway. **Department of Pediatrics and Institute for Cell Research, Karolinska Institutet, Stockholm, Sweden. Abbreviations: DHT or Sa-dihydrotestosterone, 178-hydroxy-5a-androstan-3-one; cyproterone acetate, 6-chloro-17a-acetyl-1,2a-methylene-4,6-pregnadiene-3,2O-dione; androstanediol, 3a,178-dihydroxy-5a-androstane; androstenedione or n4A, 4-androstene-3,17 dione; androstanedione, 5a-androstane-3,17-dione; ABP, androgen-binding protein; CR, cytoplasmic receptor; PAGE, polyacrylamide gel electrophoresis.

D. J. Tidal/

84

binding

to CR,

dependent nuclear

indicating

of ABP.

fractions

that

nuclear

The time-course

was essentially

uptake

and binding

are dependent

of uptake

and binding

in epididymal

the same 1 day after bilateral

ABP were present or 8 days after castration the cytoplasmic androgen

receptor

receptor

Keywords:

for androgen

in ventral

prostate

androgen-binding androgens;

but different

proteins;

from

androgen

and

supernatant

inand

when both CR and

when CR alone was present.

in rat epididymis

testosterone;

castration

on CR

et al.

It is concluded

has properties

very similar

that

to the

ABP. receptor;

dihydrotestosterone;

epididymis:

prostate;

cyproterone.

Two androgen-binding components have been demonstrated in 105,000 g supernatants of epididymal homogenates using polyacrylamide gel electrophoresis (Ritzen et al., 1971). The faster moving component is the androgenbinding protein (ABP) which is formed in the testis (French and Ritzen, I973a b; Hansson et al., 1973a; Ritzen et al., 1973). ABP is secreted into the seminiferous tubular fluid and passes through the testicular efferent ducts into the lumen of epididymis where it exists as an extracellular binding protein. In the present study, it is shown that the slower moving androgen-binding component has properties similar to the cytoplasmic receptor (CR) in rat ventral prostate and different from ABP.

MATERIALS

AND

METHODS

Materials

[ 1,2,6,7- 3H]Testosterone

(spec. act. 9 1 Ci/mmol,

[I ,2- 3Hldihydrotestosterone

(spec. act. 44 Ci/mmol) and [4-‘4C]testosterone (spec. act. 59 Ci/mmol) were obtained from New England Nuclear Corp. The radiochemical purity wa; checked by thin-layer chromatography on silica gel GF-254 in benzenemethanol (9: 1) and on aluminum oxide in diethyl ether. Cyproterone acetate was obtained from Schering-AG., Berlin. Acrylamide, N,N,N’,N’-tetramethylethylenediamine and N,N’-methylene-bisacrylamide were obtained from Mann Research Laboratories. p-Chloromercuriphenylsulfonate was obtained from Sigma Chemical Co. and was dissolved in water as a 10 mM solution before mixing with supernatant. Dry charcoal (Norit-A) obtained from Matheson, Coleman and Bell was added to the supernatants in a concentration of I mg per mg of supernatant protein. Preparation

of’ animals

Sprague-Dawley

rats weighing

3.50-450 g were castrated

either

18-24 h or

Androgen-binding

8 days prior

85

proteins in rat epididymis

to the experiment.

Sprague-Dawley

rats hypophysectomized

at

35 days of age were obtained from Hormone Assay Laboratories, Chicago, Illinois. All rats were functionally hepatectomized and eviscerated by a modification of the procedure of Ingle (1949). Hypophysectomized rats were given 200 ug of cortisol and 5% dextrose in saline (5 ml) S.C. 1 h before the evisceration and functional hepatectomy. Advantages of functional hepatectomy for studies of this type have been described by Hotta and Chaikoff (1955) and Bruchovsky and Wilson (1968). [l,2-3H]Testosterone was injected i.v. in 0.25 ml of 15 ‘A ethanol in saline, and the rats were sacrificed 3 h thereafter. Castrate rats were given 5 y0 dextrose in saline (5 ml) S.C. immediately after evisceration and functional hepatectomy. Preparation of supernatants and nuclear extracts Epididymides and ventral prostates were minced and homogenized in 3 ml of Buffer A (50 mM Tris-HCI, pH 7.4 at 4 “C, 0.32 M sucrose, 3 mM MgCI,) per gram wet weight using an all glass Dual1 homogenizer. Homogenates were centrifuged at 600 g for 10 min, and the postnuclear supernatants centrifuged at 105,000 g for I h. All procedures were carried out at O-4 “C. Nuclei were purified from the 600 g pellets by resuspension in 50 mM Tris-HCI buffer containing 3 mM MgCl, and 2.2 M sucrose and centrifugation for 1 h at 60,000 g in a Spinco SW 27 rotor. Pellets were washed once with 5 volumes of Buffer A, once with Buffer A containing 0.25% Triton X-100, and twice more with Buffer A. The washed pellets were extracted by stirring in Buffer B (50 mM Tris-HCl buffer, pH 7.4 at 4 “C containing 0.4 M KCI and 1.5 mM EDTA) (2 x pellet volume) for 30 min at 0 “C. Extracts were centrifuged at 105,000 g for 30 min and the pellet discarded. Polyacrylamide gel electrophoresis (PAGE) A modification of the procedure described by Davis (1964) was used for the preparation of 5% polyacrylamide gels (either 10 x 60 mm or 5 x 50 mm) containing 10% glycerol (Ritzen et al., 1971). A stock of solution A was prepared by diluting 48 ml of 1 M HCI, 36.3 g ofTris and 0.46 ml of N,N,N’,N’tetramethylethylenediamine to 100 ml with water. A stock of solution B was prepared by diluting 30 g of acrylamide and 0.8 g of N,N’-methylene-bisacrylamide to 100 ml with water. Solution C (0.14% of ammonium persulfate in water) was prepared prior to each experiment. Gels were prepared by mixing 2.5 parts of stock solution A, 3.85 parts of stock solution B and 2 parts glycerol with 1.65 parts water. Equal parts of this solution were mixed with equal parts of solution C and placed at 4 “C to polymerize. Gels containing 3.25% acrylamide and 0.5% agarose were prepared by a

86

D. J. TindaNet al.

modification

of the method

of Dingman

was mixed with I part solution

et al. (1968). One part of solution

A

B and 0.8 part glycerol to make the acrylamide

solution. Agarose was dissolved in water (7.7 mg/ml) by refluxing at 100 “C with vigorous stirring, and the solution was cooled slowly to 40 “C. The agarose solution (5.2 parts) was combined with the acrylamide (2.8 parts), and ammonium persulfate was added to make 0.7 mg/ml. The solution was poured into cylindrical tubes (sealed at the bottom with parafilm) and covered immediately with a layer of water. The agarose formed a gel in 20-30 min at room temperature and tubes were then placed at 4 “C overnight before use. Using agarose to strengthen the gel matrix (Uriel, 1966), it is possible to lower the acrylamide concentration so that large macromolecules can enter the gel freely. Buffer (upper and lower) contained 0.6 % Tris and 2.9% glycine in water (pH 8.6 at 4 “C). Bromphenol blue was added to the upper buffer as an electrophoretic marker, and relative mobilities (R,) of binding proteins were calculated from the bromphenol blue band. Supernatant aliquots of 500 ~1 (large gels) and 100 ~1 (small gels) were layered directly over the running gels. Electrophoresis was run at 6-8 mA/cmZ (i.e., 5 mA/tube for large gels or l-2 mA/tube for small gels) in a 4 “C cold room with the tubes immersed in lower buffer at 0 “C. Following electrophoresis, the gels were sliced into 2.3-mm sections, the sections placed in counting vials and toluene scintillation fluid added. More than 98% of the radioactivity in the gels was extracted into the toluene after standing 12 h at room temperature. Gel electrofocusing

Electrofocusing

was performed

in gels containing

3.25%

acrylamide

and

0.5 % agarose by a modification of the method described by Wrigley (1968). Carrier ampholytes (pH 3-10) were purchased from LKB-Produkter AB, Bromma, Sweden. A concentrated gel solution was prepared by mixing 3 parts solution B with 0.3 part carrier ampholytes and 0.8 part catalyst solution (1% N,N,N’,N’-tetramethylethylenediamine in water). Acrylamide solution for 12 gels (5 x 50 mm) was prepared by mixing 2.5 ml of the concentrated gel solution with 4.5 ml water and 2.0 g of sucrose. A solution of agarose (7.7 mg/ml) was prepared as described and 13 ml mixed with the acrylamide solution. Ammonium persulfate (0.7 mg/ml) was added, and the solution poured immediately into glass tubes (5 x 75 mm) sealed at the bottom with parafilm. Water (35 “C) was quickly layered on top to achieve a flat surface. When polymerization was complete, the water layer was removed and tubes were filled with ampholyte solution. The upper electrode solution was 0.1 M NaOH and the lower solution 0.2% H,SO,. Samples were layered under the protective layer of ampholytes and run at 0.2-0.3 watt/gel until a maximum

Androgen-binding

voltage

proteins

87

in rat epididymis

of 300 was reached.

Gels were sliced and each slice extracted

for 24 h

into 0.5 ml water. The pH at room temperature was determined in each vial using a Fisher microelectrode. Instagel (Packard Instrument Co.) 5 ml was added,

and the radioactivity

was counted.

Sucrose gradient centrifugation Aliquots of supernatant and nuclear extracts were applied to 4.4 ml continuous gradients of 5-20x (w/v) sucrose in Buffer B containing 10% glycerol. The gradients were centrifuged 49,000 rpm for 18 h at O-2 “C in a Spinco SW 50.1 rotor. Sedimentation coefficients of binding components were estimated by comparison with bovine serum albumin (S = 4.6). IdentiJication of Iabeled metabolites (a) Whole tissue A modification of the procedure of Stern and Eisenfeld (197 1) was used for the extraction of steroids from whole tissue. Epididymes were homogenized in 6 volumes of acetone, containing 3000 dpm [4-14C]testosterone for determination of recovery. The following unlabeled steroids (100 pg) were added as carriers : androstanediol, testosterone, DHT, androsterone, androstenedione and androstanedione. After standing 18 h at 4 “C, the extracts were passed through two layers of filter paper (Whatman No. 1) and the acetone evaporated under nitrogen to 1 ml. Steroids were extracted in 9 ml diethyl ether: water (2: 1) twice and chromatographed on silica gel plates in benzene: methanol (9: 1). Carrier steroids were detected by short wave ultraviolet light and iodine vapor and eluted separately in ethyl acetate. The radioactivity in each zone was measured by liquid scintillation counting. DHT was further separated from androsterone by thin-layer chromatography on aluminum oxide in diethyl ether, and identified by crystallization to constant specific activity. (b) Ident$cation qf protein-bound androgens Polyacrylamide eluted in 1 ml toluene overnight at room temperature. Aliquots

gel slices were were taken for

counting, and the toluene fractions containing radioactivity bound to CR were pooled. After addition of carrier steroids, toluene was evaporated under nitrogen at 50 “C, and the steroids were chromatographed on precoated thinlayer chromatography sheets of silica gel F-254 (0.25 mm) (E. Merck, Darmstadt) dichloromethane : diethyl ether (4: 1) and developed twice. Other analytical methods Radioactivity was measured in a Packard spectrometer. Scintillation cocktail was toluene

Tri-Carb containing

liquid scintillation 0.4 % 2,5-diphenyl-

88

D. J. Tidall

oxazole and concentrations

rt al.

0.01 % I ,4-bis-2-(4-methyl-%phenyloxazolyl)benzene. Protein in the supernatant fractions were measured by the method of

Lowry et al. (1951).

RESULTS Epididymal and prostate 105,000 g supernatants were labeled in vivo with [3H]testosterone I8 h after castration and analyzed by polyacrylamide gel electrophoresis. The electrophoretic mobility of the faster moving protein in 5 % polyacrylamide gels (R, = 0.57) corresponded to that of ABP (Ritzen et al., 1973), while the mobility of the slower moving protein (R, 0.38, epididymis) was similar to the androgen receptor in ventral prostate supernatant (fig. I) and is referred to as CR. Binding to both components was destroyed by incubation with the proteolytic enzyme, Pronase (I mg/ml), for 20 min at 4 “C. Binding was measured in epididymal supernatants labeled by injection of

EPIDIDYMIS

CR

PROSTATE

ABP

\

1

5 SLICE

IO NUMBER

15 SLICE

NUMBER

Fig. I. Polyacrylamide gel electrophoresis of androgen-binding components and prostatic 105,000 g supernatants. Rats were eviscerated and functionally 18 h after

castration

and injected

with

100 FCi [3H]testosterone.

injection of [“HItestosterone, the epididymis in 3 volumes of Buffer A. The homogenates nuclear

protein

were analyzed

hours

following

the

and prostates were removed and homogenized were centrifuged at 600 x to remove the crude

pellet and then at 105,000 g to obtain supernatant

Three

in epididymal hepatectomized

the cytosol

for binding

fraction.

by PAGE

Equal

concentrations

(IO 1’ 60 mm gels).

of

Androgetz-bitldiug proteitls in rat epididymis

3day

89

24 day HYPOX

HYPOX

SLICE

-+

CAPUT

-4

CAUDA

NUMBER

Fig. 2. Androgen-binding components in caput and cauda epididymis of immature rats 3 and 24 days after hypophysectomy. Rats 35 days of age were hypophysectomized and were eviscerated and functionally hepatectomized 3 and 24 days after hypophysectomy (hypox). Rats were injected with 50 WCi [3H]testosterone and killed 3 h later. Epididymes were divided into caput and cauda and homogenized in 50 mM Tris-HCI buffer (pH 7.4 at 4°C) containing 0.5 M sucrose and 3 mM MgC&. 105,OOOgsupernatants were analyzed by PAGE (5 x 50 mm gels). O---O, caput; 1,‘~--3, cauda.

[3H]testosterone 8 days following removal of the testis on one side and 1 day after removal of the testis on the opposite side. CR and ABP were both present 1 day following removal of the testis. However, on the contralateral side from which the testis had been removed 8 days earlier, CR remained after ABP had disappeared, indicating that CR is confined to the epididymis. CR also persisted in epididymal supernatants following hypophysectomy. At 3 days after hypophysectomy, CR remained Distribution

ABP was still present in epididymis

cytosol, but after 24 days only

(fig. 2).

qf CR and ABP in epididymis

CR was present in similar amounts in caput and cauda, both in the immature (3_5-day-old) (fig. 2) and mature rat (fig. 3). In contrast, ABP was present in much higher concentrations in caput than in cauda epididymis of mature rats (fig. 3) and was undetectable in cauda of the 35-day-old rat (fig. 2). The absence of binding of radioactive androgen to CR in the mature intact rat (fig. 3) was presumably due to saturation of binding sites with endogenous androgen.

90

D. J. Tindall et al.

adCASTRATE

INTACT ABP

CR

(u

8-

h * : ?

6-

? I

4-

$

-

capui --- cauda

2-

, 5 SLICE

,

IO NUMBER

I5

5

IO

I

I5

SLICE NUMBER

Fig. 3. Effect of castration on androgen binding to CR and ABP in supernatants of caput and cauda epididymis from mature rats. Left: Rats castrated 8 days before the experiment were injected daily with testosterone, 200 pg S.C. in ethanol, for 7 days after castration. Rats were eviscerated, functionally hepatectomized and sacrificed 3 h after injection of [“HItestosterone (60 BCi). 105,000 g supernatants were prepared in Buffer A and analyzed by PAGE (IO x 60 mm). Right: Rats castrated immediately before evisceration and funct~ol~al hepatectomy were injected with [3H]testosterone (60 pCi) and sacrificed 3 h later. Cytosols were prepared and analyzed by PAGE. -----, caput; - -.-- -- -, cauda.

Heat stability of binding to CR and ABP was analyzed in 105,000 g epididymal supernatants from rats castrated on one side for 8 days and on the opposite side for 1 day before labeling in vivo with [3H]testosterone (fig. 4). Comparison was made with the cytoplasmic receptor in supernatants of ventral prostate. Supernatants were incubated at 0, 25, 50 and 60 “C for 30 min, and binding was measured by PAGE. Neither binding to CR nor to ABP was influenced by heating up to 25 “C, but further increase in temperature to 50 “C caused complete destruction of binding to CR in epididymis and ventral prostate. ABP binding was not influenced by heating at 50 “C, but was destroyed at 60 “C (fig. 4). Treatment with the sul~ydryl blocking reagent, ~-chloromercuriphenylsulfonate, also had a differential effect on binding to CR and ABP. Labeled epididymal supernatants from rats hemicastrated as above were incubated with 1 mM p-chloromercuriphenylsulfonate and analyzed by PAGE (fig. 5). Binding to CR was eliminated in supernatants from the 8 day (fig. 5C) and I day castrate sides (fig. 5D). Although binding to ABP was reduced by p-chloro-

Androgen-binding

91

proteins in rat epididymis

EPIDIDYMIS

PROSTATE

5

IO 15 20

5

IO 15 20

SLICE

5

IO 15 20

5

IO 15 20

NUMBER

Fig. 4. Effect of heating on binding to CR and ABP in epididymis and on prostate CR. Upper: Epididymal 105,000 g supernatant : rats were hemicastrated and injected with testosterone, 200 ug daily for 6 days beginning on the day of castration. Seven days after hemicastration, the opposite testis was removed and 24 h later [3H]testosterone (125 uCi) was injected i.v. Homogenates were prepared in Buffer A and centrifuged at 105,000 g for I h. Aliquots of pooled epididymal 105,000 g supernatants were heated at 0, 25, 50 and 60 “C for 30 min, cooled to 0 “C, and centrifuged at 600 g for 15 min to remove protein precipitates. Aliquots were analyzed by PAGE (IO x 60 mm). C----O, I-day castrate (CR and ABP); O-- - -0, g-day castrate (CR). Lower: Prostate 105,000 g supernatant: rats 25 days old were hypophysectomized and 8 days later eviscerated, functionally hepatectomized and injected with 60 uCi [3H]testosterone. Three hours later, prostate supernatants were prepared, subjected to heat treatment and analyzed by PAGE.

mercuriphenylsulfonate

(fig. 5D),

a second,

faster

moving

peak

of bound

radioactivity appeared. This peak probably represents a smaller subunit of ABP which retains binding activity, and not a fragment of CR, since a similar peak was not observed on the contralateral side where no ABP was present (fig. 5C). Binding to prostate cytosol receptor, not shown here, was also destroyed by treatment with the sulfhydryl blocking reagent. CR and ABP exhibited diff‘erent stabilities when exposed to high concentrations of charcoal (1 mg/mg protein) for 6 h at 0 “C. After charcoal was

D. J. TidalI

9’

8 Day Castrate

SLICE Fig.

5. Effect

Rats were

fig. 4. Aliquots

and

of labeled

p-chloromercuriphenylsulfonate analyzed

1 Day Castrate

NUMBER

of p-chloromercuriphenylsulfonate

hemicastrated

by PAGE

(IO

injected

with

epididymal or with

rt ul.

on binding [3H]testosterone

105,000 buffer

to epididymal as described

g supernatant alone

(control)

? 60 mm gels). A: S-day castrate

C: R-day castrate p-chloromercuriphenylsulfonate.

D:

were

control.

for

CR

incubated 30 min

B: l-day

and ABP.

in the legend with at 25

castrate

I

to

mM

‘C and control.

1 day castrate p-chloromercuriphenyl-

sulfonate.

removed by centrifugation, the supernatants were equilibrated with 1.0 nM [3H]DHT and binding was analyzed by PAGE. Charcoal destroyed [3H]DHT binding to epididymal CR (fig. 6) as well as ventral prostate CR (fig. 6B). ABP, however, was unaffected by charcoal under these conditions and retained its [3H]DHT binding activity (fig. 6A). The isoelectric pH of CR was determined by electrofocusing on a pH 3-10 gradient. CR focused at pH 5.8 (fig. 7) which was identical to the focusing pH of prostate cytoplasmic receptor (Mainwaring, 1973). In contrast, ABP has been shown to focus as a symmetrical peak at pH 4.6-4.7 (Hansson, 1972). Androgen metabolites bound to Cl? Previous studies (Tindall et al., 1972) have demonstrated several labeled androgen metabolites in epididymal supernatant following injection of [3H]testosterone in vivo. In the nuclear fraction, however, only [3H]DHT is retained

Androgm-binding

93

proteins in rat epididymis

SLICE Fig.

6. Effect of charcoal

Supernatant after

fractions

castration

samples

extraction

from

of supernatants

epididymis

were treated

and

1.0 nM with

3.25%

gels containing

t3H]DHT,

0.5% agarose.

co:~trol

protein) and

e--O,

ABP adult

samples

in vitro. rats

18 h

were centrifuged

charcoal-treated

CR and ABP;

charcoal

from

for 6 h at 0 ‘C. Identical

for 30 min at 0 ‘C and

A: Epididymal

control;

to CR and

obtained

Charcoal-treated

from

equilibrated

on binding prostate

(I mg/mg

controls.

for 4 min at 7000 g. Aliquots

made

ventral

with dry charcoal

were kept at 0 “C as nontreated

repeatedly

NUMBER

analyzed

B: Prostate

samples

were

by PAGE CR;

in

3---c‘,

treated.

6

5 N l

h” X

Control

0 7

% ?

v) 2 a2

6, 5Ja

U

4

I

--____

5

Fig.

7. Elcctrofocusing

for 10 days,

injected

of epididymal

IO SLICE

CR. Thirty-five-day-old

with 50 HCi [3H]testosterone

supernatant was prepared and electrofocused agarose. An aliquot of supernatant was heated for 5 min before

layering

15 NUMBER

over the gels. O--O,

3

20

rats were hypophysectomized

and killed 3 h later.

Epididymal

105,000 g

in 3.25%, acrylamide gels containing 0.5% at 60 ‘C for 30 min and centrifuged at 7000 g control;

C---O,

heated;

- - -

-, PH.

94

D. J. Tindall et al.

for a prolonged

period of time (5 h). ABP has been shown

and DHT

relatively

M-‘,

with

respectively)

testosterone

high

in vitro

is present

affinity

(French

(K,

and

= 0.5

Ritzen,

in the epididymis

x

to bind testosterone

IO9 M-’

and

1.25 x

10”

1973a). Since very little t3H]-

3 h after the injection

(Tindall

et al.,

1972), it is difficult to determine from these experiments whether epididymal CR binds testosterone as well as DHT. Rennie and Bruchovsky (1972) have shown that ventral prostate receptor binds testosterone but has a much higher affinity for DHT. To identify [3H]testosterone metabolites bound to CR, supernatants were prepared from epididymis containing ABP (3 days hypophysectomized) and from epididymis completely free of ABP (24 days hypophysectomized). Rats were injected with [3H]testosterone and sacrificed 3 h after injection. Supernatants were prepared and analyzed by PAGE. Radioactivity bound to CR was eluted and the labeled androgens identified. DHT was essentially the only androgen bound to CR.

of cyprotcrone acetate on binding to CR and ABP Since cyproterone acetate has been shown to inhibit binding to the cytoplasmic receptor in rat ventral prostate (Fang and Liao, 1969a, b), experiments were carried out to determine whether this anti-androgenic steroid might have a similar effect on androgen uptake and binding to CR and/or ABP in epididymal supernatant. Cyproterone acetate was injected i.v. into 18 h castrate rats 5 min prior to the injection of [ 3H]testosterone. As seen in table 1, cyproEfect

I

Table

Total uptake of radioactive androgen in 105,000 g supernatant after cyproterone acetate treatment. Treated rats were injected with 1 mg cyproterone acetate in 0.1 ml ethanol 5 min prior to injecting labelled compounds. In experiment 1, three pairs of rats were injected i.v. with 100, 200 or 400 pCi [3H]testosterone; in experiment 2, two pairs of rats were injected i.v. with 50 FCi [3H]testosterone

and in experiment

3, two pairs of rats were injected

iv. with

31 pCi [3H]DHT. Supernatant Experiment

1 2 3

uptake

(dpm/mg

protein)

Organ “/< of control

Control

Treated

Epididymis

29,920

19,588

66

Prostate

24,803

5951

24

3171 2577

2005 1697

63 66

Epididymis Epididymis

Androgen-binding

proteins in rat epididymis

95

Table 2 Metabolism of [3H]testosterone after treatment with cyproterone acetate. Two rats in each group were eviscerated and functionally hepatectomized 18 h after castration and injected i.v. with 1 mg cyproterone acetate 5 min prior to injecting 50 pCi [3H]testosterone. One hour after injection, metabolites from whole tissue were extracted and analyzed as described in Materials and Methods. T (testosterone); DHT (5a-dihydrotestosterone); n4A (androst4-ene-3,17-dione). Zone of radioactivity (% total)

Pretreatment

0/0 Recovery

Control Cyproterone acetate

42 42

Total activity

_

~.. Water soluble 12 12

Polar

T

DHT

n4A

Others

(dpm/mg tissue)

6 13

22 9

50 49

6 10

4 7

1953 1160

terone acetate diminished the total uptake of radioactive androgen in the epididymis to only 64% of the control, whereas in the prostate the uptake was

reduced to 24% of the control. Since the effect of cyproterone acetate on cellular uptake might have been due to inhibition of testosterone conversion to DHT or inhibition of binding to receptors and/or ABP, the effects of cyproterone acetate on metabolism of [3H]testosterone and binding were investigated in vivo. As shown in table 2, cyproterone acetate had very little influence on testosterone metabolism in rat epididymis in vivo. High conversions to t3H]DHT were found both in the control and cyproterone acetate treated animals after injection of [3H]testosterone. Thus, the reduced uptake of radioactivity by epididymis in the cyproterone acetate treated rats could not have been due to lack of 13H]DHT available for binding to CR. The effect of cyproterone acetate on binding to CR and ABP was studied in the 18 h castrate rat (fig. 8). Cyproterone acetate injected 5 min before the injection of [3H]testosterone inhibited binding to CR both in the epididymis (fig. 8A) and ventral prostate (fig. 8B), but did not significantly influence binding to ABP (fig. 8A). In the same experiments, nuclei were purified through 2.2 M sucrose, extracted with 0.4 M KCl, and binding analyzed by centrifugation in 5-20x sucrose gradients. Cyproterone acetate decreased nuclear uptake and binding of radioactivity in both epididymis and prostate to the same extent that it inhibited binding to CR.

D. J. Tindali cf al.

96

A

B.

5

4 10 b_ * 3 w ” i cn ‘2 I a 0 -Control ---

acetate

5 SLICE Fig.

IO NUMBER

8. Effect of cyproterone

prostate. terone

acetate

(250 ug/lOO

15

acetate

Rats were eviscerated

testosterone

1

Cyproterone

on androgen

ABP;

containing

B: Prostate

g body

weight)

was injected

O--O,

Table Dissociation

of [3H]DHT

from

androgen-binding

ABP, androgen-binding protein; CR, dissociation rate constant was obtained by 2.303;

t,j2, half-time regression

Component

of dissociation

of epididymis injection

of [3H]-

3 h later and 0.5 ml aliquots

by PAGE.

A: Epididymis

cyproterone

acetate

CR and

treated.

3 components

in epididymis

and

prostate.

was obtained

X lo3

by dividing

coefficient

0.301

of the regression t,jz at 0 “C

by the slope

of the

line. Y

Epididymis ABP CR NR

and

Cypro-

cytoplasmic receptor; NR, nuclear receptor; Rdtss, by multiplication of the slope of the regression line

line; Y, the correlation Rdiss (min-‘)

l - - --0,

15

18 h after castration.

i.v. 5 min before

were removed

were analyzed

control;

IO NUMBER

in supernatants

hepatectomized

and prostates

4 mg of protein

CR;

binding

and functionally

(175 uCi). Epididymides

of supernatants

5 SLICE

116

5.6 min

7 12

4.0 days

0.99 0.99

2.4 days

0.99

Prostate CR

1.3

22.6 days

0.71

NR

2.2

13.4 days

0.91

Androgen-binding

proteins

97

in rrrt epididymis

L

36L 0%

Minutes

I 0

1 4

I 8

1 12

I 16

I 20

I 24

HOURS

Fig. 9. Rate of dissociation of [3H]DHT bound to cytoplasmic and nuclear receptors in rat epididymis and prostate. Comparison with dissociation of [3H]DHT from ABP. ABP was labeled in charcoal-extracted 105,000 g epididymal supernatants from intact rats by equilibration with 10 nM [3H]DHT. Receptors in cytoplasm and nuclei of epididymis and ventral prostate were labeled 8 days after bilateral castration. [3H]Testosterone (50 uCi) was injected i.v. and the rats were sacrificed 3 h later. Aliquots of 105,000 g supernatants and 0.4 M KCI extracts (Buffer B), of purified nuclei were equilibrated at 0 “C with unlabeled DHT (1000 times more than the total [‘HIDHT), and bound and free radioactivity were separated by chromatography on Sephadex G-25 columns. Bound radioactivity is plotted logarithmically on the ordinate against time on the abscissa. Time represents the period from addition of unlabeled DHT to layering of the sample over the Sephadex column. The time required for chromatography was the same for each sample. Bound radioactivity in samples containing unlabeled DHT was corrected for decreased bound radioactivity in parallel controls without unlabeled DHT. (A) U, ABP. (B) 0, epididymal CR; 0, epididymal NR; A, prostatic CR; A, prostatic NR.

Dissociation of / ‘HJDHT from CR and ABP When aliquots of the labeled epididymal supernatants were incubated with lOOO-fold excess (0.5 PM) nonlabeled DHT at 0 ‘C for 30 min, all the radioactivity bound to ABP was displaced, while that bound to CR in epididymis and prostate remained unchanged. As illustrated in fig. 9 and table 3, [3H]DHT-ABP complexes dissociated in minutes at 0 “C, while the dissociation of [3H]DHT from receptors in supernatant and nuclear extracts of epididymis and prostate was a matter of days. This almost irreversible binding of androgen at 0 “C appears to be characteristic of androgen receptors and contrasts sharply with the rapid dissociation of androgen from ABP. Sequential uptake and binding of androgen in epididymal supernatant and nuclear JLactions in presence and absence of ABP To obtain further evidence for the receptor function of CR, the time-course relationship between binding in supernatant and nuclear fractions was studied

D. J. Tindall et al.

98

in rats castrated

bilaterally

18 h (CR and ABP present)

present) before the experiment.

and 8 days (only CR

The g-day castrate group was given testosterone,

200 ug S.C. daily, for 7 days after castration. Rats were injected i.v. with [3H]testosterone and sacrificed after 0.25, 1, 3 and 5 h. Supernatant and nuclear fractions were prepared and aliquots taken for measurement of total radioactivity and for determination of binding by centrifugation in 5-20% sucrose gradients. The time course of binding in supernatant containing ABP and CR (fig. IOA) was initially somewhat more rapid than in supernatant containing

-8

2

I

3

4

5

HOURS AFTER INJECTION Fig. 8-day

10. Sequential castrate

intervals

binding

rats.

of 15 min,

centrifuged the activity

in epididymal

(A) l-day

castrate

1, 3 and

5 h before

and nuclear

sacrifice.

extracts

radioactivity gradients.

in supernatants 0, cytosol

(105,000

from

l-day

with 50 uCi of [3H]testosterone

Supernatants

and

nuclear

through 5-20% sucrose gradients and bound radioactivity under each peak. Protein determinations in supernatants

nuclei are described in Materials and 25 uCi of [3H]testosterone at intervals Bound

supernatants

rats were injected

extracts

and at were

was measured from and quantitation of

Methods. (B) I-day castrate rats were injected with of 15 min, 1, 3 and 5 h before they were sacrificed. and

nuclear

g supernatant);

extracts 0,

was measured nuclear

bound

in 5-20% radioactivity.

sucrose

CR alone

(fig. IOB); however,

the rates of increase

in nuclear

radioactivity

showed no marked difference. Binding in the supernatants increased rapidly during the first hour and thereafter either remained constant or decreased, while bound radioactivity in nuclei increased more slowly and approached a maximum level after 3 h. Total radioactivity in epididymal supernatant and nuclear fractions, not shown here, showed a close correlation to bound radioactivity. The similar patterns of uptake and binding in the presence and absence of ABP support further the concept that binding to CR alone is required for uptake of DHT by epididymal nuclei.

DISCUSSION From the present results, it is concluded that the slower moving binding protein observed by PAGE is the epididymal cytoplasmic androgen receptor (CR). Unlike ABP, CR did not disappear from epididymal supernatant following castration or hypophysectomy. Furthermore, CR was evenly distributed in taput and cauda in contrast to the intraluminal ABP, which is present in much higher concentrations in caput (Hansson et al., 1973b). The presence of CR in captit and cauda is consistent with the known fact that both of these regions of epididymis contain androgen-dependent cells. Epididymal CR resembled the cytoplasmic receptor of prostate in its electrophoretic mobility, electrofocusing pH, heat stability. and destruction of binding activity by charcoal or I mM p-chloromercuriphenylsulphonate. Hansson et al. (1973~) have recently reported a high affinity DHT-protein complex (complex I) in rat epididymal supernatant. Complex I is larger than the DHT-ABP complex and, like the cytoplasmic receptor of rat ventral prostate, is excluded from Sephadex G-200. Complex I and the DHT-CR complex reported herein are similar with respect to heat stability, and inactivation of binding by sulthydryl agents or exposure to dry charcoal. It seems likely that complex I and CR are the same protein and are identical to the 8.5-S component found by Blaquier (197la, b) in epididymal supernatants of long-term hemicastrated rats. CR further resembled prostate cytoplasmic receptor in its selective high affinity for DHT. Although DHT represented about half the total labeled metabolites in epididymis supernatant after injection of [3H]testosterone, it was essentially the only androgen bound to CR. This is in agreement with earlier studies in which binding in supernatants and nuclear extracts from castrate rats was analyzed by sucrose gradient centrifugation (Tindall et al.. 1972; Hansson et al., 1973b) or by gel filtration on Sephadex G-100 (Hansson et al.. 1973c), and 9S’x, of the bound radioactivity was identified as DHT.

100

A sharp

D. .I. Tkdall

contrast

between

the binding

specificities

of CR and

et al.

ABP was

demonstrated by studies with cyproterone acetate. Although both CR and ABP were shown to have high af%nities for DHT, the differentia1 effect of cyproterone acetate on the binding of t3H]DHT indicated CR has a much greater affinity for cyproterone acetate than does ABP. Previous studies have shown that cyproterone acetate is a potent anti-androgen which inhibits binding to cytoplasmic receptors in prostate (Fang and Liao, 1969a, b, 1971) and seminal vesicle (Stern and Eisenfeld, 1969) and blocks ~~~tr~~~eliular retention of androgen without influencing conversion of [3H]testosterone to [3H]DHT (Belham et al., 1969). Our results demonstrate a similar inhibitory effect of cyproterone acetate on androgen binding to CR in epididymis and prostate. The finding that nuclear uptake was inhibited to the same extent as binding to CR adds support to the concept that androgens must be complexed with CR prior to the translocation into the nucleus. The striking difference in rates of dissociation of [3H]DHT from epididymal CR and ABP is consistent with receptor and carrier functions for these proteins. The prostate [3H]DHT-receptor complex has been shown to dissociate extremely slowly at 0 “C (Liao et al., 1974). Androgen receptors are known to bind active androgens in the cytoplasm and to be translocated as a complex into the nucleus where they bind to chromatin. During this translocation process, the labeled androgen bound to receptor is not readily exchanged with unlabeled androgen, either in vivo (Rennie and Bruchovsky, 1973) or in vitro (Blaquier and Calandra, 1973; Liao et al., 1974). [3H]DHT-ABP complexes, on the other hand, are readily exchangeable which is consistent with a carrier function, Dissociation rates of 13H]DHT from carrier proteins such as human and rabbit TeBG are relatively rapid [t ,j2 at 0 “C of human TeBG ~; 70 min (Heyns and DeMoor, 1971; Hansson et al., 1974) and t ,12 at 0 “C of rabbit TeBG = 6 min (Hansson et al., 1974)]. The rapid dissociation of [3H]DHT from ABP is in agreement with the idea that ABP is a carrier protein within the lumen of the epididymis and releases its bound androgen to cytoplasmic receptors

in the surrounding

epithelial

ceils.

ACKNOWLEDGEMENTS Financial support was given by United States Public Health Service grants HD04466 and AM05330, World Health Organization grant #H9/181/83, and by grants from the Rockefeller Foundation, the Norwegian Research Council for Sciences and the Humanities, and the Swedish Medical Research Council.

Androgen-binding proteins in rat epididymis

101

REFERENCES Belham, J. E., Neal, G. E. and Williams, D. C. (1969) Biochim. Biophys. Acta 187, 159 Blaquier, J. A. (1971a) Biochem. Biophys. Res. Commun. 45, 1076. Blaquier, J. A. (1971b) Acta Physiol. Latinam. 21, 101. Blaquier, J. A. and Calandra, R. S. (1973) Endocrinology 93, 51. Bruchovsky, N. and Wilson, J. D. (1968) J. Biol. Chem. 243, 2012. Davis, B. J. (1964) Ann. N. Y. Acad. Sci. 121, 404. Dingman, C. W. and Peacock, A. C.(l968) Biochemistry 7, 659. Fang, S. and Liao, S. (1969a) J. Biol. Chem. 244, 6584. Fang, S. and Liao, S. (1969b) Mol. Pharmacol. 5, 428. Fang, S. and Liao, S. (1971) J. Biol. Chem. 246, 16. French, F. S. and Ritzen, E. M. (1973a) J. Reprod. Fertil. 32, 479. French, F. S. and Ritzen, E. M. (1973b) Endocrinology 93, 88. Hansson, V. (1972) Steroids 20, 575. Hansson, V., Djoseland, O., Reusch, E., Attran~adal, A. and Torgersen, 0. (1973a) Steroids 21, 457. Hansson, V., Weddington, S. C., Tindall, D. J., French, F. S., Nayfeh, S. N. and Ritzen, E. M. (197313) Biol. Reprod. 1, 90. Hansson, V., Djoseland, O., Reusch, E., Attramadal, A. and Torgersen, 0. (1973c) Steroids 22, 19. Hansson, V., Ritzen, E. M., French, F. S., Nayfeh, S. N., McLean, W. S., Tindall, D. 3. and Weddington, S. C. (1974) Endocrinology 95, 690. Heyns, W. and DeMoor, P. (1971) J. Clin. Endocrinol. Metab. 32, 147. Hotta, S. and Chaikoff, I. L. (1955) Arch. Biochem. Biophys. 56, 28. Ingle, D. J. (1949) Exp. Med. Surg. 7, 34. Liao, S. (1974) In: Biochemistry of Hormones, Eds. : C. H. Kornberg and D. C. Phillips (MTP International Review of Science), in press. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193, 265. Mainwaring, W. I. P. and Irving, R. (1973) Biochem. J. 134, 113. Rennie, P. and Bruchovsky, N. (1972) J. Biol. Chem. 247, 1546. Rennie, P. and Bruchovsky, N. (1973) J. Biol. Chem. 248, 3288. Ritzen, E. M., Nayfeh, S. N., French, F. S. and Dobbins, M. C. (1971) Endocrinology 89,143. Ritzen, E. M., Dobbins, M. C., Tindall, D. J., French, F. S. and Nayfeh, S. N. (1973) Steroids 21, 593. Stern, J. M. and Eisenfeld, A. J. (1969) Science 166, 233. Stern, J. M. and Eisenfeld, A. J. (1971) Endocrinology 88, 1117. Tindall, D. J., French, F. S. and Nayfeh, S. N. (1972) Biochem. Biophys. Res. Commun. 49, 1391. Uriel, J. (1966) Bull. Sot. Chim. Biol. (Paris) 48, 969. Wrigley, C. (1968) Science Tools 15, 17.