Inhibition of nucleic acid and chromatin binding of the rat prostate androgen receptor by pyridoxal phosphate, heparin and cibacron blue

2685

633 INHIBITION OF NUCLEIC ACID AND CHROMATIN BINDING OF THE RAT PROSTATE ANDROGEN RECEPTOR BY PYRIDOXAL PHOSPHATE, HEPARIN AND CIBACRON BLUE Eppo Mulder, Lida Vrij & John A. Foekens Department of Biochemistry (Division of Chemical Endocrinology) Medical Faculty, Erasmus University, Rotterdam, The Netherlands Received7-5-80 ABSTRACT

The androgen receptor from rat prostate binds 5a-dihydrotestosterone and related androgenic steroids at a steroid binding site and in addition shows selective binding to structures related to nucleic acids (chromatin binding site). Cytoplasmic androgen receptor, labeled with the synthetic androgen methyltrienolone (R 1881) was readily bound at 4oC by 2' ,5'-ADP-agarose, DNA-cellulose and phosphocellulose. The binding to ADP-agarose and DNA-cellulose was used as a model for study of the nucleic acid binding site of the receptor. Complete elution of androgen re'ceptors from these matrices could be obtained with low concentrations of pyridoxal phosphate (10 mM), heparin (0.2 mg/ml) and Cibacron blue (0.4 mM). Sodium molybdate (10 mM) did not interfere significantly with binding of the androgen receptor to ADPagarose and had little effect on elution of receptors from the gel. Pyridoxal phosphate, heparin and Cibacron blue in low ionic strength buffers were also shown to be very effective for the extraction of androgen receptors from nuclei of rat prostatic tissue. These results suggest gross similarities in the structure of the androgen receptor with activated forms of receptors for corticoids, estrogens and progestins with respect to nucleic acid binding.

Abbreviations: 5a-dihydrotestosterone (DHT), 17g-hydroxy5a-androstan-3-one; methyltrienolone (R 1881), 178-hydroxy17-methyl-4,9,11-estratrien-3-one; TEDG buffer, a buffer with Tris-HCl, EDTA, dithiothreitol and glycerol as defined in the method section; TED buffer, the same buffer without glycerol; Cibacron blue, C-1.61211, also known as Reactive Blue 2 and Procion Blue HB. Volume36, Ntier 6

s

TDICOXDI

December, 1980

INTRODUCTION It is generally cytoplasmic change

accepted

receptor

for nuclear

receptor

complexes

be a prerequisite Several sary

studies

step

have

shown

receptors

(7) selectively

shown

or phosphocellulose.

structural

requirements

effects.

is also

a neces-

and phosphocellu-

sites

various

that pyridoxal (6), heparin

the binding

and estradiol-receptor

progesteronecellulose

acid

inhibit

to

of the interaction

binding

(4,5), aurintricarboxylic

cron blue

is thought

and progesterone

ATP-sepharose

It was

a high

of steroid-

sites

that activation

on the nature

used.

to

a conformation

of the hormonal

of corticoid

bind

to a form with

acceptor

and chromatin

have been

hormones

(1). Binding

for expression

to DNA-cellulose,

reagents phate

("activation")

to nuclear

(2,3). In studies

between

steroid

and induce

chromatin

for binding

receptors lose

proteins

in the receptor

affinity

that

for binding

could

to DNA-

imply

of these

and Ciba-

of corticoid-,

complexes

This

phos-

certain

receptors

to

DNA. Similar receptor

information

which

testosterone formation receptor

specifically

and related

could

binds

be important

extensively

and estrogen purified.

of androgen

cellulose

steroids

because

In a previous

(9). In this report

(8). Such dn-

the labile

as well

receptors,

receptors

for the androgen

Sa-dihydrotestosterone,

androgenic

is not yet characterized

progesterone

binding

is not available

which

androgen

as corticoid, have been more

study we observed

to heparin-agarose we describe

and DNA-

the selective

interference of different chemical reagents with binding of the androgen-receptor complex to 2',5'-ADP-agarose and DNAcellulose. MATERIALS AND METHODS Materials. - 11,2,6,7-3HI-testosterone (93 Ci/mmol) was obtained from the Radiochemical Centre, Amersham, U.K. and 117a-methyl-3HI-methyltrienolone (87 Ci/mmol) was obtained from New England Nuclear, Dreieich, Germany. Pyridoxal-5phosphate was obtained from Boehringer, Mannheim, Germany. Cibacron Blue F3G-A, heparin- and ATP-agarose (adenosine-5'triphosphate attached through ribose hydroxyls with a 6carbon spacer to agarose, 2.2 pmol/ml) were obtained from Sigma, U.S.A. ADP-agarose (2',5'-ADP-sepharose-4B, 2 pmol/ml) was obtained from Pharmacia, Sweden. Phosphocellulose (cellulose phosphate) was obtained from Whatman Inc., U.K. Protamine sulfate was bought from Organon, Oss, The Netherlands. DNA-cellulose (150 pg/g) was prepared according to Alberts and Herrick (10). Cytoplasmic and nuclear receptors. - Cytoplasmic and nuclear receptors were prepared from rat prostates one day after castration. A buffer (pH 7.4) containing 0.01 M Tris-HCl, 1.5 mM EDTA, 1.5 mM dithiothreitol, with 10% glycerol (TEDG buffer) was used for preparation of cytosol. Prostatic tissue was homogenized in 3 volumes TEDG buffer and the 100,000 x g cytosol fraction was isolated (11). Cytosol fractions were incubated for 2 h at OoC with 10 nM labeled methyltrienolone. In control experiments a lOO-fold excess non-radioactive methyltrienolone was added together with 13HI-methyltrienolone. Cytosols with radiolabeled androgen receptors were stored at -200C for periods up to 2 weeks. For binding to the different matrices cytosol (8 ml) was incubated with thoroughly washed gel (2 ml) for 2 h at 6oC under continuous gentle agitation. For preparations of nuclear pellets prostate tissue was incubated for 1 h at 37OC in Eagle's minimal essential medium with 20 nM 13Hltestosterone. The tissue was homogenized as above and the 700 x g pellet was prepared. The pellet was washed with TEDG buffer with 0.2% Triton X-100 and subsequently two times with TEDG buffer. Nuclear receptors were extracted as indicated in the legend of Table III. Estimation of androgen receptors by protamine sulfate precipitation. - Androgen receptors were estimated essentially as described by Chamness -et al. (12). In albumincoated tubes 250 ~1 protamine sulfate (1 mg/ml), 200 ~1 TEDG buffer and 50 ~1 sample were mixed and the protamine sulfate precipitate was isolated by centrifugation. The precipitate was washed several times with TEDG buffer and dissolved in Soluene (Packard Instruments) for estimation of radioactivity

S

TXIROXDD

by liquid scintillation counting in a Searle Isocap/300 cooled liquid scintillation system model 6879. Protein determination. Peterson (13).

- Protein

RESULTS The androgen

receptor

rats one day after trienolone, receptor drogenases lone

Table

cytosol

I. The amount

amount

with

with

steroid

experiments

with

eluted

in the 0.4 M KC1 fraction

radioactive

sucrose

(15,16) and chromatography that

50-80%

eluates

was

sis prevents

The percentage

probably

a complete

2',5'-ADP-agarose recovery

studies

receptors

androgen

it was

LH-20

analysis

purification

by

(17) showed

receptor

of labeled

10%

from experi-

in the eluates

recovery

were

less than

centrifugation

but dissociation

the highest

of receptors

In recent gesterone

higher,

of radio-

Additional

on sephadex

of the radioactivity

receptor-bound.

was

gradient

with

containing

the amount

only.

in

from the

in the eluates

steroid

gel el'ectrophoresis,

are shown

cytosol

activity

agar

steroid

ATP-

from the matrices

non-labeled

with

dehy-

of methyltrieno-

calculated

eluted

of radioactivity

for this

2',5'-ADP-agarose,

was

of

13HI-methyl-

affinity

studies

excess

ments

with

high

lOO-fold

of the amount

to

cytosol

by 3-hydroxysteroid

of binding

of receptor

In control

labeled

and phosphocellulose

of radioactive

0.4 M KCl.

was

receptor

DNA-cellulose

in the prostate

to attack

(14). Results

labeled

agarose,

present

androgen

and not liable

according

AND DISCUSSION

castration

a synthetic

was estimated

was in the

during

analy-

receptors.

With

and a good

obtained. shown

are only bound

that corticoid

and pro-

to DNA-cellulose

or ATP-

s

TDEOSDII

637

Table I. Binding of cytoplasmic androgen receptors to different matrices. % receptor recovered

purification factor

ATP-agarose

67 19

47 43

Phosphocellulose

82

23

DNA-cellulose

30

39

DNA-cellulose, limited washing procedure '1

52

38

ADP-agarose

Prostate cytosol (4 ml) labeled with /3HI-methyltrienolone was incubated with the different matrices (1 ml) for 2 h at 6oC under continuous gentle agitation. The mixture was transferred to small glass columns (10 mm diameter) and washed successively with 10 ml of TEDG buffer, of TEDG buffer with 0.05 M KC1 and of TEDG buffer with 0.1 M KCl. The receptors were eluted with 10 ml of 0.4 M XCl. Corrections for non-specifically bound steroid were obtained from arallel experiments with cytosol labeled with 1Ii3 WI -methyltrienolone in the presence of a lOO-fold excess of the non-radioactive compound. The "purification factor" was estimated as the ratio of the amount of receptor per mg protein of the 0.4 M KC1 eluate versus the amount of receptor per mg protein of the original cytosol. ') Washing of the gel with 0.05 and 0.1 M KC1 containing buffers was omitted.

agarose after a temperature-dependent activation step and that sodium molybdate was effective in preventing this process f18,19). The presence of 10 mM sodium molybdate decreased receptor binding

to

13% under the incubation

conditions as given in Table I. In the present study androgen receptors from prostate cytosol, freshly prepared in TEDG buffer without sodium molybdate at O°C, were bound by ADP-agarose and the other

matrices.

Receptors

similarly.

However,

tion profiles

in cytosol

stored

differences

were

on sucrose

cytosol

sedimenting

cytosol

mainly

at 3.5-4

observed

gradients,

mainly

at -2OOC

behaved in sedimenta-

receptors

from fresh

at 8-9 S and those

from stored

Formation

S (Fig. 1, left panel).

of slower

of the of proteolytic in detail

by WiLson

gradation

et al. --

to proteolytic to the

be involved.

do not

a process. The effect

of

on the by

androgen

a nucleotide-containing

investigated

mainly

Fig.

the selective

2 shows

ADP-agarose receptor

purification step.

was

The sucrose

labeled phosphate

with

ADP-agarose binding

and the removal

protein

of

generally gradient

of androgen

washing

obtained

receptors

amounts

steps.

in a single

sedimentation

profile

of the

from ADP-agarose

with

showed

a peak

of radioactivity

sedimenting

S, comparable

of the slower

sedimenting

(Fig. 1). The effects

heparin,

Cibacron

of the receptor-bound

pyridoxal

to the sedimentation

at value

form of the original_ cytoplasmic

receptor

blue

of non-

adsorption

eluted

3.5-4

by

A fifty-fold

receptor

approximately

was

and DNA-cellulose.

of considerable

in different

backbone

of pyridoxal

and sodium radioactivity

molybdate

phosphate, on the elution

from ADP-agarose

and

S Cytosoi

TDEOXDI

639

receptor

DPA

Eluoh

ADP-agarosa

DPEJ

x10’ !

S: 7.2

4.6 3.6

S: 7.2

4.6

4

3

2

lb fraction

i0

number

Fig. 1. - Profiles of 3H-radioactivity after sucrose gradient centrifugation of androgen receptors labeled with \3HI-methyltrienolone. Left panel: Cytosol receptors in TEDG buffer centrifuged T----.----lmmedlately after preparation ("fresh"); after storage for one week at -2OOC ("1 week -2OoC"); control experiment with an excess of a lOO-fold non-radioactive methyltrienolone, with identical results for both fresh and stored cytosol ("control"). Right -- 7" panel: ----- Receptors eluted from ADP-agarose with 10 mM pyrldoxal phosphate in 50 mM borate buffer pH 8 (see Fig. 2). A 200 ~1 portion of the different fractions was centrifuged at l°C in linear 5-20% sucrose gradients in TEDG buffer for 18 h at 310,000 x gav in a Beckman SW-60 rotor (left panel) or for 2 h 45 min at 370,000 x gav in a Beckman VTi-65 rotor (right panel). The sedimentation markers indicate positions of y-globulin (7.2 S), bovine serum albumin (4.6 S) and ovalbumin (3.6 S).

S

640

receptor

TEIROXDII

protein--

-

Fig.

2. - Elution with pyridoxal phosphate of methyltrienolone labeled androgen receptor from 2',5'-ADPagarose. Labeled rat prostate cytosol was incubated with ADP-agarose as described in the method section and the ADP-agarose (2 ml) was transferred to a small column and washed with TEDG buffer, pH 7.4 (l), TEDG buffer with 0.05 M KC1 (2), TEDG buffer with 0.1 M KC1 (3), 0.1 M phosphate buffer, pH 7.5 (4), 10 mM pyridoxal phosphate in 50 mM borate buffer, pH 8 (5) and 0.4 M KC1 in TEDG buffer, pH 7.4 (6). In a control experiment with cytosol labeled in the presence of an excess non-radioactive methyltrienolone no radioactive steroid remained bound to the gel after washing with phosphate buffer (line not shown in the figure).

DNA-cellulose and Cibacron phate.

are shown blue

Heparin

0.02 mg/ml, 10,000). receptors

was

are more already

or 0.002 mM

Sodium

in Table

effective active

(assuming

molybdate

from ADP-agarose

II. On a molar

was

base

than pyridoxal

heparin phos-

in a concentration

of

a molecular

of

less effective

and DNA-cellulose.

weight

for removal

of

S

TB&OXDI

Table II. Effect of pyridoxal phosphate, heparin, Cibacron blue and sodium molybdate on the elution of androgen receptors from ADP-agarose and DNAcellulose. matrix

concentration

reagent

% receptor eluted

ADP-agarose 1KlM

84 19

Heparin

0.2 mg/ml 0.02 mg/ml

90 38

Cibacron blue

0.4 mM 0.04 mM

95 64

10 mM

24

10 mM

14

Pyridoxal phosphate

Sodium molybdate

10 mM

DNA-cellulose Pyridoxal phosphate Heparin

0.2 mg/ml

75

Cibacron blue

0.4 mM

78

Prostate cytosol labeled with 13H\-methyltrienolone was incubated with the gels for 2 h at 6oC and washed with buffers l-4 as indicated in Fig. 2. Fractions of the gels (0.5 ml, containing approximately 400 fmol receptor) were incubated for 15 min at 4oC with 1 ml of the different reagents. The suspension was centrifuged for 10 min at 800 x g and the supernatant collected. This incubation procedure was repeated twice. The amounts of receptor in the pooled supernatants were estimated by protamine sulfate precipitation. The values are means of duplicate experiments.

Over 80% of the steroid eluted from the gels could be precipitated with protamine sulfate. Only after heating to 60°C for 20 min or incubation overnight at 15'C with excess nonradioactive steroid, no radioactivity could be demonstrated in the protamine sulfate

precipitates. This indicates the

presence of a saturable binding protein that is denaturated at 60°C and capable of exchange of labeled and non-labeled

S

642

steroid

at 15'C. Both criteria

from non-specific reagents acids

only

was

androqen

with

effective

isolated

labeled

reagents

nuclei

in vitro -~

labeled

data

in Table

blue

and heparin

blue

The specific separate

of tissue

receptors

methyltrienolone

between

used

receptors

binding

The

(8).

Cibacron are as for the

nuclear in the

and DNA-cellu-

pyridoxal

and the binding nucleic

region

13H/-

in the cell nucleus.

of the different

from the androgen

with

(0.4 M KC11

to ADP-agarose

containing

binding

buffer

a similarity

to chromatin

and heparin

chromatin

rats were

androqen

from prostate

suggest

to

androgen

phosphate,

salt solutions

of the interaction

effect

receptors

minces

nuclear

strength

receptors

or the matrices

study

not

matrices,

/3HI-5a-dihydrotestosterone

experiments

the binding

were

of one day castrated

this condition

of androgen

The nature

to the

of

from these

For this

in a low ionic

of androgen

lose with

nuclei

tissue

binding

of androgen

III show that pyridoxal

These

pellets.

Cibacron

as well.

as concentrated

extraction

to nucleic

and DNA-cellulose

binding

with

with

of receptors

by incubation

Under

testosterone.

binding

with

in prostatic

effective

the different

of receptors

interfering

for removal

cell

are mainly

receptors

of methyltrienolone

to ADP-agarose

interfere

receptors

Therefore

binding

binding

androgen

not affected.

receptors

but could

distinguish

proteins.

interfered

The different

apart

binding

on the gel, whereas

receptor

only

TDROIDCS

of receptors acids

reagents

in the present

demonstrate

a

molecule

The synthetic study

to

is not known.

of the receptor

site.

phosphate,

androgen

has a high

affi-

S

T~ROLDI

643

Table III. Extraction of labeled receptors from nuclear pellets obtained from rat prostates incubated in vitro. -extraction medium

fmol/lOO mg prostate

Buffer

8+

2

0.4 M KC1

53 2 10

0.6 M KC1 (repeated twice)

692

3

Cibacron blue, 0.4 mM

45+

3

Heparin, 0.2 mg/ml

602

9

Pyridoxal phosphate, 10 mM

532

6

6+

3

Sodium molybdate, 10 mM

Prostates obtained from one day castrated rats were incubated with 2.10-B M 13Hf-testosterone and in a parallel experiment with an additional amount of 2.10-6 M non-radioactive testosterone. A thoroughly washed nuclear pellet was prepared and extracted for 1 h at 4oC. The amount of radioactive steroid extracted was estimated and corrected for non-specifically bound steroid with the values obtained from the parallel incubation. The extraction media used were: TED buffer, pH 8.4; 0.4 M KC1 in TED buffer, pH 8.4; 0.6 M KC1 in TED buffer, pH 8.4 the extraction procedure was repeated twice and the extracts were combined; Cibacron blue 0.4 mM in TED buffer, pH 8.4; heparin 0.2 mg/ml in TED buffer, pH 8.4: pyridoxal phosphate 10 mM in 50 mM borate buffer, pH 8.4; sodium molybdate 10 mM in TED buffer, pH 8.4. The values are means + S.D. of triplicate experiments.

nity for the prostate androgen receptor and did not dissociate from the receptor under the influence of the different reagents at 4OC. From comparable studies with .corticoid and progesterone receptors it has been suggested that the effects of pyridoxal phosphate might involve essential lysine

E-aIdnO-grOUpS

in the binding region of the receptor

(4,5). The effects of Cibacron blue F3G-A are often explained as effects on proteins containing a 'dinucleotide fold" (21)

(many dehydrogenases and kinasesf, although not all proteins binding to Cibacron blue show this super-secondary structure (22). It is noteworthy that the sulfate groups in Cibacron blue and the phosphate groups in 2',5'-ADP-agarose show a similar spatial orientation

(21). In general Cibacron blue

might act as a substance showing both localized cation exchange properties and some hydrophobic interactions. The lower optimal concentrations of heparin and Cibacron blue compared to pyridoxal phosphate required for interference with receptor binding to chromatin might indicate the involvement of a spatial arrangement of more than one negatively

charged

group separated by a less polar region. In conclusion, the present study shows that androgen receptors interact with structures containing nucleic acids in a similar way as corticoid, progesterone and estradiol-17% receptors. For these receptors an "activation" process is required before binding to the nucleic acid matrix occurs (23). For the cytoplasmic androgen receptor prepared as described in this report, we always observed, irrespective of a distinct activation step, binding to the different matrices. This is not necessarily an essential difference between the various receptor systems, but might indicate a difference in the rate of formation of an "activated" form. A similar effect has been observed for the estrogen receptor for which binding to DNA-cellulose at 4OC was shown to accelerate the conversion of the 4 S to the 5 S (activated) form of the receptor (24).

We wish to thank Dr. H.J. van der Molen for his continuous interest in our work and constructive criticism of this manuscript. REFERENCES Nature 278, 752-754 (1979). P. J. Biol. Chem,

A. and Foley, R. Kalimi, M., Colman, P. and Feigeison, 5:Munck, 250, 1080-1086 (1975).

MilE, J.B. and Taft, D.O. Biochemistry x, 173-177 (1978). 4. Cake, M.H., Di Sorbo, D.M. and Litwack, G. J. Biof. Chem. 253, 4886-4891 (1978). 5. Nishigori.,. and Toft, D. J. Biol. Chem. 254, 91559161 (1979). 6. Moudgil., V.K. and Weekes, G.A. FEBS Lett, 2, 324-326 (1978). 7. Kumar, S-A., Beach, T.A. and Dickerman, H.W. Proc. Natl. Acad. Sci. USA 76, 2199-2203 (5979). 8. Mainwaring, W.I.P. The Mechanism of Action of Androgens, Springer-Verlag, New York (1977). 9. Mulder, E., Foekens, J.A., Petexs, M.J. and van der Molen, H.J. FEBS Lett. 97, 260-264 (1979). 10. Alberts, B.M. and Herrick, G. Methods Enzymol. 21, 198. 217 (1971). l.1,Hpiisaeter, P.A. Biochim. Biophys. Acta 3x7, 492-499 (1973). 12. Chamness, G.C., Huff, K. and McGuire, W.L. Steroids 25, 627-635 (1975). 13. Peterson, G.L. Anal. Biochem. 83, 346-356 (197'7). 14. Bonne, G. and Raynaud, J.P. Steroids z, 497-507 (1976). 15. Mulder, E., Peters, M.J., van Beurden, W.M.O. and van der Molen, H-3. FEBS Lett. 47, 209-211 f1974). 16. Mulder, E., Peters, M.J., de Vxies, J., van der Mofen, H.J.r gstgaard, K., Eik-Nes, K.B. and Gftebro, R. Molec. Cell. Endocr. 11, 309-323 (1978). 17. Sirett, D.A.N. and Grant, J.K. J. Endocr. 77, 101-110 (1978). 18. Leach, X.L., Dahmerr M.K.# Hammond, N-D., Sancho, J.J. and Pratt, W.B. J. 3iol. Chem. 254, 11884-11890 (1979). 19. Toft, D. and Nishigori, H. J. Steroid Biochem. 11, 413~ 416 (1979). 20. Wilson, E.M. and French, F.S. J. Biol. Chem. 254, 6310.. 6319 (1979). 21. Thompson, S.T., Cass, K.H. and Stellwagen, 33. Proc. Natl. Acad. Sci. USA 72, 669-672 (1975). 22. Beissner, R.S., Quiocho, F.A. and Rudolph, F.B. J. Mol. Biol. 134, 847-850 (1979). 23. Bailly, A., Le Fevre, B., Savouxet, J.-F. and Milgrom, E. J. Biol. Chem. 255, 2729-2734 (1980). 24. Yamamoto, K.R. and Alberts, 3.M. PXOC * Natl. Acad. Sci. USA 2, 2105-2109 (1972). 3.