The nuclear matrix is the site of glucocorticoid receptor complex action in the nucleus

Vol. 137, No. 2, 1986

BlOCHEMlCALANDBlOPHYSlCALRESEARCH

COMMUNICATIONS

Pages 640-648

June13.1986

THENUXEAR

MATRIX

IS THE SITE OF UUCOCORTICOIO RECEPTOR CCWLEX ACTION IN THE NUCLEUS

Ted M. Kirsch, Fels Research Temple LhiVerSity Received

April

23,

Andrea

Miller-Diener

Institute School

of Medicine,

and Gerald Litwack*

and Oepartment

of Biochemistry

Philadelphia,

PA 19140

1986

of highly purified glucocorticoid receptor complexes to nuclear was evaluated. Extraction of purified nuclei with 2M potassium chloride and brief deoxyribonuclease digestion leaves a matrix struttUre Containing 1% of nuclear DNA and 612% of nuclear proteins. The nuclear matrix retained two binding sites for receptor complexes, a high affinity, low capacity site and a low affinity, high capacity site. These sites have affinities and capacities consistent with those reported for binding of these complexes to intact nuclei. More extensive deoxyribonuclease treatment of the matrix resulted in a marked reduction of high affinity complex binding. Furthermore, the DNA binding form of the receptor complex but not the unactivated receptor complex bound to DNA fibers anchored to nuclear matrix as visualized by 18 nm gold particle receptor complexes. The data suggest that the nuclear matrix is the major site for coordinating Q 1986 Academic Press, Inc. glucocorticoid hormone action in the nucleus. Binding

matrix

The nuclear

matrix has been proposed to have specific functions in ONA is spatially arranged on this three organizing nuclear processes. dimensionalstructureas supercoiledloopsanchoredtothe matrixattheir bases. The site ofattachment allows for orderly transcription and replication of DNA to occur (l-3). Actively transcribed genes under hormonal control have been localized to the nuclear matrix, and nuclear forms of androgen and estrogen receptor are tightly associated with this structure (4-7). The unactivated glucocorticoid holoreceptor complex (9-10s) is localized in the cytoplasm and in the presence of steroid, the complex is rapidly activated to a smaller form (45) which translocates to the nucleus specific sive

This receptor complex (4s) (8-9). DNA sequences of enhancer regions genes

and

the

complexes

are

has been

shown

to bind

of glucocorticoia responknown to be potent inducers of

*To whom correspondence should be addressed. TA, triamcinolone acetonide Iba-Fluoro-llS, 21 Abbreviations Used: dihydroxy 16a, 17-kl-methylethylidene) bis (OXY)] pregna-1,4-dieneyes, 2- (Nmorpholino)ethane-sulfonic acid; DTT, 3,20-dionel; dithiothreitol; CNase, deoxyribonuclease.

0006-291X/86 $1.50 Copyright All rights

0 1986 by Academic Press. of reproduction in any form

Inc. reserved.

640

to

Vol. 137, No. 2, 1986

8lOCHEMlCALANDBlOPHYSlCALRESEARCH

COMMUNICATIONS

several well characterized proteins (10). Previously, these receptor complexes have been reported to bind to nuclei, nuclear membranesand chromatin (11-U). In this study we report the association of highly purified activated glucocorticoid receptor complexes with the nuclear matrix, a major site for DNA organization and active gene transcription. MATERIALS

ANDMMOOS

Preparation of Receptor: Inactivated glucocorticoid receptor complexes were purified from rat liver cytosol by affinity chromatography (16). The receptor complexes were activated at 25-C for 30 min and isolated from MAE-cellulose by ion-exchange exactly as described (16). Specific binding of 3il-TA (44 ci/mmol; New England Nuclear, Boston, MA) was measured by hydroxylapatite adsorption and quantification of the activated receptor complexes were determined by CNA-cellulose binding The purity and integrity of the assay exactly as described (16).

receptor complexes were monitored by SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE) during purification as described previously (16). Isolation of Nuclei and Nuclear Matrix: Sucrose purified nuclei were prepared essentially as described previously (13). Nuclear matrices were prepared as described by Buttyan et al. (6) except that the matrices were resuspended in Buffer A, (10 mN Mes, 0.1 n+l DTT, 10% glycerol, pH 6.8). Treatment of Matrices: Brief MJase I (5 ug/mL; Worthington; electrophoretically pure) digestion of matrices was performed for 30 min at 22-C, according to Buttyan et. al. (6). A separate extensive DNase I (1 mg/mL) digestion of matrices for 1 h at 22-C was performed to determine ONasesensitivity of receptor complex binding sites. Exhaustive DNase treated matrices were compared to brief CNasetreated matrices in parallel binding assays. DNase

Nuclear matrices were resuspended in Buffer A (50 ul = DNA; 52 ug protein). Activated glucocorticoid receptor complex (5000 fold purified; 8-400 fmol/mL) in Buffer A was added to matrices and the volume was adjusted to 0.8 ti with Buffer A, in an Eppendorf tube. Incubation was for 45 min at 22’C and was terminated by centrifugation (15,000 xg, 3 min). The supernatant was removed by capillary aspiration and the bottom of the tube was cut and placed into a 7 ml scintillation vial containing 5 ml Scintiverse II cocktail. (Fisher, Pittsourgh, PA). Scintillations were quantitated with an Intertechnique SL-30 scintillation counter and dpm were computed using an external standard (3H efficiency = 35%). Higher concentrations of receptor complexes were achieved by concentration using a low binding Novacell pressure concentrator (Filtron, Clinton, t#.; 10 kDa cut off). Recovery was 85-90% with no alteration in functional Binding

Assay:

9 ug matrix

proper ties of receptor . Aggregation of receptor (4% of preparation) was determined by centrifugation of receptor alone and data were correctea accordingly. Presaturacion (5 min at 25'C) with receptor labeled with radioinert steroid (300 fold excess) competed 3H-TA complex binding (3G-50%) and data were ad juste0 accordingly. This competition precludes nonspecific trapping by a fixed volume of nuclear matrix. Total dpm added to each tube were corrected for the amount of receptor (30-45%) capable

receptor

of binding associated

to an excess

3H-steroid

of CNA cellulose

(45 min at 22-C)

was monitored by hydroxylapatite

min at 22-C) Replicates of bound receptor binding 5%. The structural integrity of the

and

assay

(45

641

to matrix did not vary more than receptor complex

glucocorticoid

Vol. 137, No. 2, 1986

BIOCHEMICAL

AN0 BIOPHYSICAL

RESEARCH COMMUNICATIONS

bound to nuclear matrix was determined by SDS-PAGE and scanning an ultroscan laser densitometer (LKB, Gaithersburg, MO).

with

Gold-Receptor Complex Binding to Nuclear Matrices: Colloidal gold (18 nm particle size) was prepared and the minimum stabilizing concentration of receptor was determined as described (17) and gold-receptor complexes were resuspended in Buffer A. Nuclear matrices (450 pg ONA) were incubate0 with 200 fmol untransformed or DNA-binding transformed gold-receptor complexes for 45 min at 22-C in an Eppendorf tube. The matrices were centrifuged at 1000 x g and gently washed with Buffer A. Functions of M\IA binding of receptor complexed to gold were monitored by DNA-cellulose assay and for steroid binding site by hydoxylapatite assay, respectively. Electron Microscopy: Nuclear matrices were pelleted and prefixed with 2% (v/v> glutaraldehyde in phosphate buffer (O.lM sodium phosphate buffer pH 7.4) for 90 min at 4-C. The pellet was washed in phosphate buffer containing 0.22 M sucrose for 2 h at 4'C then post fixed in 1% (w/v) osmium tetroxide in phosphate buffer for 90 min at 4-C. The pellet was then dehydrated in an ethanol series (50-10046) and embedded in Epon 812. Thin sections (17 nm) were cut and examined in a Philips300 electron microscope (Philips, Newark, NJ.). Protein was determined by a modified DNA was determined by a sensitive fluorometric

Miscellaneous:

(

)

gifd;er

method assay

of Lowry of Hine-

(19).

RESULTS Sucrose purified nuclei from the liver of adult Sprague Oawley rats were a homogeneous population with membranes intact. Nuclear matrices isolated from this population, retained a residual laminar membrane with associated nuclear pore complexes, internal fibrogranular network and nucleoli. The nuclear matrix consistently contained 1.5-2.5% of the total nuclear DNA and 6-7% of the total nuclear protern. Highly purified activated glucocorticoid receptor complexes bound to liver nuclear matrices rapidly (5 min at 25-C) and in a saturable The binding of complexes was stable and constant for 45 manner. min at 22-C. A sigmoidal saturation curve indicated two binding sites for receptor complexes present in the nuclear matrix (Fig. 1). The = 0.98) and the high capacity, sites were non-cooperative (Hill coeff. low affinity site was determined to have a Kd = 1.2 + 0.1 x 10m8 M and Bmax = 5 nm (X + SEM; N = 3 curves) representing 27,500 sitesjhaploid genome as determined by LIGAND computer program run on an IBM-PC (20). The computer algorithm also identified a high affinity, low capacity site for receptor complexes on the nuclear matrix having a Kd = 1.4 + 0.2 x lo-lo M and ENMIX= 17 pM representing 1,500 sites/haploid genome which is presented in the computer generated graph in Figure 2. Unactivated purified receptor complexes did not bind to nuclear matrix nor did free steroid (100 nM 3H-TA; 44 Ci/mmol) show specific 642

Vol

137,

BIOCHEMICAL

No. 2, 1986

AND

TOTAL

Figure

1:

Saturation to nuclear fmol/ml) Methods.

curve

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

CPMOLI

of

glucocorticoid receptor complex binding matrix. Hormone receptor complexes (8-400 were added to nuclear matrices as described in

B/F 0.1117

+

:

11 +

I

+ l 2

0.0931

+

:

* + + 0.0744

:

*3 +

t4 * +

o.o!56

+

:

+

+ +

0.0372

+

+ +

I--;---;-;-----;--;---;-w;

0. oooo

OAMENO

2.34E-12

4.67E-12

7.0lE12 iaM

Figure

+

2:

9.3X-12

643

l.)(K-11

1.64E11

rHa.1

generated Scatchard plot glucocorticoid receptor complex nuclear matrix (Bmax = 17 PM). computer generated curve. CompUter

l.lTE-11

of high b’

affinity

sites ; “f+Jy

for

Vol.

137, No. 2, 1986

AND BIOPHYSICAL

BIOCHEMICAL

67 +

10

20

30

4330 + +

40

50

60

RESEARCHCOMMUNICATIONS

20 +

70

80

90

100

1

RF WI Figure

3:

Densitometric glucocorticoid to nuclear are aligned bound to

scan

of

SDS-PAGE

analysis

of

highly

purified

receptor complex (A); receptor complex bound matrix (B); nuclear matrix alone (C). The scans to demonstratepositional integrity of receptor matrix. Equal relative intensity (scale included for subjective comparisononly) and position of receptor suggests that no degradation has occured.

.

Figure 4:

Direct

localization of high affinity receptor binding sites using activated receptor-colloidal gold probe. Activated receptor-gold complexes are visualized bound to 3-4 nm Cilamentous structures associated with the nuclear matrix (A); A less dense region of the matrix clearly shows

(arrows) the orientation of receptor-gold probe along matrix-associated MA threads (B); Vlactivated receptorgold complexesdo not bind to DNA fibers associated with Dar = 0.25 urn. the nuclear matrix (C). X 76,000; 644

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Vol. 137, No. 2, 1986

Figure

binding

to

min).

the

nuclear

Ten percent

or receptor

bound

characteristic (Fig.

of

matrix

4--continued.

under

the

conditions

crosslinked

SOS-PAGE analysis

to

matrix

nuclear

purified

intact

revealed a single glucocorticoid

45

receptor

o'r' activated

94 kDa band

receptor

complex

capacity

of

the

high

acceptor was reduced by

affinity

(Bmax = 7pM) when 65% of the remaining matrix associated

half

removed by exhaustive ONase digestion, but Furthermore, these high affinity

low

unaffected.

as described of

(22-C,

3). The binding

of

tested

by 8ekers

et al.

were visualized

(21)

affinity

Mase

CNA was

binding

was

threads,

sensitive by direct

binding

receptor colloidal gold complexes (Fig. 4). The CNA binding form the receptor bound to these threads whereas the untransformed

non-WA

binding

receptor

vated

form of to

the

receptor after

CNA-cellulose

did

not.

complexing

Binding to

of

the

colloidal

actigold

ranged from 28-45% and of this population 50-85% bound to nuclear matrices , representing receptor concentrations in the picomolar range. Attempts to visualize the high capacity sites on the nuclear matrix failed probably due to the low affinity of these binding regions. DISCUSSION Cytoplasmic estimated

concentration

of

glucocorticoid

to be 60,000 receptors

per cell. 645

receptors Binding

has been

studies

have

Vol. 137, No. 2, 1986

BIOCHEMICALAND

BIOPHYSICALRESEARCH

suggested that the nucleus contains about sites which mask the 1000 sites representing DNA sequences cated

to the

(11).

40,000 nuclear glucocorticoid

The majority of cytosolic within 30 min after the

nucleus

COMMUNICATIONS

acceptor specific

receptors

are translo-

administration

of gluco-

to adrenalectomized rats (9). Our binding data demonstrated rapid saturable binding of purified receptor complexes to nuclear matrices. A requirement for a functional DNA binding domain appears pre-requisite for binding of receptor to the nuclear matrix, corticoids

The

non-ONA

binding

unactivated

form

glucocorticoid receptor nor did free glucocorticoid of

bind to nuclear matrices to matrix sites under the conditions tested. It is significant to note that nuclear forms of both androgen and estrogen receptors which may be tightly associated with the nuclear matrix (1,7) can readily rebind steroid, whereas the activated glucocorticoid receptor complexes have a slow rate of reassociation with unbound steroid. This is consistent with a model for recycling of glucocorticoid complexes complex

did

steroid

bind

to

cytoplasm

the

not

glucocortlcoid

and

may explain

receptors

why tightly

were not identified

associated

unoccupied

here (9).

sites for glucocorticoid receptors have been previously in intact nuclei, nuclear envelopes and chromatin (11-X). These sites were located on the nuclear matrix, a structure which retains only 1% of the total genomic ONA and 10% of total nuclear protelns . The acceptors identified were consistent in affinity and capacity with those sites attributed to intact nuclei or other subnuclear Acceptor

determined

fragments,

After chromatin binding

nuclear

translocation, the exact and ONA are not known. --In vitro evidence membrane

of complexes

to specific

enhancer

responsive active genes, A recent study polymerase (22-23). mone

lished These

the attachment transcriptional

transcripts enhancer

the

sequences

increasing

and sequences

nucleoskeleton

by Jackson

of transcriptional units included

active of

genes. active

allowing

initiation

events

strongly suggests upstream from horsites

and

to

for

Cook (24)

RNA

estab-

elements to a nucleoskeleton. active RHA polymerase, nascent

Furthermore genes

binding

may mediate

interaction

with

it

was postulated attachment fixed

of

that DNA to

transcriptional

units.

Our data strongly suggest that high affinity binding of glucocorticoid receptor complexes in the nucleus is restricted to DNA attached These matrix-M anchorage points have been to the nuclear matrix. shown to be the sites of active hormone responsive genes and my be the locus of glucocorticoid responsive genes as Well. 646

BIOCHEMICAL

Vol. 137, No. 2, 1986 We have highly

recently

determined

receptor

purified

dependent nuclear

protein

kinase

gene expression

that

the

(25).

reactions

by altering

activated has

complexes activity

phosphorylation

and alteration

AND BIOPHYSICALRESEARCH DNA binding

associated It

is

well

the affinities

of binding

(26).

form

of

glucocorticoid established

are important to the

of UNA conformation

COMMUNICATIONS

that

regulation

of

proteins for WA

Conceivably,

the binding

of

receptor complexes to high affinity sites may involve subsequent phosphorylation of histone or non-histone proteins. In turn, this couid DNA superstructural

affect

especially

conformation,

if

mammalian

topoisomerase II were a target of phosphorylation, and this enzyme has recently been shown to regulate sites of initiation by RNA polymerase (27). reported here suggest

The data cept

for

been

shown

glucocorticoid to be the

action

for the first

in

the nucleus.

site of active

preservation of high affinity receptor complexes in the matrix

The

major locus

of action

time, a unifying conThe nuclear matrix has

genes and transcriptional acceptor

infers

for glucocorticoids

units.

for glucocorticoid this structure is the

sites that

in the nucleus.

ACKNOWLEDGEMENTS The

authors

Bodine

and

thank

Kiyoko

wish

to

acknowledge

Drs. James Keen, for the

Inoue

Matthev

electron

the

helpful

Chestnut

comments and

of

Peter

Tom Schmidt.

We

microscopy and Cassandra Wooten

by NIH grants ~~13531, expert typing. AM32870and Core Grant CA12227 to the Fels Research Institute. This

for

work

was

supported

--REFERENCES

1.

Barrack, E.,

2.

Hormones Vogelstein,

3.

79-85. Goldberg, G.I.,

(ed.

and Coffey, G. Litwack) B., Pardoll,

D.S. (1983) in Biochemical Actions of 10, pp. 23-90, Academic Press, New York. D.M. and Coffey, D.S. (198D1 a, 22,

Collier, I. and Cassel, A. (1983) Proc. Nat'1 Sci. U.S.A. 80, 6887-6891. Ciejek, E.M., Tsai, M-J. and O'Malley, B.W. (1983) Nature 306, 607-609. Robinson, S.I., Nelkin, B.D. and Vogelstein, B. (1982) Cell 28, 00 ,nr Il-I”O. Buttyan, R., Olsonn, C.A., Sheard, 8. and Kallos, J. (1983) -J. Biol. Chem. 258, 14366-14370. Rennie, P.S., Bruchovsky, N. and Cheng, H. (1983) -J. Biol. Chem. 258, 7623-7630. Cake, M.H. and Litwack, G. (1975) in Biochemical Actions of Hormones (ed. G. Litwack) 3, pp. 317-390, Academic Press, New --

Acad. 4. 5. 6. 7. 8.

York. 9. 10. 11.

Schmidt, T. and Litwack, G. (1982) Phys. Rev. 62, 1131-1192. Ringold, G.M. (1985) Ann. Rev. Pharmcol. Toxicol. 25, 529-566. Higgins, S.J., Baxter, J.D. and Rousseau, G.G. (1979) in Glucocorticoid Action (ed. Baxter, J.D. and Rousseau, G.G.) 136-160, Springer Verlag, New York. 647

pp.

Vol. 137, No. 2, 1986

BIOCHEMICALAND

12.

SimonS Jr.,

13.

Tomkins, Parchman,

Oevel. 14. 15. 16. 17.

18. 19.

20. 21. 22. 23. 24. 25. 26. 27.

BIOPHYSICALRESEARCH

COMMUNICATIONS

S.S., Martinez, H.M., Garcea, R.L., Baxter, J.O. and G.M. (1976) J. Biol. Chem. 251, 334-343. G.L., Cake, M.H. and Litwack, G. (1978) Mech. Ageing

7, 227-240.

m P. and Von Holt, C. (1981) Biochemistry 20, 2900-2908. Hamana, K. and Iwai, K. (1978) J. Biochem. 83, 279-286. A.S. and Litwack, G. (1985) Webb, M.L.. Miller-Oiener. Biochemist :iv_ 24, 1946-1952. Demey, J. (1984) EMSA Bulletin 14, 54-66. Pulley, J.D. and Grieve, P.A. (1975) Anal. Biochem. 64, 136-141. Hinegardner, R.T. (1971) Anal. Biochem. 39, 197-201. Munson, P.J. and Rodbard,ml980)1. Biochem. 107, 220-239. Bekers, A.M., Gijzen, H.J., Taalman, R. and Wanka, F. (1981) JUltrastruc. Res. 75, 352-362. Johnson. L.K.. &xter. J.O. and Rousseau. G.G. (1979) in Glucocorticoib Hormone Action (ed. Baiter, J;D. and Rousseau, G.G.) pp. 305-326, Springer Verlag, New York. Yamamoto, K.R. (1985) Ann. Rev. Genet. 19, 209-52. Jackson, D.A. and Cook, P.R. (1985) EM80 Journal 4, 919-925. Miller-Dlener, A., Schmidt, T.J. and Litwack, G. (1985) Proc. Natl. Acad., Sci, U.S.A 82, 4003-4007. Jungmann, R.A. and Kranias, E.G. (1977) Int. J. Biochem. 8, 819-830. Kmiec, E.B., Ryoji, M. and Worcel, A. (1986) Proc. Natl. Acad Sci. U.S.A. 83, 1305-1309.

648