Vol. 185, No. 3, 1992 June 30, 1992

ESTROGEN

BIOCHEMICAL

AND BIOPHYSICAL

RECEPTOR-INDUCED

VITELLOGENIN

RESEARCH COMMUNICATIONS Pages 944-952

BENDING

A2 GENE HORMONE

OF THE

RESPONSE

XENOPUS

ELEMENT

Michele Sabbah, Sophie Le Ricousse, Gerard Redeuilh, and Etienne-Emile INSERM

Received

U33, Lab. Hormones,

May 4,

Baulieu

94276 Le Kremlin-BicCtre

Cedex, France

1992

SUMMARY: DNA bending is increasingly proposed as an essential step for the establishment of the multiprotein complexes required for transcription initiation. Polyamines and metallic cations, known to promote DNA-bending, enhance the binding of purified estrogen receptor (ER) to the estrogen response element (ERE) of the Xenopus vitellogenin A2 gene. Using both circular permutation electrophoretic mobility and cyclization assays, we provide evidence that ER bends the DNA at the estrogen response element. The same bending occurs as a result of estrogen receptor protein binding independently of its conformational changes induced by hormone or anti-hormone. We suggest a role of the observed DNA bending in estrogen-regulated transcription. o 1992~~~~~~~~ prBss,Inc.

The estrogen binds to cis-acting

receptor (ER) is a ligand regulated transcription factor that DNA elements, termed estrogen response elements (EREs),

present in the promoter of target genes (1). The intimate mechanism by which the receptor modulates the rate of transcription is unknown. It probably involves receptor

interaction

chromatin

with

structure.

independently

In vitro,

differently

difference

factors

the estrogen

and/or

mobility

changes

in DNA

agonist or antagonist

shift assays show that the ER-DNA

mobilities

change of the complex (2). To understand better the structure

suggests a ligand-induced and conformation

(2,3).

complexes

when the receptor binds to an agonist or an antagonist.

in electrophoretic

or

receptor binds to EREs as a dimer

of the absence or presence of hormone,

Data from electrophoretic migrate

transcription

This

conformational

of ER-ERE complexes,

we have tested the influence of polyamines and several metallic cations, known to alter the DNA structure, on the binding of ER to an oligonucleotide containing the palindromic ERE sequence of the Xenopus vitellogenin A2 gene (AB-ERE). We have investigated whether binding of ER in either the absence or presence of hormone

or antagonist

permutation

gel shift

receptor, liganded 0006-291X/92 Copyright All rights

alters

the DNA

assay to demonstrate

structure. that

We have used a circular the binding

of the estrogen

or not, induces a bend at the ERE of the vitellogeain

$4.00 0 1992 by Academic Press, of reproduction in any form

Inc. reserved.

944

A2 gene.

Vol.

185,

No.

3,

1992

MATERIALS

AND

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

METHODS

Plasmids and probes preparation: The oligonucleotide AB-ERE used in the band shift experiments was: 5’-TCGAGCTCAAAGTCAGGTCACAGTGACCTGATCAAAGTI’G-3’. It contains the palindromic hormone response element of the Xenopus vitellogenin A2 gene. It was labeled by end filling with the Klenow fragment of DNA polymerase I (Boehringer). Plasmid pERE-Bend resulted from the insertion of this 40 bp oligonucleotide at the Sal1 site of the pBend2 vector that contains olylinker direct repeats on each side of the Sal1 site (4). The DNA probes, use B in the circular permutation experiments, were prepared by cutting, with the appropriate restriction enzymes, either the plasmid or a fragment encompassing the direct repeats. This fragment was obtained by polymerase chain reaction amplification, using primers complementary to flanking sequences, and purified by electxophoresis on polyacrylamide gel. The phosphatase-treated probes were labeled with [y32P]-ATP and T4 polynucleotide kinase (Boehringer) using standard procedures. The probe used in the cyclization assay was excised from plasmid pBLcat2 ERE (5) by cleavage with PvuII. It was ligated to a dodecameric NcoI linker (Boehringer), cleaved with NcoI to generate cohesive ends, purified on gel, and labeled with [y32Pl-ATP, as above, after phosphatase treatment. Electrophoretic mobility shift assays: Assays were performed as previously described (2). Binding reactions were carried out. in 20 pl binding buffer (10 mM Tris-HCl pH 7.4, 1 mM DTI’, 5% glycerol, 0.2 pg BSA and 1 pg poly (dI-dC)). In some experiments (see the results section), the binding buffer contained polyamines or metallic cations. The ER was purified as previously described (6). Cyclization assays: Labeled DNA (50 PM) was incubated with ER (150 PM) in 200 pl ligation buffer (20 mM Tris-HCl pH 7.6, 100 mM KCl, 0.5 mM D’IT, 5 mM MgC12, 1 mM ATP). Ligation was started by addition of the T4 DNA li ase (10 U/ml) (Boehringer). Aliquots (25 pl) were withdrawn at the indicate % times and the reaction quenched adjusting the concentration to 50 mM EDTA and treatment at 65°C for 10 min. Separation of ligation products was performed by electrophoresis in 4% polyacrylamide gel. The fraction of linear DNA was determined at each time by densitometric scanning of the autoradiograms or by direct gel scanning using a multitrace-master model LB-85 (Berthold Analytical Instruments).

RESULTS Polyamines

and metallic

We investigated

the influence

known to alter strongly its DNA

target

cations

. Figure

enhance

the binding

of polyamines

the DNA structure

of ER to DNA

and several metallic

(7,8), on the in vitro binding

1 presents the results from an electrophoretic

cations, of ER to mobility

shift experiment, in which the oligonucleotide probe AB-ERE was incubated with ER in the presence of various concentrations of spermine, spermidine, or of several metallic cations also known to strongly stabilize or promote DNA bending. In this experiment, a concentration of ER was used, such that only a slight ER-AZERE complexes were detectable (Fig. 1, lane 1). A dramatic increase in the amount of ER-AO-ERE complexes was observed in the presence of polyamines. Maximal complexes

formation

was reached in the presence of 0.5 mM spermine 945

or 1 mM

Vol.

185,

No.

3.

BIOCHEMICAL

1992

Spermine '01

05

2

3

1

AND

BIOPHYSICAL

Spermidine 3"Ol

1

3"l

7

8

9

5

COMMUNICATIONS

ZnClp

CaCl2

M&b

05

RESEARCH

lo"1

5

IO"50

12 13

14 15

100 200'

FEZ? DNA-

1

4

5

6

10

11

16

17 18

Fi re 1. Effects of polyamines and metallic cations on the DNA binding of ER. T*P]-labeled ERF probe (0 4 ng) was incubated with the indicated concentrations of spermine, spermidine or chloride salts of M 2 + , Ca2 + (all in mM) and Zn2+ (in pM) prior to addition of ER (10 fmo k es). Protein-DNA complexes were separated from protein-free DNA by native electrophoresis in 6% polyacrylamide gel, and visualized by autoradiography. The control (lane 1) was run on the same gel.

spermidine. Addition of Caz+ or Mg2+ also induced a significant increase in the amounts of ER-DNA complexes. The maximal effect was seen for 5-10 mM of Mg2+ and 1 mM of Caz + . In contrast, in the presence of Z&+ , a cation very potent to induce DNA bending, binding of ER to AB-ERE was undetectable. This result can be related to previous observations showing that an excess of transition metals might interfere with the tetrahedral coordination of zinc, altering the conformation of the zinc finger domain of the ER, and thus inhibiting its DNA binding activity (9). These results clearly show that polyamines or divalent metallic cations strongly increase the affinity of ER for its cognate DNA, probably as a result of an alteration of the DNA motif configuration. Although polyamines were reported to stabilize the estrogen receptor protein, they were also reported to increase the affinity of the ER towards synthetic non-specific DNA containing alternating purine-pyrimidine sequences by promoting the conversion of the polynucleotide to the left-handed Z-DNA form (10,111. To test if the alteration of the DNA structure motif results from a A-to-Z DNA conformation transition, we have performed experiments in the presence of distamycin. This drug has an intrinsic twist that favors its insertion into the minor groove of the DNA (12). It fails to bind left-handed Z-DNA; in fact, binding of distamycin and related molecules to DNA favors A-to-B and Z-to-B helix transitions. Distamycin incubated with DNA at concentrations between 2 pM to 2 946

Vol.

185, No. 3, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

1234 Figure 2. Effect of distamycin on the DNA-binding of ER. The [y32P]-labeled ERE probe was incubated with increasing concentration of distamycin (lane 2: 2 pM, 3: 2 nM, 4: 2pM) prior to addition of ER (70 fmoles). Protein-DNA complexes were separated as in Fig. 1. The control (lane 1) was run on the same gel.

pM prior

to addition

of ER did not interfer

with

the formation

of ER-DNA

complexes. Figure 2 shows results obtained in the presence of Mg2+, which demonstrate that distamycin neither changed the amount nor the electrophoretic

of the complexes. Similar results were obtained in the absence of polyamines or metallic cations and in the presence of polyamines (data not shown.) Three conclusions can be drawn from these results: 1) Effect of polyamines mobility

results from their capacity to induce DNA bending and not from the conversion

of

the polynucleotide to the left-handed Z-DNA conformation; 2) Contacts of ER with the DNA minor groove are not required for its high affinity binding to ERR. This agrees with data recently reported on NMR studies of peptide containing the DNA-binding domain of ER and X-ray crystal analysis of GR-DNA binding domain complex (13,14); and 3) Binding of ER to linear DNA is energetically less favorable than binding to bent DNA. ER induces

bending

at the A2 ERE motif

To test if the binding of purified ER to ERB induces a bend, we have used a circular permutation electrophoretic mobility shift assay. It has been demonstrated that, in gel electrophoresis, bent DNA fragments migrate slower than a linear DNA fragment of same size and base composition. This behavior is also observed for protein-induced DNA bending (15). We have utilized set of DNA probes, of same length (157 bp) and base composition, containing a single AS-ERR located at different

positions

within

the DNA fragment 947

(Fig. 3A). Results

of the

Vol.

185, No. 3, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

A A

B

C

DEF

G

H

A

C

B

DEF

G

H

-

A

-

B C DEFP G

ABCDEFGH

Fimre probes

“/\

Ab

3. (A) Structure were generated

of DNA probes used for circular permutation analysis. The cleavage at sites within two by restriction endonuclease

direct repeated sequences flanking the ER bindin site (dark box). Restriction endonucleases used are: Mlu I (A), Nhe I (B), Xho I (8 ).,Eco RV (D), Sma I (E), Nru I (F), Kpn I (G) and Bam HI (H). (B) Electrophoresls mobility shift assay. The [yJ2P]-labeled

DNA

fragments

were

incubated

with

the ER and subjected

to

electrophoresis on 6% poly acrylamide gel. Monoclonal antibody H222 anti-ER (1 pg) included in the binding reaction mixture containing the Xho probe was used as control of protein specificity (lane Ab) (C). The mobility of the ER-DNA complexes is plotted against the position of the center of the ERE binding site within the 157 bp fragments. The bend center is located around position 82 bp as determined by an extrapolation of the flanks of the curve (15,17).

electrophoretic fragments

mobility

displayed

shift

assay are shown

the same electrophoretic

in Fig. 3B. All mobility

naked

regardless

what

DNA the

position of the AB-ERE binding site was. This indicates the absence of intrinsic curvature in the DNA fragments. In contrast, migration of ER-DNA complexes was clearly dependent

on the location of the ER-binding

site within

the fragment.

The fastest migration of the complex occured when the binding site approached one of the ends of the DNA fragment. When the binding site was more central, the complexes migrated slower. This variation in electrophoretic mobilities, upon the position of the ER binding site within the DNA fragment, suggests that ER binding

induces a DNA bend. 948

Vol.

185,

No.

3,

BIOCHEMICAL

1992

There is substantial

AND

BIOPHYSICAL

the effect of hormone-free

receptor

antagonist ICI 164,384 on the ER-induced dependence of the electrophoretic mobility, the DNA fragment,

between

DNA

caused by protein-induced

investigated

the

electrophoretic reduced

or receptor

shift

bending

complexed

and an increased

or untwisting

dependence

with

difference

only

flexibility

of the DNA,

on the

assay. Running

did not affect significantly

the

circular

by 3% (data

permutation

not shown).

the electrophoretic

mobilities

of receptor-DNA

complexes

of the we have

the gel at 25°C instead mobility

conclude that the position-dependent differences in electrophoretic ER-DNA complexes are consistent with DNA bending. Relative

to

We have

DNA bending. In all cases a similar upon the position of the ERE within

melting

temperature

mobility

the mobility

temperature

agonist, or antagonist.

was observed (data not shown).

To distinguish DNA

COMMUNICATIONS

evidence that, in vitro, the receptor binds specifically

the ERE in the presence or absence of hormone, examined

RESEARCH

of 4°C

Since

the

pattern,

we

mobilities

were plotted

versus

of the

position of the center of the ERE within the 157 bp fragment (Fig.30 The bending angle (a) was calculated using the empirical equation pM/pE = co&2 (16) where pM is the relative mobility of the complex with ER bound at the center of the DNA fragment and pE is the relative mobility of the complex with ER bound at the end of the DNA fragment. According to this equation the calculated bending angle is 50+8” and the bending occurred, within the experimental error (5 bp), near the center of the ERE. An alternate method to study protein-induced DNA bending has been developed (18). If a protein induces a bend in a linear DNA fragment, it will decrease the distance between cohesive ends and hence increase the probability ring closure. This can be monitored

under defined conditions

in which the rate of

intramolecular

ligation

the equilibrium

fraction of DNA molecules with cohesive ends. We determined

rate

of cyclization

pBLcat2ERE

of a DNA fragment

of the 484 bp PvuII

of

by T4 DNA ligase is proportional fragment

derived

from

to the

plasmid

(5), in the absence or presence of various amounts of ER. This linear

DNA fragment contained an ERE binding site close to its center and cohesive ends were generated by addition of a linker (see methods). The products of cyclization were resolved by electrophoresis

on 4% polyacrylamide

gels. Results are presented

in Fig. 4. The ratio L+jLo (Lt = linear DNA remaining at time t and La = linear DNA at time 0) is plotted as a function of time. A 3-fold molar excess of ER over binding sites increased the rate of cyclization approximately 1.5 fold. Higher concentrations of receptor did not raise further the rate of cyclization but interfered

with cyclization,

presumably

through

non-specific

ER binding

to DNA

termini. Thus only a narrow range of ER concentrations can be used to measure the rate of cyclization. Temperature, between 10 and 25°C did not significantly affect the magnitude of circularization rate enhancement, resulting from ERinduced DNA bending, and revealed the absence of ER-induced DNA untwisting. 949

Vol.

185, No. 3, 1992

BIOCHEMICAL

AND BIOPHYSICAL

Time

RESEARCH COMMUNICATIONS

(min)

Figure 4. Li ase-mediated DNA cyclization assay. Kinetics of circularization of the 484 bp f ragment (50 pM) were performed at 25°C in the absence (0) or presence (0) of 150 pM of ER.

DISCUSSION

This study provides evidence that binding of ER alters the structure of its target DNA. The alteration is most likely to be a bend or a kink of the target DNA located within the ER.E rather than a local DNA-flexibility due to melting or untwisting. The enhancing effect of polyamines or metallic cations on the binding of ER to ERE is presumably due to salt effects on charges repulsion of the DNA phosphate backbone. When charges are neutralized, the energy requirement for bending is lowered and thus the binding of the receptor to DNA is enhanced. This indicates that binding of ER to linear DNA is energetically less favorable than binding to bent DNA. DNA bending may be only the most energetically DNA configuration for interaction with the receptor and have no role in a regulatory function. However, the release of the free energy stored in the bend can be involved in the dissociation of receptor from DNA. Several studies have demonstrated that, in vitro, the receptor binds to ERR with similar affinity, regardless the presence of agonist or antagonist ligands (2,3). However, in electrophoretic mobility shift assays, all ER-antagonists complexes bound to the ERR migrate slower than ER-estradiol or ER-agonist complexes bound to the same ERJZ. In the absence of ligand intermediate shifted complexes are observed. These modifications in electrophoretic mobilities have been attributed to a ligand-induced conformational change of the complexes that occurs probably within the ligand-binding domain of the ER (2, 19, 20). The presence or absence of ligand do not influence the magnitude of the DNA- bend angle. This suggests that the ligand-induced conformational modification of ER changes neither

the affinity

of its DNA-binding

domain

for the target

DNA

nor

the magnitude of the induced-bend angle. These results indicate that the bending occurs as a result of ER protein binding to its cognate DNA target independently of its ligand-dependent conformation. 950

Vol.

185,

No.

3,

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

What is the significance of DNA-bending in the mechanism of hormone action? Transcription initiation requires the establishement of multiproteic DNA complexes in which protein-protein interactions are involved (21). In these complexes the DNA conformation is likely to be important. It is generally considered that protein-induced DNA bending can facilitate interaction between proteins bound to separate sites on DNA. Transcriptional transactivation function of ER is induced by hormone, probably through ligand-dependent transconformation within the hormone binding domain (TAF-2) (22). This transconformation is not achieved in the presence of antagonists. In this context, it is conceivable that DNA bend may bring the receptor domain, involved in transcriptional activation (TAF-2), into close vicinity of other regulatory factor(s) enabling protein- protein interaction to occur. This paper was ready for publication when a similar study appeared (23), using the Xenopus laevis ER DNA-binding domain and finding 34” for the bending angle of a slightly different ERE DNA.

ACKNOWLEDGMENTS

We thank Dr. Sanker Adhya (NIH, Bethesda) for the gift of plasmid pBend2; Dr. Helene Richard-Foy for discussion of the manuscript; Philippe Astier for technical help and Jean-Claude Lambert for art work. This work was partially supported by the Centre National de la Recherche Scientifique.

REFERENCES

k 3. 4. 2 i: 9. ::* 12: ::: 15.

Beam, M. (1989) Cell, 56,335-344. Sabbah, M., Gouilleux, F., Sola, B., Redeuilh, G., and Baulieu, E.E. (1991) Proc. Natl. Acad. Sci. USA 88,390-394. Murdoch, F.E., Meier, D.A., Furlow, J.D., Grunwald, K.A.A., and Gorski, J. (1990) Biochemistry 29,8377-8385. Kim, J., Zwieb, C., Wu, C., and Adh a, S. (1989) Gene 85,15-23. Klock, G., Strahle, U., and Schiitz, ii . (1987) Nature 329,734-736. Redeuilh, G., Montcharmont, B., Secco, C., and Baulieu, E.E. (1987) J. Biol. Chem. 262,6969-6975. Laudon, H.C., and Griffith, J.D. (1987) Biochemistry 25,3759-3762. Fenerstein, B.G., Pattabiramau, N., and Marton, L.J. (1990) Nucl. Acids Res. 18,1271-1282. Thiescu, H.J., and Bach, C. (1991) B&hem. Biophys. Res. Commun. 176, 551-557. Thomas, T., and Kiang, D.T. (1987) Cancer Res. 47,1799-1804. Thomas, T., and Kiang, D.T. (1988) Nucl. Acids Res. 16,4705-4720. Kopka, L.M., Yoon, C., Goodsell, D., Pjura, P., and Dickerson, R.E. (1985) Proc. Natl. Acad. Sci. USA 82,1376-1380. Schwarbe, J.W.R., Neuhans, D., and Rhodes, D. (1990) Nature 348,458461. Luisi, B.F., Xu, W.X., Otwinowski, Z., Freedman, L.P., and Yamamoto, K.R. (1991) Nature 352,497~505. Wu, H.M., and Crothers, D.M. (1984) Nature 308,509-513. 951

Vol.

16. 17. 18. 19. 20. 21. 22. 23.

185, No. 3, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

Thompson, J.F., and Laudy, A. (1988) Nucl. Acids Res. 16,9687-9705. Gartenberg, M.R., and Crothers, D.M. (1988) Nature 333,824-829. Kotharz, D., Fritsch, A., and But, H. (1986) EMBO J. 5,799-803. Pakdel, F., and Katzenellenbogen, B.S. (1992) J. Biol. Chem. 267,3429-3437. Danielian, P.S., White, R., Lees, J.A., and Parker, M.G. (1992) EMBO J. 11, 1025-1033. Mitchell, P.J., and Tjian, R. (1989) Science 245,371-378. Webster, N.J.G., Green, S., Jin, J.R., and Chambon, P. (1988) Cell 54, 199207. Nardulli, A.J., and Shapiro, D.J. (1992) Mol. Cell. Biol. 12,2037-2042.

952