Proc. Natl. Acad. Sci. USA
Vol. 90, pp. 11117-11121, December 1993
Genomic binding-site cloning reveals an estrogen-responsive gene
that encodes a RING finger protein
(transcription factor/target gene/zinc finger/estrogen-responsive element)
SATOSHI INOUE*, AKIRA OluMo*, TAKAYUKI HOSOIt, SHIGERU KONDO*, HIDEo TOYOSHIMA*,
TAKASHI KONDO*, AKIRA IKEGAMIt, YASUYOSHI OUCHIt, HAJIME ORIMOt, AND MASAMI MURAMATSU*§
*Department of Biochemistry, Saitama Medical School, 38 Moro-Hongo, Moroyama-machi, Iruma-gun, Saitama, 350.04, Japan; and tDepartment of Geriatrics
and *3rdDepartmentof Internal Medicine, FacultyofMedicine, TheUniversityofTokyo, Hongo, Bunkyo-ku, Tokyo 113, Japan
Communicated by Rachmiel Levine, August 12, 1993
were isolated from human genomic DNA by using a recombi-
nant ER protein. Using one of these fragments as a probe, we
have identified an estrgen-responsive gene that encodes a
putative zinc finger protein. It has aRING finger motifpresent
in afamily ofapparentDNA-binding proteins and is designated
estrogen-responsive finger protein (efp). efp cDNA contains a
consensus estrogen-responsive element at the 3' untranslated
region that can act as a downstream estrogen-dependent en-
hancer. Moreover, efp is regulated by estrogen as demon-
strated at both the mRNA and the protein level in ER-positive
cells derivedfrommammary gland. These data suggest thatefp
may represent an estrogen-responsive transcription factor that
mediates phenotypic expression ofthe diverse estrogen action.
Thus, the genomic binding-site cloning may be applicable for
isolation of the target genes of other transcription factors.
Estrogen receptor (ER)-binding fragments
Estrogen is a hormone that is secreted from the ovary and
causes the development of female organs. It regulates
growth, differentiation, and the function of target cells. It is
assumed that the estrogen receptor (ER), a member of the
steroid/thyroid hormone receptor superfamily (1, 2), medi-
ates this actionby bindingligand dependently to the estrogen-
responsive element (ERE) that exists in the enhancer region
of target genes, regulating their transcription directly. How-
ever, in contrast to the diverse estrogen action on a variety
of organs, tissues, and cells, relatively few genes are known
to respond to ER. Those include vitellogenin, prolactin, pS2,
ovalbumin, and the progesterone receptor (cited in ref. 3).
Important genes that regulate the growth and differentiation
of female organs such as mammary gland and uterus, for
example, in response to estrogen have not yet been identi-
fied. Although greaterthan one-third ofhuman breast cancers
are known to be responsive to estrogen and hence to anti-
hormone treatment (4), the mechanism of the growth pro-
motion ofthese cancer cells by ER is still unclear. Moreover,
the ER has been identified in various nuclei of brains,
implicating some roles of estrogen in the central nervous
system (5). Estrogen, ER, and the estrogen-responsive genes
must play important roles in a number ofvital organs besides
These circumstances prompted us to isolate more estro-
gen-responsive genes to understand the molecular physiol-
ogy of estrogen action. Recently, we developed a method to
isolate ER-binding fragments from human genomic DNA (3).
All these fragments contained consensus EREs, some of
which showed estrogen-dependent enhancer activity. We
then isolated several more ER-binding fragments and used
them as probes for identification of target genes adjacent to
The publication costs ofthis article were defrayed in part by page charge
payment. This article must therefore be hereby marked "advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
genomic ER-binding sites. A gene that encodes a protein,
referred to as the estrogen-responsive finger protein (efp),
containing zinc finger motifs and that has been shown to be
regulated by estrogen has now been isolated.
MATERIALS AND METHODS
Isolation of ER-Binding Fragments and efp cDNA Clones.
The EO fragment was isolated from human genomic DNA as
described (3). Briefly, high molecular weight DNA from
HeLa cells (10 &g) was digested with Pst I and BamHI, and
the nitrocellulose filter binding selection was performed with
the recombinant DNA-binding domain of the ER (10 pmol).
The DNA trapped by the filter (-10 ng) was eluted, cloned
into the plasmid vectorpUC18 (PstI-BamHI), and amplified.
Then, the plasmid DNA (10 pg) was again incubated with the
DNA-binding domain of ER (10 pmol), and the selection
cycle was repeated five times.
AgtlO and AZAPII cDNA libraries were prepared from
poly(A)+ RNA of human placenta, transfected into Esche-
richia coli C600hfl or XL1-Blue, and screened by hybridiza-
tion with the 32P-labeled EO fragment (see Results). Clone
AC1, which had the largest insert including a long 3' untrans-
lated region (accession no. D21205), and an overlapping AC3
clone, which had the longest open reading frame, were
further characterized. Both strands of the cDNA insert of
AC3 were completely sequenced by the dideoxynucleotide
method (6) with Sequenase (United States Biochemical).
Northern Blot Analysis and Immunological Procedure. For
Northern blot analysis, extraction ofRNA, fractionation on
formaldehyde/agarose gels, and hybridization conditions
were as described (7). The hybridization probe was the
32P-labeled EO, ERcDNA (8) or ,B-actincDNA fragment. The
IgG fraction of a rabbit anti-DDVRNRQQDVRMTANRK-
VEQ antiserawas prepared as anti-efp antibody (by courtesy
ofMedical and Biological Laboratories, Ina, Japan) and used
at 1:1000 dilution for immunoblotting. The cDNA insert of
AC3 was cloned into the EcoRI site of the pSSRa expression
vector (9) with the SRa promoter in the sense orientation to
construct pSSRacefp. Either pSSRa (10 ug) or pSSRacefp
(10 ug) was transfected into COS-7 cells by the calcium-
phosphate precipitation method (10). Nuclear extracts were
prepared as described (11), and immunoblotting using the
anti-efp antibody was performed as described (12). In some
experiments, anti-efp antibody was preincubated with anti-
gen (0.1 mg/ml) overnight at 4°C.
Abbreviations: efp, estrogen-responsive fingerprotein; ER, estrogen
receptor; ERE, estrogen-responsive element; CAT, chlorampheni-
col acetyltransferase; FBS, fetal bovine serum; vitERE, Xenopus
vitellogenin gene ERE.
§To whom reprint requests should be addressed.
$The sequence reported in this paper has been deposited in the
GenBank data base (accession no. D21205).
Biochemistry: Inoue et al.
;kC3 -ORF:630 AA
;&ci ~~~~~~~~~~~.pl AB
Proc. Natl. Acad. Sci. USA 90 (1993)
by the solid box and compared with vitERE and consensus ERE (conERE) sequences. The open reading frame is indicated by the open box,
and the potential polyadenylylation site is denoted by a star. (B) Northern blot analysis of human placenta poly(A)+ RNA (5 ug) with the EO
probe. Migration positions of ribosomal RNA markers are shown on the right. (C) Estrogen-dependent enhancer activity of the EO fragment.
The reporter plasmids, tk-cat-EO, vitERE-tk-cat, and tk-cat (2 ug each), were cotransfected with (+) or without (-) 0.1 pg ofthe ERexpression
vectorpSV2RcER(ER) intoCOS-7 cells. They were incubated in the presence (+) orabsence (-) of10nM 17,B-estradiol (E2), andCAT activities
(A) Restriction map ofthe EO fragment and the cDNAs (AC1, AC3) cloned thereby as probe. The ERE sequence (EREO) is indicated
Ceil Culture and Chloramphenicol Acetyltransferase (CAT)
Assay. HBL-100 cells (13) were grown to a subconfluent state
in Dulbecco's modified Eagle's medium containing 10% fetal
bovine serum (FBS). One day prior to estrogen treatment, the
medium was changed to phenol red-free Dulbecco's modified
Eagle's medium containing 10% dextran-coated charcoal-
CGC GGG TGC AGC AGT TGT GTC CCG ACC CCT GGG AGC GCC ATG GCA GAG CTG TGC
CCC CTG GCC GAG GAG CTG TCG TGC TCC ATC TGC CTG GAG CCC TTC AAG GAG CCG
GTC ACC ACT CCG TGC GGC CAC AAC TTC TGC GGG TCG TGC CTG AAT GAG ACG TGG
GCA GTC CAG GGC TCG CCA TAC CTG TGC CCG CAG TGC CGC GCC GTC TAC CAG GCG
CGA CCG CAG CTG CAC AAG AAC ACG GTG CTG TGC AAC GTG GTG GAG CAG TTC CTG
CAG GCC GAC CTG GCC CGG GAG CCA CCC GCC GAC GTC TGG ACG CCG CCC GCC CGC
GCC TCT GCA CCC AGC CCG AAT GCC CAG GTG GCC TGC GAC CAC TGC CTG AAG GAG
GCC GCC GTG AAG ACG TGC TTG GTG TGC ATG GCC TCC TTC TGT CAG GAG CAC CTG
CAG CCG CAC TTC GAC AGC CCC GCC TTC CAG GAC CAC CCG CTG CAG CCG CCC GTT
CGC GAC CTG TTG CGC CGC AAA TGT TCC CAG CAC AAT CGG CTG CGG GAA TTT TTC
TGC CCC GAG CAC AGC GAG TGC ATC TGC CAC ATC TGC CTG GTG GAG CAT AAG ACC
TGC TCT CCC GCG TCC CTG AGC CAG GCC AGC GCC GAC CTG GAG GCC ACC CTG AGG
CAC AAA CTA ACT GTC ATG TAC AGT CAG ATC AAC GGG GCG TCG AGA GCA CTG GAT
GAT GTG AGA AAC AGG CAG CAG GAT GTG CGG ATG ACT GCA AAC AGA AAG GTG GAG
CAG CTA CAA CAA GAA TAC ACG GAA ATG AAG GCT CTC TTG GAC GCC TCA GAG ACC
ACC TCG ACA AGG AAG ATA AAG GAA GAG GAG AAG AGG GTC AAC AGC AAG TTT GAC
ACC ATT TAT CAG ATT CTC CTC AAG AAG AAG AGT GAG ATC CAG ACC TTG AAG GAG
GAG ATT GAA CAG AGC CTG ACC AAG AGG GAT GAG TTC GAG TTT CTG GAG AAA GCA
TCA AAA CTG CGA GGA ATC TCA ACA AAG CCA GTC TAC ATC CCC GAG GTG GAA CTG 1026
treated FBS (3). The EO fragmentwas cloned into the Sma I
site of pBLCAT2 (tk-cat) (14) in the sense orientation(5'-+
3') to construct tk-cat-EO. Theoligonucleotidecontaining the
wild-type ERE oftheXenopus viteliogenin geneA2 enhancer
(vitERE) was synthesized and inserted at the upstream
position ofpBLCAT2 to construct vitERE-tk-cat(3). CAT
AAC CAC AAG CTG ATA AAA GGC ATC CAC CAG AGC ACC ATA GAC CTC AAA AAC GAG 1080
CTG AAG CAG TGC ATC GOG CGG CTC CAG GAG CTC ACC CCC AGT TCA GGT GAC CCT 1134
GGA GAG CAT GAC CCA GCG TCC ACA CAC AAA TCC ACA CGC CCT GTG AAG AAG GTC 1188
TCC AAA GAG GAA AAG AAA TCC AAG AAA CCT CCC CCT GTC CCT GCC TTA CCC AGC 1242
AAG CTT CCC ACG TTT GGA GCC CCG GAA CAG TTA GTG GAT TTA AAA CAA GCT GGC 1296
TTG GAG GCT GCA GCC AAA GCC ACC AGC TCA CAT CCG AAC TCA ACA TCT CTC AAG 1350
GCC AAG GTG CTG GAG ACC TTC CTG GCC AAG TCC AGA CCT GAG CTC CTG GAG TAT 1404
TAC ATT AAA GTC ATC CTG GAC TAC AAC ACC GCC CAC AAC AAA GTG GCT CTG TCA 1458
GAG TGC TAT ACA GTA GCT TCT GTG GCT GAG ATG CCT CAG AAC TAC COG CCG CAT 1512
CCC CAG AGG TTC ACA TAC TGC TCT CAG GTG CTG GGC CTG CAC TGC TAC AAG AAG 1566
GGG ATC CAC TAC TGG GAG GTG GAG CTG CAG AAG AAC AAC TTC TGT G0G GTA GGC 1620
ATC TGC TAC GGA AGC ATG AAC CGG CAG G0C CCA GAA AGC AGG CTC GGC CGC AAC 1674
AGC GCC TCC TGG TGC GTG GAG TGOG TTC AAC ACC AAG ATC TCT GCC TOG CAC AAT 1728
AAC GTG GAG AAA ACC CTG CCC TCC ACC AAG GCC ACG CGG GTG GCC GTO CTT CTC 1782
AAC TGT GAC CAC GGC TTT GTC ATC TTC TTC GCT GTT GCC GAC AAG GTC CAC.CTG 1836
ATG TAT AAG TTC AGG GTG GAC TTT ACT GAG GCT TTG TAC CCG GCT TTC TOGGTA 1890
TTT TCT GCT GGT CCC ACA CTC TCC ATC TGC TCC CCC AAG TAG GCA GOC TGOT AGG 1944
CAC TTG GGC TGOA CMG CCT GCA GAA GTC CCA AGA CCC TAG TGA AAA TAC AGC AGG 1998
CAG AAC TCT CCT TOG ATA ATT CCC CCA AGA GGT CCC CAA GGA TTG GGA GCA TGG 2052
GTG ATT GTG TTG TOG GCG AGG AGG CGT TTC CAC CCC CTG GTGCCT ATC AGG GCA 2160
TGA CCT ACT CCC CAT TGT TCT GGA AAT CTC CAG GCT GCT GGG CAG CTG GOC 2214
AGC TGG GCA GAG CTC TGG GAA GTG AAG TCA TGA OTG CCC GAT TCC TCT TAGAGA 2268
AAA TCC ATA 0CC TTC AGA TCT TOG TGT TTT GAA
AGC TGG CGGGAG GGT GGG AGG TGG GAT TTA GCC AGO AAA G00
GTG AGA 2106
regions, RING finger (R), Bl box (B1), B2 box (B2), and coiled-coil domains(C),are shown on theright. Cysteineand histidine residues that
may be involved in metal finger formation are circled.
Nucleotide sequence and deduced amino acid sequence (one-letter symbols)ofefpdetermined from the AC3 cDNA clone. Four
Biochemistry: Inoue et al.
assay was performed as described (3). Briefly, 1 x 106 COS-7
cells were plated in 60-mm Petri dishes and maintained in
Eagle's minimal essential medium containing 10% FBS for 24
h. One hour prior to transfection, the medium was replaced
with phenol red-free Eagle's minimal essential medium con-
taining 10% dextran-coated charcoal-treated FBS. Cells were
transfected by the calcium phosphate precipitation method
(10) with 0.1 ,g ofpSV2RcER (8), 2 ,gof reporter plasmids,
and 2 ,g of the pCH110 /3-galactosidase expression vector
(Pharmacia), used as an internal control to normalize for
variations in transfection efficiency. In some experiments,
pSV2RcER was omitted. The total amount of DNA trans-
fected was made up to 20Mgwith carrier DNA pGEM3Zf(-)
(Promega). After 12 h of incubation, the cells were divided
into two dishes and cultured further in the absence or
presence of 10 nM 17,3estradiol for 24 h. Cell extracts were
assayed for protein concentration and CAT activity (15). The
experiment was carried out four times, and a representative
pattern is shown.
Isolation of Estrogen-Responsive Gene Fragments and
cDNA. Using one of the ER-binding fragments named EO
(Fig. 1A) as aprobe, we detected positive signals by Northern
blot analysis in human placenta mRNA (Fig. 1B). This
suggested the existence of a transcribed region adjacent to
this genomic DNA fragment. We then screened human
placenta cDNA libraries with the EO probe and found >10
positive cDNAs out of 500,000 plaques. Restriction mapping
and partial sequencing indicated that all the clones were
derived from the same RNA. As shown in Fig. 1A, AC1 had
the largest insert containing a poly(A) tail, and AC3 had the
longest open reading frame (see below), a part of which
overlapped the 5' region of AC1. AC1 contained the complete
EO fragment, and AC3 contained the Pst I-EcoRI fragment of
EO as an exon at the 3' untranslated region. The EO fragment
contains the consensus ERE sequence (EREO), which is
compared with the vitERE sequence (Fig. 1A). To confirm
the estrogen-dependent enhancer activity of this region, the
EO fragment was inserted into a downstream position of a
reporter vector having a herpes simplex virus thymidine
kinase promoter to construct the tk-cat-EO. The reporter
plasmids were cotransfected with or without an ER expres-
sion vector into COS-7 cells. The CAT activity was stimu-
lated significantly only in the presence of both the ER
expression vector and17p-estradiol(Fig. 1C). The estrogen-
dependent enhancer activity of tk-cat-EO was thus demon-
strated. Northern blot analysis of human placenta mRNA
showed positive bands of 6 kb and 10 kb (Fig. 1B), and the
lower band corresponded to the size of the cDNA picked up
by AC1 and AC3. The 10-kb band may correspond to a high
molecular mRNA precursor, a splicing variant, or an mRNA
with alternative end. Alternatively, the 10-kb band may be
derived from another gene with strong homology to efp in this
Structure of efp. Fig. 2 shows the nucleotide sequence and
predicted amino acid sequence of the AC3 cDNA. Interest-
ingly, the predicted protein contains a zinc finger motif called
the RING finger (16, 17) and is named efp. The nucleotide
sequence surrounding the putative initiation codon closely
resembles the consensus sequence of Kozak (18). The pre-
dicted efp protein consists of 630 amino acids, with a calcu-
lated relative molecular mass (Mr) of 70,986. Computer-
assisted analysis and data base search show that efp contains
structurally characteristic regions: a coiled-coil region and
cysteine-rich regions including a RING finger, a Bl box, and
a B2 box (17). There is no signal peptide sequence or a
transmembrane region, suggesting that efp is an intracellular
Proc. Natl. Acad. Sci. USA 90 (1993)
efp belongs to a family of nuclear proteins containing a
RING finger motif (Fig. 3A). These proteins not only have a
common variant zinc finger structure but also appear to be
related to cell regulation and differentiation proteins (see
Discussion). These proteins are assumed to bind with Zn2+
and then to DNA using the zinc finger domains (17) (Fig. 3B).
Some members of this family possess a second CH domain,
the B-box domain, downstream of the RING finger (17, 19)
(Fig. 3C). Interestingly, efp, PML (20-22), and T18 (23)
contain two B box motifs and appear to form a subgroup as
shown in Fig. 3D. Furthermore, all members of the B-box-
containing family possess a predicted coiled-coil domain
downstream of the B box (Fig. 3D).
ML DCGHNICCACLARC WGT
SA DCNHSFCRACITLN YESNRNTDGKGNCPVCRVP
SI ECGHSFCQECISQV GKG
RCDTFPCMHRFCI PCMKTW MQL
TRCK ESA DFWCFECEQLLCAKCFEA
SS -A/ Ro
PFHKKEQL KLYCETCDKLTCRDC QLLE H
EKHR EPL KLYCEEDQMPICVVCDRSRE H
AQHG EKL RLFCRKDMMVICWLCERSQE H
AVHG ERL HLFCEKDGKALCWVCAQSRK H
SEHD ERL KLYCKDDGTLSCVICRDSLK H
SQHN RLR EFFCPEHSECICHIC
DZinc Finger Domain
B: rfp, rpt-1, SSA/Ro,
RING finger domain. Conserved cysteine and histidine residues are
denoted in bold type. In the T18 oncogene, the number 38 in
parentheses indicates the position of 38 residues omitted to maintain
the alignment. (B) Possible metal ion coordination ofthe RING finger
motif. (C) Alignment and comparison of proteins containing the B
box domain. (D) Schematic representation of the B-box-containing
proteins. The solid boxes represent the RING finger, the open boxes
represent the B boxes, and the wavy-lined boxes represent the
coiled-coil domain. The brackets represent a gap. Members of each
subgroup (A and B) are shown.
(A) Alignment and comparison of proteins containing the
Proc. Natl. Acad. Sci. USA 90 (1993)
- L1 )
tracts (10jg)were prepared from COS-7 cells transfected with the
mock expression vector (lane 1), from HBL-100 cells (lane 2), and
from COS-7 cells transfected with the AC3 expression vector (lanes
3 and 4). Anti-efp antibody (lanes 1-3) or anti-efp antibody preincu-
bated with the antigen (lane 4) was used for immunoblotting. Migra-
tion positions of the molecular size markers (in kDa) are shown on
Immunoblotting using anti-efp antibody. Nuclear ex-
Identification of the efp Protein in the Cell. To detect the
specific efp product, a polyclonal antibody against a partial
peptide sequence ofthe efp protein was prepared. By immu-
noblotting, this anti-efp antibody detected a specific band of
70 kDa in the nuclear extracts ofHBL-100 cells derived from
the human mammary gland (Fig. 4, lane 2). The molecular
size of this band calculated from the relative mobility corre-
sponded to the predicted Mr ofefp. Moreover, the size ofthe
band of the natural product agreed with that in COS-7 cells
transfected with the AC3 expression vector (Fig. 4, lane 3).
The band was not detected in COS-7 cells transfected with
the control expression vector (Fig. 4, lane 1). The specific
band in COS-7 cells transfected with the AC3 expression
vector was blocked by preincubation ofthe anti-efp antibody
with the synthetic peptide that was used for immunization
(Fig. 4, lane 4). Immunostaining of COS-7 cells transfected
with the efp expression vector demonstrated the nuclear
localization of the efp products (data not shown).
Estrogen Responsiveness ofefp Expression. To demonstrate
that efp is actually regulated by estrogen in vivo, we treated
the HBL-100 cells with estrogen and followed the efp mRNA
level by Northern blot analysis (Fig. SA). The level of efp
mRNA was elevated from 2 h after estrogen treatment,
reached a peak (3.5 times) at 10 h, and then returned to the
initial level by 20 h. The mRNA level of ER and ,B-actin did
not change. Immunoblotting analysis showed that the efp
protein was also increased, reaching the highest level at 10 h
and then decreasing by 20 h (Fig. 5B). These results confirm
the conclusion obtained by mRNA analysis that efp is regu-
lated in vivo by estrogen.
In this study, we used genomic binding-site cloning to isolate
the estrogen-responsive gene efp, whose ERE is located in an
exon corresponding to the 3' untranslated region ofmRNA.
CAT assay has shown that it can act as a downstream
estrogen-dependent enhancer in the presence of ER in the
cell. A number ofenhancers are known to exist in introns or
even in 3' untranslated regions of mRNA-e.g., the case of
K-fgf (24). It is noteworthy that some target genes of Dro-
sophila transcription factor Ultrabithorax (Ubx) are located
adjacent to the Ubx-binding sites in genomic DNA (25), and
one of the binding sites is present at the 3' region of a target
gene (26). Here, we have shown that thegenomic binding-site
cloning is potentially useful to obtain the direct target genes
of mammalian transcription factors.
This study has shown that one ofthe target genes ofER is
a zinc finger protein having a new class of motif, the RING
finger. Members of the RING finger family are putative
DNA-binding proteins, some of which are implicated in
transcriptional regulation, DNA repair, and site-specific re-
combination. PML is a putative transcription factor that was
found fused to the retinoic acid receptor a in promyelocytic
leukemia translocations (20-22). rfp is a proposed regulator
of spermatogenesis (27), rpt-1 is a potential transcription
factor that regulates expression of the interleukin 2 receptor
gene and human immunodeficiency virus 1 genes (28). Pos-
terior sexcomb (Psc) and suppressortwo ofzesta [Su(z)2] are
- -- Xs
treatment and analyzed by Northern blot hybridization using efp, ER, and 3-actincDNAprobes. Migration positionsofribosomalRNA markers
are shown on the right. (B) efp protein is regulated by estrogen. Nuclear extracts (25 umg)werepreparedfrom the HBL-100 cells at the indicated
hours after 17,B-estradiol treatment and analyzed by immunoblotting using anti-efp antibody. Migration positionsof molecular size markers(in
kDa) are shown on the right. (C) A model for the estrogen action throughtheestrogen-responsive transcriptionfactor. In this model, an
ER-regulated transcription factor mediates and possibly amplifiestheestrogeneffect.
(A) efp is regulated by estrogen. Poly(A)+ RNA (5 pg)waspreparedfrom the HBL-100 cells at the indicated hours after 171-estradiol
Biochemistry:Inoue et al.
Proc. Natl. Acad. Sci. USA 90 (1993)11121 Download full-text
Drosophila polycomb group (Pc-G) genes of which mamma-
lian homologs include the bmi-J protooncogene (29, 30).
RAD18 is a yeast protein required for DNA repair (31), and
RAG-1 is a recombination-activating gene product (32).
efp possesses the B box and coiled-coil domain character-
istic ofa subfamily ofthe RING finger family. This subfamily
includes efp, PML, T18, rfp, rpt-1, SS-a/Ro (33), and xnf7
(19). Three of the seven subfamily members, PML, rfp, and
T18, have transformation capabilities when found in chro-
mosomal translocations. In each of these translocations,
RING finger, B box, and coiled-coil domains are retained
when fused to other proteins, suggesting that these domains
play an important role in cell transformation. The coiled-coil
region in which the negatively charged residues are exposed
on one side of the helix is known to function as a transacti-
vation domain (34). Alternatively, these domains may play a
role in protein-protein interaction, including dimer forma-
tion. The fact that efp has both the potential DNA-binding
and the dimerization-transactivation domains strongly sug-
gests that it belongs to one of the transcription factors.
Estrogen exerts a wide variety of effects on different
organs, butER, the putative sole mediator ofestrogen action,
was found as a single molecular species. To achieve the
diversity of estrogen action, we may postulate a second
mediator of estrogen action, the ER-regulated transcription
factor. By this model, the estrogen-responsive transcription
factor can mediate and amplify the estrogen action, forming
a cascade of gene regulation and providing diverse and
specific pathways in each target organ. efp is a good candi-
date for this model (Fig. 5C). The short response time ofefp
to estrogen (within 2 h) is compatible with this model. The
progesterone receptor gene, ofwhich the promoter region is
responsive toER (35), is anotherexample that fits this model.
The structure and function of the ER have been intensively
studied, but the whole mechanism of estrogen action is still
poorly understood. We propose the possibility that the sec-
ond mediators such as estrogen-responsive transcriptional
regulators have an important implication in the mechanism of
We thank Mr. Inagaki (Pharmaceutical Laboratory, KIRIN, Mae-
bashi, Japan) for kindly providing synthetic peptides; Dr. Tamai
(Medical and Biological Laboratories, Ina, Japan) for preparing
antibodies to synthetic peptides; Drs. H. Hamada, M. Hashimoto,T.
Shimazaki, R. Sakai, S. Kato, T. Nishimura, and T. Matsuse for
helpful discussion; and Ms. M. Goto and H. Yamaguchi for expert
technical assistance. The HBL-100 cells were generously supplied by
the RIKEN Cell Bank. This work was supported by grants from the
Ministry of Education, Science and Culture, Japan.
Evans, R. M. (1988) Science 240, 889-895.
Green, S. & Chambon, P. (1988) Trends Genet. 4, 309-314.
Inoue, S., Kondo, S., Hashimoto, M., Kondo, T. & Mura-
matsu, M. (1991) Nucleic Acids Res. 19, 4091-4096.
Harris, J. R., Hellman, S., Henderson, I. C. & Kinne, D. W.
(1987) Breast Diseases (Lippincott, Philadelphia).
Simerly, R. B., Chang, C., Muramatsu, M. & Swanson, L. W.
(1990) J. Comp. Neurol. 294, 76-95.
Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl.
Acad. Sci. USA 74, 5463-5467.
Hashimoto, M., Shoda, A., Inoue, S., Yamada, R., Kondo, T.,
Sakurai, T., Ueno, N. & Muramatsu, M. (1992) J. Biol. Chem.
Koike, S., Sakai, M. & Muramatsu, M. (1987) Nucleic Acids
Res. 15, 2499-2513.
Takebe, Y., Seiki, M., Fujisawa, J.-I., Hoy, P., Yokota, K.,
Arai, K.-I., Yoshida, M. & Arai, N. (1988) Mol. Cell. Biol. 8,
Graham, F. L. & van derEb, A. J. (1973) Virology 52,456-467.
Schreber, E., Matthias, P., Muller, M. M. & Schaffner, W.
(1989) Nucleic Acids Res. 17, 6419.
Hashimoto, M., Nakamura, T., Inoue, S., Kondo, T., Yamada,
R., Eto, Y., Sugino, H. & Muramatsu, M. (1992) J. Biol. Chem.
Gaffney, E. V. (1982) Cell Tissue Res. 227, 563-568.
Luckow, B. & Schutz, G. (1987) Nucleic Acids Res. 15, 5490.
Gorman, C. M., Moffat, L. F. & Howard, B. H. (1982) Mol.
Cell. Biol. 2, 1044-1051.
Freemont, P. S., Hanson, I. M. & Trowsdale, J. (1991) Cell64,
Reddy, B. A., Etkin, L. D. & Freemont, P. S. (1992) Trends
Biol. Sci. 17, 344-345.
Kozak, M. (1987) Nucleic Acids Res. 15, 8125-8148.
Reddy, B. A. & Etkin, L. D. (1991) Nucleic Acids Res. 19,
de The, H., Lavau, C., Marchio, A., Chomienne, C., Degos, L.
& Dejean, A. (1991) Cell 66, 675-684.
Kakizuka, A., Miller, W. H., Umesono, K., Warrell, R. P.,
Frankel, S. R., Murty, V. V. V. S., Dmitrovsky, E. & Evans,
R. M. (1991) Cell 66, 663-674.
Goddard, A. D., Borrow, J., Freemont, P. S. & Solomon, E.
(1991) Science 254, 1371-1374.
Miki, T., Fleming, T. P., Crescenzi, M., Molloy, C. J., Blam,
S. B., Reynolds, S. H. & Aaronson, S. A. (1991) Proc. Natl.
Acad. Sci. USA 88, 5167-5171.
Curatola, A. M. & Basilico, C. (1990) Mol. Cell. Biol. 10,
Gould, A. P., Brookman, J. J., Strutt, D. I. & White, R. A. H.
(1990) Nature (London) 348, 308-312.
Grabe, G., Aragnol, D., Laurenti, P., Garzino, V., Charmot,
D., Berenger, H. & Pradel, J. (1992) EMBO J. 11, 3375-3378.
Takahashi, M., Inaguma, Y., Hiai, H. & Hirose, F. (1988) Mol.
Cell. Biol. 8, 1853-1856.
Patarca, R., Schwartz, J., Singh, R. P., Kong, Q.-T., Murphy,
E., Anderson, Y., Sheng, F.-Y. W., Singh, P., Johnson, K. A.,
Guamagia, S. M., Durfee, T., Blattner, F. & Cantor, H. (1988)
Proc. Natl. Acad. Sci. USA 85, 2733-2737.
van Lohuizen, M., Frasch, M., Wientjens, E. & Berns, A.
(1991) Nature (London) 353, 353-355.
Brunk, B. P., Martin, E. C. & Adler, P. N. (1991) Nature
(London) 353, 351-353.
Jones, J. S., Weber, S. & Prakash, L. (1988) NucleicAcidsRes.
Schatz, D. A., Oettinger, M. A. & Baltimore, D. (1989) Cell59,
Chan, E. K. L., Hammel, J. C., Buyon, J. P. & Tan, E. M.
(1991) J. Clin. Invest. 87, 68-76.
Giniger, E. & Ptashne, P. (1987)Nature (London) 330, 670-672.
Kastner, P., Krust, A., Turcotte, B., Stropp, U., Tora, L.,
Gronemeyer, H. & Chambon, P. (1990)EMBO J. 9, 1603-1614.
Biochemistry:Inoue et al.