The estrogen receptor-interacting protein HPIP increases estrogen-responsive gene expression through activation of MAPK and AKT

The estrogen receptor-interacting protein HPIP increases estrogen-responsive

gene expression through activation of MAPK and AKT

Xiaohui Wang

a,1

, Zhihong Yang

a,1

, Hao Zhang

a,1

, Lihua Ding

a

, Xiru Li

b

, Cui Zhu

a

,

Yiqiong Zheng

b

, Qinong Ye

a,

⁎

a_{Department of Molecular Oncology, Beijing Institute of Biotechnology, Beijing 100850, PR China} b_{Department of General Surgery, PLA General Hospital, Beijing 100853, PR China}

Received 6 November 2007; received in revised form 9 January 2008; accepted 22 January 2008 Available online 12 February 2008

Abstract

Estrogen receptors (ERα and ERβ) are estrogen-regulated transcription factors that play important roles in the development and progression of

breast cancer. The biological function of ERs has been shown to be modulated by ER-interacting proteins. However, the ER-interacting proteins

that not only activate MAPK and AKT, two important growth regulatory protein kinases, but also increase growth related estrogen-responsive

gene expression remain unknown. Here, we report that hematopoietic PBX-interacting protein (HPIP) interacts both with ER

α and with ERβ, and

increases ERα target gene expression through activation of MAPK and AKT and enhanced ERα phosphorylation. ERβ inhibits ERα target gene

expression, possibly by competition of ER

β with ERα for binding to HPIP, and by a decrease in available ERα for HPIP binding through the

interaction of ER

β with ERα. Furthermore, HPIP increases breast cancer cell growth. These data suggest that HPIP may be an important regulator

in ER signaling and that the relative ratio of ERβ to ERα may be important for HPIP function.

© 2008 Elsevier B.V. All rights reserved.

Keywords: Estrogen receptor; Hematopoietic PBX-interacting protein (HPIP); Interaction; Mitogen-activated protein kinase (MAPK); Protein kinase B/AKT

1. Introduction

Estrogen receptor (ER) plays an important role in the

development and progression of breast cancer

[1–3]

. ER has

two subtypes, ERα and ERβ

[4–6]

. They belong to steroid/

thyroid hormone superfamily of transcription factors. ERs

contain N-terminal estrogen-independent and C-terminal

estro-gen-dependent activation function domains (AF1 and AF2,

respectively). The DNA-binding domain (DBD) of the ERs is

centrally located. The ligand-binding domain, overlapping AF2,

shows 58% homology between ERα and ERβ. The DBD is

identical between the two receptors except for three amino acids.

However, the N-terminus containing the AF1 domain of ERβ

has only 28% homology with that of ERα. ERα and ERβ have

similar binding affinities for estrogen and their cognate

DNA-binding site because of the high degree of sequence homology

they share in their ligand- and DNA-binding domains

[7]

.

The classic effects of estrogens are mediated through binding

to ERs, which regulate gene expression by recruiting cofactors

and binding to an estrogen-responsive element (ERE) in the

nucleus

[8,9]

. Other actions of estrogens involve interaction of

ERs with other transcription factors, such as activating protein 1

(AP-1) and GC-box binding protein (SP-1), which in turn bind to

DNA and regulate transcription

[10,11]

. These events are

generally termed estrogen genomic effects because of the

involvement of gene transcription. Some estrogen-responsive

genes, such as pS2 and cathepsin D, two ERα target genes, have

been identified and well characterized by examination of the

estrogen genomic effects

[12,13]

. However, little is known about

ERβ target genes. Recently, estrogen has been shown to induce

rapid non-genomic pathways through interaction with membrane

receptors, especially ERs

[14–17]

. The rapid non-genomic actions

of estrogens include activation of mitogen-activated protein

kinase (MAPK), protein kinase B/AKT, and intracellular second

Biochimica et Biophysica Acta 1783 (2008) 1220–1228

www.elsevier.com/locate/bbamcr

⁎ Corresponding author. Department of Molecular Oncology, Beijing Institute of Biotechnology, 27 Tai-Ping Lu Road, Beijing 100850, PR China. Tel.: +86 10 6818 0809; fax: +86 10 6824 8045.

E-mail address:[email protected](Q. Ye).

1_{These authors contribute equally to this work.}

0167-4889/$ - see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.bbamcr.2008.01.026

messengers

[18–20]

. ERα can interact with the regulatory subunit

of phosphatidylinositol-3-OH kinase (PI3K), resulting in the

activation of AKT

[21]

. AKT phosphorylates ERα at serine 167,

thereby increasing transcriptional activation by ER, as determined

by estrogen-responsive reporter assay

[22]

. The MAPK pathway

can also interact with ER. MAPK/ERK1/2 can phosphorylate

ER

α at serine 118 to turn on receptor transactivation

[23

–25]

.

Although many ER-interacting proteins have been shown to

regulate estrogen-responsive gene expression, the

ER-interact-ing proteins that not only activate both MAPK and AKT, two

important growth regulatory protein kinases, but also stimulate

expression of estrogen-responsive proteins remain unknown. In

this study, we have identified and characterized an ERα- and

ERβ-interacting protein, hematopoietic pre-B cell leukemia

transcription factor (PBX)-interacting protein (HPIP). HPIP was

shown to be a corepressor for the transcription factor PBX

[26,27]

. The function of HPIP, however, is largely unknown.

Here, we show that HPIP increases ERα target gene expression

through activation of MAPK and AKT and enhanced ER

α

phosphorylation. ER

β can compete with ERα for binding to

HPIP and inhibit the expression of the ER

α target gene pS2.

2. Materials and methods

2.1. Plasmids

The reporter construct pS2-Luc and expression vector for ERα and ERβ have been described previously[28–30]. The ERα (S167A) mutant construct was made by recombinant PCR. The FLAG-tagged HPIP expression plasmid was cloned into a pcDNA3 vector linked with FLAG at the amino terminus by PCR using mammary cDNA library (Clontech) as the template. Plasmids encoding GST-fusion proteins were prepared by amplification of each sequence by standard PCR methods, and the resulting fragments were cloned in frame into pGEX-KG (Amersham Pharmacia Biotech) using appropriate sites. All of the constructs were confirmed by sequencing.

2.2. Stable transfection

Stable transfection of expression vector for FLAG-HPIP or HPIP siRNA into MCF7 cells were performed using Lipofectamine 2000 (Invitrogen). Transfected cells were selected in 500μg/ml G418 (Invitrogen) for approximately 2 months. Pooled clones or individual clones were screened by Western blot using FLAG or anti-HPIP antibody. Similar results were observed with individual clones or pooled clones.

2.3. Yeast two-hybrid assay

The bait plasmid pGBKT7-ERβ-AF2 (255–504) was used to screen a human mammary gland cDNA library fused to the GAL4 activation domain in pACT2 (Clontech) according to the manufacturer's instructions. Transformants were plated on synthetic medium lacking tryptophan, leucine, adenine and histidine but containing 1 mM 3-aminotriazole. Approximately one million transformants were screened. The candidate clones were rescued from the yeast cells and reintroduced back to the same yeast strain to verify the interaction between the candidates and the ERβ AF2 bait. The specificity of the interaction was determined by comparing the interactions between the candidates and various bait constructs. The unrelated prey plasmid pACT2-lamin and the empty vector pACT2 were used as negative controls.

2.4. Antibody production

The GST-HPIP (1–137) fusion protein was expressed in bacteria and purified using glutathione-Sepharose 4B beads according to the manufacture's protocol (Amersham Pharmacia). To generate polyclonal HPIP antibody, the purified GST-HPIP (1–137) protein was injected subcutaneously into each of

two BALB/c female mice. Sera from the immunized mice were collected and purified by affinity chromatography according to the manufacture's instructions (Pierce).

2.5. Luciferase assay

MCF7 cells were maintained in DMEM medium (Invitrogen) supplemented with 10% fetal bovine serum (FBS). For transfection, cells were seeded in 24-well plates containing phenol red-free DMEM medium (Invitrogen) supplemented with 5% charcoal-stripped FBS (Hyclone). The cells were transfected using Lipofecta-mine 2000 (Invitrogen) with 0.2μg of pS2-Luc, 1.0 μg of the expression vector for HPIP or HPIP siRNA, and 0.1μg of β-galactosidase reporter as an internal control. After treatment with 10 nM 17β-estradiol (E2), 100 nM ICI182,780, 40 μM PD98059, or 50μM LY294002 for 24 h, the cells were harvested. Cell extracts were analyzed for luciferase andβ-galactosidase activities as described previously[28].

2.6. GST pull-down assay

The GST alone and GST-fusion proteins were expressed in bacteria and bound to glutathione-Sepharose beads according to the manufacturer's instructions (Amersham Pharmacia). The expression plasmid for the ERα or ERβ was used for in vitro transcription and translation in the TNT system (Promega). The35S-labeled ERα or ERβ was incubated with approximately 1μg of GST-fusion protein bound to glutathione-Sepharose beads in 500 μl binding buffer (50 mM Tris–HCl, pH7.5, 150 mM NaCl, 1 mM EDTA, 0.3 mM DTT, 0.1% NP-40 and protease inhibitor tablets from Roche) at 4 °C. The beads were precipitated, washed four times with binding buffer, eluted by boiling in SDS sample buffer and analyzed by SDS-PAGE. After electrophoresis, radiolabeled proteins were visualized by autoradiography.

2.7. Coimmunoprecipitation

For transfection-based coimmunoprecipitation assays, 293T cells were transfected with the indicated plasmids using Lipofectamine 2000 (Invitrogen), lysed in 500μl lysis buffer (50 mM Tris at pH8.0, 500 mM NaCl, 0.5% NP-40, 1 mM DTT and protease inhibitor tablets from Roche), and immunoprecipitated with anti-FLAG agarose beads (Sigma-Aldrich) overnight at 4 °C. The beads were washed four times with the lysis buffer, and eluted in SDS sample buffer. The eluted proteins were separated by SDS-PAGE, followed by Western blotting with the indicated antibodies according to the standard procedures.

For detecting interaction of endogenous ERα or ERβ with HPIP, human breast cancer MCF7 cells were lysed in 500μl lysis buffer and immunopre-cipitated with anti-ERα, anti-ERβ or control serum (Santa Cruz). After extensive washing with the lysis buffer, the immunoprecipitates were resolved by SDS-PAGE, followed by Western blot analysis using anti-HPIP.

2.8. siRNA generation

siRNA was designed using the web-based insert design tool atwww.ambion. com/techlib/misc/psilencer_converter.html. The pSilencer2.1-U6 neo vector (Ambion)-based siRNA was made according to the manufacture's instructions. The cDNA target sequences of siRNAs for HPIP were AGCAGCTTGGAT-CAGGGAG corresponding to the coding region 405–423 relative to the first nucleotide of the start codon. Plasmid pSilencer2.1-U6 neo Negative Control (Ambion) was used as a control vector. Transient transfection of the vector-based siRNA into MCF7 cells was performed using Lipofectamine 2000 according to the manufacture's instructions (Invitrogen).

2.9. Cell growth assay

Cells were seeded in 24-well plate with 10% FBS–DMEM. The medium was changed at day 3. Cell growth was analyzed by crystal violet assay as described previously[29]. Briefly, cells were fixed by 1% glutaraldehyde for 15 min and stained with 0.5% crystal violet for 15 min at room temperature. Plates were washed with distilled water several times and air-dried. The dye was eluted by Sorenson's solution for 30 min at room temperature with constant shaking. A mircroplate reader was used to read aliquots of eluant at 590 nm.

2.10. Statistical analysis

Statistical significance was determined by Student's t-test. A p valueb0.05 was considered statistically significant.

3. Results

3.1. Identification of HPIP as an ERβ-interacting protein

To identify proteins that interact with ERβ, we screened a

human mammary cDNA library using amino acids 255–504 of

ERβ containing part of hinge region and entire AF2 domain as

bait in the yeast two-hybrid system. HPIP was identified as an

ERβ-interacting protein. The specificity of this interaction was

confirmed in a direct two-hybrid binding assay (

Fig. 1

A).

Transformation of yeast cells with HPIP in pACT2 vector

together with the GAL4 DNA-binding domain (DBD) in

pGBKT7 vector or together with an unrelated protein, lamin C,

fused to the GAL4 DBD instead of ER

β(255–504), did not

activate the ade, his and lacZ reporter genes, suggesting the

specific interaction of HPIP with ER

β.

To determine whether HPIP specifically interacts with amino

acids 255–504 of ERβ, different ERβ mutants were created for

yeast two-hybrid experiments. HPIP did not interact with

ERβ(1–145) containing the AF1, ERβ(131–324) containing

the DBD and hinge region, and ERβ(490–530) (

Fig. 1

B).

3.2. Generation of HPIP-specific antibody and detection of

HPIP protein

Since HPIP antibody is not commercially available, HPIP

polyclonal antibody was developed. To examine if the HPIP

antibody we produced can recognize HPIP protein, 293T cells

were transfected with expression vector for FLAG-tagged HPIP

or its empty vector control. Immunoblotting of whole cell

extracts with anti-FLAG or anti-HPIP demonstrated a specific

single band of a molecular weight of ~ 90 kDa (

Fig. 1

C).

Immunoblotting with anti-HPIP showed a single band in 293T

cells transfected with the control vector, suggesting that 293T

cells expressed endogenous HPIP (

Fig. 1

C, right panel). Due to

the small size of the FLAG tag (18 amino acids), we could not

clearly distinguish FLAG-HPIP from endogenous HPIP. Using

the specific anti-HPIP antibody, HPIP protein with varying

levels was detected in all of the breast cancer cell lines tested

(

Fig. 1

D).

3.3. HPIP interacts both with ERα and with ERβ in mammalian

cells

Since the amino acid sequence of ER

β(255–504) we used in

the yeast two-hybrid shares a high degree of similarity with the

corresponding region of ERα, the possibility that HPIP could

interact with ERα and ERβ was investigated by

coimmuno-precipitation. 293T cells were transfected with FLAG-HPIP

and ERα or ERβ, and cultured in the absence or presence of

10 nM 17β-estradiol (E2). FLAG-HPIP was

immunoprecipi-tated (IP) from cell lysates by anti-FLAG and analyzed for ERα

or ERβ binding by immunoblotting (IB). FLAG-HPIP could

be coimmunoprecipitated in the presence of ERα or ERβ

(

Fig. 2

A and B).

E2 increased the interaction of FLAG-HPIP with ERα or

ERβ. Endogenous HPIP was also found to be specifically

coimmunoprecipitated with endogenous ER

α or ERβ (

Fig. 2

C

and D). These data strongly suggest that HPIP interacts with

ER

α or ERβ in vivo.

3.4. Mapping of interaction regions of ERα and ERβ in HPIP

To define which regions of HPIP binds to ERα or ERβ, GST

pull-down experiments were performed. ERα and ERβ bound

three HPIP fragments (HPIP(138–220), HPIP(328–561) and

Fig. 1. Identification of HPIP and preparation of HPIP-specific antibody.

(A) Identification of HPIP as an ERβ-interacting protein by the yeast two-hybrid system. AH109 yeast cells were transformed with different plasmids (bait and prey) and grown on SD/-Trp-Leu-His-Ade. +, grown within 96 h;−, no growth within 96 h. Positive colonies were tested forβ-galactosidase activity. +, turned blue within 2 h;−, did not turn blue within 2 h. (B) Mapping of the HPIP interaction region in ERβ. AH109 cells were transformed with different constructs and analyzed as in (A). (C) Characterization of anti-HPIP mouse antibody by immunoblotting. The lysates from 293T cells transfected with FLAG-HPIP or its empty vector were prepared and the proteins detected with anti-FLAG (left panel) or anti-HPIP (right panel) antibody. (D) Expression of HPIP in different human cell lines. HPIP expression was detected by Western blot with anti-HPIP using proteins from indicated human cell lines.

HPIP(562–731)), whereas two HPIP fragments (HPIP(1–137)

and HPIP(221–367)) failed to interact with ERα and ERβ

(

Fig. 3

). Since full-length GST-HPIP was readily degraded, the

GST-HPIP was not used in the GST pull-down experiments.

3.5. HPIP increases estrogen-responsive reporter gene transcription

To determine whether HPIP–ER interaction affects ER

transcriptional activity, ERα- and ERβ-positive MCF7 breast

cancer cells were cotransfected with pS2-Luc, an

estrogen-responsive promoter, and FLAG-HPIP or HPIP small

interfer-ing RNA (siRNA). As shown in

Fig. 4

A and B, overexpression

of HPIP increased the reporter activity, whereas reduction of

endogenous HPIP with HPIP siRNA decreased the activity.

Taken together, these data indicate that HPIP enhances the pS2

reporter gene transcription.

3.6. HPIP increases ERα target gene expression

To corroborate the results of the luciferase reporter assay, the

effect of HPIP on the expression of endogenous

estrogen-responsive genes were determined. The E2-deprived MCF7

cells stably expressing either the empty vector or HPIP were

treated with E2 and then harvested for immunoblotting. As

expected, E2 stimulated the expression of pS2 and cathepsin D

(

Fig. 4

C). Importantly, stable transfection of HPIP further

Fig. 3. Mapping of ERα/ERβ interaction region in RBPMS.35_{S-labelled in vitro}

translated ERα or ERβ was incubated with GST-HPIP(1–137), GST-HPIP (138–220), GST-HPIP(221–367), GST-HPIP(328–561) or GST-HPIP(562– 731), or with GST. The bound proteins were subjected to SDS-PAGE followed by autoradiography. SDS-PAGE analysis of the purified GST-fusion proteins is shown at the bottom.

Fig. 2. HPIP interacts with ERα and ERβ in mammalian cells. (A and B) Interaction of overexpressed HPIP with overexpressed ERα or ERβ in mammalian cells. 293T cells were transfected with expression plasmids as indicated in the presence or absence of 10 nm E2. Immunoprecipitation (IP) was performed using anti-FLAG monoclonal antibody, and immunoblotted (IB) with anti-ERβ (A) or anti-ERα (B) antibody. (C and D) Interaction of endogenous HPIP with endogenous ERα or ERβ in vivo. MCF7 cells, cultured in serum-free medium for 3 days, were treated without and with 10 nm E2 for 2 h. Cell lysates were immunoprecipitated with anti-ERβ (C), anti-ERα (D) or preimmune control serum. The precipitates were analyzed by immunoblot using anti-HPIP.

enhanced the expression of pS2 and cathepsin D. Opposite

effects were observed with MCF7 cells stably transfected with

HPIP siRNA (

Fig. 4

D).

Fig. 4. HPIP regulates estrogen-responsive gene expression. (A and B) Overexpression of HPIP or knockdown of endogenous HPIP modulates pS2 reporter activity. MCF7 cells were cotransfected with 0.2 µg of reporter pS2-Luc and 1 µg of the expression vector for FLAG-tagged HPIP (A) or HPIP siRNA (B). Cells were treated with or without 10 nm E2 and analyzed for luciferase activity. Expression levels of HPIP were shown by immunoblotting with anti-FLAG (A) or anti-HPIP (B) at the bottom. GAPDH was used as a loading control. (C and D) overexpression of HPIP or knockdown of endogenous HPIP modulates estrogen-responsive protein expression. MCF7 cells stably transfected with HPIP (C) or HPIP siRNA (D) or corresponding empty vector were cultured in phenol red-free DMEM and treated with control (0.1% ethanol) vehicle or 10 nM E2 for 24 h. The cells were harvested and lysed, and conditioned medium was concentrated using a 3-kDa membrane. The concentrate was used for Western blot analysis of the expression of pS2 and cathepsin D, and the whole cell lysate was used for Western blot analysis of the expression of HPIP and GAPDH.

Fig. 5. HPIP regulates estrogen-responsive protein expression through activation of MAPK and AKT. (A and B) HPIP activates MAPK and AKT in an ER-dependent manner. HPIP (A) or HPIP siRNA (B) stably expressed MCF7 cells were treated with E2 or E2 plus ICI182,780. Cells were harvested for immunoblotting with antibodies to phospho-MAPK/ERK1/2 (phos-ERK), ERK1/2, phospho-AKT (phos-AKT) or AKT. (C) Phosphorylation of MAPK and AKT is responsible for HPIP stimulation of pS2 expression. HPIP stably expressed MCF7 cells were treated with E2 or E2 plus PD98059 or E2 plus LY294002. Cell lysates were blotted with the indicated antibodies.

3.7. HPIP increases ERα target gene expression through

activation of MAPK and AKT

Since HPIP is predominantly located in the cytoplasm (data

not shown), the potential non-genomic effect of HPIP was

determined using the above HPIP or HPIP siRNA-transfected

stable cell lines. Western blot analysis showed that E2 rapidly

increased phosphorylation of MAPK/ERK1/2 and AKT in

empty vector-transfected MCF7 cells (

Fig. 5

A). HPIP further

enhanced phosphorylation of MAPK and AKT in

HPIP-transfected cells. This activation of MAPK and AKT could be

abolished by ICI182,780, an ERα antagonist, indicating that

HPIP-mediated activation of MAPK and AKT is ER-dependent

(

Fig. 5

A). Consistent with the overexpression experiments,

stable knockdown of HPIP with HPIP siRNA in MCF7 cells

reduced MAPK and AKT phosphorylation (

Fig. 5

B).

To investigate whether activation of MAPK and AKT is

responsible for the enhancement of estrogen-responsive gene

expression by HPIP, PD98059 and LY294002, which are

MAPK and PI3K/AKT inhibitors, respectively, were used to

treat MCF7 cells stably transfected with either HPIP or empty

vector. As expected, the MAPK and AKT inhibitors decreased

phosphorylation of ERK1/2 and AKT, respectively (

Fig. 5

C).

The pS2 protein expression was impaired, but not abolished, by

the inhibition of ERK1/2 and AKT, in the empty vector and

HPIP stably transfected MCF7 cells at comparable levels. These

data suggest that HPIP regulates pS2 expression at least in part

through activated MAPK and AKT.

3.8. HPIP increases ERα target gene transcription through

enhanced ERα phosphorylation

Although ERβ phosphorylation is largely unknown, ERα

has been shown to be phosphorylated at serine 118 and serine

167 by MAPK and AKT, respectively. Since HPIP can activate

MAPK and AKT, we determined whether HPIP increases

phosphorylation of ERα. As shown in

Fig. 6

A, overexpression

of HPIP in MCF7 cells by stable transfection with HPIP

increased phosphorylation of ERα at serine residue, whereas

stable knockdown of HPIP with HPIP siRNA decreased ERα

phosphorylation. Furthermore, ectopic expression of HPIP

enhanced phosphorylation of ERα at serine 167, while

reduction of endogenous HPIP with HPIP siRNA reduced

phosphorylation of ERα at serine 167. Importantly, in the pS2

reporter assay, knockdown of HPIP reduced wild-type

ERα-mediated pS2 reporter transcription greater than the pS2

reporter transcription mediated by mutant ERα in which serine

167 was mutated to alanine (

Fig. 6

B). These data indicate that

phosphorylation of ER

α at serine 167 is important for the

enhancement of ER

α target gene transcription by HPIP.

3.9. Reduction of ERα binding to HPIP through ERβ competition

and interaction of ERβ with ERα

Since HPIP interacts both with ERα and with ERβ, we

investigated whether ERβ could affect the interaction of ERα

with HPIP. In the presence of increasing levels of ERβ, ERα

interaction with HPIP progressively decreased, whereas ERβ

interaction with HPIP increased, indicating that ER

β can

compete with ER

α for binding to HPIP (

Fig. 7

A).

To test the possibility that ER

β could decrease ERα binding

to HPIP through interaction of ERβ with ERα,

coimmunopre-cipitation experiments were again performed. With the

increas-ing amounts of ERβ, ERα interaction with HPIP progressively

decreased, whereas ERβ interaction with ERα increased,

suggesting that ERβ interacts with ERα and thus decreases

ERα interaction with HPIP (

Fig. 7

B).

3.10. ERβ negatively regulates ERα target gene expression

Next, we examined whether ERβ could affect ERα target

gene expression. For this purpose, we generated stable

transfectants of MCF7 breast cancer cells with FLAG-tagged

ER

β expression vector. MCF7 cells stably transfected with the

FLAG-tagged ER

β expressed FLAG-ERβ (

Fig. 7

C).

Expres-sion of the FLAG-ER

β inhibited the expression of the ERα

target gene pS2.

3.11. HPIP regulates breast cancer cell growth

To test the effect of HPIP on breast cancer cell growth, the

growth rate of MCF7 cells stably transfected with HPIP or empty

Fig. 6. HPIP modulates estrogen-responsive gene transcription through increased ERα phosphorylation. (A) HPIP increases ERα phosphorylation. HPIP or HPIP siRNA stably expressed MCF7 cells were immunoprecipitated with ERα antibody and blotted with antibodies to serine (phos-Ser) and phospho-ERα (serine 167) or ERα. (B) Mutation of serine 167 in ERα to alanine impairs HPIP function. MCF7 cells were cotransfected with 0.2 µg of reporter pS2-Luc and 1 µg of HPIP siRNA construct, together with 50 ng wild-type ERα or mutant ERα. Cells were treated with or without E2 and analyzed for luciferase activity.

vector, or with HPIP siRNA or control siRNA, was examined.

MCF7 cells stably transfected with HPIP grew faster than those

with empty vector (at day 5, p

b0.05), whereas MCF7 cells

stably transfected with HPIP siRNA grew more slowly than

those with control siRNA (at day 5, pb0.05) (

Fig. 8

). These

results indicate that HPIP increases breast cancer cell growth.

4. Discussion

In this study, we demonstrated that HPIP physically and

functionally interacts with ER. HPIP increases

estrogen-responsive protein expression in breast cancer cells.

Interest-ingly, HPIP activates MAPK and AKT, which are involved in

the increased expression of estrogen-responsive proteins. To the

best of our knowledge, this is the first evidence for an

ER-interacting protein that not only activates MAPK and AKT but

also potentiates estrogen-responsive gene expression.

While our project was under investigation, Manavathi et al.

reported that HPIP regulates ERα-dependent activation of

MAPK and AKT by recruiting Src kinase and p85 subunit of

PI3K to E2–ERα complex

[31]

. Consistent with the results

reported by Manavathi et al., we also demonstrated that HPIP

can activate MAPK and AKT. However, we have got a lot of

new results. First, we showed the interaction of ERβ with HPIP

Fig. 8. HPIP modulates breast cancer cell growth. FLAG-HPIP (A) or HPIP siRNA (B) stably expressed MCF7 cells were cultured in regular medium. At the indicated times, cells were harvested, and the cell number was determined by crystal violet assay.

Fig. 7. Reduction of ERα binding to HPIP through ERβ competition and interaction of ERβ with ERα, and negative regulation of ERα target gene expression by ERβ. (A) Competition between ERβ and ERα for HPIP binding. The indicated expression constructs were cotransfected into 293T cells. Cell lysates were immunoprecipitated with anti-FLAG antibody, followed by immunoblotting with the indicated antibodies. (B) Reduction of ERα binding to HPIP through interaction of ERβ with ERα. 293T cells were cotransfected with the indicated plasmids and analyzed as in (A). (C) ERβ inhibition of ERα target gene expression. MCF7 cells stably transfected with FLAG-ERβ or empty vector were grown in phenol red-free DMEM and treated with control (0.1% ethanol) vehicle or 10 nM E2 for 24 h. The cells were harvested and lysed for Western blot analysis as described in the legend toFig. 4C and D.

besides the interaction of ERα with HPIP demonstrated by

Manavathi et al. Second, we successfully generated a

HPIP-specific antibody and present evidence of physical interaction

between endogenous HPIP and endogenous ERα/ERβ in

human breast cancer cells using the antibody. HPIP protein is

expressed at varying levels in breast cancer cell lines we

examined. Third, we indicated that HPIP can increase

ex-pression of ERα target genes, including pS2 and cathepsin D,

and activation of MAPK and AKT is responsible for the

increased estrogen-responsive protein expression, suggesting an

important role of HPIP in estrogen signaling. Fourth, we

demonstrated that HPIP increases phosphorylation of ERα at

serine residue, such as serine 167, whose phosphorylation is

involved in ERα transcriptional activity regulated by HPIP.

Fifth, we indicated that ERβ inhibits ERα target gene

expres-sion, possibly by competition of ERβ with ERα for binding to

HPIP, and by a decrease in available ERα for HPIP binding

through interaction of ERβ with ERα. Finally, our study

showed that HPIP can promote breast cancer cell growth in

cultured cells.

Although both our study and that of Manavathi et al. showed

the interaction of HPIP with ERα and HPIP activation of MAPK

and AKT, there are some discrepancies between our study and

that reported by Manavathi et al. In our study, both ERα and ERβ

bound three HPIP regions, amino acids 138–220, 328–561, and

562–731, with the first region showing relatively strong

interaction. However, Manavathi et al. reported that amino

acids 1–277 of HPIP, which contains the region 138–220 we

examined, do not interact with ERα. The discrepancy between

our study and that of Manavathi et al. may result from the different

HPIP fragments used in GST pull-down experiments. Although

we showed the interaction of ERα and ERβ with amino acids

138–220 of HPIP, both ERα and ERβ do not interact with the

region 1

–138 of HPIP. Combined with the report by Manavathi

et al., these results suggest that the region 1

–138 of HPIP may

inhibit the interaction of HPIP with ER in the context of the region

1–277 of HPIP. Manavathi et al. reported that overexpression of

HPIP in MCF7 inhibits ERα transcriptional activity, as measured

by analysis of the ERE-Luc reporter, a synthetic

estrogen-responsive reporter. Manavathi et al. have not reported the effect

of knockdown of HPIP with HPIP siRNA in MCF7 cells on ERα

transcriptional activity. They also have not shown the effect of

HPIP on the expression of endogenous estrogen-responsive

genes, such as pS2. In our study, overexpression of HPIP in

MCF7 slightly increased ERα transcriptional activity analyzed by

the pS2-Luc reporter, a natural estrogen-responsive reporter,

while knockdown of endogenous HPIP with HPIP siRNA

dra-matically reduced the reporter activity. More importantly, our

study showed that overexpression of HPIP in MCF7 cells

enhanced the pS2 protein expression, whereas reduction of

endo-genous HPIP with HPIP siRNA decreased the pS2 protein

expression, further confirming the results of the reporter assay.

The discrepancy between our study and that of Manavathi et al.

could be due to a different reporter used.

It has been shown that the ligand-independent activity of

ERα is a result of ERα phosphorylation at multiple sites (serine

104, serine 106, serine 118, serine 167, serine 236 and tyr-537)

by multiple kinases

[19]

. For example, serine 118 is a target of

activated MAPK in response to EGF or insulin-like growth

factor (IGF) treatment. However, the identity of the kinase

involved in serine 118 phosphorylation in response to E2 is

controversial. Several kinases, such as AKT and p90 ribosomal

S6 kinase (RSK), have been reported to be responsible for

serine 167 phosphorylation. Our study showed that HPIP

activates MAPK and AKT and increases ERα phosphorylation

at serine residue. Furthermore, we demonstrated that serine 167

phosphorylation is important for HPIP activation of ERα

transcriptional activity because a mutant ERα in which serine

167 is changed to alanine reduced the HPIP function. Although

HPIP can stimulate MAPK and AKT, which are important cell

survival factors, we cannot exclude the possibility that HPIP can

also activate other kinases. It will be interesting to precisely

determine what kinds of kinases are involved in HPIP-mediated

activation of ER signaling.

Like ERα, the more recently discovered ERβ has been

shown to be an important determinant in breast cancer. Contrary

to ER

α, ERβ expression generally decreases during breast

tumorigenesis

[32

–34]

Most studies have indicated that ER

β

expression is an indicator of a favorable prognosis for breast

cancer patients

[35–38]

. Although many ERα-regulated target

genes, such as pS2 and cathepsin D, have been identified and

well characterized, little is known for ERβ-regulated

down-stream genes. We demonstrated for the first time that ERβ can

inhibit ERα target protein expression. Generally speaking, ERα

expression levels are higher than those of ERβ in most breast

cancers, and the majority of breast cancers express ERβ

together with ERα

[39–41]

. Estrogen via ERα induces

pro-liferation and inhibits apoptosis, whereas ERβ opposes the

proliferative effect of ERα in cultured cells

[42]

. When ERβ is

cotransfected with ERα, it inhibits ERα-mediated

transcrip-tional activity on an ERE reporter gene, suggesting that ER

β is

a negative regulator of ER

α

[43]

. Our study indicated that ER

β

competes with ER

α for binding to HPIP and inhibits the

expression of the ERα-regulated gene pS2. Therefore, the

relative ratio of ERβ to ERα should be an important factor

responsible for HPIP function. Conceivably, cells with lower

ERβ levels and higher HPIP levels might readily develop breast

cancer and have more malignant phenotype. It will be very

interesting to investigate clinical significance of HPIP in breast

cancer.

Acknowledgements

This work was supported by National Natural Science

Foundation (30530320, 30625035 and 30428012), Medicine

and Health Research Foundation of PLA (06J021) and the Major

State Basic Research Development Program (2007CB914603).

References

[1] A.T. Ferguson, N.E. Davidson, Regulation of estrogen receptorα function in breast cancer, Crit. Rev. Oncog. 8 (1997) 29–46.

[2] C.M. Klinge, Estrogen receptor interaction with activators and co-repressors, Steroids 65 (2000) 227–251.

[3] S. Saji, E.V. Jensen, S. Nilsson, T. Rylander, M. Warner, J.A. Gustafsson, Estrogen receptorsα and β in the rodent mammary gland, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 337–342.

[4] A. Aranda, A. Pascual, Nuclear hormone receptors and gene expression, Physiol. Rev. 81 (2001) 1269–1304.

[5] D.P. McDonnell, J.D. Norris, Connections and regulation of the human estrogen receptor, Science 296 (2002) 1642–1644.

[6] J.D. Yager, N.E. Davidson, Estrogen carcinogenesis in breast cancer, N. Engl. J. Med. 354 (2006) 270–282.

[7] J.A. Gustafsson, An update on estrogen receptors, Semin. Perinatol. 24 (2000) 66–69.

[8] J.G. Moggs, G. Orphanides, Estrogen receptors: orchestrators of pleiotropic cellular responses, EMBO Rep. 2 (2001) 775–781.

[9] V.C. Jordan, Selective estrogen receptor modulation: concept and consequences in cancer, Cancer Cell 5 (2004) 207–213.

[10] W. Porter, B. Saville, D. Hoivik, S. Safe, Functional synergy between the transcription factor Sp1 and the estrogen receptor, Mol. Endocrinol. 11 (1997) 1569–1580.

[11] C. Teyssier, K. Belguise, F. Galtier, D. Chalbos, Characterization of the physical interaction between estrogen receptorα and JUN proteins, J. Biol. Chem. 276 (2001) 36361–36369.

[12] S.R. Johnston, G. Saccani-Jotti, I.E. Smith, J. Salter, J. Newby, M. Coppen, S.R. Ebbs, M. Dowsett, Changes in estrogen receptor, progesterone receptor, and pS2 expression in tamoxifen-resistant human breast cancer, Cancer Res. 55 (1995) 3331–3338.

[13] D. Coradini, E. Biganzoli, P. Boracchi, E. Bombardieri, E. Seregni, G. De-Palo, G. Martelli, G. Di-Fronzo, Effect of steroid receptors, pS2 and cathepsin D on the outcome of elderly breast cancer patients: an explor-atory investigation, Int. J. Cancer 79 (1998) 305–311.

[14] E.R. Levin, Cellular functions of plasma membrane estrogen receptors, Steroids 67 (2002) 471–475.

[15] M. Razandi, A. Pedram, I. Merchenthaler, G.L. Greene, E.R. Levin, Plasma membrane estrogen receptors exist and functions as dimmers, Mol. Endocrinol. 18 (2004) 2854–2865.

[16] C.M. Revankar, D.F. Cimino, L.A. Sklar, J.B. Arterburn, E.R. Prossnitz, A transmembrane intracellular estrogen receptor mediates rapid cell signal-ing, Science 307 (2005) 1625–1630.

[17] K. Moriarty, K.H. Kim, J.R. Bender, Minireview: estrogen receptor-mediated rapid signaling, Endocrinology 147 (2006) 5557–5563. [18] C.K. Osborne, R. Schiff, S.A. Fuqua, J. Shou, Estrogen receptor: current

understanding of its activation and modulation, Clin. Cancer Res. 7 (2001) 4338–4342.

[19] D.A. Lannigan, Estrogen receptor phosphorylation, Steroids 68 (2003) 1–9. [20] R.X. Song, Z. Zhang, R.J. Santen, Estrogen rapid action via protein complex formation involving ER and Src, Trends Endocrinol. Metab. 16 (2005) 347–353.

[21] T. Simoncini, A. Hafezi-Moghadam, D.P. Brazil, K. Ley, W.W. Chin, J.K. Liao, Interaction of oestrogen receptor with the regulatory subunit of phosphatidylinositol-3-OH kinase, Nature 407 (2000) 538–541. [22] M.B. Martin, T.F. Franke, G.E. Stoica, P. Chambon, B.S.

Katzenellenbo-gen, B.A. Stoica, M.S. McLemore, S.E. Olivo, A. Stoica, A role for Akt in mediating the estrogenic functions of epidermal growth factor and insulin-like growth factor I, Endocrinology 141 (2000) 503–511.

[23] P.B. Joel, A.M. Traish, D.A. Lannigan, Estradiol and phorbol ester cause phosphorylation of serine 118 in the human estrogen receptor, Mol. Endocrinol. 9 (1995) 1041–1052.

[24] S. Kato, H. Endoh, Y. Masuhiro, T. Kitamoto, S. Uchiyama, H. Sasaki, S. Masushige, Y. Gotoh, E. Nishida, H. Kawashima, D. Metzger, P. Chambon, Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase, Science 270 (1995) 1491–1494.

[25] G. Bunone, P.A. Briand, R.J. Miksicek, D. Picard, Activation of the unliganded estrogen receptor by EGF involves the MAP kinase pathway and direct phosphorylation, EMBO J. 15 (1996) 2174–2183.

[26] C. Abramovich, W.F. Shen, N. Pineault, S. Imren, B. Montpetit, C. Largman, R.K. Humphries, Functional cloning and characterization of a novel nonhomeodomain protein that inhibits the binding of PBX1–HOX complexes to DNA, J. Biol. Chem. 275 (2000) 26172–26177.

[27] C. Abramovich, E.A. Chavez, P.M. Lansdorp, R.K. Humphries, Functional characterization of multiple domains involved in the subcellular localiza-tion of the hematopoietic Pbx interacting protein (HPIP), Oncogene 21 (2002) 6766–6771.

[28] L. Ding, J. Yan, J. Zhu, H. Zhong, Q. Lu, Z. Wang, C. Huang, Q. Ye, Ligand-independent activation of estrogen receptorα by XBP-1, Nucleic Acids Res. 31 (2003) 5266–5274.

[29] H. Zhang, X. Xie, X. Zhu, J. Zhu, C. Hao, Q. Lu, L. Ding, Y. Liu, L. Zhou, Y. Liu, C. Huang, C. Wen, Q. Ye, Stimulatory cross-talk between NFAT3 and estrogen receptor in breast cancer cells, J. Biol. Chem. 280 (2005) 43188–43197.

[30] J. Han, L. Ding, B. Yuan, X. Yang, X. Wang, J. Li, Q. Lu, C. Huang, Q. Ye, Hepatitis B virus X protein and the estrogen receptor variant lacking exon 5 inhibit estrogen receptor signaling in hepatoma cells, Nucleic Acids Res. 34 (2006) 3095–3106.

[31] B. Manavathi, F. Acconcia, S.K. Rayala, R. Kumar, An inherent role of microtubule network in the action of nuclear receptor, Proc. Natl Acad. Sci. U. S. A. 103 (2006) 15981–15986.

[32] A. Bardin, N. Boulle, G. Lazennec, F. Vignon, P. Pujol, Loss of ERβ expression as a common step in estrogen-dependent tumor progression, Endocr. Relat. Cancer 11 (2004) 537–551.

[33] J. Matthews, J.A. Gustafsson, Estrogen signaling: a subtle balance between ERα and ERβ, Mol. Interv. 3 (2003) 281–292.

[34] V. Speirs, P.J. Carder, S. Lane, D. Dodwell, M.R. Lansdown, A.M. Hanby, Oestrogen receptor β: what it means for patients with breast cancer, Lancet. Oncol. 5 (2004) 174–181.

[35] V. Speirs, Oestrogen receptorβ in breast cancer: good, bad or still too early to tell? J. Pathol. 197 (2002) 143–147.

[36] T.A. Hopp, H.L. Weiss, I.S. Parra, Y. Cui, C.K. Osborne, S.A. Fuqua, Low levels of estrogen receptor β protein predict resistance to tamoxifen therapy in breast cancer, Clin. Cancer Res. 10 (2004) 7490–7499. [37] S. Saji, M. Hirose, M. Toi, Clinical significance of estrogen receptorβ in

breast cancer, Cancer Chemother. Pharmacol. 56 (2005) 21–26. [38] A.M. Shaaban, C. Jarvis, F. Moore, C. West, A. Dodson, C.S. Foster,

Prognostic significance of estrogen receptorβ in epithelial hyperplasia of usual type with known outcome, Am. J. Surg. Pathol. 29 (2005) 1593–1599. [39] S.A. Fuqua, R. Schiff, I. Parra, W.E. Friedrichs, J.L. Su, D.D. McKee, K. Slentz-Kesler, L.B. Moore, T.M. Willson, J.T. Moore, Expression of wild-type estrogen receptorβ and variant isoforms in human breast cancer, Cancer Res. 59 (1999) 5425–5428.

[40] S.A. Fuqua, R. Schiff, I. Parra, J.T. Moore, S.K. Mohsin, C.K. Osborne, G.M. Clark, D.C. Allred, Estrogen receptorβ protein in human breast cancer: correlation with clinical tumor parameters, Cancer Res. 63 (2003) 2434–2439.

[41] S. Sommer, S.A. Fuqua, Estrogen receptor and breast cancer, Semin. Cancer Biol. 11 (2001) 339–352.

[42] L.A. Helguero, M.H. Faulds, J.A. Gustafsson, L.A. Haldosen, Estrogen receptors (ERα) and (ERβ) differentially regulate proliferation and apoptosis of the normal murine mammary epithelial cell line HC11, Oncogene 24 (2005) 6605–6616.

[43] Y. Omoto, H. Eguchi, Y. Yamamoto-Yamaguchi, S. Hayashi, Estrogen receptor (ER)β1 and ERβcx/2 inhibit ERα function differently in breast cancer cell line MCF7, Oncogene 22 (2003) 5011–5020.