A

crmriAilbimi©

K

1

wcrm

oB3

0d

Journal of Steroid Biochemistry & Molecular Biology 102 (2006) 22–31

Steroid receptor coactivator 2 is essential for progesterone-dependentuterine function and mammary morphogenesis: Insights from the

mouse—implications for the human�

Atish Mukherjee a, Paula Amato b, D. Craig Allred c, Rodrigo Fernandez-Valdivia a,Jonathan Nguyen a, Bert W. O’Malley a, Francesco J. DeMayo a, John P. Lydon a,∗

a Department of Molecular and Cellular Biology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, United Statesb Department of Obstetrics and Gynecology, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, United States

c Breast Center at Baylor College of Medicine and the Methodist Hospital, Baylor College of Medicine, One Baylor Plaza,Houston, TX 77030, United States

bstract

While the indispensability of the progesterone receptor (PR) in female reproduction and mammary morphogenesis is acknowledged, theoregulators preferentially recruited by PR to mediate its in vivo effects have yet to be fully delineated. To further parse the roles of steroideceptor coactivator (SRC)/p160 family members in P-dependent physiological processes, genetic approaches were employed to generate aouse model (PRCre/+SRC-2flox/flox) in which SRC-2 function was ablated specifically in cell-types that express the PR. Fertility evaluation

evealed that while ovulation occurred normally in the PRCre/+SRC-2flox/flox mouse, uterine function was markedly affected. Absence of SRC-2n PR positive uterine cells contributed to an early block in embryo implantation, a phenotype not shared by knockouts for SRC-1 or -3.lthough the PRCre/+SRC-2flox/flox uterus could mount a partial decidual response, removal of SRC-1 in the PRCre/+SRC-2flox/flox uterus resulted

n a complete block in decidualization, confirming that uterine SRC-2 and -1 are both required for P-initiated transcriptional programs whichead to full decidualization. In the case of the mammary gland, whole-mount and histological analyses revealed the absence of significantranching morphogenesis in the hormone-treated PRCre/+SRC-2flox/flox mammary gland, reinforcing an important role for mammary SRC-2

n cellular proliferative events that require PR. Based on the above and the observation that SRC-2 is expressed in many of the uterine andammary cell-lineages in the human as observed in the mouse, we suggest that further investigations are warranted to gain additional insights

nto SRC-2’s involvement in normal (and possibly abnormal) uterine and mammary cellular responses to progestins.2006 Elsevier Ltd. All rights reserved.

Human

nsmt

eywords: SRC2; Progesterone receptor; Uterus; Mammary gland; Mouse;

. Introduction

The progesterone receptor knockout (PRKO) mouse, inhich both isoforms (PR-A and -B) were abrogated, show-

ased the importance of P as a master coordinator of femaleeproductive biology [1]. Ablation of PR not only compro-ised uterine function but severely undermined operational

� Lecture presentation at the 17th International Symposium of the Journalf Steroid Biochemistry & Molecular Biology, ‘Recent Advances in Steroidiochemistry and Molecular Biology’ (Seefeld, Tyrol, Austria,1 May–3 June 2006).∗ Corresponding author. Tel.: +1 713 798 3534; fax: +1 713 790 1275.

E-mail address: jlydon@bcm.tmc.edu (J.P. Lydon).

onhtepi

P

960-0760/$ – see front matter © 2006 Elsevier Ltd. All rights reserved.oi:10.1016/j.jsbmb.2006.09.007

ormalcy of the hypothalamo-pituitary-ovarian axis. Paralleltudies also revealed a critical involvement for P signaling inammary epithelial proliferation, an essential cellular event

hat allows pregnancy-induced mammary morphogenesis toccur in the adult. Importantly, the PRKO exhibited a sig-ificant reduction in mammary tumor susceptibility [2–4],ighlighting an important role for P signaling in mammaryumorigenesis. These findings broadly agree with severalxperimental investigations using the rodent model [5–8],

revious human observational studies [9–11], and recent clin-cal trials [12,13].

Apart from disclosing novel cellular principles by whichcontrols proliferative and differentiative programs that are

chemist

mstttrcsobiuprm

tmfb21hodlrPtP

som1osipau[ag

[patPtmmmta

eSet[

m2tmgthsplvi

atrP[rPSwrmicpciSt

2

2

ma2iPti[

A. Mukherjee et al. / Journal of Steroid Bio

andatory for target tissue morphogenesis and tumorigene-is, two important questions have arisen from these investiga-ions concerning PRs mechanism of action in a given targetissue: (1) what are the downstream molecular pathwayshat transduce the P signal to an appropriate physiologicalesponse? And (2) which coregulators (coactivators and/ororepressors) are selectively recruited in PR mediated tran-criptional programs that induce or suppress the expressionf these molecular pathways? While significant progress haseen made to reveal the downstream targets of PR actionn the mouse [14–20], identification of the primary coreg-lators specifically involved in PR dependent physiologicalrocesses is only now being addressed due in part to theecent development of powerful genetic approaches in theouse.Previous in vitro studies demonstrated that PR mediated

ranscription is dependent on coordinate interactions withembers of the steroid receptor coactivator (SRC/p160) gene

amily [21]. The SRC family is composed of three mem-ers: SRC-1 (ERAP140/ERAP160/NcoA-1), SRC-2 (TIF-/GRIP-1/NcoA-2) and SRC-3 (p/CIP/RAC3/AIB1/TRAM-/ACTR/NcoA-3); reviewed in [22]. The three coactivatorsave been shown to interact with the ligand binding domainf PR to potentiate its transactivational function in a ligand-ependent manner. Depending on physiological and cellu-ar context, we posit that one or more SRC members areecruited to the promoter/enhancer region of a subset ofR target gene(s), the transcription of which manifests (at

he molecular level) a particular physiological response toexposure.Establishing whether one or more SRC family members

erve as a coactivator in physiological processes that dependn PR was achieved by examining knockout (KO) mouseodels for each SRC member. Although viable, the SRC-

KO female revealed a significant reduction in the abilityf its uterus to undergo decidualization [23]. Apart fromupporting an essential role for this coactivator in a uter-ne remodeling event that is dependent on P signaling, theartially decidualized uterus of the SRC-1KO suggesteddditional coactivators are required to achieve full decid-alization. In the case of the SRC-1KO mammary gland23], stunted ductal elongation and dichotomous branchingt puberty implicated a role for SRC-1 in mammary morpho-enetic effects that rely on estrogen (E) signaling.

A uterine defect was not observed in the SRC-3KO model24]; however, the SRC-3KO mammary gland displayed aartial defect in pregnancy-associated ductal side-branchingnd alveologenesis, a mammary defect that correlates to a cer-ain extent with the PRKO mammary phenotype [1]. Like theRKO [2], the SRC-3KO mammary gland is less susceptible

o mammary tumorigenesis [25] and suggests that SRC-3ay be preferentially recruited by a select subset of PR-

ediated transcriptional programs that underlie P-inducedammary morphogenesis and tumorigenesis. Collectively,

he use of coactivator KOs in studies of female reproductivend mammary gland biology suggest that while SRC-1 has

SS2o

ry & Molecular Biology 102 (2006) 22–31 23

volved as an important coactivator for uterine PR action,RC-3 is selected for a subgroup of mammary PR mediatedffects; recent reporter studies with the PR activity indica-or (PRAI) model would further support this supposition26].

In contrast to the KOs for SRC-1 and -3, the SRC-2KOodel (referred to as the transcriptional intermediary factorKO or TIF2−/− from hereon) registers severe reproduc-

ive defects in the both sexes [27]. The TIF2−/− male isarkedly subfertile with developmental defects in spermato-

enesis and an age-dependent testicular degeneration. Impor-antly, initial tests of the TIF2−/− female revealed a severeypofertility phenotype due to significant placental hypopla-ia. Follow up investigations have demonstrated that TIF2−/−ups (male and female) are significantly underrepresented initters derived from TIF2+/− intercrosses (unpublished obser-ations); TIF2−/− females resulting from such crosses arenfertile.

Because of the severity of the TIF2−/− reproductive defectnd the possibility that SRC-2 (TIF2) may play an impor-ant coactivator role in PR mediated physiological processesequired for the maintenance of fertility, we created a novelRCre/+ SRC-2flox/flox bigenic in which the PRCre/+ knockin28] was crossed with the SRC-2flox/flox model (previouslyeferred to as the TIF2 floxed (L2) version [27]). Because theRCre/+ SRC-2flox/flox bigenic enables postnatal ablation ofRC-2 specifically in cell lineages that express the PR, weere able to avoid the embryonic, reproductive, and recently

eported metabolic [29] defects associated with the TIF2−/−odel and examine the requirement of this coactivator specif-

cally in PR dependent transcriptional programs in the adoles-ent and adult. Unlike SRC-1 and SRC-3, whose coactivatorroperties in female reproductive biology are primarily spe-ialized for P-initiated transcriptional programs operativen the uterus and mammary gland, respectively, we revealRC-2 as an indispensable PR coactivator in both target

issues.

. Insights from the mouse

.1. Generation of the PRCre/+SRC-2flox/flox mouse

To reveal those P-dependent reproductive and mam-ary responses that require SRC-2, cre-loxP engineering

pproaches were used to create a mouse model (PRCre/+SRC-flox/flox) in which SRC-2 expression is abrogated specificallyn cell-lineages that score positive for PR expression. TheRCre/+SRC-2flox/flox bigenic was generated by crossing of

he PRCre/+ knockin [28] with a mouse model (SRC-2flox/flox)n which exon 11 of the SRC-2 gene was flanked by lox P sites27] (Fig. 1A and Ref. [30] for more details); exon 11 of the

RC-2 gene encodes the nuclear interacting domain (NID).imilar to a wild type (WT) sibling, the uterus of PRCre/+SRC-flox/flox mouse showed normal PR expression in all cell-typesf the uterus (compare Fig. 1B with C). Even though the

24 A. Mukherjee et al. / Journal of Steroid Biochemistry & Molecular Biology 102 (2006) 22–31

Fig. 1. Creation of the PRCre/+SRC-2flox/flox mouse. Panel (A): to ablate SRC-2 expression in PR specific cell lineages, the SRC-2flox/flox mutation was introducedinto the PRCre/+ genetic background to generate the PRCre/+SRC-2flox/flox bigenic (see [30]). The endogenous PR promoter drives Cre mediated excision offloxed exon 11 of the SRC-2 gene only in all cell-lineages that score positive for PR expression. Panels (B) and (C) show PR immunohistochemical staining ofuteri obtained from ovariectomized SRC-2flox/flox and PRCre/+SRC-2flox/flox mice, respectively, whereas panels (D) and (E) display SRC-2 immunohistochemicals ectivelyu ll as ther ght [200

PfrWWt2uSt

2

we2

taining of uteri obtained from SRC-2flox/flox and PRCre/+SRC-2flox/flox, respterus (panel E). The luminal and glandular epithelial compartments as weespectively. Scale bar in panel (B) applies to all panels. From [30] (Copyri

RCre/+ knockin mutation creates a genotype heterozygousor the intact PR allele [28], the PR immunohistochemicalesult indicates PR levels are not significantly reduced from

T levels in the PRCre/+SRC-2flox/flox uterus (Fig. 1B and C).hile SRC-2 expression was detected in all uterine cell-types

hat express PR in the WT (Fig. 1D), as anticipated, SRC-

expression was not observed in the PRCre/+SRC-2flox/flox

terus (Fig. 1E). Importantly, this result showed that uterineRC-2 was ablated in PR positive cells in accordance with

he genetic design in Fig. 1A.

aePp

. Note the absence of uterine SRC-2 expression in the PRCre/+SRC-2flox/flox

stromal and myometrial compartments are denoted by LE, GE, S, and M,6] Molecular and Cellular Biology).

.2. The PRCre/+SRC-2flox/flox female is infertile

Although female and male PRCre/+SRC-2flox/flox neonatesere represented at the expected Mendelian frequency and

xhibited normal postnatal development, the PRCre/+SRC-flox/flox female was shown to be infertile (Table 1). During

6-month period, normal size litters were produced at the

xpected frequency by WT, SRC-2flox/+, SRC-2flox/flox, andRCre/+SRC-2flox/+ mothers. Despite exhibiting copulatorylugs at normal frequency, the PRCre/+SRC-2flox/flox failed

A. Mukherjee et al. / Journal of Steroid Biochemist

Table 1The PRCre/+SRC2flox/flox female is infertile

Genotype Number ofmice tested

Mean numberof pups/litter ± S.D.a

Mean numberof litters/mouse ± S.D.a

WT 8 7.1 ± 0.8 4.6 ± 0.2SRC-2flox/+ 8 6.8 ± 0.6 4.4 ± 0.4SRC-2flox/flox 8 6.9 ± 0.8 4.8 ± 0.7PRCre/+SRC-2flox/+ 7 6.9 ± 0.9 4.5 ± 0.2PRCre/+SRC-2flox/flox 7 0 0

The SRC-2flox/+, SRC-2flox/flox, PRCre/+SRC-2flox/+ represent mice withintermediate genotypes that were used to generate the PRCre/+SRC-2flox/flox

mouse and serve as additional controls along with WT. From [30] (Copyright[2006] Molecular and Cellular Biology).

a S.D. refers to the standard deviation from the mean.

tPnistnMPt2s

[

Fig. 2. An implantation and decidualization defect in the PRCre/+SRC-2flox/flox uterua WT (1) mouse (5.5 d.p.c.). Implantation sites are revealed by the localized retentiin similarly treated uteri taken from PRCre/+SRC-2flox/flox (2) mice at 5.5 d.p.c. Thmice analyzed is tabulated. Panel (B) shows the overt morphological response ofand PRCre/+SRC-2flox/flox (2) mice; for both genotypes, the right (R) uterine horn restandard deviation (S.D.)) of stimulated (L) to control (R) horn for SRC-2flox/flox

uterine tissue obtained from untreated adult virgin SRC-2flox/flox (1) and PRCre/+SRSRC-1 and SRC-3 between the two genotypes (loading control is �-actin). From [3

ry & Molecular Biology 102 (2006) 22–31 25

o produce litters during this time-frame; in contrast, theRCre/+SRC-2flox/flox male exhibited normal fertility (dataot shown). The fact that the PRCre/+SRC-2flox/flox males fertile contrasts with the TIF2−/− male which reveals aevere hypofertility phenotype [27]; the difference in pheno-ypes between the two models highlights the tissue-selectiveature of SRC-2 ablation in the PRCre/+SRC-2flox/flox model.oreover, the absence of a metabolic phenotype in the

RCre/+SRC-2flox/flox mouse further underscores the pheno-ypic differences between the TIF2−/− and PRCre/+SRC-flox/flox genotypes; TIF2−/− mice have previously been

hown to exhibit defects in energy homeostasis [29,31].

Because (1) SRC-2 is expressed in the murine ovary30] and (2) the PRKO ovary fails to ovulate [1], the

s. Panel (A): arrows show the location of implantation sites in the uterus ofon of Chicago blue dye. Importantly, implantation sites were not observede average number of implantation sites per genotype per total number ofthe left (L) uterine horn to a deciduogenic stimulus for SRC-2flox/flox (1)presents the unstimulated control. Panel (C): the average weight ratios (±and PRCre/+SRC-2flox/flox uteri are shown. Panel (D): western analysis ofC-2flox/flox (2) mice reveals no difference in expression levels for uterine

0] (Copyright [2006] Molecular and Cellular Biology).

26 A. Mukherjee et al. / Journal of Steroid Biochemistry & Molecular Biology 102 (2006) 22–31

Table 2The PRCre/+SRC2flox/flox mouse ovulates normally

Genotype Number ofmice tested

Meannumberof eggs

Mean numberof fertilizedeggs

Percentfertilized

SP

F

gea2Tstdi

2c

2n2eteicecooita(srwdaScytui1sp

Pu

Fig. 3. The PRCre/+SRC-2flox/flox SRC-1KO trigenic uterus fails to undergodecidualization. Panel (A): the stimulated left (L) uterine horn of the SRC-2flox/flox mouse (1) displays full decidualization (the right (R) horn is theunstimulated horn). By contrast, the similarly treated PRCre/+SRC-2flox/flox

SRC-1KO trigenic uterus (2) fails to launch a decidual response in the left(da[

tittStrtffad

2

tiutiauPreat

RC-2flox/flox 5 13 5 38.4RCre/+SRC-2flox/flox 8 15.75 5.5 34.9

rom [30] (Copyright [2006] Molecular and Cellular Biology).

onadotropin-treated PRCre/+SRC-2flox/flox mouse wasxamined to determine whether ovulation could occur in thebsence of ovarian SRC-2. Surprisingly, the PRCre/+SRC-flox/flox ovary was shown to ovulate normally (Table 2).he average yield of oocytes collected in the oviduct of theuperovulated PRCre/+SRC-2flox/flox mouse was comparableo that observed in the similarly treated SRC-2flox/flox sibling,emonstrating ovarian-derived SRC-2 (like SRC-1 and -3)s not required for PR mediated follicular rupture.

.3. PRCre/+SRC-2flox/flox uterine function isompromised

The absence of implantation sites in the PRCre/+SRC-flox/flox uterus (5.5 days post-coitum (d.p.c.)) demonstratedot only that a uterine defect underlies the PRCre/+SRC-flox/flox infertility phenotype but that this defect blocks thearly progression of multi-stage uterine cellular processeshat establish the maternofetal interface (Fig. 2A). Followingmbryo apposition, attachment and subsequent trophoblastnvasion, the uterus undergoes a decidual reaction that isompletely dependent on P-signaling [1]. In the absence ofmbryo implantation, however, an artificial decidual responsean be induced in the uterus from an appropriately EP treatedvariectomized mouse [1]. Using this approach, the uterusf the SRC-2flox/flox mouse exhibits a full decidual responsen the left (L) stimulated horn (Fig. 2B and C). However,he PRCre/+SRC-2flox/flox uterus consistently displayed only

partial decidual response to the deciduogenic stimulusFig. 2B and C), suggesting not only that PR mediated tran-cription requires SRC-2 to mount a full uterine decidualesponse but that other coactivators are required in concertith SRC-2 in P-signaling pathways that ensure completeecidualization. Because the SRC-1KO uterus also exhibitspartial decidual response (Section 1 and Ref. [23]) and thatRCs (through increased expression) have been shown toompensate for the absence of another [23], Western anal-sis was undertaken to rule out the possibility that the par-ial decidual response exhibited by the PRCre/+SRC-2flox/flox

terus may be indirectly linked to a parallel reduction in uter-ne SRC-1. Fig. 2D clearly demonstrates that uterine SRC-

levels are not altered in the PRCre/+SRC-2flox/flox mouseupporting the conclusion that the partial decidual response

henotype is directly attributable to loss of uterine SRC-2.

Because the partial decidual response displayed by theRCre/+SRC-2flox/flox bigenic phenocopies the limited decid-al response exhibited by the SRC-1KO uterus [23] we posit

eP(a

L) uterine horn. Panel (B) graphs the normalized weight ratios (± standardeviation (S.D.)) of stimulated (L) to control (R) horn for SRC-2flox/flox (1)nd the PRCre/+SRC-2flox/flox SRC-1KO trigenic (2). From [30] (Copyright2006] Molecular and Cellular Biology).

hat uterine SRC-2 and SRC-1 may have coevolved to facil-tate full decidualization in response to P. To test this asser-ion, the PRCre/+SRC-2flox/flox mutation was introduced intohe SRC-1KO germline to create the PRCre/+SRC-2flox/flox

RC-1KO trigenic model. Fig. 3A and B shows that whilehe SRC-2flox/flox uterus can exhibit a complete decidualesponse, the uterus from the PRCre/+SRC-2flox/flox SRC-1KOrigenic failed to display even a partial decidual response. Theailure of the trigenic uterus to elicit a decidual response rein-orces the proposal that, as coactivators, SRC-1 and SRC-2re necessary and sufficient in ensuring complete P-inducedecidualization.

.4. SRC-2 exerts a selective coactivator role in uterus

Since ER is expressed in the same uterine cell-typeshat express PR [30] and that SRC-2 has been implicatedn ER mediated signaling [32,33], the PRCre/+SRC-2flox/flox

terus was tested to determine whether loss of SRC-2 inhis cell-type compromises established ER-mediated signal-ng events that lead to uterine luminal epithelial prolifer-tion. Compared to uteri from ovariectomized hormonallyntreated mice, uteri from E-treated ovariectomized WT andRCre/+SRC-2flox/flox mice showed an equivalent proliferativeesponse to E [30]. Furthermore, the increase in E-inducedpithelial proliferation in uteri from both genotypes wasccompanied by the classic uterotropic response, revealinghat SRC-2 is not required for unopposed E-induced uterine

pithelial proliferation. Moreover, SRC-2 is not required for-suppression of E-induced uterine epithelial proliferationdata not shown); note: for postmenopausal hormone ther-py, progestins have been routinely included along with E

A. Mukherjee et al. / Journal of Steroid Biochemist

Fig. 4. The PRCre/+SRC-2flox/flox mammary gland exhibits severeimpairments in branching morphogenesis. Panels (A) and (B) showwhole-mounts of inguinal mammary glands from EP-treated SRC-2flox/flox

and PRCre/+SRC-2flox/flox mice, respectively (LN denotes lymph node).Panels (C) and (D) are higher magnifications of regions in panels (A)and (B), respectively. Compared to the EP-treated SRC-2flox/flox gland,note the significant reduction in ductal side-branching and alveologenesis(black arrow) in the EP-treated PRCre/+SRC-2flox/flox gland. Panels (E) and(F) are hematoxylin and eosin (HE) stained sections of glands shown inpanels (A) and (B), respectively; compared to the EP-treated SRC-2flox/flox

gland, the marked decrease in epithelial content in the similarly treatedPRCre/+SRC-2flox/flox gland is indicated by the arrow. In contrast to theEP-treated SRC-2flox/flox gland (panel (G) (multiple arrows)), note the sig-nificantly lower number of luminal epithelial cells scoring positive for BrdU

tpticsmd

2a

mTStmmstwdUggflmmtPeo

3

ss2sAi

i(sPmmiSaB

ry & Molecular Biology 102 (2006) 22–31 27

o counteract the adverse hyperproliferative effects of unop-osed E action in the endometrium. Moreover, the observa-ion that SRC-2 is not required for P-suppression of uter-ne PR expression (data not shown) further supports theonclusion that SRC-2 has evolved to serve only a sub-et of P initiated transcriptional programs in the uterus,ainly those that are required for embryo implantation and

ecidualization.

.5. SRC-2 is required for ductal side-branching andlveologenesis in the mammary gland

Similar to the PRKO [1], the PRCre/+SRC-2flox/flox mam-ary gland develops normally to adulthood (data not shown).his observation suggests that, like PR, mammary derivedRC-2 is not required for early post-natal development of

he mammary gland. Because SRC-2 is expressed in mam-ary PR positive cells [30], we reasoned that this coactivatoray have a role in PR-mediated signaling that leads to ductal

ide-branching and alveologenesis in the adult gland. To testhis prediction, the PRCre/+SRC-2flox/flox mouse was treatedith a standard 3-week EP treatment regimen which inducesuctal side-branching and alveologenesis in the WT gland.nlike the hormone-treated WT, the PRCre/+SRC-2flox/flox

land failed to elicit full ductal side-branching and alveolo-enesis in response to EP treatment (Fig. 4A–F). As reportedor the PRKO phenotype, the PRCre/+SRC-2flox/flox epithe-ium did not undergo proliferation even in the presence of

ammary PR (Fig. 4G–I). Despite normal levels of mam-ary SRC-1 and -3 in the PRCre/+SRC-2flox/flox (Fig. 4I),

hese data underscore an essential role for SRC-2 function in-induced mammary ductal side-branching and alveologen-sis, cellular processes that normally occur with pregnancynset.

. Implications for the human

Consistent with previous reports [34], we have demon-trated that the transactivational potency of human PR isignificantly enhanced with increasing levels of human SRC-

(Fig. 5A), suggesting that this coactivator is critical forteroid-dependent physiological processes in the human.s further support for this proposal, immunohistochem-

stry demonstrates that SRC-2 is expressed in a number

ncorporation (arrow) in the EP-treated PRCre/+SRC-2flox/flox gland (panelH)). Panel (I): the average percentage of mammary epithelial cells (±S.D.)coring positive for BrdU staining in the EP-treated SRC-2flox/flox andRCre/+SRC-2flox/flox glands is tabulated. The inset displays a Western forammary SRC-1 and -3 for SRC-2flox/flox (1) and PRCre/+SRC-2flox/flox (2)ice. Significant alterations in the levels of SRC-1 and -3 are not detected

n the PRCre/+SRC-2flox/flox mammary gland (�-actin is a loading control).cale bars in panels (A), (C), (E), and (G) apply to panels (B), (D), (F)nd (H), respectively. From [30] (Copyright [2006] Molecular and Cellulariology).

28 A. Mukherjee et al. / Journal of Steroid Biochemistry & Molecular Biology 102 (2006) 22–31

Fig. 5. SRC-2 is expressed in steroid hormone responsive tissues of the human Panel (A) graphically shows that the increase in transactivational functionalityof human PR-B is dependent on increased levels of human SRC-2 (red bars ± S.D.) in the presence of P ligand; Note: in the absence of ligand this increase is notobserved (blue bars). As described previously [44], human PR-B; SRC-2 (both cloned into pCR3.1) and the luciferase reporter pGRE.E1b.LUC were transientlycotransfected into HeLa cells in the presence or absence of 10−7 M progesterone; results are expressed in relative light units (RLU). Panel (B) shows SRC-2is expressed in the majority of epithelial cells of the human prostate (black arrow), an established cellular target for androgen receptor action [35]; Note: thestromal compartment scores negative for SRC-2 expression (blue arrow). Panel (C) shows a representative example of a normal type 1 terminal ductal lobularunit (TDLU) of the human breast in which SRC-2 expression is restricted to the epithelial compartment (black arrow). Panel (D) is a higher magnification ofthe region indicated by the black arrow in panel (C). Note that SRC-2 expression is restricted to a subset of epithelial cells of the TDLU (black arrow indicatesan epithelial cell scoring positive for SRC-2 expression whereas the red arrow highlights an epithelial cell which is negative for SRC-2 expression; blue arrowdenotes a stromal cell which is negative for SRC-2 expression). The spatial pattern of mammary SRC-2 expression resembles that previously reported for ER-�and PR expression in the human breast [38]. Panels (E) and (F) show sections of the glandular epithelial compartment (with associated stroma) of the humanendometrium stained for PR and SRC-2 expression, respectively. Note that PR and SRC-2 are located in the nucleus of the same cell-types in the glandularepithelium and stromal compartment (black and blue arrows, respectively). The red arrow in panels (E) and (F) indicates a stromal cell that is negative forPR and SRC-2 expression, respectively; scale bar in panel (E) applies to panel (F). Endometrial biopsies were obtained by endometrial pipelle from normallycycling women, aged between 18 and 35 years, during the mid-secretory (luteal) phase of the menstrual cycle (days 20–24 which is based on the ideal 28 daysc the dau stochemb e sampli ferred t

otht

ycle in which day 1 represents the first day of menstrual flow and day 14rinary luteinizing hormone (LH) surge. Human SRC-2 and PR immunohiy our group [30,45]. With institutional review board approval, human tissunterpretation of the references to color in this figure legend, the reader is re

f target tissues of the human that are directly responsiveo steroid hormone signaling (Fig. 5B–F). In the case ofuman prostate (Fig. 5B), SRC-2 expression is restricted tohe epithelial compartment, a direct target-site for androgen

sari

y of ovulation). Cycle phase was determined relative to the timing of theical detection was performed by established methods previously reported

es were obtained from Baylor College of Medicine affiliated hospitals. (Foro the web version of the article.)

ignaling [35]; previous studies have shown SRC-2 to bepotent coactivator for the androgen receptor [36,37]. Of

elevance to studies described here, nuclear immunoreactiv-ty for SRC-2 is clearly detected in a subset of epithelial

chemist

cmot2iioiAogbgbriSnaanii

2wic2[ieiirtfiibleiwad

lripumtb

4

irrbmf

cti2ff

Solrpb

PifSaSmce

tagstmsmiiiiwa

tod

A. Mukherjee et al. / Journal of Steroid Bio

ells in the normal human breast (Fig. 5C and D); like theouse [30], SRC-2 is not expressed in the stroma. More-

ver, the punctate spatial expression pattern for SRC-2 inhe human mammary epithelium draws parallels with SRC-’s expression pattern in the murine mammary gland [30]n which SRC-2 colocalizes with PR expression. Interest-ngly, PR has also been shown to exhibit a punctate patternf expression in the human breast [38], whether PR colocal-zes with SRC-2 in the human breast remains to be examined.ddressing this latter issue will be important as segregationf PR positive mammary epithelial cells from cells under-oing proliferation is now recognized as a conserved featureetween normal rodent and human mammary tissue and sug-ests a paracrine mechanism for PR action in the normalreast of both species, reviewed in [39]. Based on similaregional expression patterns for mammary SRC-2 and PRn the rodent and human coupled with the observation thatRC-2 abrogation in the mouse results in a mammary phe-otype similar to the PRKO suggests that SRC-2 may haven important role not only in PRs paracrine mechanism ofction in the normal breast but also in an autocrine mecha-ism which has been proposed for steroid hormone signalingn some breast cancers of rodent and human origin, reviewedn [39].

The most overt phenotype displayed by the PRCre/+SRC-flox/flox female is an infertility defect, the underlying cause ofhich is an inability to develop a receptive uterus for embryo

mplantation (Table 1 and (Figs. 2 and 3)). Immunohisto-hemical studies have clearly demonstrated that uterine SRC-and PR colocalize to the same cell-types in the murine uterus

30], supporting a direct role for SRC-2 in P-initiated signal-ng cascades that are essential for preparing the uterus formbryo implantation and subsequent decidualization. Whilet is not known whether SRC-2 exhibits a similar functionn the human uterus, immunohistochemical analysis clearlyeveals that SRC-2 and PR localize to many of the same cell-ypes of the human uterus (Fig. 5G–F) that have been reportedor the mouse [30]. From a clinical standpoint, recurrentmplantation failure is now recognized as a key limiting factorn the establishment of pregnancy either by natural means ory assisted reproductive technologies (ART) [40]. Althoughittle is known regarding the role of SRC-2 in the humanndometrium, one report has described abnormal elevationsn SRC-2 levels in endometrial biopsies from infertile womenith polycystic ovarian syndrome (PCOS) [41], suggestingpossible role for this coactivator in human reproductive

isorders.Considering the pervasive use of progestins (and to a

esser extent antiprogestins) in the management of femaleeproductive health, coregulators (such as SRC-2) that aredentified as essential for PR mediated responses to a givenrogestin treatment modality could represent novel molec-

lar markers for several areas of diagnostic reproductiveedicine as well as future targets for intervention in the

reatment of gynecologic pathologies and abnormal uterineleeding.

hsui

ry & Molecular Biology 102 (2006) 22–31 29

. Summary and conclusions

In this report, we have evaluated the possible autonomous,nteracting or redundant coactivator roles of SRC-2 in femaleeproductive and mammary developmental processes thately on P signaling. To attain this goal, a PRCre/+SRC-2flox/flox

igenic was generated to avoid the complex reproductive andetabolic defects of the TIF2−/− through ablation of SRC-2

unction only in cell-lineages that express the PR.We reveal that absence of SRC-2 in PR positive uterine

ells results in an inability of the PRCre/+ SRC-2flox/flox mouseo develop a receptive uterus for blastocyst implantation. Themplantation defect not only explains why the PRCre/+SRC-flox/flox female is infertile but markedly distinguishes this KOrom KOs for SRC-1 and -3 which do not exhibit implantationailure or an infertility phenotype for that matter.

The partial decidual response exhibited by the PRCre/+

RC-2flox/flox uterus indicates that uterine PRs dependencyn SRC-2 precipitously diversifies to other coactivators fol-owing the implantation process. The absence of a decidualesponse in the PRCre/+ SRC-2flox/floxSRC-1KO trigenic sup-orts this conclusion in which PR requires not only SRC-2ut also SRC-1 to mediate a full decidual reaction.

The observation that the mammary gland of the adultRcre/+ SRC-2flox/flox failed to exhibit ductal-side branch-

ng and alveologenesis in response to hormone stimulationurnishes strong support for a critical coactivator role forRC-2 in PR-mediated signal transduction pathways thatre required for parity-induced mammary morphogenesis.imilar to the PRKO, the basis of the PRcre/+ SRC-2flox/flox

ammary phenotype is an inability of the luminal epithelialompartment to undergo proliferation in response to hormonexposure.

Remarkably, the PRcre/+ SRC-2flox/flox mammary pheno-ype was not compensated by SRC-3; this coactivator (as wells SRC-1) exists at normal levels in the PRcre/+ SRC-2flox/flox

land. Although SRC-3 has been shown to be involved interoid-induced mammary morphogenesis [24] as well asumorigenesis [25,42,43], our data suggest that SRC-2 and -3

ay operate separately in the mammary epithelial cell. Irre-pective of the functional interrelationships between mam-ary SRC-2 and other members of the SRC family, our stud-

es reveal SRC-2 to be an important coactivator for P signalingn the mammary epithelial cell. Considering P’s importancen mammary tumorigenesis [2–4] and SRC-2’s role in P-nitiated mammary epithelial proliferation [30], future studiesill test whether SRC-2, like SRC-3 [25,42,43], can act as

n oncogene in the mammary gland.In conclusion, identification of tissue-specific coregula-

ors that are selectively co-opted by PR in vivo representsne of the next important challenges in understanding whyifferent target tissues display different responses to the same

ormone, in this case P. In the studies described herein,tate-of-the-art murine genetic approaches have afforded usnprecedented access into deciphering SRC-2’s importancen a number of target tissue responses that rely on progestin

3 chemist

ecgg

A

tISMoT4R

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

0 A. Mukherjee et al. / Journal of Steroid Bio

xposure and have underscored the key involvement of thisoactivator in a subset of PR-mediated transcriptional pro-rams that are necessary for normal uterine and mammaryland function.

cknowledgements

We thank Drs. Pierre Chambon and Martine Gehin, Insti-ut Clinique de la Souris (ICS-IGBMC), BP10142, 67404LLKIRCH Cedex France, for kindly providing the floxedRC-2 mouse model and Dr. Jun Qin, Baylor College ofedicine, for the SRC-2 antibody. The technical assistance

f Jie Li, Yan Ying, and Jie Han is gratefully acknowledged.his research was supported by NIH and private grants HD-2311 (F.J.D.), and CA-07730 and Susan G. Komen Breastesearch Cancer Program (J.P.L.).

eferences

[1] J.P. Lydon, F.J. DeMayo, C.R. Funk, S.K. Mani, A.R. Hughes, C.A.Montgomery Jr., G. Shyamala, O.M. Conneely, B.W. O’Malley, Micelacking progesterone receptors exhibit pleiotropic reproductive abnor-malities, Genes Develop. 9 (1995) 2266–2278.

[2] J.P. Lydon, G. Ge, F.S. Kittrell, D. Medina, B.W. O’Malley, Murinemammary gland carcinogenesis is critically dependent on progesteronereceptor function, Cancer Res. 59 (1999) 4276–4284.

[3] R.T. Chatterton Jr., J.P. Lydon, R.G. Mehta, E.T. Mateo, A. Pletz,V.C. Jordan, Role of the progesterone receptor (PR) in susceptibilityof mouse mammary gland to 7,12-dimethylbenz[a]anthracene-inducedhormone-independent preneoplastic lesions in vitro, Cancer Lett. 188(2002) 47–52.

[4] D. Medina, F.S. Kittrell, A. Shepard, A. Contreras, J.M. Rosen, J.Lydon, Hormone dependence in premalignant mammary progression,Cancer Res. 63 (2003) 1067–1072.

[5] J.W. Jull, The effects of oestrogens and progesterone on the chemicalinduction of mammary cancer in mice of the IF strain, J. Path. Bact. 68(1954) 547–559.

[6] S.P. Robinson, V.C. Jordan, Reversal of antitumor effects of tamox-ifen by progesterone in the 7,12-dimethylbenzanthracene-induced ratmammary carcinoma model, Cancer Res. 47 (1987) 5386–5390.

[7] I.H. Russo, J. Russo, Progestagens and mammary gland development:differentiation versus carcinogenesis, Acta Endocrinol. 125 (1991)7–12.

[8] C.M. Aldaz, Q.Y. Liao, M. La Bate, D.A. Johnston, Medroxyproges-terone acetate accelerates the development and increases the incidenceof mouse mammary tumors induced by dimethylbenzanthracene, Car-cinogenesis 17 (1996) 2069–2072.

[9] G.A. Colditz, K.M. Egan, M.J. Stampfer, Hormone replacement therapyand risk of breast cancer: results from epidemiologic studies, Am. J.Obstet. Gynecol. 168 (1993) 1473–1480.

10] C. Schairer, J. Lubin, R. Troisi, S. Sturgeon, L. Brinton, R.Hoover, Menopausal estrogen and estrogen–progestin replacementtherapy and breast cancer risk [see comments], JAMA 283 (2000)485–491.

11] R.K. Ross, A. Paganini-Hill, P.C. Wan, M.C. Pike, Effect of hormonereplacement therapy on breast cancer risk: estrogen versus estrogen

plus progestin, J. Natl. Cancer Inst. 92 (2000) 328–332.

12] J.E. Rossouw, G.L. Anderson, R.L. Prentice, A.Z. LaCroix, C. Kooper-berg, M.L. Stefanick, R.D. Jackson, S.A. Beresford, B.V. Howard, K.C.Johnson, J.M. Kotchen, J. Ockene, Risks and benefits of estrogen plusprogestin in healthy postmenopausal women: principal results from

[

ry & Molecular Biology 102 (2006) 22–31

the women’s health initiative randomized controlled trial, JAMA 288(2002) 321–333.

13] V. Beral, Breast cancer and hormone-replacement therapy in the millionwomen study, Lancet 362 (2003) 419–427.

14] S.K. Das, I. Chakraborty, B.C. Paria, X.N. Wang, G. Plowman,S.K. Dey, Amphiregulin is an implantation-specific and progesterone-regulated gene in the mouse uterus, Mol. Endocrinol. 9 (1995)691–705.

15] H. Lim, L. Ma, W.G. Ma, R.L. Maas, S.K. Dey, Hoxa-10 regulates uter-ine stromal cell responsiveness to progesterone during implantation anddecidualization in the mouse, Mol. Endocrinol. 13 (1999) 1005–1017.

16] C. Brisken, A. Heineman, T. Chavarria, B. Elenbaas, J. Tan, S.K. Dey,J.A. McMahon, A.P. McMahon, R.A. Weinberg, Essential function ofWnt-4 in mammary gland development downstream of progesteronesignaling, Genes Develop. 14 (2000) 650–654.

17] Y.P. Cheon, Q. Li, X. Xu, F.J. DeMayo, I.C. Bagchi, M.K. Bagchi,A genomic approach to identify novel progesterone receptor regulatedpathways in the uterus during implantation, Mol. Endocrinol. 16 (2002)2853–2871.

18] N. Takamoto, B. Zhao, S.Y. Tsai, F.J. DeMayo, Identification of Indianhedgehog as a progesterone-responsive gene in the murine uterus, Mol.Endocrinol. 16 (2002) 2338–2348.

19] P.M. Ismail, F.J. DeMayo, P. Amato, J.P. Lydon, Progesterone inductionof calcitonin expression in the murine mammary gland, J. Endocrinol.180 (2004) 287–295.

20] J.W. Jeong, K.Y. Lee, I. Kwak, L.D. White, S.G. Hilsenbeck, J.P. Lydon,F.J. DeMayo, Identification of murine uterine genes regulated in aligand-dependent manner by the progesterone receptor, Endocrinology146 (2005) 3490–3505.

21] N.J. McKenna, B.W. O’Malley, Combinatorial control of gene expres-sion by nuclear receptors and coregulators, Cell 108 (2002) 465–474.

22] D.M. Lonard, B.W. O’Malley, Expanding functional diversity of thecoactivators, Trends Biochem. Sci. 30 (2005) 126–132.

23] J. Xu, Y. Qiu, F.J. DeMayo, S.Y. Tsai, M.J. Tsai, B.W. O’Malley, Par-tial hormone resistance in mice with disruption of the steroid receptorcoactivator-1 (SRC-1) gene, Science 279 (1998) 1922–1925.

24] J. Xu, L. Liao, G. Ning, H. Yoshida-Komiya, C. Deng, B.W.O’Malley, The steroid receptor coactivator SRC-3 (p/CIP/RAC3/AIB1/ACTR/TRAM-1) is required for normal growth, puberty, femalereproductive function, and mammary gland development, Proc. Natl.Acad. Sci. U.S.A. 97 (2000) 6379–6384.

25] S.Q. Kuang, L. Liao, S. Wang, D. Medina, B.W. O’Malley, J. Xu, Micelacking the amplified in breast cancer 1/steroid receptor coactivator-3are resistant to chemical carcinogen-induced mammary tumorigenesis,Cancer Res. 65 (2005) 7993–8002.

26] S.J. Han, F.J. DeMayo, J. Xu, S.Y. Tsai, M.J. Tsai, B.W. O’Malley,Steroid receptor coactivator (SRC)-1 and SRC-3 differentially modu-late tissue-specific activation functions of the progesterone receptor,Mol. Endocrinol. 20 (2006) 45–55.

27] M. Gehin, M. Mark, C. Dennefeld, A. Dierich, H. Gronemeyer, P.Chambon, The function of TIF2/GRIP1 in mouse reproduction is dis-tinct from those of SRC-1 and p/CIP, Mol. Cell. Biol. 22 (2002)5923–5937.

28] S.M. Soyal, A. Mukherjee, K.Y. Lee, J. Li, H. Li, F.J. DeMayo, J.P.Lydon, Cre-mediated recombination in cell lineages that express theprogesterone receptor, Genesis 41 (2005) 58–66.

29] J.W. Jeong, I. Kwak, K.Y. Lee, L.D. White, X.P. Wang, F.C. Brunicardi,W. O’Malley, B.F.J. DeMayo, The genomic analysis of the impact ofsteroid receptor coactivators (SRCs) ablation on hepatic metabolism,Mol. Endocrinol. 20 (2006) 1138–1152.

30] A. Mukherjee, S.M. Soyal, R. Fernandez-Valdivia, M. Gehin, P. Cham-bon, F.J. DeMayo, J.P. Lydon, B.W. O’Malley, Steroid receptor coac-

tivator 2 is critical for progesterone-dependent uterine function andmammary morphogenesis in the mouse, Mol. Cell. Biol. 26 (2006)6571–6583.

31] F. Picard, M. Gehin, J. Annicotte, S. Rocchi, M.F. Champy, B.W.O’Malley, P. Chambon, J. Auwerx, SRC-1 and TIF2 control energy

chemist

[

[

[

[

[

[

[

[

[

[

[

[

[

A. Mukherjee et al. / Journal of Steroid Bio

balance between white and brown adipose tissues, Cell 111 (2002)931–941.

32] A. Cvoro, C. Tzagarakis-Foster, D. Tatomer, S. Paruthiyil, M.S. Fox,D.C. Leitman, Distinct roles of unliganded and liganded estrogen recep-tors in transcriptional repression, Mol. Cell 21 (2006) 555–564.

33] X. Wei, H. Xu, D. Kufe, MUC1 oncoprotein stabilizes and activatesestrogen receptor alpha, Mol. Cell 21 (2006) 295–305.

34] K. Hofman, J.V. Swinnen, G. Verhoeven, W. Heyns, Coactivation of anendogenous progesterone receptor by TIF2 in COS-7 cells, Biochem.Biophys. Res. Commun. 295 (2002) 469–474.

35] Z. Culig, H. Klocker, G. Bartsch, A. Hobisch, Androgen receptors inprostate cancer, Endocr. Relat. Cancer 9 (2002) 155–170.

36] C.A. Berrevoets, A. Umar, J. Trapman, A.O. Brinkmann, Differentialmodulation of androgen receptor transcriptional activity by the nuclearreceptor co-repressor (N-CoR), Biochem. J. 379 (2004) 731–738.

37] X. Ye, S.J. Han, S.Y. Tsai, F.J. DeMayo, J. Xu, M.J. Tsai, B.W.O’Malley, Roles of steroid receptor coactivator (SRC)-1 and transcrip-tional intermediary factor (TIF) 2 in androgen receptor activity in mice,Proc. Natl. Acad. Sci. U.S.A. 102 (2005) 9487–9492.

38] R.B. Clarke, A. Howell, C.S. Potten, E. Anderson, Dissociationbetween steroid receptor expression and cell proliferation in the humanbreast, Cancer Res. 57 (1997) 4987–4991.

39] R. Fernandez-Valdivia, A. Mukherjee, B. Mulac-Jericevic, O.M. Con-neely, F.J. DeMayo, P. Amato, J.P. Lydon, Revealing progesterone’s

[

ry & Molecular Biology 102 (2006) 22–31 31

role in uterine and mammary gland biology: insights from the mouse,Semin. Reprod. Med. 23 (2005) 22–37.

40] E.R. Norwitz, D.J. Schust, S.J. Fisher, Implantation and the survival ofearly pregnancy, N. Engl. J. Med. 345 (2001) 1400–1408.

41] C.W. Gregory, E.M. Wilson, K.B. Apparao, R.A. Lininger, W.R. Meyer,A. Kowalik, M.A. Fritz, B.A. Lessey, Steroid receptor coactivatorexpression throughout the menstrual cycle in normal and abnormalendometrium, J. Clin. Endocrinol. Metab. 87 (2002) 2960–2966.

42] M.I. Torres-Arzayus, J. Font de Mora, J. Yuan, F. Vazquez, R. Bronson,M. Rue, W.R. Sellers, M. Brown, High tumor incidence and activationof the PI3K/AKT pathway in transgenic mice define AIB1 as an onco-gene, Cancer Cell 6 (2004) 263–274.

43] S.L. Anzick, J. Kononen, R.L. Walker, D.O. Azorsa, M.M. Tanner, X.Y.Guan, G. Sauter, O.P. Kallioniemi, J.M. Trent, P.S. Meltzer, AIB1,a steroid receptor coactivator amplified in breast and ovarian cancer,Science 277 (1997) 965–968.

44] D.M. Lonard, S.Y. Tsai, B.W. O’Malley, Selective estrogen receptormodulators 4-hydroxytamoxifen and raloxifene impact the stability andfunction of SRC-1 and SRC-3 coactivator proteins, Mol. Cell. Biol. 24

(2004) 14–24.

45] S. Lee, S.K. Mohsin, S. Mao, S.G. Hilsenbeck, D. Medina, D.C. Allred,Hormones, receptors, and growth in hyperplastic enlarged lobular units:early potential precursors of breast cancer, Breast Cancer Res. 8 (2005)1–9.