Liver Receptor Homologue-1 (LRH-1) Regulates Expression of Aromatase in Preadipocytes*

Estrogen biosynthesis from C19steroids is catalyzed by aromatase cytochrome P450. Aromatase is expressed in breast adipose tissue through the use of a distal, cytokine-responsive promoter (promoter I.4). Breast tumors, however, secrete soluble factors that stimulate aromatase expression through an alternative proximal promoter, promoter II. In other estrogenic tissues such as ovaries, transcription from promoter II requires the presence of the Ftz-F1 homologue steroidogenic factor-1 (SF-1); adipose tissue, however, does not express SF-1. We have explored the hypothesis that in adipose tissue, an alternative Ftz-F1 family member, liver receptor homologue-1 (LRH-1), substitutes for SF-1 in driving transcription from promoter II. In transient transfection assays using 3T3-L1 preadipocytes, promoter II reporter constructs were modestly (2–3-fold) stimulated by either treatment with activators of protein kinases A or C (PKA/C) or by cotransfection with LRH-1. In combination, these treatments synergistically activated promoter II (>30-fold). Induction by LRH-1 (but not by PKA/C) required an AGGTCA motif at −130 base pairs, to which LRH-1 bound in gel shift assays. Activity of GAL4-LRH-1 fusion proteins was not altered by activators of PKA or PKC. Quantitative real-time PCR revealed that LRH-1 (but not SF-1) is expressed in the preadipocyte fraction of human adipose tissue at levels comparable with that of liver. Differentiation of cultured human preadipocytes into mature adipocytes was associated with a time-dependent induction of peroxisome proliferator-activated receptor-γ (PPARγ), and rapid loss of LRH-1 and aromatase expression. We conclude that LRH-1 is a preadipocyte-specific nuclear receptor that regulates expression of aromatase in adipose tissue. Alterations in LRH-1 expression and/or activity in adipose tissue could therefore have considerable effects on local estrogen production and breast cancer development.


INTRODUCTION 4
inhibitors act in a global fashion and inhibit estrogen action or synthesis in all sites of production. This has the potential to result in bone loss and other sequelae of estrogen insufficiency such as cognitive dysfunction and hepatic steatosis with prolonged treatment (25)(26)(27). Thus there is a clear need for more specific, tissueselective anti-estrogens.
The increased CYP19 expression in response to breast tumor derived factors is associated with a switch in promoter usage from the normal adipose-specific promoter I. 4 to the cAMP-responsive promoter II (28)(29)(30). Since these two promoters are regulated by different cohorts of transcription factors and coactivators, it follows that the differential regulation of CYP19 expression via alternative promoters in disease-free and cancerous breast adipose tissue may permit the development of selective aromatase modulators (SAMs) 1 , that target the aberrant over-expression in cancerous breast, while sparing estrogen action in other sites of synthesis such as normal adipose tissue, bone and brain (31). A more complete understanding of the mechanisms regulating CYP19 transcription from promoter II in breast adipose tissue is a prerequisite for the development of such SAMs.
In classic steroidogenic tissue such as ovary and testis, promoter II is regulated by Steroidogenic Factor-1 (SF-1 / Ad4BP / NR5A1 (32) which binds to a nuclear receptor half-site (NRE) within the promoter to mediate basal transcription and, in part, cAMP-induced transcription (12). Although adipose tissue does not express SF-1 (our unpublished observations, and Figure 2 herein), the NRE within promoter II has been shown to bind other negative and positive transcription factors in adipose stromal cells and breast cancer cell lines including COUP-TF and ERRα (33). Thus 1 The abbreviations used are: SAM(s), Selective Aromatase Modulator(s); LRH-1, Liver 5 the activity of promoter II in adipose tissue is controlled, at least in part, by the balance of stimulatory and inhibitory transcription factors binding to this site.
Whereas SF-1 expression is mainly restricted to steroidogenic tissues of the reproductive axis (38), LRH-1 is expressed at high levels in liver where it regulates expression of genes involved in cholesterol metabolism and bile acid synthesis including cholesterol 7αhydroxylase (CYP7A) (34,39,40), sterol 12α-hydroxylase (CYP8B1) (41) and the cholesteryl ester transfer protein (42). LRH-1 is also expressed in the pancreas, ovary, intestine and colon (43), and has been reported to regulate adrenal expression of 11β-hydroxylase (44). It has however been reported to be absent from adipose tissue (43).
In the current study we show that LRH-1 can bind to and strongly activate promoter II of the CYP19 gene. Importantly, LRH-1 is expressed at high levels in the undifferentiated stromal compartment of human adipose tissue (the site of CYP19 expression) but not in differentiated adipocytes, and therefore represents a physiologically relevant regulator of estrogen biosynthesis in breast.

EXPERIMENTAL PROCEDURES
Plasmids PII-516 is a CYP19 promoter II / luficerase construct containing -516 / +17 nt of human CYP19 promoter II, and has been described previously (12,45  were preincubated on ice for 10 min before addition of probe. In some experiments protein extracts were heated to 37°C for 3 minutes either before or after addition of probe.

Western Blotting
Protein lysates of human mammary fat and adipose stromal cells were generated using a lysis buffer (137 mM NaCl, 0.5% nonidet P-40, 10 mM Tris-HCl pH 7.5, and Protease Inhibitor Cocktail tablet (Roche)), separated on a 10% SDS-polyacrylamide gel (15 µg / lane) and transferred to nitrocellulose filter. Filters were probed with a CPF (LRH-1) antibody (Santa Cruz), then stripped and re-probed with a β-tubulin antibody to confirm equal protein loading. Detection was by ECL Plus Western Blotting Detection System (Amersham Pharmacia).

Reverse transcription-PCR
Total RNA was prepared from freshly isolated mouse tissues or various cell lines using the QiaAMP RNA Blood Mini kit (Qiagen). First strand cDNA synthesis from 250 ng total RNA was performed using AMV reverse transcriptase (Roche) primed by random hexamers. PCR reactions were carried out using the following primer sets Real time PCR amplification of LRH-1 and 18S was performed on the LightCycler (Roche) using SYBR Green reaction mix (Roche) and the primers described above.
cDNA samples were diluted 1:20 in water immediately before use. Experimental samples were quantified by comparison with standards of known concentration (0.01 -100 fg / µl)

Regulation of CYP19 transcription by nuclear receptors
To identify receptors that could bind and activate promoter II though the NRE, 3T3L1 mouse preadipocytes were cotransfected with a luciferase reporter construct harboring 516 nt of CYP19 promoter II (12,45) and expression vectors encoding various nuclear receptors (Fig 1). Luciferase activity was not significantly altered by cotransfection with ERRα, NGFIB, Nurr1, Nor1 (Fig 1), or either ERα or ERβ in the presence or absence of ligand (not shown). In contrast, both SF-1 and LRH-1 stimulated promoter II activity 9-fold and 6-fold, respectively. Although the stimulatory effect of SF-1 on CYP19 promoter II in gonadal cells has been well characterized (38), the role of LRH-1 in CYP19 transcription has not previously been investigated. In contrast, expression of SF-1 mRNA was restricted to mouse adrenal and human adrenocortical H295R cells. We also investigated SF-1 and LRH-1 expression in human primary breast cancer tissue ( Fig   2B). All seven breast cancer specimens examined expressed LRH-1, whereas SF-1 was undetectable. Thus LRH-1, but not SF-1, is expressed in breast adipose and cancer tissues. To quantify LRH-1 mRNA expression in these tissues we performed real-time PCR ( Fig 3A). Expressed as the ratio of LRH-1 molecules:18S molecules per µg total RNA, LRH-1 expression in adipose tissue was approximately 20% that of liver.

LRH-1 is expressed in breast adipose and cancer tissue
However, the relative expression in primary cultured adipose stromal cells was much higher -approximately 11-fold higher than in whole adipose tissue, and 2.5-fold higher than in liver. The 61 KDa LRH-1 protein was also readily detectable by Western blotting in isolated preadipocytes, but not in whole adipose tissue (Fig 3B).
Thus although LRH-1 mRNA levels in adipose tissue are relatively low compared to liver, LRH-1 mRNA and protein expression is enriched in the adipose stromal cell compartment. This suggests that LRH-1, like CYP19, may be a marker of the undifferentiated preadipocyte phenotype.
To test this hypothesis, primary cultured human preadipocytes were induced to differentiate into adipocytes by a twelve-day incubation in adipogenic medium ( Fig   3C). Under such conditions lipid droplets became visible after 6 days, and by day 12 approximately 50% of cells exhibited abundant lipid accumulation (Fig 3C, lower panel). The mature adipocyte phenotype was confirmed by a rapid and sustained expression of PPARγ ( Fig 3D, upper panel). LRH-1 mRNA was readily detectable in untreated preadipocytes but was dramatically reduced following 3 days of culture in adipogenic medium, and undetectable at days 9 and 12. CYP19 mRNA expression also displayed a time-dependent decrease with progression of differentiation.
Therefore, LRH-1 is expressed at high levels in human preadipocytes, but not mature adipocytes. Since this expression profile mirrors that of CYP19, LRH-1 is a potential physiological regulator of CYP19 expression in breast adipose stromal cells.

Regulation of CYP19 promoter II by LRH-1
Aromatase activity and CYP19 mRNA expression are strongly induced by prostaglandin E 2 derived from breast cancer cells and / or macrophages infiltrating the tumor site (24). PGE 2 binds to EP1 and EP2 receptors linked to PKC and PKA signaling pathways, activation of which together maximally stimulates CYP19 expression via promoter II (24). To assess the effect of these pathways on LRH-1 induced CYP19 transcription, 3T3L1 cells were co-transfected with the CYP19 promoter II reporter construct and increasing concentrations of LRH-1 expression construct. Cells were then incubated in the presence or absence of the adenylyl cyclase activator forskolin and the PKC activator PMA for 8 hours (Fig 4). We next explored the contribution of the promoter II NRE to transcriptional regulation by FSK, PMA and LRH-1 (Fig 6). 3T3-L1 cells were transfected with LRH-1 and either a wild-type promoter II reporter construct (pII-516) or a promoter construct harboring a mutation in the NRE (AGGTCA → AaaTCA (pII-516mNRE)). Under control conditions (Fig 6 upper panel), LRH-1 increased promoter II activity 2-fold.
This modest stimulation was abolished when the NRE was mutated. In the presence of FSK+PMA (Fig 6 lower panel) activity of pII-516 increased approximately 4-fold.
LRH-1 cotransfection increased luciferase activity by a further 6-fold. Mutation of the NRE did not affect the ability of the promoter to respond to FSK+PMA; however, LRH-1 did not increase activity of this construct. These data suggest that the NRE is required for induction of promoter II by LRH-1, but not by FSK+PMA, and further support the hypothesis that LRH-1 acts as a basal transcription factor whereas hormone-induced transcription occurs through other, non-LRH-1 mechanisms.

Binding of LRH-1 to the promoter II NRE
To ascertain whether LRH-1 derived from adipose stromal cells is capable of binding to the promoter II NRE, a synthetic oligonucleotide probe encompassing this sequence was prepared and used in electrophoretic mobility shift assay (EMSA). In the presence of adipose stromal cell nuclear extracts, two specific protein-DNA  (35,44,49).
Unlike other nuclear receptors, the DNA binding activity of human and rat LRH-1 is rapidly and irreversibly disrupted by incubation at 37°C. Once bound to DNA, however, LRH-1 is resistant to heat treatment (35,49). We therefore explored the thermal sensitivity of adipose stromal cell protein binding and compared it to that of LRH-1 or SF-1 (Fig 7B). Adipose stromal cell nuclear extracts were incubated at 22°C or 37°C for 3 minutes either before or after addition of the DNA binding probe, and subjected to EMSA. No alteration in the pattern of protein / DNA complexes was observed on incubation at 22°C before or after addition of probe (lanes 1&2, 5&6, 9&10 before, but not after, probe addition (lanes 7&8). In contrast, SF-1 binding was unaffected by heat treatment (lanes 11&12). The thermal sensitivity of adipose stromal cell nuclear protein binding to the promoter II NRE therefore resembles that of human LRH-1 (49). Note that the relative insensitivity of in vitro translated mouse LRH-1 to heat treatment likely arises from divergence in the amino acid sequences of mouse and human LRH-1 between the DBD and AF2 domain, the region known to confer thermal instability to human LRH-1 (35,49). Together, these data provide strong evidence that LRH-1 is a component of the protein / DNA complexes formed between adipose stromal cell nuclear extracts and the promoter II NRE. Since the overexpression of aromatase that occurs in adipose tissue of breast cancer patients arises through aberrant transcription from promoter II, we and others have sought to understand the mechanisms controlling the tissue-specific regulation of promoter II (12,28,33,45,47,51).
In the current study we show that LRH-1 is expressed in adipose tissue, and can bind and activate promoter II. There are several important implications of these findings.
Firstly, that LRH-1 and aromatase are co-expressed in preadipocytes provides a mechanism whereby aromatase expression in adipose tissue can be maintained in the absence of SF-1. Secondly, the regulation of promoter II by LRH-1 imparts a new level of control of aromatase expression in breast adipose through changes in LRH-1 expression or activity, which could in turn be modified by hormonal regulation, ligand binding or co-regulator recruitment. Finally, the tight confinement of LRH-1 expression to the preadipocyte fraction of human adipose tissue raises the possibility that LRH-1 may participate in more general control of preadipocyte function and / or adipose differentiation. These points are discussed below.
We hypothesized that since breast preadipocytes do not express SF-1, other nuclear receptors may bind the SF-1 site within promoter II to contribute to basal or hormoneinduced transcription. Indeed, a previous study implicated the orphan receptor ERRα in regulation through this site in breast cancer cell lines (33). We think it unlikely by guest on October 5, 2017 http://www.jbc.org/ Downloaded from however that ERRα contributes to transcription from promoter II in preadipocytes for the following reasons: ERRα was a very weak transactivator in transient transfection assays in SK-BR3 breast cancer cells (~2-fold stimulation of promoter II (33)). We were also unable to detect ERRα expression in human preadipocytes (not shown), and in our hands ERRα did not activate promoter II in transfection assays (Fig 1).
Finally, ERRα expression is positively correlated with adipocyte differentiation (52) whereas aromatase expression is down-regulated on adipocyte differentiation. In contrast, as shown here, LRH-1 is expressed at high level in preadipocytes, strongly transactivates promoter II, and, like aromatase, is rapidly down-regulated upon adipocyte differentiation. For these reasons we propose that LRH-1 is a physiological regulator of aromatase expression in preadipocytes.
Little is known about the regulation of LRH-1 expression or activity. The LRH-1 promoter contains several conserved elements that are thought to confer liver- noteworthy that the LRH-1 promoter contains at least 2 consensus binding sites for the basic helix-loop-helix leucine zipper protein SREBP (53), which itself is closely associated with the process of adipocyte differentiation (56,57).
The rapid loss of LRH-1 expression in differentiating adipocytes might suggest an inhibitory effect of PPARγ or other adipogenic agents on LRH-1 expression.
Preliminary experiments in our laboratory using ligands for PPARγ and / or RXR have not, however, supported this notion (data not shown). An alternative hypothesis would be that LRH-1 inhibits the expression and / or action of PPARγ. In this role, LRH-1 would then play a critical role in adipocyte differentiation. Consistent with this, LRH-1 activity is inhibited by the small heterodimer partner (SHP) (39,40), an atypical orphan receptor that lacks a DNA binding domain (58). SHP is expressed in adipose tissue and markedly potentiates the activity of PPARγ (59). Although this occurred through direct interactions between SHP and PPARγ, involvement of LRH-1 in this process and in adipocyte differentiation in general warrants further investigation.
In conclusion we have shown that LRH-1 is a preadipocyte-specific nuclear receptor that can stimulate CYP19 transcription. Alterations in LRH-1 expression and / or activity have great potential to influence aromatase expression and estrogen production in adipose tissue. In particular, elucidation of the mechanisms regulating LRH-1 expression in normal and tumor-bearing breast adipose tissue will be an important area of future research.      Human adipose stromal cell nuclear extracts or in vitro transcribed / translated LRH-1 or SF-1 were incubated at 22°C or 37°C for 3 minutes either before (22b, 37b) or after (22a, 37a) addition of radiolabeled probe. DNA / protein complexes were visualised as above.
by guest on October 5, 2017