Expression of aromatase P450 (P450arom), which catalyzes the formation of estrogens, is aberrantly increased in adipose fibroblasts surrounding breast carcinomas, giving rise to proliferation of malignant cells. Aromatase in human adipose tissue is primarily expressed in undifferentiated fibroblasts under the control of several distinct and alternatively used P450arom promoters. In tumor-free breast adipose tissue, P450arom is usually expressed at low levels via a distal promoter (I.4), whereas in the breast adipose tissue bearing a tumor, P450arom is increased through the activation of two proximal promoters, II and I.3. Because the in vivo activation of P450arom promoter II is a key event responsible for aberrantly high P450arom expression in breast tumors, we studied the molecular basis for the enhancement of P450arom promoter II using human adipose fibroblasts (HAFs) in primary culture treated with T47D breast cancer cell-conditioned medium (TCM) as a model system. Upon treatment with TCM, HAFs displayed a striking induction of P450arom mRNA levels via promoter II usage. This effect appeared to be specific for malignant breast epithelial cells, because conditioned media from breast cancer cell lines T47D and MCF-7 induced promoter II activity, whereas normal breast epithelial cells or liver or prostate cancer cell lines did not produce such an effect. Although treatment with a cyclic AMP analogue also caused a switch in the promoter use from I.4 to II in cultured HAFs, TCM-induced promoter II use was found to be mediated via a cyclic AMP-independent pathway. Use of serial deletion mutants of the promoter II 5′-flanking sequence revealed the presence of critical cis-acting elements in the −517/−278 bp region, which regulate the baseline activity. TCM caused a 5.7-fold induction of the −517-bp promoter II construct, whereas site-directed mutagenesis of a CCAAT/enhancer binding protein (C/EBP) binding site (−317/−304 bp) abolished both baseline and TCM-induced activities. Ectopic expressions of C/EBPα and C/EBPβ, but not C/EBPδ, significantly induced promoter II activity. Moreover, we demonstrated the presence of both C/EBPβ and C/EBPδ but not C/EBPα in a DNA-protein complex formed by the nuclear extract from TCM-treated HAFs and a probe containing this critical C/EBP binding element (−317/−304 bp). Finally, treatment of HAFs with TCM strikingly induced C/EBPβ expression, whereas this did not affect the levels of C/EBPα or C/EBPδ transcripts. In conclusion, malignant breast epithelial cells secrete factors, which induce aromatase expression in adipose fibroblasts via promoter II. This is, at least in part, mediated by a TCM-induced up-regulation and enhanced binding of C/EBPβ to a promoter II regulatory element.

The conversion of C19 steroids to estrogens by P450arom3 takes place in a number of human cells, e.g., the ovarian granulosa cell (1), skin, and adipose fibroblasts (2, 3). Aromatase expression in the adipose tissue is limited to fibroblasts and is not detected in significant quantities in the fully differentiated and lipid-filled adipocytes (2, 3). Aromatase activity in adipose fibroblasts has long been implicated in the pathophysiology of breast cancer growth (4, 5, 6, 7). Estrogen produced in breast adipose tissue acts locally to promote the growth of tumor (8). Thus, the relationship between adipose stroma and breast cancer is unique in that the adipose fibroblast provides structural and functional support for cancer growth. O’Neill et al.(6) demonstrated that the breast quadrant displaying the highest level of aromatase activity was consistently involved with tumor. Subsequently, we found the highest levels of P450arom transcripts in adipose tissue from the quadrant bearing a tumor (7). In the same study, tumor-bearing quadrants contained the highest fibroblast:adipocyte ratios. It follows then that the breast quadrant with the highest fibroblast content contains the highest levels of P450arom transcripts. The clinical relevance of these observations has been exemplified by the successful treatment of breast carcinomas with potent aromatase inhibitors (9, 10, 11).

Expression of the human P450arom (CYP19) gene is under the control of several distinct and partly tissue-specific promoters (12, 13). Three of these promoters (I.4, I.3, and II) are used in adipose tissue. Interestingly, in disease-free breast adipose tissue, P450arom is usually expressed at low levels via a distal promoter (I.4), whereas in the adipose tissue of the breast bearing a tumor, P450arom expression is increased through the activation of two proximal promoters, II and I.3 (14, 15, 16). In addition to these in vivo observations, treatments of HAFs in culture with various hormones switch promoter use. For example, glucocorticoids plus cytokines induce P450arom expression via promoter I.4 in cultured primary HAFs, whereas treatment with a cAMP analogue switches the promoter use to II and I.3 (12, 13). We hypothesized that malignant breast epithelial cells interact with the surrounding adipose tissue fibroblasts to activate promoters II and I.3. The data presented in this report will serve to reconcile the in vivo and in vitro observations summarized above (12, 13, 14). We report a novel epithelial-stromal interaction, which favors the induction of P450arom expression in HAFs by malignant epithelial cells via promoter II.

We and others have shown previously that breast cancer cells could stimulate aromatase expression in HAFs, which was suggestive of cross-talk between malignant epithelial cells and surrounding HAFs to favor estrogen production in breast tumors (17, 18, 19). We demonstrated recently that medium conditioned with malignant epithelial cells inhibited the differentiation of HAFs to mature adipocytes via the suppression of the essential adipogenic transcription factors C/EBPα and PPARγ. C/EBPβ and C/EBP/δ, on the other hand, were up-regulated in these undifferentiated murine fibroblasts treated with TCM (20). TCM-induced decreases in C/EBPα or PPARγ were sufficient to completely inhibit adipogenic differentiation of 3T3-L1 cells in our hands (20). This was in agreement with reports published previously (21, 22). On the other hand, we were intrigued by the TCM-induced increases in C/EBPβ and C/EBP/δ mRNA levels in 3T3-L1 murine cells (20). In contrast to C/EBPα or PPARγ, ectopic expressions of C/EBPβ or C/EBPδ were not sufficient to induce adipocyte differentiation in the absence of C/EBPα or PPARγ (23). Thus, we hypothesize that breast cancer-induced increases in C/EBPβ or C/EBP/δ levels do not affect adipocyte differentiation but may serve to increase aromatase expression in adipose fibroblasts surrounding the cancer. We used herein a model whereby human TCM is added to primary HAFs to understand the roles of C/EBP isoforms in the up-regulation of aromatase expression in undifferentiated fibroblasts. We chose to study the activation of promoter II, because work from three different laboratories demonstrated that the activity of this promoter was up-regulated in vivo in breast stroma bearing a carcinoma (14, 15, 16).

Cell Cultures.

Human adipose tissues were obtained at the time of surgery from women undergoing reduction mammoplasty following a protocol approved by the Institutional Review Board for Human Research of the University of Illinois at Chicago. For primary HAF cultures, adipose tissues were minced and digested with collagenase B (1 mg/ml) at 37°C for 2 h. Single-cell suspensions were prepared by filtration through a 75-μm sieve. Fresh cells were suspended in DMEM/F12 containing 10% FBS in a humidified atmosphere with 5% CO2 at 37°C. Twelve to 24 h after the attachment of fibroblasts, culture medium was removed, and cell medium was changed at 48-h intervals until the cells became confluent. Before total RNA or nuclear proteins were extracted from HAFs, these cells were cultured in either serum-free DMEM/F12, DMEM/F12 containing 10% FBS, serum-free DMEM/F12 containing Bt2cAMP (0.5 mm) together with PDA (100 nm), DMEM/F12 containing 10% FBS plus DEX (250 nm), or DMEM/F12 conditioned with malignant or benign cells. All treatments were continued for 48 h.

T47D cells purchased from American Type Culture Collection (Rockville, MD) were initially grown in RPMI 1640 with 10% FBS containing 0.02 mm HEPES, whereas MCF-7 cells, prostate cancer cell line PC-3, and hepatocellular carcinoma cell line HepG2 (American Type Culture Collection) were grown in MEM with 10% FBS. Human normal mammary epithelial cells purchased from Clonetics, Inc. (Walkersville, MD) were grown in fully supplemented MEGM medium (Clonetics). Before shipment, these cells were passed twice and demonstrated to contain immunoreactive cytokeratins 14 and 18. In our hands, these cells were alive and dividing every 48–72 h. Cell-conditioned media from T47D, MCF-7, PC-3, HepG2, or normal mammary epithelial cells were collected to be used subsequently as treatments on HAFs. To collect conditioned media, cells were initially grown to confluence and switched to DMEM/F12 for a 12-h washout period; then, cells were incubated in DMEM/F12 for 24 h to allow accumulation of secreted factors in the medium.

RT-PCR Amplification.

Amplification of the untranslated 5′ ends of P450arom transcripts from HAFs under various treatments was accomplished with exon-specific oligonucleotide pairs as described below. Five μg of DNase I-treated total RNA were used for reverse transcriptase reaction. Five μl of reverse transcriptase mixture were amplified using PCR. For the amplification of total P450arom transcripts, 5′-end sense primer from coding exon II (5′-TTG GAA ATG CTG AAC CCG AT-3′) and 3′-end antisense primer complimentary to coding exon III (5′-CAG GAA TCT GCC CTG GGG AT-3′) were used. To amplify promoter-specific 5′-untranslated sequences, primers for promoter II-specific sequence (5′-GCA ACA GGA GCT ATA GAT-3′) and exon I.4 (5′-GTA GAA CGT GAC CAA CTG G-3′) were used as 5′-end sense primers, together with an antisense primer complimentary to the coding exon III (5′-ATT CCC ATG CAG TAG CCA GG-3′). PCR conditions were as follows: denaturing at 95°C for 30 s, annealing at 55°C for amplification of promoter II-specific sequence or 58°C for amplification of exon I.4 and the coding region for 40 s, and extension at 72°C for 40 s for 30 cycles. GAPDH was chosen as an endogenous marker to check the integrity of cDNA. A 5′-end sense primer (5′-CGG AGT CAA CGG ATT TGG TCG TAT-3′) and a 3′-end antisense primer (5′-AGC CTT CTC CAT GGT GGT GAA GAC-3′) were used for amplifying a 306-bp-long sequence in GAPDH mRNA. PCR conditions were the same as those used for amplification of promoter II-specific fragments, except for the number of cycles (21) and the quantity of reverse transcriptase mixture (0.5 μl). This RT-PCR method was described previously in greater detail (14).

Determination of Intracellular cAMP.

HAFs were plated in six-well, 35-mm culture dishes. After reaching confluence, HAFs were cultured either in serum-free DMEM/F12, DMEM/F12 containing 10% FBS, DMEM/F12 containing 10% FBS and forskolin (10 μm), DMEM/F12 containing 10% FBS and DEX (250 nm), or T47D cell conditioned DMEM/F12. Measurements were performed in triplicate replicates, and treatments were carried for 0, 12, 24, and 48 h. HAFs were lysed in a 0.1 m HCl solution after the removal of medium. Cell lysis mixture was centrifuged, and the supernatant was then used directly in the cAMP assay using Direct Cyclic AMP Enzyme Immunoassay kit (Assay Design, Inc., Ann Arbor, MI), following the protocol supplied by the vendor. Briefly, 50 μl of the pink-neutralizing reagent were added into each well, except for the total activity and blank wells. Samples (100 μl) were then added to appropriate wells. Fifty μl of the conjugate were added into each well, followed by the addition of 50 μl of the yellow antibody. After incubating at room temperature for 2 h on a shaker at 500 rpm, the plate was washed three times with 200 μl of washing buffer, followed by the addition of substrate solution 200 μl to each well. The stop solution (50 μl) was then added to each well, and the absorbance was read at 405 nm with correction to 570 nm. Results were obtained by plotting on the standard curve.

Transient Transfections and Luciferase Assays.

HAFs in primary culture were transfected using Lipofectamine Plus (Life Technologies, Inc., Grand Island, NY) with the following plasmids: (a) 1 μg of modified PGL3-Basic Luciferase reporter plasmid that contains serial deletion mutants of P450arom promoter II; (b) 0.2 μg of pcDNA3 expression plasmid (Invitrogen, Carlsbad, CA), which contains the cDNA of either C/EBPα (human), C/EBPβ (rat) or C/EBPδ (rat); and (c) 5 ng of pRL-CMV Renilla luciferase control reporter vectors that contain the cDNA encoding Renilla luciferase (Promega Corp., Madison, WI) as an internal control for transfection efficiency. The day before transfection, HAFs in primary culture were seeded into 35-mm dishes at 2 × 105 cell/dish. The transfection solution was made of 200 μl of OPTI-MEM I reduced-serum medium containing PLUS reagent (8 μl), precomplexed DNA (1.2 μg), and 5 μl of Lipofectamine reagent. After transfection for 6 h in transfection solution at 37°C in 5% CO2, medium was changed to antibiotic-free DMEM/F12 containing 10% FBS for overnight recovery. Cells were then switched to medium conditioned by normal breast epithelial cells or T47D cells for another 48 h. Luciferase and Renilla luciferase assays were performed using a dual-luciferase reporter assay system kit (Promega). Results are presented as the average of data from triplicate replicates and expressed as the ratio to the internal standard Renilla luciferase. The empty luciferase vector PGL3-Basic was arbitrarily assigned a unit of 1, and the rest of the results were expressed as multiples of the PGL3-Basic vector.

Northern Blotting.

Total RNA was isolated from HAFs in primary culture growing in: (a) DMEM/F12; (b) normal breast epithelial cell conditioned medium; or (c) TCM. Twenty μg of total RNA were used. cDNA probes for C/EBPδ, C/EBPβ, and C/EBPα were prepared from plasmids kindly provided by Drs. Steve McKnight (University of Texas Southwestern Medical Center, Dallas, TX), Gokhan Hotamisligil (Harvard Medical School, Boston, MA), and Gretchen Darlington (Baylor College of Medicine, Houston, TX).

Site-directed Mutagenesis.

To generate serial plasmids bearing mutated consensus-binding sequences for transcription factors of C/EBPs, SF-1 and CREB, site-directed mutagenesis was performed using the GeneEditor in vitro site-directed mutagenesis system (Promega), per the manufacturer’s instructions. A −517-bp promoter II/PGL3-Basic construct containing wild-type −517/−16 bp of P450arom promoter II 5′-flanking DNA was used as a template for site-directed mutagenesis. Briefly, DNA template (0.5 pmol) was denatured and annealed with mutagenic and selection oligonucleotides. Mutant strand was synthesized in the reaction mixture containing 1× synthesis buffer, 5 units of T4 DNA polymerase, and 2 units of T4 DNA ligase at 37°C for 90 min. The mutagenesis reaction mixture was then used to transform BMH 71-18 mutS competent cells. These transformed competent cells were incubated in a medium containing GeneEditor antibiotic selection mix overnight to select the desired mutant plasmids. The plasmids isolated from the BMH 71-18 mutS were transformed into JM109 competent cells. The transformed JM109 competent cells were grown overnight on the LB plates containing ampicillin and GeneEditor antibiotic selection mix to further select the mutated plasmids. The mutation of binding consensus was confirmed by DNA sequencing. Consensus binding sequences for mutation and primers used were depicted in Table 1.

EMSA.

The nuclear extracts used for EMSA were prepared as described previously (24). Briefly, cells were grown to confluence and cultured in either DMEM/F12 only or T47D cell conditioned DMEM/F12 for 48 h. Cells were then scraped from the dishes. The cell pellet was resuspended in cold buffer A [10 mm HEPES (pH 7.4), 1.5 mm MgCl2, 10 mm KCl, 9.5 mm DTT, 10 μg/ml leupeptin, 100 μg/ml pepstatin, 2 μg aprotinin, 0.5 mm and phenylmethylsulfonyl fluoride]. The cell pellets were homogenized on ice. Once >90% of the cell membranes were broken, the lysate was centrifuged for 2 min at 700 × g. After the supernatant was removed, the nuclear pellet was resuspended in buffer C [20 mm HEPES (pH 7.4), 420 mm NaCl, 1.5 mm MgCl2, 0.2 mm EDTA, 0.5 mm DTT, and 20% glycerol] and incubated on ice for 30 min with intermittent mixing. After centrifugation at 60,000 rpm for 5 min at 4°C, the supernatant was snap-frozen in liquid nitrogen. Protein concentrations were determined by a modified Bradford assay (Bio-Rad, Hercules, CA), and nuclear extracts were stored at −80°C.

Double-stranded oligonucleotides were obtained through annealing sense and antisense sequences. The double-stranded oligonucleotide probes were end-labeled with [γ-32P]ATP using T4 kinase. EMSAs were performed as described previously (24). Briefly, 5 μg of nuclear extract were incubated with the radiolabeled double-stranded oligonucleotide probe for 15 min at room temperature in a reaction buffer containing 20 mm HEPES (pH 7.6), 75 mm KCl, 0.2 mm EDTA, 20% glycerol, and 2 μg of poly(deoxyinosinic-deoxycytidylic acid) as a nonspecific competitor. Protein-DNA complexes were resolved on 6% nondenaturing polyacrylamide gels. EMSAs were performed after the addition of 0.5 μl of an antibody against C/EBPα, C/EBPβ, C/EBPδ, or CREB to the binding reaction, followed by a 30-min incubation on ice before electrophoresis. All antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). We used the following double-stranded probes. C/EBP binding site probe (5′-GAA GAA GAT TGC CTA AAC AA-3′) represents an identical 20-bp-long sequence (−303/−322) in the promoter II regulatory region of the P450arom gene. Mutated C/EBP binding site probe (5′-GAA GAA Gcc cGC CTg gtC AA-3′) contains a mutated version of C/EBP binding motif that does not interact with any of the C/EBP isoforms.

Aromatase Expression in HAFs Is Stimulated by Breast Cancer Cell Conditioned Medium via the P450arom Promoter II.

Aromatase expression in human tissues is under the control of alternatively used and partially tissue-specific promoters. The coding region of P450arom transcripts and, thus, the translated protein, however, are identical in each tissue site of expression. In the breast adipose tissue of disease-free women, P450arom expression is expressed at low levels via a distal promoter I.4, whereas in the adipose tissue of breast bearing a tumor, P450arom expression is increased through activation of an ovarian-type proximal promoter II (14, 15, 16). We attempted to verify these in vivo data by the following in vitro experiments. Because the use of each alternative promoter gives rise to a P450arom transcript with an untranslated 5′-end unique for that particular promoter, we used exon-specific RT-PCR to determine total and promoter-specific P450arom transcript levels in HAFs in primary culture treated with TCM. As expected, Bt2cAMP plus PDA stimulated P450arom transcript levels primarily via activation of promoter II, whereas DEX plus serum activated promoter I.4. Most importantly, we found that TCM stimulated P450arom transcript levels via P450arom promoter II activation, which was demonstrated previously in vivo in adipose tissue of the breast bearing a tumor (Fig. 1,A). To address the specificity of effects of T47D cells on HAFs, we used media conditioned with either the MCF-7 breast cancer cell line or normal breast epithelial cells (NCM) to treat HAFs. Medium conditioned with MCF-7 cells but not with normal epithelial cells induced P450arom transcript levels via promoter II (Fig. 1,B). Furthermore, other malignant cell lines HepG2 and PC-3 failed to activate P450arom gene transcription via promoter II, which demonstrated that the stimulatory effect produced by T47D and MCF7 cells was specific for breast cancer (Fig. 1 C). These results demonstrate that malignant breast epithelial cells in culture produce specific factors, which stimulate aromatase expression via promoter II.

Activation of P450arom Promoter II by Breast Cancer Cell Conditioned Medium Is Not cAMP Dependent.

Because both TCM and cAMP analogues induce aromatase expression in HAF via promoter II, we sought to determine whether this effect of TCM is mediated via increased formation of cAMP in HAFs. Therefore, we first measured the intracellular levels of cAMP in HAFs treated with forskolin, FBS, FBS plus DEX, or TCM. Contrary to our expectations, treatment with TCM decreased intracellular levels of cAMP at 12-, 24-, and 48-h time points (Fig. 2,A). FBS or FBS plus DEX also decreased cAMP levels, whereas treatment with the adenylate cyclase inducer forskolin (10 μm) gave rise to a striking increase in cAMP levels in HAFs (positive control; Fig. 2,A). On the other hand, addition of the adenylate cyclase inhibitor SQ 22,536 to the culture medium 0.5 h before the treatment with TCM for 48 h did not inhibit promoter II activation (negative control; Fig. 2 B). These results indicate that the activation of P450arom promoter II by TCM is mediated via a cAMP-independent pathway.

Regulation of P450arom Promoter II Activity in Primary HAFs.

We determined the genomic regions critical for the regulation of baseline levels of promoter II activity in HAFs (Fig. 3). Use of serial deletion mutants of promoter II fused to luciferase reporter gene demonstrated that the −517/−278-bp region contained critical stimulatory elements (Fig. 3).

The C/EBP Binding Sequence (−317/−304 bp) Is Essential for the Breast Cancer Cell-induced Activation of Promoter II.

We identified two C/EBP binding sites in the −517/−278 bp region using the TFSEARCH database (Fig. 4).4 The −278/−100-bp region contains two SF-1 sites and a CRE. One of these SF-1 binding sites (−136/−124 bp) and CRE (−211/−197 bp) were shown previously to be critical for cAMP-induced promoter II activity in ovarian granulosa cells and endometriosis-derived stromal cells (24, 25). We determined the effect of TCM on the activity of the −517-bp promoter II/luciferase construct, because this construct showed the highest baseline activity (Fig. 3). Treatment with TCM for 48 h induced the activity of the −517-bp construct by 5.7-fold (Fig. 5). Site-directed mutagenesis of five potentially important cis-acting elements demonstrated that CRE (−211/−197 bp) and a C/EBP binding site (−317/−304 bp) were essential for TCM induction of promoter II activity. In particular, mutation of the −317/−304-bp C/EBP binding site completely abolished both baseline and TCM-induced activities. Mutation of the two SF-1 sites or the −350/−337-bp C/EBP binding site did not effect TCM induction of promoter II activity (Fig. 5). We found that NCM did not change the activity of the −517-bp construct in comparison with incubation with DMEM/F12 only (data not shown). Therefore, promoter II activity in NCM-treated HAFs is similar to the baseline level. In this particular experiment illustrated in Fig. 5, we determined the TCM fold induction of promoter in comparison with NCM treatment.

Induction of Promoter II Activity in HAFs by Factors Derived from T47D Breast Cancer Cells Is Mediated by C/EBPβ.

Fig. 6 depicts the effects of the adipogenic transcription factors, C/EBPα, C/EBPβ, and C/EBPδ, on the activity of the −517-bp promoter II construct in HAFs. Ectopic expressions of C/EBPβ (3.5-fold) and C/EBPα (2.5-fold) stimulated promoter II activity, whereas C/EBPδ did not have any significant effect (Fig. 6).

Thus far, these results were indicative of TCM induction of promoter II activity via a C/EBP binding site (−317/−304 bp). Ectopic expressions of C/EBPα and C/EBPβ significantly stimulated the −517-bp promoter II construct (Fig. 6). To determine whether C/EBPα or C/EBPβ mediates TCM induction of promoter II, EMSA was used using an oligonucleotide probe (−322/−303 bp) containing the −317/−304-bp C/EBP binding site, nuclear extracts from HAFs incubated with or without TCM, and supershifting antibodies against C/EBPα, C/EBPβ, and C/EBPδ. This C/EBP binding site (−317/−304 bp) was chosen to be included in the probe, because this element was found to be critical for TCM activation of promoter II (Fig. 5). We identified two specific complexes (1 and 2) as verified by wild-type and mutated cold competitors in TCM-treated HAFs (Fig. 7). Antibodies against both C/EBPβ and C/EBPδ supershifted complex 1, indicating the presence of C/EBPβ and C/EBPδ. On the other hand, antibodies against C/EBPα or CREB did not eliminate or supershift any of these complexes. To further investigate whether the activation of P450arom promoter II is mediated by C/EBPβ, we demonstrated that the effects of TCM and C/EBPβ were not additive. TCM stimulated the −517 construct by 6-fold, whereas the addition of C/EBPβ to TCM did not further increase this induction, which was suggestive that the effects of TCM on promoter II were, at least in part, mediated by C/EBPβ (Fig. 7 B).

These experiments were suggestive that TCM induction of promoter II activity was mediated by C/EBPβ but not by C/EBPα or C/EBPδ, because C/EBPα does not bind to the regulatory element at −317/−304 bp, which is critical for TCM stimulation of promoter II. Although C/EBPδ binds to this site, ectopic expression of C/EBPδ does not increase promoter II activity. To confirm this conclusion, we determined the effects of TCM on the mRNA levels of C/EBP isoforms in HAFs. Treatments with TCM or NCM did not change the mRNA levels of C/EBPα or C/EBPδ. On the other hand, only TCM induced C/EBPβ expression in HAFs strikingly (Fig. 8). Thus, we conclude that TCM induction of P450arom promoter II in HAF is mediated, at least in part, by the induction of the expression of C/EBPβ, which binds to the −317/−304 bp region in this promoter.

The understanding of the molecular mechanisms that are responsible for aberrant P450arom expression in tumor-bearing breast adipose tissue may provide insights into the etiology of breast cancer and lead to the identification of molecular targets for the development of novel treatment strategies. Investigators from at least four different laboratories have demonstrated strikingly increased levels of aromatase activity and P450arom mRNA in breast adipose tissue containing a tumor compared with breast tissue from disease-free women (6, 7, 14, 15, 16). It was also consistently found that up-regulation of promoter II activity was responsible, in part, for increased aromatase expression in breast cancer (14, 15, 16). Although it was suggested that promoter II up-regulation by breast tumors might be mediated by prostaglandin E2 and cAMP, no direct evidence to support this concept has been provided to date (26). Another report, on the other hand, supports our findings regarding the effect of MCF-7 breast cancer cells on switching the promoter use from I.4 to II (27). The downstream signal transduction events or the specificity of breast epithelial cell types, however, has not been characterized in this article (27). We herein present data to support that malignant breast cells induce aromatase expression via promoter II using a cAMP-independent mechanism. A key event is the binding of the adipogenic factor C/EBPβ to a specific cis-acting element upstream of promoter II to activate its transcription. We have used an in vitro system to support our conclusion. The use of malignant and normal breast epithelial cell conditioned media with clear and consistent biological effects on fibroblasts and the use of positive and negative controls for cell types and various components of signal transduction pathways, however, offset the disadvantages of using an in vitro system and permit the performance of useful mechanistic experiments. Additionally, our conclusions are supported by in vivo data from human breast cancer specimens, which showed down-regulation of C/EBPα but persistent expression of C/EBPβ and C/EBPδ proximal to malignant cells (20). On the basis of these data, we suggest the following model. Breast cancer cells secrete cytokines that selectively down-regulate essential adipogenic factors, which inhibit the differentiation of fibroblasts to mature adipocytes. Estrogen production in these fibroblasts maintained in the undifferentiated state by malignant cells is further enhanced by tumor-derived factors, which exist in the T47D or MCF-7 cell conditioned media. These factors act via a cAMP-independent pathway to increase C/EBPβ expression in adipose fibroblasts and enhance the binding of C/EBPβ to a specific promoter II regulatory sequence. The end result is increased local estrogen concentration in the breast tumor.

We do not know yet the identities of the unknown factors in TCM that increase C/EBPβ expression and P450arom promoter II activity in adipose fibroblasts. Cytokines such as TNF-α and IL-6 were shown to increase the transcriptional activity of C/EBPβ (28, 29). It is not clear, however, whether these cytokines increase C/EBPβ expression or promoter II activity. Our preliminary findings and previous publications demonstrated that these cytokines (IL-11, IL-6, and TNFα) do not activate P450arom promoter II, which is up-regulated in vivo in breast tumors. Instead, these substances activate promoter I.4, which is not up-regulated in tumors (17, 30, 31, 32).5 Therefore, these cytokines by themselves probably do not account for the in vivo up-regulation of aromatase expression in breast tumors. Our efforts will continue to identify these unknown factors originating from malignant cells to induce aromatase expression in the adipose fibroblast via promoter II.

Fig. 1.

A and B, breast cancer cells stimulate P450arom expression via promoter II in primary HAFs. HAFs in primary culture were incubated under various conditions for 48 h [No treatment, serum-free DMEM/F12; TCM; DEX + FBS (10%), DEX (250 nm) and FBS; Bt2cAMP + PDA, 0.5 mm Bt2cAMP + 100 nm PDA; 10% FBS, DMEM/F12 + 10% FBS; MCF-7-CM, MCF-7 cell conditioned medium; NCM]. Total RNA was subjected to exon-specific RT-PCR to amplify promoter-specific untranslated first exons. GAPDH was amplified to control the integrity of RNA. A, total P450arom transcript levels were up-regulated by treatments with T47D-CM and Bt2cAMP + PDA via the use of promoter II. DEX + FBS treatment, on the other hand, robustly induced P450arom transcripts via another promoter, promoter I.4. B, medium conditioned by another breast cancer cell line, MCF-7, also induced P450arom transcripts via promoter II usage, whereas NCM did not induce P450arom transcript levels at all. C, media conditioned by prostate cancer cell line PC-3 and hepatocellular carcinoma cell line HepG2 failed to induce promoter II or the levels of total P450arom transcripts (Coding region).

Fig. 1.

A and B, breast cancer cells stimulate P450arom expression via promoter II in primary HAFs. HAFs in primary culture were incubated under various conditions for 48 h [No treatment, serum-free DMEM/F12; TCM; DEX + FBS (10%), DEX (250 nm) and FBS; Bt2cAMP + PDA, 0.5 mm Bt2cAMP + 100 nm PDA; 10% FBS, DMEM/F12 + 10% FBS; MCF-7-CM, MCF-7 cell conditioned medium; NCM]. Total RNA was subjected to exon-specific RT-PCR to amplify promoter-specific untranslated first exons. GAPDH was amplified to control the integrity of RNA. A, total P450arom transcript levels were up-regulated by treatments with T47D-CM and Bt2cAMP + PDA via the use of promoter II. DEX + FBS treatment, on the other hand, robustly induced P450arom transcripts via another promoter, promoter I.4. B, medium conditioned by another breast cancer cell line, MCF-7, also induced P450arom transcripts via promoter II usage, whereas NCM did not induce P450arom transcript levels at all. C, media conditioned by prostate cancer cell line PC-3 and hepatocellular carcinoma cell line HepG2 failed to induce promoter II or the levels of total P450arom transcripts (Coding region).

Close modal
Fig. 2.

A and B, activation of P450arom promoter II by TCM is not cAMP dependent. HAFs were incubated under various conditions [No treatment, serum-free DMEM/F12; 10% FBS, DMEM/F12 + 10% FBS; Forskolin, 10 μm forskolin; DEX+ 10% FBS, 250 nm DEX + 10% FBS; TCM; NCM; TCM + SQ22,536, TCM + 100 μm SQ 22,536; No RT, no reverse transcriptase reaction mixture for negative control]. HAFs were then sampled at 0, 12, 24, and 48 h for intracellular cAMP assay or at 48 h for semiquantitative RT-PCR. A, TCM did not increase the intracellular cAMP levels at 12, 24, and 48 h as in treatments with FBS and DEX-FBS. On the other hand, forskolin, an adenylate cyclase inducer, strikingly increased cAMP levels, as expected. Bars, SE. B, SQ 22,536, an adenylate cyclase inhibitor, could not eliminate the TCM-induced induction of P450arom promoter II-specific transcripts.

Fig. 2.

A and B, activation of P450arom promoter II by TCM is not cAMP dependent. HAFs were incubated under various conditions [No treatment, serum-free DMEM/F12; 10% FBS, DMEM/F12 + 10% FBS; Forskolin, 10 μm forskolin; DEX+ 10% FBS, 250 nm DEX + 10% FBS; TCM; NCM; TCM + SQ22,536, TCM + 100 μm SQ 22,536; No RT, no reverse transcriptase reaction mixture for negative control]. HAFs were then sampled at 0, 12, 24, and 48 h for intracellular cAMP assay or at 48 h for semiquantitative RT-PCR. A, TCM did not increase the intracellular cAMP levels at 12, 24, and 48 h as in treatments with FBS and DEX-FBS. On the other hand, forskolin, an adenylate cyclase inducer, strikingly increased cAMP levels, as expected. Bars, SE. B, SQ 22,536, an adenylate cyclase inhibitor, could not eliminate the TCM-induced induction of P450arom promoter II-specific transcripts.

Close modal
Fig. 3.

Regulation of P450arom promoter II activity in primary HAFs. Luciferase plasmids containing the 5′-flanking region of human P450arom promoter II with serial deletions (−100, −140, −214, −278, −517, and −694 bp) were transfected into HAFs. pCMV Renilla was used as an internal control for transfection efficiency. Promoter II activity was normalized to pCMV Renilla and was represented as the average of data from triplicate replicates; bars, SE. The empty luciferase vector PGL3-Basic was arbitrarily assigned a unit of 1, and the rest of the results were expressed as multiples of the PGL3-Basic vector. We conclude that the −517/−278-bp region contains critical stimulatory elements, which regulate the baseline activity.

Fig. 3.

Regulation of P450arom promoter II activity in primary HAFs. Luciferase plasmids containing the 5′-flanking region of human P450arom promoter II with serial deletions (−100, −140, −214, −278, −517, and −694 bp) were transfected into HAFs. pCMV Renilla was used as an internal control for transfection efficiency. Promoter II activity was normalized to pCMV Renilla and was represented as the average of data from triplicate replicates; bars, SE. The empty luciferase vector PGL3-Basic was arbitrarily assigned a unit of 1, and the rest of the results were expressed as multiples of the PGL3-Basic vector. We conclude that the −517/−278-bp region contains critical stimulatory elements, which regulate the baseline activity.

Close modal
Fig. 4.

Potential cis-acting elements located in the P450arom promoter II region. A computer-assisted search revealed two C/EBP binding sites at −350/−337 bp and −317/−304 bp, located within the −517/−278 bp region of promoter II. Additionally, two previously identified SF-1 sites and a CRE are present within the −278/−100-bp region. The percentages depict the homology to the consensus sequences.

Fig. 4.

Potential cis-acting elements located in the P450arom promoter II region. A computer-assisted search revealed two C/EBP binding sites at −350/−337 bp and −317/−304 bp, located within the −517/−278 bp region of promoter II. Additionally, two previously identified SF-1 sites and a CRE are present within the −278/−100-bp region. The percentages depict the homology to the consensus sequences.

Close modal
Fig. 5.

A C/EBP binding site (−317/−304 bp) is essential for both basal and TCM-induced activation of promoter II. TCM induced the activity of the −517-bp construct by 5.7-fold as compared with NCM treatment. Site-directed mutagenesis of five potentially important cis-acting elements demonstrated that a CRE (−211/−197 bp) and a C/EBP binding site (−317/−304 bp) were essential for TCM induction of promoter II activity. In particular, mutation of the −317/−304-bp C/EBP binding site completely abolished both baseline and TCM-induced activities. Mutations of the two SF-1 sites or another C/EBP binding site (−350/−337 bp) did not affect TCM induction of promoter II activity. Bars, SE.

Fig. 5.

A C/EBP binding site (−317/−304 bp) is essential for both basal and TCM-induced activation of promoter II. TCM induced the activity of the −517-bp construct by 5.7-fold as compared with NCM treatment. Site-directed mutagenesis of five potentially important cis-acting elements demonstrated that a CRE (−211/−197 bp) and a C/EBP binding site (−317/−304 bp) were essential for TCM induction of promoter II activity. In particular, mutation of the −317/−304-bp C/EBP binding site completely abolished both baseline and TCM-induced activities. Mutations of the two SF-1 sites or another C/EBP binding site (−350/−337 bp) did not affect TCM induction of promoter II activity. Bars, SE.

Close modal
Fig. 6.

The effects of adipogenic transcription factors, C/EBPα, C/EBPβ, and C/EBPδ, on the activity of the −517-bp promoter II construct in HAFs. Mammalian expression vectors of C/EBPs were cotransfected into HAFs, together with the −517-bp promoter II construct. Ectopic expressions of C/EBPβ (3.5-fold) and C/EBPα (2.5-fold) stimulated promoter II activity, whereas C/EBPδ did not have any significant effects. Bars, SE.

Fig. 6.

The effects of adipogenic transcription factors, C/EBPα, C/EBPβ, and C/EBPδ, on the activity of the −517-bp promoter II construct in HAFs. Mammalian expression vectors of C/EBPs were cotransfected into HAFs, together with the −517-bp promoter II construct. Ectopic expressions of C/EBPβ (3.5-fold) and C/EBPα (2.5-fold) stimulated promoter II activity, whereas C/EBPδ did not have any significant effects. Bars, SE.

Close modal
Fig. 7.

The effects of TCM on promoter II is mediated by C/EBPβ. A, C/EBPβ and C/EBPδ bind to the C/EBP site (−317/−304 bp) upstream of promoter II in TCM-treated HAFs. EMSA was used with an oligonucleotide probe (−322/−303 bp) containing the −317/−304-bp C/EBP binding sequence, nuclear extracts from HAFs incubated with or without TCM, and supershifting antibodies against C/EBPα, C/EBPβ, C/EBPδ, and CREB. We identified two specific complexes (1 and 2) as verified by wild-type (WT) and mutated (Mut) cold competitors in TCM-treated HAFs. Antibodies against C/EBPβ or C/EBPδ supershifted complex 1, indicating the presence of C/EBPβ and C/EBPδ in this complex. On the other hand, C/EBPα or CREB did not eliminate or supershift any of these complexes. B, TCM induced the activity of the −517-bp promoter II construct by six-fold, whereas the addition of C/EBPβ to TCM did not increase the promoter II activity any further. Bars, SE.

Fig. 7.

The effects of TCM on promoter II is mediated by C/EBPβ. A, C/EBPβ and C/EBPδ bind to the C/EBP site (−317/−304 bp) upstream of promoter II in TCM-treated HAFs. EMSA was used with an oligonucleotide probe (−322/−303 bp) containing the −317/−304-bp C/EBP binding sequence, nuclear extracts from HAFs incubated with or without TCM, and supershifting antibodies against C/EBPα, C/EBPβ, C/EBPδ, and CREB. We identified two specific complexes (1 and 2) as verified by wild-type (WT) and mutated (Mut) cold competitors in TCM-treated HAFs. Antibodies against C/EBPβ or C/EBPδ supershifted complex 1, indicating the presence of C/EBPβ and C/EBPδ in this complex. On the other hand, C/EBPα or CREB did not eliminate or supershift any of these complexes. B, TCM induced the activity of the −517-bp promoter II construct by six-fold, whereas the addition of C/EBPβ to TCM did not increase the promoter II activity any further. Bars, SE.

Close modal
Fig. 8.

C/EBPβ transcripts are induced by TCM in HAFs. HAFs were treated by NCM or TCM, or left untreated (No treatment) for 48 h. Twenty μg of total RNA isolated from each sample were then used for Northern blot analysis. The 28S RNA fraction was included to demonstrate the presence of comparable amounts of total RNA in each lane. TCM profoundly increased the expression of C/EBPβ, whereas the expression patterns of C/EBPα and C/EBPδ were not altered by TCM treatment.

Fig. 8.

C/EBPβ transcripts are induced by TCM in HAFs. HAFs were treated by NCM or TCM, or left untreated (No treatment) for 48 h. Twenty μg of total RNA isolated from each sample were then used for Northern blot analysis. The 28S RNA fraction was included to demonstrate the presence of comparable amounts of total RNA in each lane. TCM profoundly increased the expression of C/EBPβ, whereas the expression patterns of C/EBPα and C/EBPδ were not altered by TCM treatment.

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

Supported by United States Army Medical Research and Materiel Command Grant DAMD17-97-l-7025 and National Cancer Institute Grant CA67167 (to S. E. B.).

3

The abbreviations used are: P450arom, aromatase P450; HAF, human breast adipose fibroblast; cAMP, cyclic AMP; C/EBP, CCAAT/enhancer binding protein; PPAR, peroxisome proliferator-activated receptor; TCM, T47D cell conditioned medium; Bt2cAMP, dibutyryl cAMP; PDA, phorbol diacetate; DEX, dexamethasone; FBS, fetal bovine serum; NCM, normal mammary epithelial cell conditioned medium; RT-PCR, reverse transcription-PCR; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; SF, steroidogenic factor; CREB, cAMP response element binding protein; EMSA, electrophoresis mobility shift analysis; TNF, tumor necrosis factor; IL, interleukin; CMV, cytomegalovirus.

4

Internet address: http://www.blast.genome.ad.jp/sit/TFSEARCH.

5

J. Zhou, B. Gurates, S. Yang, S. Sebastian, and S. E. Bulun, unpublished observations.

Table 1

Primers used for site-directed mutagenesis

Mutated consensus sequencea5′-Phosphorylated and mutagenic primersa
C/EBP binding site (−350/−337 bp) 5′-GGG AGA TTG CCT TTT TGT ccc GAA ATT GAT TTG GCT TC-3′ 
TTGTTTTGAAATT→TTGTcccGAgggT 5′-ATT GCC TTT TTG Tcc cGA ggg TGA TTT GGC TTC AAG GG-3′ 
C/EBP binding site (−317/−304 bp) 5′-TGG CTT CAA GGG AAG AAG ccc GCC TAA ACA AAA CCT GCT G-3′ 
AGATTGCCTAAACA→AgcccGCCTggtCA 5′-CAA GGG AAG AAG ccc GCC Tgg tCA AAA CCT GCT GAT GAA G-3′ 
SF-1 binding site (−263/−251 bp) 5′-GAC TCC ACC TCT GGA ATG gGa aTT ATT TTC TTA TAA TTT GGC-3′ 
ATGAGCTTTATTT→ATGgGaaTTATTT  
SF-1 binding site (−136/−124 bp) 5′-GGA ACC TGA GAC TCT ACC Acc cTC AGA AAT GCT GCA ATT CAA GC-3′ 
AGGTCAGAAA→cccTCAGAAA  
CRE (−211/−197 bp) 5′-GGC TTT CAA TTG GGA ATG gAa tTC ACT CTA CCC ACT CAA GGG CA-3′ 
TGCACGTCACTCT→TGgAatTCACTCT  
Mutated consensus sequencea5′-Phosphorylated and mutagenic primersa
C/EBP binding site (−350/−337 bp) 5′-GGG AGA TTG CCT TTT TGT ccc GAA ATT GAT TTG GCT TC-3′ 
TTGTTTTGAAATT→TTGTcccGAgggT 5′-ATT GCC TTT TTG Tcc cGA ggg TGA TTT GGC TTC AAG GG-3′ 
C/EBP binding site (−317/−304 bp) 5′-TGG CTT CAA GGG AAG AAG ccc GCC TAA ACA AAA CCT GCT G-3′ 
AGATTGCCTAAACA→AgcccGCCTggtCA 5′-CAA GGG AAG AAG ccc GCC Tgg tCA AAA CCT GCT GAT GAA G-3′ 
SF-1 binding site (−263/−251 bp) 5′-GAC TCC ACC TCT GGA ATG gGa aTT ATT TTC TTA TAA TTT GGC-3′ 
ATGAGCTTTATTT→ATGgGaaTTATTT  
SF-1 binding site (−136/−124 bp) 5′-GGA ACC TGA GAC TCT ACC Acc cTC AGA AAT GCT GCA ATT CAA GC-3′ 
AGGTCAGAAA→cccTCAGAAA  
CRE (−211/−197 bp) 5′-GGC TTT CAA TTG GGA ATG gAa tTC ACT CTA CCC ACT CAA GGG CA-3′ 
TGCACGTCACTCT→TGgAatTCACTCT  
a

Lowercase represents the mutated base pairs.

We are grateful for the expert editorial assistance of Dee Alexander.

1
McNatty K. P., Baird D. T., Bolton A., Chambers P., Corker C. S., Maclean H. Concentration of oestrogens and androgens in human ovarian venous plasma and follicular fluid throughout the menstrual cycle.
J. Endocrinol.
,
77
:
77
-85,  
1976
.
2
Ackerman G. E., Smith M. E., Mendelson C. R., MacDonald P. C., Simpson E. R. Aromatization of androstenedione by human adipose tissue stromal cells in monolayer culture.
J. Clin. Endocrinol. Metab.
,
53
:
412
-417,  
1981
.
3
Price T., Aitken J., Head J., Mahendroo M., Means G. D., Simpson E. R. Determination of aromatase cytochrome P450 messenger RNA in human breast tissues by competitive polymerase chain reaction (PCR) amplification.
J. Clin. Endocrinol. Metab.
,
74
:
1247
-1252,  
1992
.
4
Hemsell D. L., Grodin J. M., Brenner P. F., Siiteri P. K., MacDonald P. C. Plasma precursors of estrogen.
J. Clin. Endocrinol. Metab.
,
38
:
476
-479,  
1974
.
5
James V. H., Reed M. J., Lai L. C., Ghilchik M. W., Tait G. H., Newton T. J., Goldham N. G. Regulation of estrogen concentrations in human breast tissues.
Ann. NY Acad. Sci.
,
595
:
227
-235,  
1990
.
6
O’Neill J. S., Elton R. A., Miller W. R. Aromatase activity in adipose tissue from breast quadrants: a link with tumor site.
Br. Med. J.
,
296
:
741
-743,  
1988
.
7
Bulun S. E., Price T. M., Mahendroo M. S., Aitken J., Simpson E. R. A link between breast cancer and local estrogen biosynthesis suggested by quantification of breast adipose tissue aromatase cytochrome P450 transcriptions using competitive polymerase chain reaction after reverse transcription.
J. Clin. Endocrinol. Metab.
,
77
:
1622
-1628,  
1993
.
8
Yue W., Wang J. P., Hamilton C. J., Demers L. M., Santen R. J. In situ aromatizaton enhances breast tumor estradiol levels and cellular proliferation.
Cancer Res.
,
58
:
927
-932,  
1998
.
9
Brodie A. Aromatase and its inhibitors–an overview.
J. Steroid Biochem. Mol. Biol.
,
40
:
255
-261,  
1991
.
10
Santen R. J. Clinical use of aromatase inhibitors in human breast carcinoma.
J. Steroid Biochem. Mol. Biol.
,
40
:
247
-253,  
1991
.
11
Santen R. J. Estrogen synthesis inhibitors from “off the rack” to haute couture.
J. Clin. Endocrinol. Metab.
,
77
:
316
-318,  
1993
.
12
Mahendroo M. S., Mendelson C. R., Simpson E. R. Tissue-specific and hormonally controlled alternative promoters regulate aromatase cytochrome P450 gene expression in human adipose tissue.
J. Biol. Chem.
,
268
:
19463
-19470,  
1993
.
13
Zhao Y., Mendelson C. R., Simpson E. R. Characterization of the sequences of the human CYP19 (aromatase) gene that mediate regulation by glucocorticoids in adipose stromal cells and fetal hepatocytes.
Mol. Endocrinol.
,
9
:
340
-349,  
1995
.
14
Agarwal V. R., Bulun S. E., Leitch M., Rohrich R., Simpson E. R. Use of alternative promoters to express the aromatase cytochrome P450 (CYP19) gene in breast adipose tissues of cancer-free and breast cancer patients.
J. Clin. Endocrinol. Metab.
,
81
:
3843
-3849,  
1996
.
15
Zhou C., Zhou D., Esteban J., Murai J., Siiteri P. K., Wilczynski S., Chen S. Aromatase gene expression and its exon I usage in human breast tumors.
J. Steroid Biochem. Mol. Biol.
,
59
:
163
-171,  
1996
.
16
Harada N. Aberrant expression of aromatase in breast cancer tissues.
J. Steroid Biochem. Mol. Biol.
,
61
:
175
-184,  
1997
.
17
Nichols J. E., Bulun S. E., Simpson E. R. Effects of conditioned medium from different cultured cell types on aromatase expression in adipose stromal cells.
J. Soc. Gynecol. Investig.
,
2
:
45
-49,  
1995
.
18
Singh, A., Purohit, A., Coladham, N. G., Ghilchik, M. W., and Reed, M. J. Biochemical control of breast aromatase. Breast Cancer Res. Treat., 49 (Suppl. 1): S9–S14; discussion, S33–S37, 1998.
19
Miller W. R., Mullen P., Sourdaine P., Watson C., Dixon J. M., Telford J. Regulation of aromatase activity within the breast.
J. Steroid Biochem. Mol. Biol.
,
61
:
193
-202,  
1997
.
20
Meng L., Zhou J. F., Sasano H., Suzuki T., Zeitoun K. M., Bulun S. E. Tumor necrosis factor α and interleukin-11 secreted by malignant breast epithelial cells inhibit adipocyte differentiation by selectively downregulating C/EBPα and PPARγ: mechanism of desmoplastic reaction.
Cancer Res.
,
61
:
2250
-2255,  
2001
.
21
Hu E., Tontonoz P., Spiegelman B. M. Transdifferentiation of myoblasts by the adipogenic transcription factors PPARγ and C/EBPα.
Proc. Natl. Acad. Sci. USA
,
92
:
9856
-9860,  
1995
.
22
Zhang B., Berger J., Hu E., Szalkowski D., White-Carrington S., Spiegelman B. M., Moller D. E. Negative regulation of peroxisome proliferator-activated receptor-γ gene expression contributes to the antiadipogenic effects of tumor necrosis factor-α.
Mol. Endocrinol.
,
10
:
1457
-1466,  
1996
.
23
Cao Z., Umek R. M., McKnight S. L. Regulated expression of three C/EBP isoforms during adipose conversion of 3T3-L1 cells.
Genes Dev.
,
5
:
1538
-1552,  
1991
.
24
Zeitoun K., Takayama K., Michael M. D., Bulun S. E. Stimulation of aromatase P450 promoter II activity in endometriosis and its inhibition in endometrium are regulated by competitive binding of steroidogenic factor-1 and chicken ovalbumin upstream promoter transcription factor to the same cis-acting element.
Mol. Endocrinol.
,
13
:
239
-253,  
1999
.
25
Michael M. D., Michael L. F., Simpson E. R. A CRE-like sequence that binds CREB and contributes to cAMP-dependent regulation of the proximal promoter of human aromatase P450 (CYP 19) gene.
Mol. Cell. Endocrinol.
,
134
:
147
-156,  
1997
.
26
Zhao Y., Agarwal V. R., Mendelson C. R., Simpson E. R. Estrogen biosynthesis proximal to a breast tumor is stimulated by PGE2 via cyclic AMP, leading to activation of promoter II of the CYP 19 (aromatase gene).
Endocrinology
,
137
:
5739
-5742,  
1996
.
27
Harada, N., and Honda, S. Molecular analysis of aberrant expression of aromatase in breast cancer tissues. Breast Cancer Res. Treat., 49 (Suppl. 1): S15–S21; discussion S33–S37, 1998.
28
Baer M., Williams S. C., Dillner A., Schwartz R. C., Johnson P. F. Autocrine signals control CCAAT/enhancer binding protein β expression, localization, and activity in macrophages.
Blood
,
92
:
4353
-4365,  
1998
.
29
Yin M., Yang S. Q., Lin H. Z., Lane M. D., Chatterjee S., Diehl A. M. Tumor necrosis factor α promotes nuclear localization of cytokine-inducible CCAAT/enhancer binding protein isoforms in hepatocytes.
J. Biol. Chem.
,
271
:
17974
-17978,  
1996
.
30
Crichton M. B., Nichols J. E., Zhao Y., Bulun S. E., Simpson E. R. Expression of transcripts of interleukin-6 and related cytokines by human breast tumors, breast cancer cells, and adipose stromal cells.
Mol. Cell. Endocrinol.
,
118
:
215
-220,  
1996
.
31
Singh A., Purohit A., Duncan L. J., Mokbel K., Ghilchik M. W., Reed M. J. Control of aromatase activity in breast tumors: the role of the immune system.
J. Steroid Biochem. Mol. Biol.
,
61
:
185
-192,  
1997
.
32
Zhao Y., Nichols J. E., Valdez R., Mendelson C. R., Simpson E. R. Tumor necrosis factor-α stimulates aromatase gene expression in human adipose stromal cells through use of an activating protein-1 binding site upstream of promoter I.4.
Mol. Endocrinol.
,
10
:
1350
-1357,  
1996
.