¹Department of Cell and Cancer Biology, Medicine Branch, National Cancer Institute, NIH, Rockville, Maryland, MD 20850, USA

²Department of Microbiology and Immunology, Eastern Virginia Medical School, Norfolk, Virginia, VA 2350, USA

³Department of Medicine, University of Texas Health Science Center at San Antonio, San Antonio, Texas, TX 78284, USA

Correspondence to: Michael J Birrer, 9610 Medical Center Drive, Room 300, Rockville, Maryland, MD 20850, USA

^aCurrent address: Department of Radiological Health Sciences, Colorado State University, Fort Collins, Colorado, CO 80523, USA

^bCurrent address: Department of Cell Biology, Parke-Davis Research, Ann Arbor, Michigon, MI 48105, USA

^cCurrent address: Departments of Clinical Cancer Prevention and Gastrointestinal Medical Oncology, University of Texas, MD Anderson Cancer Center, Houston, Texas, TX 77030, USA

^dCurrent address: Department of Digestive Diseases, University of Texas, MD Anderson Cancer Center, Houston, Texas, TX 77030, USA

We have previously demonstrated decreased Jun/AP-1 activity in the breast cancer cell line MCF-7 when compared to normal or immortalized mammary epithelial cells. In this paper, we overexpress Jun in MCF-7 cells (MCF7Jun) and demonstrate that it results in diverse biologic and biochemical changes, which mimic those seen clinically in breast cancer. Overexpression of Jun causes significant alterations in the composition of AP-1, decreased junB and increased fra-1 expression and results in an increased biologic aggressiveness. MCF7Jun cells exhibit increased cellular motility, increased expression of a matrix degrading enzyme MMP-9, increased in vitro chemoinvasion and tumor formation in nude mice in the absence of exogenous estrogens. Furthermore, MCF7Jun cells are unresponsive to the growth stimulating effects of estrogen and growth inhibitory effects of tamoxifen. Analysis of the estrogen receptor (ER) expression and activity showed that the MCF7Jun cells have no detectable ER. MCF-7 cells overexpressing mutant forms of cJun were responsive to the growth stimulatory effects of estrogen indicating that full-length cJun is required to acquire the estrogen-independent phenotype in breast cancer cells.

cJun; AP-1; breast cancer; tumorigenicity; invasion; hormone resistance

The proto-oncogene c-jun encodes a 39 kDa nuclear phosphoprotein, cJun, which is a component of the mammalian transcription factor activator protein-1 (AP-1) (Bohmann et al., 1987; Angel et al., 1988). The AP-1 complex was first isolated as a factor that binds DNA sequences found in the promoter regions of genes stimulated by phorbol esters such as 12-O-tetradecanoylphorbol-13-acetate (TPA) (Angel et al., 1987). AP-1 complexes are formed by dimers of the Jun family members (cJun, JunB, JunD) or heterodimers of the Jun family members with the Fos family members (cFos, FosB, Fra-1, Fra-2). cJun is also known to interact with other transcription factors. cJun interacts with the NF-B subunit p65 to enhance binding to B and AP-1 sites (Stein et al., 1993). Jun dimers and Jun/Fos heterodimers also form complexes with the T-cell transcription factor NFAT (Jain et al., 1993). The AP-1 and ATF/CREB transcription factors can form selective cross-family heterodimers with distinct DNA binding specificities. Cross-talk between AP-1 and several steroid receptors such as estrogen receptor (ER), glucocorticoid and retinoic acid receptors can either enhance or repress AP-1 activity through direct or indirect protein - protein interactions (Yang-Yen et al., 1990; Uht et al., 1997; Vig et al., 1994). In earlier studies, TPA or cJun and cFos were found to inhibit estrogen-dependent estrogen receptor transcriptional activity in a variety of cell lines but the mechanism involved may be complex (Tzukerman et al., 1991). Endogenous ER mRNA was found to be rapidly down regulated in MCF7 breast cancer cells in response to TPA treatment (Tzukerman et al., 1991). Transiently transfected c-jun or c-fos was also found to inhibit estrogen receptor transcriptional activity in human breast cancer cells (Doucas et al., 1991).

The role of cJun in cellular transformation has been defined previously in rodent and avian model systems (Vogt et al., 1990; Suzuki et al., 1991; Castellazzi et al., 1990, Dong et al., 1994). Deregulated expression of cJun can lead to malignant transformation of immortalized rat fibroblasts while transformation of primary rat embryo cells required co-expression of an activated c-Ha-ras (Alani et al., 1991; Schutte et al., 1989). In chicken embryo fibroblasts, cJun can induce cellular transformation by itself (Suzuki et al., 1991; Castellazzi et al., 1990). Stable expression of a trans-activation suppressing deletion mutant of c-jun in malignant mouse epidermal cell lines inhibited tumor formation in nude mice (Domann et al., 1994). In human cancer, the role of c-jun/AP-1 is less clear. Increased constitutive levels of c-jun and c-fos mRNA and AP-1 levels have been reported for drug-resistant cells (such as etoposide resistant human leukemia K562 cells) compared to drug-sensitive parental lines (Ritke et al., 1994). In the human breast adenocarcinoma cell line MCF-7, mitogenic stimulation by insulin or other insulin-like growth factors (IGFs) leads to increased c-jun or c-fos expression and AP-1 activity. In a study of non-small cell lung primary and metastatic tumors, cJun is found to be overexpressed in 31% of the cases (Szabo et al., 1996). Histologically atypical areas from the surrounding lung were also found to express cJun while normal airway and alveolar epithelia did not have detectable expression, suggesting that cJun expression is altered early during human lung carcinogenesis (Szabo et al., 1996).

We have reported previously the variable levels of expression of jun and fos family members and AP-1 DNA binding and transactivating activity in a panel of human breast cancer cell lines (Chen et al., 1996). In a following study using normal proliferating breast cells and breast cell lines at different stages of transformation, we observed that AP-1 expression and activity decreases as the breast epithelial cell becomes more transformed (Smith et al., 1997). Our studies suggest that transformed cells may become less dependent on AP-1-mediated signal transduction pathways for continued growth which is reflected in their lower AP-1 expression and transactivating activity compared to normal proliferating breast epithelial cells. High AP-1 activity in breast cancer cells may also be inhibitory to their growth since TPA, a potent activator of AP-1, has been shown to inhibit the growth of breast cancer cells (Darbon et al., 1986).

The aim of this study was to determine the effect of overexpression of cJun in a human breast cancer cell line. We co-transfected MCF-7 cells with a eukaryotic expression vector containing the human c-jun gene and a plasmid containing the neomycin (neo) resistance gene. G418-resistant clones were isolated and characterized. Western blot analysis demonstrated variable levels of cJun expression over the low level background in multiple clones (Yang et al., 1997). Three clones expressing high levels of cJun (MCF7Jun) (clones 2-16, 2-31 and 2-33) were randomly chosen along with three neomycin resistant control clones (clone 7-1, 7-2 and 7-3) (Figure 1a). We analysed their AP-1 DNA binding activity by a gel retardation assay using a labeled oligonucleotide containing a consensus AP-1 sequence and found that all MCF7Jun clones demonstrated high levels of AP-1 DNA binding activity (data not shown). Control neo clones as well as the parental cells (WT) have minimal levels of AP-1 DNA binding (Chen et al., 1996).

To demonstrate the transactivation of AP-1 regulated genes, we analysed the expression of vimentin, a known AP-1 target (Angel and Karin, 1991). Northern blot analysis demonstrated significant up-regulation of expression of vimentin mRNA in MCF7Jun clones (Figure 1b). Transcriptional transactivation assay using a collagenase AP-1 reporter construct has also shown increased AP-1 activity for MCF7Jun clones (data not shown) (Yang et al., 1997). To determine the effect of cJun overexpression on the expression level of the jun and fos gene family members, Northern blot analysis was performed. As shown in Figure 1b, overexpression of cJun altered the expression of several members of these gene families. junB mRNA expression decreased fivefold while junD and c-fos decreased slightly from their already minimal levels. fra-1 expression was undetectable in MCF-7 vector control cells but became easily detectable in MCF7Jun clones.

The biologic consequences of cJun over-expression were evident in changes in cell morphology in vitro. MCF7Jun cells are much larger than control neo cells with a larger cytoplasmic to nuclear ratio (Figure 1c). In addition, all MCF7Jun clones grew in a less compact fashion than the neomycin resistant control cells with multiple floating cells evident in the culture. We did not detect increased apoptosis in either the adherent or non-adherent MCF7Jun cells (data not shown).

MCF-7 cells have minimal ability to invade through an artificial basement membrane (matrigel) in vitro (Holst-Hansen et al., 1996; Lambert et al., 1997) whereas, treatment with estrogen increased this invasive capacity (Clarke et al., 1989). Analysis of the chemotactic ability of MCF7Jun clone 2-33 demonstrated that overexpression of cJun results in increased motility of these cells (Table 1). Clone 2-33 exhibited a 5 - 30-fold increased motility compared to control clone 7-1 in response to undiluted NIH3T3 conditioned media. Due to the strong motility observed for 2-33, we compared invasive and chemotactic capacities of 2-33 and 7-1 using 1 : 20 dilution of the NIH3T3 conditioned media as chemoattractant and 0.2 mg/ml matrigel for coating the membrane. The chemomigratory response of clone 2 - 33 cells to the diluted NIH3T3 conditioned media was reduced to 2.8-fold compared to clone 7-1. However, under these conditions, clone 2-33 was 25-fold more invasive than clone 7-1. In an attempt to determine the mechanisms of this increased invasion, we analysed supernatants from the transfected clones for the presence of matrix degrading enzymes. Zymogram analysis showed an active enzyme of 92 kDa only in the MCF7Jun cells (Figure 2a). Western analysis of cellular extracts from the different clones revealed this enzyme to be type IV collagenase (MMP-9), a known AP-1 regulated gene (Figure 2b).

MCF-7 cells are non-tumorigenic in nude mice unless exogenous estrogen is provided to the mice (Clarke et al., 1990). We tested the cJun overexpressing clones for tumor formation in normal and ovariectomized nude mice by injecting 1´10⁷ cells subcutaneously into each mouse. As shown in Table 2, none of the control cell lines or parental MCF-7 cells gave rise to tumors after 3 months, while 15/18 and 13/18 mice injected with MCF7Jun cells produced tumors with or without ovariectomy respectively (Student's t-test, 2-tailed: P<0.01). In addition, all three MCF7Jun clones produced tumors in a rapid fashion (>100 mm³ by 6 weeks).

MCF-7 WT cells express the estrogen receptor (ER) and proliferate in response to estrogen while their growth is inhibited by the estrogen antagonist tamoxifen in vitro (Bignon et al., 1989). We observed that the MCF7Jun cells grew slower than their control clones in complete media. Determination of their doubling time in culture showed that the average doubling times of MCF-7 WT and neo clones is 36.9±2.3 h while MCF7Jun clones is 44.8±3.1 h (P<0.01) (Table 3). Since normal growth media has estrogen (from serum) and estrogen-like substances (phenol red), we tested our clones in estrogen-depleted and estrogen-supplemented media. Upon removal of estrogen from the media, MCF7 parental and control neo clones slowed their growth by 3.6-fold (mean doubling time of 131.7±15 h) (P<0.01) while the doubling time of MCF7Jun clones only increased by 1.5-fold compared to their growth in regular media (66.2±8.8 h) (P<0.01). This result indicates that cJun overexpression may partially replace the effect of estrogen stimulation observed in the neo controls. While addition of estrogen stimulated the growth of the parental and control cells by an average of 2.5-fold, the growth of MCF7Jun cells did not change upon estrogen supplementation. The average doubling times of the control cells in estrogen supplemented media is 2.5-fold less compared to estrogen depleted media while MCF7Jun clones have similar doubling times with or without estrogen (Table 3).

Furthermore, the effect of tamoxifen alone (10 nM) or estrogen (1 nM) and tamoxifen (10 nM) together on cell growth of all the clones was determined. As shown in Figure 3, tamoxifen inhibits the growth of the control neo clones. The growth observed in the control cells upon addition of estrogen is diminished by co-treatment with tamoxifen. There was no observed growth inhibition by tamoxifen alone or tamoxifen plus estrogen for MCF7Jun clones. These results show that MCF7Jun cells are estrogen-independent for growth in vitro and have also become insensitive to inhibition by tamoxifen.

Due to the observed estrogen-unresponsiveness of the MCF7Jun clones, the expression level of estrogen receptor activity was determined. Using a radioligand binding assay to quantitate ER protein, results show that while the control clones have ER protein levels ranging from 128 to 316 fmol ER/mg total protein, all three MCF7Jun clones do not express ER protein (Table 4). Using a more sensitive assay, ER transcriptional activity was analysed by transfection of a reporter plasmid containing an estrogen-responsive element (ERE) into all of the cell lines. The transfected cells were either untreated or treated with estrogen (100 nM) in the media. While the parental and control cells have estrogen-inducible ER transcriptional activity (up to threefold), there was no detectable ER transcriptional activity in the MCF7Jun cell lines (Figure 4a). To determine the mechanism of down-regulation of ER, we analysed ER mRNA expression. Northern blot analysis of polyA RNA showed that ER mRNA was detected only in the control neo cell line (7-1) and not in the MCF7Jun clone 2-33 (Figure 4b). This suggests that the down-regulation of ER is at the mRNA level due to transcriptional regulation or post-transcriptional processing of the ER mRNA.

To study the mechanism by which cJun overexpression produces an estrogen-independent phenotype, we transfected MCF7 cells with deletion mutants of cJun and established stable clones. Cell lines 1 - 3 and 1 - 5 express cJun 1-286 [Jun 287-331: leucine zipper domain (LZD) mutant], 2-2 and 2-3 express Jun A-D²⁶⁵ [(DNA binding domain mutant (DBD)], and Tam 6 and Tam 9 express cJun Tam67 [Jun 3-122: transactivation domain (TAD) mutant] (data not shown) (Brown et al., 1996). The TAD mutant is able to form homodimers with full-length cJun and heterodimers with cFos. Both DBD mutant and LZD mutant are unable to bind DNA although the DBD mutant is able to inhibit Jun/Jun homodimers from binding DNA while the LZD mutant is unable to specifically inhibit cJun dimerization and binding to DNA (Brown et al., 1996). Analysis of the growth of these cJun clones with or without estrogen in their media showed estrogen induced proliferation in all the control and cJun mutant cell lines by an average of 2.61±0.46-fold compared to MCF7Jun clone 2-33 (0.96±0.05) (Figure 5). Only the clone expressing the full-length cJun (2-33) did not respond to estrogen. Statistical analysis using Student's t-test showed that the difference in growth for all the stable clones studied were significant (P<0.05) compared to MCF7Jun clone 2-33. These data suggests that expression of full length cJun is required to acquire estrogen-unresponsiveness in breast cancer cells.

In this study, we demonstrate that increased expression of cJun results in: (1) altered expression of other AP-1 complex genes and increased AP-1 activity; (2) changes in cellular morphology, increased expression of matrix degrading enzymes, and increased motility and invasiveness of the cells in vitro; (3) tumor formation in nude mice in the absence of exogenous estrogen; and (4) unresponsiveness to both estrogen and tamoxifen in cell culture.

The increased cJun expression in MCF7Jun cells results in an enrichment of the AP-1 complex with cJun resulting in increased AP-1 transcriptional activity. This increase in turn alters the expression of other AP-1 members which are known to be AP-1 regulated genes (junB and fra-1) (Deng and Karin, 1993; Yoshioka et al., 1995). We demonstrate that junB expression decreases fivefold. In most cells examined, JunB antagonizes cJun transcriptional activity and as such any decrease in its level would enhance cJun activity resulting in a further increase in AP-1 transcriptional activity (Deng and Karin, 1993). fra-1, another AP-1 regulated gene, is upregulated in the MCF7Jun cells. This increase in fra-1 may be compensatory since Fra-1 is frequently inhibitory to cJun activity (Yoshioka et al., 1995). The net result of these effects is quantitative and qualitative changes in the AP-1 complex. The quantitative change can be seen by the DNA binding assay (increase in binding activity), increase in AP-1 transcriptional activity using an AP-1 luciferase construct (data not shown) and upregulation of two well-characterized `downstream' AP-1 regulated genes, vimentin and type IV collagenase. Similar changes in the AP-1 complex have been seen in immortalized mammary epithelial cells expressing a cJun-estrogen receptor (ER) protein although no upregulation of collagenase genes was noted (Fialka et al., 1996). The qualitative changes in the AP-1 complex (changes in protein composition) may produce alterations in DNA binding specificity and affinity. These changes can be determined only when `downstream' target genes and their regulatory elements are identified.

These changes in the AP-1 complex result in a number of important biochemical and biologic changes. The morphological changes observed in vitro, in MCF7Jun cells (loose aggregates of cells, floating cells) are consistent with alterations in cell surface proteins that are required for tight cell-to-cell contact. In a study by Fialka et al. (1996), expression of c-JunER protein in mammary epithelial cells induced a reversible dissociation of tight cell contacts and loss of epithelial polarity due to destabilization of adherens junctions. The phenotype of MCF7Jun cells is similar confirming that the alterations in cellular morphology is a direct effect of cJun and not contributed by the ER (HBD). MCF7Jun cells have an increased ability to invade through a matrigel gel coated membrane in vitro. This increased invasive capacity of the MCF7Jun cells correlates with the increased expression of matrix degrading enzymes as seen on a zymogram gel. Western blot analysis demonstrated that this activity resulted from the increased expression of type IV collagenase. Multiple studies have shown that increased expression of matrix metalloproteinases is associated with invasiveness of tumor cells due to degradation of basement membrane (Sato and Seiki, 1993; Thompson et al., 1993). While we can not eliminate the posibility that the increased invasiveness demonstrated by MCF7Jun cells is not the result of some other metalloproteinase, it would appear highly likely that type IV collagenase mediates this invasiveness in our study. This is the only activity seen on the zymogram gel and type IV collagenase is known to be an AP-1 regulated gene (Sato and Seiki, 1993). A variant MCF-7 cell line selected for tamoxifen resistance has been reported as displaying increased AP-1 activity and increased collagenase expression (Dumont et al., 1996).

Increased expression of cJun in MCF-7 cells produces different effects upon cell growth depending upon the culture conditions. In complete media (phenol red containing, serum supplemented), cJun decreased cell growth. This decrease results from a loss of estrogen receptor expression (and its corresponding transcriptional activity) since these cJun inhibitory effects are observed only in the presence of estrogen (E2). In estrogen-free media (phenol red free media with charcoal stripped serum), cJun provides a growth stimulus similar to but quantitatively different from that seen with estrogen. Normally, estrogen binds to its receptor and the E2-ER complex binds to response elements in target genes leading to transcription and expression of specific genes, some of which are important for the growth of the cells. By overexpressing cJun, increased AP-1 activity may increase expression of the same or a subset of the estrogen-regulated target genes required by the cell for growth in estrogen. It has been proposed that the ER can interact with the AP-1 complex and activate AP-1 regulated genes (Uht et al., 1997). Thus, the requirement for ER activity can be partially bypassed. Conversely, cJun growth stimulation could result from the regulation of an independent group of `downstream' genes that substitute for the ER regulated genes. cJun is known to regulate a large number of genes, some of which stimulate cellular growth (Angel and Karin, 1991). Finally, cJun overexpression could have induced secondary changes that occurred during selection, which produced growth advantages to the clones.

The effects of cJun overexpression on the tumorigenicity of MCF-7 cells were rather surprising. It is well established that MCF-7 cells will form tumors in nude mice only under supra-physiologic levels of (exogenous derived) estrogen (Clarke et al., 1989Clarke et al., 1989). However, all MCF7Jun cell lines formed large tumors in a rapid fashion in nude mice. To ensure that cJun is not amplifying an estrogen derived signal from the physiologic levels of estrogen present in these mice, we tested the MCF7Jun cells in ovariectomized nude mice. Even in these animals, the MCF7Jun cells produced tumors of similar size over approximately the same time course. To our knowledge, this is the first reported example of cJun overexpression inducing human cancer cells to form tumors in nude mice. c-jun has been demonstrated to transform primary avian or rodent cells as a single gene or in cooperation with an activated ras gene (Alani et al., 1991; Schutte et al., 1989). The mechanism by which cJun transforms cells remains unclear but presumably involves the regulation of gene expression. Whether this is a direct effect of cJun or in part due to cJun-induced secondary alterations in these clones remain to be determined. The requirement of all portions of the cJun protein (transactivation domain, DNA binding domain, dimerization domain) to inhibit estrogen responsiveness suggests that deregulated expression (activation or repression) of `downstream' genes must be critical to this process.

In this study, increased cJun/AP-1 activity correlates with decreased ER activity. This relationship has been observed before and is proposed to result from several possible mechanisms. Inhibition of estrogen receptor activity by c-jun and c-fos is well documented and thought to be the result of direct interaction of these transcription factors and the estrogen receptor (Doucas et al., 1991). This inhibitory activity mapped to the glycine rich region (between amino acid 147 - 220) within the cJun protein (Doucas et al., 1991). This `cross talk' at the level of proteins is similar to that described for the inhibition of glucocorticoid induced transcription by AP-1 (Vig et al., 1994). This inhibition is reciprocal, does not require DNA binding, and requires protein - protein interaction (Vig et al., 1994). Additionally, inhibition of ER activity by AP-1 has been proposed to result from the `squelching out' of common cofactors used by the estrogen receptor and components of the AP-1 complex (Tzukerman et al., 1991; Doucas et al., 1991). We demonstrate in this paper a novel mechanism of `cross-talk' between the estrogen receptor and cJun. Overexpression of cJun downregulates the ER gene expression presumably at the transcriptional level. All three components of cJun (transactivation domain, DNA binding domain, and dimerization domain) are required for this effect, strongly arguing that the downregulation of the ER occurs due to transcriptional control. This effect on the ER expression by cJun may be an additional mechanism by which TPA and/or PKC downregulates the ER and inhibits the growth of breast cancer cells (Darbon et al., 1986; Dickson, 1995; Saceda et al., 1991). The precise mechanism by which cJun accomplishes this remains to be delineated. The action of cJun could be through a direct interaction with cis elements within the ER promoter. An AP-1 element has recently been identified, but was characterized as a positive enhancing element (Tang et al., 1997). Conversely, cJun could be acting through an intermediate gene product or interference with other positively acting transcription factors. Recently, a transcription factor ERF-1 found to be involved in the regulation of ER gene transcription has been identified as a member of the AP2 family (McPherson et al., 1997).

The observed MCF7Jun cells phenotype is similar to that observed clinically in many patients with breast cancer. Many patients with advanced breast cancer initially responded to hormonal therapy (tamoxifen) but eventually recur with tumors that are now hormone unresponsive (Clarke et al., 1990; Leonessa et al., 1991). In addition, these tumors tend to be more aggressive, invasive, and chemoresistant. The mechanism(s) involved remain unclear but the results presented here suggest that upregulation of the transcriptional complex AP-1 may be important. In a recent study, increased AP-1 DNA binding activity was observed in tamoxifen-resistant primary human breast tumors (Johnston et al., 1999). In vivo selection of tamoxifen resistant breast cancer cells in nude mice has been shown to give rise to cells with high levels of AP-1 transcriptional activity and diminished ER expression (Clarke et al., 1989Clarke et al., 1989; Dumont et al., 1996). Thus, alterations of AP-1 complex members may contribute to the phenotype of relapsed breast cancer. We propose that increases in cJun/AP-1 activity may be important as a progression event in breast carcinogenesis.

Acknowledgements

LM Smith and SC Wise are co-first authors for this study. We thank Dr M Karin and Dr P Chambon for their generous gifts of plasmids. We also thank Amy Guzzone for her technical expertise.

Alani R, Brown P, Binetruy B, Dosaka H, Rosenberg R, Angel P and Karin M. (1991). Mol. Cell. Biol. 11, 6286-6295. MEDLINE

Angel P, Allegretto EA, Okino ST, Hattori K, Boyle WJ, Hunter T and Karin M. (1988). Nature 332, 166-171. MEDLINE

Angel P, Imagawa M, Chiu RJ, Stein B, Inmbra RJ, Rahmsdorf HJ, Jonat C, Herrlich P and Karin M. (1987). Cell 49, 729-739. MEDLINE

Angel P and Karin M. (1991). Biochim. Biophys. Acta 1072, 129-157. MEDLINE

Bignon E, Pons M, Crastes de Paulet A, Dore JC, Gilbert J, Miquel JF, Pons M and Raynaud JP. (1989). J. Med. Chem. 32, 2092-2103. MEDLINE

Bohmann D, Bos TJ, Admon A, Nishimura T, Vogt PK and Tjian R. (1987). Science 238, 1386-1392. MEDLINE

Brown PH, Kim SH, Wise SC, Sabichi AL and Birrer MJ. (1996). Cell Growth Differ. 7, 1013-1021. MEDLINE

Castellazzi M, Dangy J-P, Mechta F, Hirai S-I, Yaniv M, Samarut J, Lassailly A and Brun G. (1990). Oncogene 5, 1541-1547. MEDLINE

Chen TK, Smith LM, Gebhardt DK, Birrer MJ and Brown PH. (1996). Mol. Carcinogen. 15, 215-226.

Clarke R, Brunner N, Katzenellenbogen BS, Thompson EW, Norman MJ, Koppi C, Paik S, Lippman ME and Dickson RB. (1989). Proc. Natl. Acad. Sci. USA 86, 3649-3653. MEDLINE

Clarke R, Brunner N, Thompson EW, Glanz P, Katz D, Dickson RB and Lippman ME. (1989). J. Endocrinol. 122, 331-340. MEDLINE

Clarke R, Dickson RB and Brunner N. (1990). Ann. Oncol. 1, 401-407. MEDLINE

Darbon J, Valette A and Bayard F. (1986). Biochem. Pharmacol. 35 (16), 2683-2686. MEDLINE

Deng T and Karin M. (1993). Genes Dev. 7, 479-490. MEDLINE

Dickson R. (1995). Molecular Biology in Cancer Medicine. R Kurzrock and M Talpaz (eds). Martin Dunitz Ltd: London, pp. 241-272.

Domann FE, Levy JP, Birrer MJ and Bowden GT. (1994). Cell Growth Differ. 5, 9-16. MEDLINE

Dong Z, Birrer MJ, Watts RG, Matrisian LM and Colburn NH. (1994). Proc. Natl. Acad. Sci. USA 91, 609-613. MEDLINE

Doucas V, Spyrou G and Yaniv M. (1991). EMBO J. 10 (8), 2237-2245. MEDLINE

Dumont JA, Bitonti AJ, Wallace CD, Baumann RJ, Cashman EA and Cross-Doersen DE. (1996). Cell Growth Differ. 7, 351-359. MEDLINE

Fialka I, Scharwz H, Reichmann E, Oft M, Busslinger M and Beug H. (1996). J. Cell Biol. 132, 1115-1132. MEDLINE

Holst-Hansen C, Johannessen B, Hoyer-Hansen G, Romer J, Ellis V and Brunner N. (1996). Clin. Exp. Metastasis 14, 297-307. MEDLINE

Jain J, McCaffrey PG, Miner Z, Kerppola TK, Lambert J, Verdine GL, Curran T and Rao A. (1993). Nature 365, 352-355. MEDLINE

Johnston SR, Lu B, Scott GK, Kushner PJ, Dowsett M and Benz CC. (1999). Clin. Cancer Res. 5, 251-256. MEDLINE

Lambert PA, Somers KD, Kohn EC and Perry RR. (1997). Surgery 122, 372-379. MEDLINE

Leonessa F, Boulay V, Wright A, Thompson EW, Brunner N and Clarke R. (1991). Acta Oncol. 31, 115-123.

McPherson LA, Baichwal VR and Weigel RJ. (1997). Proc. Natl. Acad. Sci. USA 94, 4342-4347. MEDLINE

Ritke MK, Bergoltz VV, Allan WP and Yalowich JC. (1994). Biochem. Pharmacol. 48, 525-533. MEDLINE

Saceda M, Knabbe C, Dickson RB, Lipmann ME, Bronzert D, Lindsey RK, Gottardis MM and Martin MB. (1991). J. Biol. Chem. 266, 17809-17814. MEDLINE

Sato H and Seiki M. (1993). Oncogene 8, 395-405. MEDLINE

Schutte J, Minna JD and Birrer MJ. (1989). Proc. Natl. Acad. Sci. USA 86, 2257-2261. MEDLINE

Smith LM, Birrer MJ, Stampfer MR and Brown PH. (1997). Cancer Res. 57, 3046-3054. MEDLINE

Stanier RY, Adelberg EA and Ingraham JL. (1976). The Microbial World. Prentice Hall, Inc.: New Jersey, pp. 276-277.

Stein B, Baldwin Jr AS, Ballard DW, Greene WC, Angel P and Herrlich P. (1993). EMBO J. 12 (10), 3879-3891. MEDLINE

Suzuki T, Hashimoto Y, Sato H, Nishina H and Iba H. (1991). Jpn. J. Cancer Res. 82, 58-64. MEDLINE

Szabo E, Riffe ME, Steinberg SM, Birrer MJ and Linnoila RI. (1996). Cancer Res. 56, 305-315. MEDLINE

Tang Z, Treilleux I and Brown M. (1997). Mol. Cell. Biol. 17 (3), 1274-1280. MEDLINE

Tzukerman M, Zhang X-K and Pfahl M. (1991). Mol. Endocrinol. 91, 1983-1992.

Thompson EW, Brunner N, Torri J, Johnson MD, Boulay V, Wright A, Lippman ME, Steeg PS and Clarke R. (1993). Clin. Exp. Metastasis 11, 15-26. MEDLINE

Uht RM, Anderson CM, Webb P and Kushner PJ. (1997). Endocrinol. 138 (7), 2900-2908.

Vig E, Barrett TJ and Vedeckis V. (1994). Mol. Endocrinol. 8, 1336-1346. MEDLINE

Vogt PK and Bos TJ. (1990). Adv. Cancer Res. 55, 1-35. MEDLINE

Yang L, Kim H, Munoz-Medellin D, Reddy P and Brown PH. (1997). Cancer Res. 57, 4652-4661. MEDLINE

Yang-Yen HF, Chambard JC, Sun YL, Smeal T, Schmidt TJ, Drouin J and Karin M. (1990). Cell 62, 1205-1215. MEDLINE

Yoshioka K, Deng T, Cavigelli M and Karin M. (1995). Proc. Natl. Acad. Sci. USA 92, 4972-4976. MEDLINE

Figure 1 (a) Western blot analysis of cJun expression in MCF-7 stable transfectants. Equal amounts of protein (as determined by the Bio-Rad colorimetric assay) from nuclear extracts of exponentially growing cells were used and cJun protein was detected using rabbit polyclonal anti-cJun antibody (Ab-1 from Oncogene Science). MCF7 WT is the parental MCF-7 cell line. (b) Gene expression in MCF-7 cells transfected with c-jun expression vector or neomycin vector control. A Northern blot with RNA from control neo and MCF7Jun clones were serially hybridized to probes for jun and fos gene family members and an AP-1 regulated gene, vimentin. Equal loading was determined by ethidium bromide staining of 28S and 18S ribosomal RNA. (c) Effect of cJun overexpression on cell morphology. The morphology of MCF7 stable clones grown in culture is shown at 200´magnification

Figure 2 (a) Enhanced protease expression of MCF7Jun cells. Parental and stable transfectants of MCF-7 cells were grown in serum free media for 48 h. Media was harvested, standardized to equivalent cell number and loaded onto a Zymogram gel (Novex, San Diego, CA, USA). The gel was stained with Coomasie and clear areas indicate the location of protease activity (arrow). (b) Western blot analysis of type IV collagenase/MMP-9 expression in MCF7Jun cells. Total cellular extracts from parental and stable transfectants of MCF-7 cells were electrophoresed, transfered to nitrocellulose membrane and incubated with an anti-MMP9 rabbit polyclonal antibody. The 92 kDa band (arrow) was detected using enhanced chemiluminescence (Amersham)

Figure 3 Effects of estrogen and tamoxifen on MCF7Jun and control cells. Cells were seeded in 96 well plates at a concentration of 4.3´10³ cells/well in four different treatments (vehicle, 1 nM estrogen, 10 nM tamoxifen, 1 nM estrogen+10 nM tamoxifen) in phenol-red free media with 10% charcoal stripped fetal calf serum. Cell density was quantitated for 7 - 10 days by MTT assay

Figure 4 (a) Estrogen receptor transcriptional activity in MCF7Jun and control cells. Cell lines were transiently transfected with an ERE-CAT construct, divided into two cultures and treated with either 100 nM estrogen or vehicle. Cells were harvested 4 days after treatment and CAT activity was quantitated. The cells were co-transfected using calcium-phosphate transfection with 10 g reporter plasmid containing the CAT gene linked to a promoter which contains an estrogen responsive element (ERE) and 5 g of CMV-Gal (Clontech) DNA. The cells were harvested, lysed and transfection efficiency was determined by measuring -galactosidase activity in cell lysates. After normalizing for transfection efficiency, CAT activity was measured using thin layer chromatography and quantitated using the PhosphorImager (Molecular Dynamics). (b) Analysis of estrogen receptor RNA in control and c-jun transfected MCF-7 cell line. Northern blot analysis using poly(A)⁺ RNA from control (7-1) and cJun overexpressor (2-33) was performed using ³²P-labeled ER cDNA as probe. Ribosomal RNA (28S and 18S) were used as molecular weight markers

Figure 5 Estrogen-induced cellular proliferation of MCF-7 stable transfectants expressing cJun mutants. Cells were grown in 96 well plates in complete media or estrogen depleted media for 5-6 days. Results are expressed as fold increase of growth in complete media over growth in estrogen depleted media. The cell lines are as follows: 7-1, 1-2, and 4-1 are controls; 2-33, cJun overexpressor; 1-3 and 1-5 express cJun 1-286 (Jun 287-331: leucine zipper mutant); 2-2 and 2-3 express JunA-D²⁶⁵ (DNA binding mutant); Tam6 and Tam9 express cJun Tam67 (transactivation domain mutant). There was a statistically significant difference between clone 2-33 and all other clones using Student's t-test (P<0.05)

 Chemotactic ability of MCF-7 clones

 Nude mouse tumorigenicity of MCF7Jun clones

 Effect of estrogen on doubling time of MCF-7 stable clones

 Radioligand binding assay for estrogen receptor (ER)