C-Jun Has Multiple Enhancing Activities in the Novel Cross Talk Between the Androgen Receptor and Ets Variant Gene 1In Prostate Cancer

c-Jun Has Multiple Enhancing Activities in the Novel Cross Talk between the Androgen Receptor and Ets Variant Gene 1in Prostate Cancer

Changmeng Cai, Chen-Lin Hsieh, and Lirim Shemshedini

Department of Biological Sciences, University of Toledo, Toledo, Ohio

Abstract matrix (reviewed in ref. 3). Through this transactivation process, The multiple transcriptional roles of c-Jun are shown in a AP-1 is able to directly induce the expression of a wide variety of novel cross-talk between the androgen receptor (AR) genes, many of which mediate cell cycle progression (reviewed and its new target gene, Ets variant gene 1(ETV1).In this in ref. 2). A second transcriptional function of c-Jun is report, we show that c-Jun can mediate AR induction of transrepression, a process by which c-Jun is able to inter- ETV1expression independent of c-Jun transactivation fere with the activities of other transcription factors, including function. Interestingly, c-Jun can transactivate the nuclear receptors (reviewed in ref. 4). Coactivation is a third cloned ETV1promoter also in the absence of ligand- major function for c-Jun that, like transactivation, positively activated AR, suggesting two mechanisms by which regulates transcription. However, unlike transactivation, coac- c-Jun can induce ETV1expression. In addition, both tivation does not depend on dimerization with Fos family wild-type c-Jun and a transactivation-deficient mutant members (5) or direct DNA binding (6), and has been best can enhance the transcriptional activity of ETV1, as characterized on the androgen receptor (AR), a nuclear receptor measured by both reporter gene assay and endogenous that mediates the biological activities of androgens (7). expression of matrix metalloproteinase genes, well- Among the multiple physiologic functions of the AR, its known targets of Ets proteins. Overexpression of the important role in prostate cancer has been a focus of both basic c-Jun mutant protein also led to increased prostate and clinical research. The development of prostate cancer, just cancer cell invasion. Immunoprecipitation and like the normal prostate gland, is dependent on androgen immunocytochemistry experiments showed signaling through the AR. In addition, prostate cancer copurification and colocalization of c-Jun with AR or progression is dependent on androgens and AR (reviewed in ETV1, suggesting that c-Jun acts on AR or ETV1 via ref. 8). Indeed, it was recently shown that progression to a physical association. Collectively, these results, hormone-refractory prostate cancer, a lethal form of the disease, together with a parallel overexpression of ETV1, is induced by the overexpression only of AR (9), a genetic c-Jun, and AR in prostate tumors, imply that c-Jun change that is often seen in late-stage prostate cancer. plays a pivotal role in the pathway that connects Interestingly, AR transcriptional functions remain important ligand-activated AR to elevated ETV1expression, as the cancer cells transition from a hormone-dependent to a leading to enhanced expression of matrix hormone-independent stage, and the AR adapts to be activated metalloproteinases and prostate cancer cell invasion. by mechanisms that do not depend on androgens (8). (Mol Cancer Res 2007;5(7):725–35) AR has several supportive functions in prostate cancer, including inhibiting apoptosis and inducing cell proliferation and invasiveness (8). Several AR-regulated genes have been Introduction identified that mediate the antiapoptotic and proproliferative Originally identified as the oncogenic form v-Jun that causes effects of this nuclear receptor (10). We have recently identified sarcomas in chickens (1), c-Jun represents a protein with multiple Ets variant gene 1 (ETV1) as a novel androgen-regulated gene transcriptional functions. The best studied and understood that mediates prostate cancer cell invasion.1 ETV1 is a member function of c-Jun depends on its ability to heterodimerize with of the Ets family of transcription factors (reviewed in ref. 11). c-Fos to form the activator protein-1 (AP-1) transcription factor Ets proteins have been reported to be involved in intestinal (reviewed in ref. 2). AP-1 represents a transcription factor tumors (12), gastric cancer (13), and breast cancer metastasis complex that is activated by a broad range of external cell stimuli, (14, 15). Most interestingly, ETV1 and its Ets relative ERG including growth factors, chemokines, and the extracellular were recently identified to be targets for translocation with TMPRSS2, an androgen-regulated gene, in a subset of prostate cancers (16). Our previous data show that ETV1 androgen Received 12/20/06; revised 4/6/07; accepted 4/26/07. regulation or overexpression in prostate tumors does not depend Grant support: NIH and Ohio Cancer Research Associates. 1 The costs of publication of this article were defrayed in part by the payment of on the TMPRSS2 translocation, suggesting that the intact page charges. This article must therefore be hereby marked advertisement in ETV1 plays an important role in prostate cancer. accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Requests for reprints: Lirim Shemshedini, Department of Biological Sciences, University of Toledo, Toledo, OH 43606. Phone: 419-530-1553; Fax: 419-530- 7737. E-mail: [email protected] 1 C. Cai, C-L. Hsieh, J. Omwancha, Z. Zheng, S-Y. Chen, L-L. Baert, L. Copyright D 2007 American Association for Cancer Research. Shemshedini. ETV1 is a novel androgen-regulated gene that mediates prostate doi:10.1158/1541-7786.MCR-06-0430 cancer cell invasion. Mol Endocrinol. In press.

In this report, we show that AR induction of ETV1 gene gene expression (19). We have recently identified ETV1 as a expression in prostate cancer cells is mediated by c-Jun novel androgen-regulated gene.1 To determine if changes in coactivation. Interestingly, the c-Jun enhancement activity was endogenous c-Jun expression levels could affect androgen also observed on ETV1 transcriptional activity, as measured by induction of ETV1 expression, LNCaP stable cell lines were both reporter gene assay and endogenous gene expression, used that express either antisense c-Jun (AJ81 cells) or leading to elevated prostate cancer cell invasion. Using c-Jun(Ala63/73) (M37 cells; ref. 19). We have previously shown immunoprecipitation and immunocytochemistry studies, we that AJ81 cells have greatly reduced endogenous c-Jun can detect a physical interaction and a nuclear colocalization expression and androgen-dependent proliferation, whereas between endogenous c-Jun and AR or ETV1 in prostate cancer M37 cells express c-Jun(Ala63/73), a mutant of c-Jun that is cells. Lastly, ETV1 and AR overexpression in prostate tumors is fully active in AR coactivation but deficient in AP-1 trans- paralleled by overexpression of c-Jun, suggesting coordinated activation, and display significantly elevated proliferation (19). expression patterns among these three proteins in prostate cancer. As shown in Fig. 1A, dihydrotestosterone (DHT)-induced expression of ETV1 mRNA is markedly enhanced in M37 cells Results as compared with C14 cells, which are stably transfected with c-Jun Coactivation Enhances Androgen-Induced Expres- an empty vector (19). In contrast, the DHT effect on ETV1 sion of ETV1 expression is almost completely eliminated in AJ81 cells, which Our previous results show that c-Jun can act as a coactivator have drastically reduced levels of endogenous c-Jun (19). These for AR transcriptional activity (17, 18) and androgen-dependent data were confirmed by quantitative PCR, which showed that

FIGURE 1. c-Jun coactivates AR-induced expression of ETV1. The LNCaP stable cell lines C14, M37, and AJ81 were analyzed for DHT-induced expression of ETV1 mRNA by semiquantitative RT-PCR (A) or real-time quantitative PCR (B) or protein by Western blotting (C). The ETV1 mRNA level is represented as a ratio of GAPDH, and real-time PCR values are relative to first condition, and this was set to 1. M37 cells were transfected with control or c-Jun siRNA and monitored for expression of ETV1 mRNA (D) or protein (E). F. LNCaP cells were cotransfected with ETV1-Luc and c-Jun or c-Jun(Ala63/73) and measured for ETV1 promoter activity by luciferase assay. ‘‘Vector’’ is transfection with empty vector as a control. G. LNCaP stable cell lines were transfected with the Fes3ÂWT-Luc reporter plasmid. Where indicated, cells were treated with ethanol (À) or 100 nmol/L DHT (+). The luciferase activity is represented relative to the activity of either empty vector (F) or LNCaP without DHT (G), and each activity was set to 1. *, P < 0.05, statistical significance in the DHT positive effect in M37 cells or in the negative effect in AJ81 cells in expression of ETV1 mRNA (B) or luciferase activity (G) or c-Jun activity on the ETV1 promoter (F).

Table 1. Expression of ETV1 and Other Androgen-Regulated Genes in LNCaP Cells Using Oligonucleotide Gene Microarray Analysis

Gene Name DHT-Induced Expression

C14 M37 M37 vs C14

ETV1 (U17163) 708 (P = 0.0002) 1555 (P = 0.000037) I (P = 0.000004) VEGF (M63978) 899 (P = 0.0001) 1367 (P = 0.000006) I (P = 0.000092) sGCa1 (Y15723) 1154 (P = 0.000001) 1317 (P = 0.000019) NC (P = 0.5) Cyclin B (M25753) 1245 (P = 0.000074) 1052 (P = 0.003) NC (P = 0.85) PLZF (AF060568) 500 (P = 0.000001) 460 (P = 0.000001) NC (P = 0.87) Nkx3.1 (U80669) 374 (P = 0.0011) 253 (P = 0.0012) NC (P = 0.84)

NOTE: LNCaP (C14) and M37 cells were grown for 2 d in the presence (+) or absence (À) of 100 nmol/L DHT. I (increase), D (decrease), and NC (no change) represent the effect of DHT on gene expression in the same cell line (C14 or M37) or comparison of C14 and M37 cells in the presence of DHT.

DHT-induced levels of ETV1 are significantly elevated in M37 transfected Fes promoter exhibited a small but reproducible cells and inhibited in AJ81 cells, as compared with C14 cells DHT-induced increase in activity (Fig. 1G). Importantly, this (Fig. 1B). Western blotting was used to measure the expression promoter yielded higher DHT induction in M37 cells than in of ETV1 protein, showing higher expression in M37 cells and C14 cells and no DHT effect in AJ81 cells (Fig. 1G). lower expression in AJ81 cells than in C14 cells (Fig. 1C). To Interestingly, M37 cells exhibited significantly higher Fes determine if the increased ETV1 expression in M37 cells is due promoter activity than LNCaP cells in the absence of DHT. to the elevated levels of Jun protein in these cells, c-Jun small These data suggest that DHT-activated AR can induce the interfering RNA (siRNA) was transfected into M37 cells, activity of an ETV1-regulated reporter plasmid and that the resulting in reduced endogenous c-Jun expression (both wild- c-Jun coactivation function may be important in this effect. type and mutant; Fig. 1D). Importantly, the c-Jun siRNA also We have previously published that the androgen-induced resulted in a marked reduction in DHT-induced ETV1 mRNA expression of prostate-specific antigen (PSA), hKLK2, and (Fig. 1E) and protein (Fig. 1D) expression. Collectively, these TMPRSS2 is elevated in M37 cells as compared with C14 cells data suggest that the c-Jun coactivation function on AR is an (19), suggesting that c-Jun can mediate AR transactivation of important regulator of ETV1 expression in prostate cancer cells. these androgen-regulated genes. The gene microarray approach To determine if the c-Jun(Ala63/73) mutant protein overex- was used not only to identify ETV1 as a novel androgen- pressed in M37 cells is able to directly transactivate ETV1, we regulated gene1 but also to measure the potential role of c-Jun used an ETV1-Luc reporter plasmid in a transfection experi- coactivation function on other androgen-regulated genes. This ment in LNCaP cells. This reporter plasmid contains a 1-kb analysis showed significantly increased androgen-induced ETV1 promoter fragment that is activated by AR and is able to expression in M37 cells, as compared with C14 cells, for some recruit AR in a ligand-dependent manner.1 As shown in Fig. 1F, genes including the three genes mentioned above (data not transfected wild-type c-Jun was able to strongly activate the shown) and VEGF (Table 1). Interestingly, other genes, cloned ETV1 promoter, whereas c-Jun(Ala63/73) was almost including soluble guanylyl cyclase a1(sGCa1; ref. 21), cyclin completely inactive. Western blotting confirmed similar B, promyelocytic leukemia zinc finger (PLZF), and Nkx3.1, expression of the two c-Jun proteins.2 This finding suggests were not altered in their expression (Table 1). These data show that the increased ETV1 expression in M37 cells is likely due to that c-Jun coactivation of AR-regulated gene expression is gene elevated AR activity in response to c-Jun coactivation. It is specific. interesting to note, however, that c-Jun(Ala63/73) can weakly enhance ETV1 mRNA expression in absence of DHT (Fig. 1B), c-Jun and AR Coassociate and Colocalize in LNCaP whereas it has no effect on the ETV1 promoter under the Cells same conditions (Fig. 1F), suggesting that the endogenous Our data above show that c-Jun stimulates AR trans- ETV1 promoter may harbor an element(s) through which activation of ETV1. To determine if there is a physical 63/73 c-Jun(Ala ) can act independent of ligand-activated AR, and association between c-Jun and AR, immunoprecipitation this is absent from the 1-kb ETV1 genomic fragment found in experiments were done with endogenous proteins found in the ETV1-Luc reporter plasmid. LNCaP cells. When an anti–c-Jun antibody was used to ETV1 transcriptional activity was also measured using the immunoprecipitate endogenous c-Jun in LNCaP cells, c-Jun Â ETV1-regulated luciferase (Luc) reporter plasmid Fes3 WT- and AR proteins were copurified (Fig. 2A). Neither c-Jun nor Luc, which contains three copies of an Ets-responsive element AR was purified with a nonspecific immunoglobulin G from the Fes promoter (20). This promoter was strongly (Fig. 2A). These findings show an endogenous interaction 2 induced by transfected ETV1 in LNCaP cells. When it was between c-Jun and AR in LNCaP cells and confirm previous tested for DHT regulation in the LNCaP cell lines, the data (17). Our success with the immunoprecipitation experiments above in LNCaP cells prompted us to investigate subnuclear colocalization of endogenous AR and c-Jun. This was done by 2 C. Cai and L. Shemshedini, unpublished results. immunofluorescence with anti-AR monoclonal and anti–c-Jun

polyclonal antibodies. As shown in Fig. 2B, androgen treatment multiple genomic sites. This is consistent with our microarray stabilized endogenous AR and caused nuclear localization in (Table 1) and reverse transcription-PCR2 (RT-PCR; ref. 19) LNCaP cells, as previously reported (22). Just as with ligand- results, showing that c-Jun can coactivate multiple AR- activated AR, endogenous c-Jun is also nuclear (Fig. 2C). regulated genes. Our data show that VEGF and ETV1 exhibit When the two immunofluorescence images were merged, a elevated expression in M37 cells, whereas other genes including fraction of AR was found to colocalize with c-Jun, as shown by Nkx3.1 are not changed in their expression (Table 1). yellow staining. These results show an in vivo nuclear Our confocal images also show that AR and c-Jun are found colocalization between AR and c-Jun. To determine if the on other sites without colocalization (Fig. 2D), suggesting that changes in c-Jun expression alter the colocalization with AR, these two transcription factors can also act independent of one M37 cells were also studied. As shown in Fig. 2C, the another, as would be expected on the Nkx3.1 genomic site as expression of c-Jun(Ala63/73) led to increased intensity of the well as others. c-Jun signal in M37 cells as compared with C14 cells, whereas To determine if c-Jun interaction with AR can lead to AR levels were unaltered. Importantly, when the c-Jun and AR recruitment of c-Jun to the ETV1 promoter, chromatin images were merged, there is significantly more nuclear immunoprecipitation was used. In our earlier work, we have colocalization of these two proteins in M37 cells than in C14 shown by chromatin immunoprecipitation that liganded AR cells (Fig. 2C), strongly suggesting that the overexpressed was recruited to an ETV1 promoter region harboring a putative c-Jun(Ala63/73) can colocalize with ligand-activated AR. These androgen-responsive element.1 The chromatin immunopre- data collectively show that endogenous c-Jun colocalizes with cipitation assay was repeated to determine if c-Jun can be co- endogenous AR in LNCaP nuclei and this colocalization is recruited to this same ETV1 promoter region. The recruitment directly proportional to the levels of c-Jun protein. of c-Jun to the ETV1 promoter is weak in the absence of DHT To more closely examine the colocalization of AR and c-Jun and, importantly, markedly stronger in the presence of DHT, in M37 cells, the confocal image of one cell was amplified. As especially at 24 h (Fig. 2E). Because PSA gene expression is Fig. 2D shows, AR and c-Jun are colocalized on multiple also enhanced by c-Jun coactivation (19), the same chromatin nuclear sites, suggesting that these two proteins act together on immunoprecipitation experiment showed androgen-induced

FIGURE 2. Endogenous c-Jun and AR coassociate and colocalize in LNCaP cells. A. Whole-cell extracts were prepared from LNCaP cells grown in 10% serum and subjected to immunoprecipitation (IP) with an anti – c-Jun antibody. Western blotting was used to detect AR (top) or c-Jun (bottom). The negative control immunoprecipitation was done with an immunoglobulin G (IgG) antibody. ‘‘Input’’ represents 20% of extract used in the immunoprecipitation experiments. B to D. Immunocytochemistry was used to measure subcellular localization of AR in LNCaP cells treated with 100 nmol/L DHT (+) or ethanol (À; B) or AR and c-Jun in LNCaP stable cell lines treated with 100 nmol/L DHT (C and D). Magnification, Â400 (B and C); Â1,000 (D). E. Chromatin immunoprecipitation assay was done with LNCaP cells treated with or without DHT for the indicated times to measure c-Jun recruitment to the ETV1 1-kb promoter, PSA promoter, or sGCa1 promoter.

FIGURE 3. c-Jun stimulates the activities of ETV1-regulated promoters. LNCaP cells were transfected with expression plasmids for ETV1, c-Jun, or c-Jun(Ala63/73) in different combinations and Fes3ÂWT-Luc (A), MMP1-Luc (B), or MMP1(mAP1)-Luc (C). D. The same transfection as in (B) except that c-Jun(M14) mutant was used in place of c-Jun(Ala63/73) and the AP-1 – responsive reporter plasmid TRE3-tk-Luc (19) was also used. E. LNCaP cells were transfected with c-Jun or c-Jun(Ala63/73) and control or ETV1 siRNA together with MMP1-Luc. In all cases above, ‘‘Vector’’ is transfection with empty vector as a control and the luciferase activity is represented relative to the activity of Vector, and this activity was set to 1. F. Different LNCaP stable cell lines were transfected with MMP-Luc or Fes3ÂWT-Luc reporter plasmids. The luciferase activity is represented relative to the activity in C14 cells, and this activity was set to 1. *, P < 0.05, statistical significance in the negative effect of ETV1 siRNA (E) or of luciferase activity in M37 cells (F). binding of c-Jun to the PSA promoter that was similar to the strongly enhance ETV1 activation of this promoter, with an ETV1 promoter (Fig. 2E). As a negative control, we used the activity similar to wild-type c-Jun (Fig. 3A). Similar but weaker promoter for sGCa1, whose expression is not affected by effects were seen on matrix metalloproteinase 1 (MMP1)-Luc c-Jun2 (see Table 1). Whereas the sGCa1 promoter can recruit (Fig. 3B), a reporter plasmid driven by the MMP-1 gene liganded AR (21), no c-Jun recruitment was detected (Fig. 2E). promoter (24). Hence, c-Jun enhances ETV1 transactivation These results suggest that c-Jun is recruited to only those AR- independent of promoter specificity. regulated promoters that respond to c-Jun coactivation and that The MMP-1 promoter harbors a consensus AP-1 site (25), this c-Jun recruitment depends on ligand-activated AR. whereas the Fes promoter does not.3 This makes it possible that the c-Jun activity on the MMP-1 promoter may depend on c-Jun Regulates the Expression of ETV1-Induced Matrix c-Jun binding to this AP-1 site. To examine this possibility, Metalloproteinase Genes we used two approaches. First, the AP-1 site on MMP-1 was Because previous evidence suggests that c-Jun can stimulate mutated to yield the reporter MMP1(mAP1)-Luc (25). This the activity of PEA3 proteins (12, 23), we hypothesized that promoter was completely unresponsive to either transfected c-Jun can enhance ETV1 transactivation. This hypothesis was ETV1 or c-Jun (Fig. 3C), an expected finding in view of the first tested using the ETV1 reporter plasmid Fes3ÂWT-Luc. As shown in Fig. 3A, transfected ETV1 or c-Jun similarly induced (f10-fold) Fes promoter activity, whereas the c-Jun(Ala63/73) mutant was weaker. However, this c-Jun mutant was able to 3 J-L. Baert, personal communication.

overlapping nature of the ETV1 and AP-1 binding sites (25). ETV1 siRNA significantly decreased MMP-1 and MMP-9 Therefore, we used an alternative approach, in which the c-Jun expression (Fig. 4C), indicating an essential transactivation role mutant M14, deficient in dimerization and thus DNA binding for ETV1 in M37 cells. The siRNA effects on ETV1 expression (26), was used. As shown in Fig. 3D, M14 is able to in M37 cells were confirmed by RT-PCR (Fig. 4C) or Western cooperatively activate the wild-type MMP-1 promoter as well blotting (Fig. 4E). as wild-type c-Jun protein, whereas it has no activity on an The reduction of endogenous c-Jun by siRNA also resulted AP-1 responsive promoter (Fig. 3D), showing that the in decreased levels of MMP-1 and MMP-9 (Fig. 4D), suppor- cooperation of c-Jun with ETV1 does not depend on the ting our hypothesis that the expression of c-Jun(Ala63/73) DNA-binding function of c-Jun. mutant in M37 cells is responsible for the increased ETV1 To show that the c-Jun enhancement of the activity of the transactivation in these cells. AJ81 cells, which have reduced ETV1-regulated promoter relies on the endogenous ETV1 in levels of c-Jun (9), expressed markedly lower MMP-7 and LNCaP cells, we cotransfected c-Jun or c-Jun(Ala63/73) mutant MMP-13 mRNA levels (Fig. 4F), consistent with a cooperative together with ETV1 siRNA to block endogenous ETV1 function of endogenous c-Jun on ETV1 in LNCaP cells. expression (Fig. 3E). ETV1 siRNA significantly attenuated the c-Jun positive effect on the MMP-1 promoter, suggesting c-Jun Associates and Colocalizes with ETV1 that c-Jun depends on endogenous ETV1 for activation of the Our data above show that c-Jun stimulates both AR MMP-1 promoter. To further confirm these results, we used our transactivation of ETV1 and subsequent ETV1 transactivation two LNCaP stable cell lines that express either c-Jun(Ala63/73) of MMP genes, indicating that c-Jun can cooperate with both mutant (M37 cell line) or antisense c-Jun (AJ81 cell line). As transcriptional activators. To determine if there is a physical seen in Fig. 3F, M37 cells had significantly higher activity association between c-Jun and ETV1, immunoprecipitation from both the MMP-1 and Fes promoters than did C14 cells, experiments were done with endogenous proteins found in whereas AJ81 cells had markedly lower activity. Collectively, LNCaP cells. When ETV1 was immunoprecipitated with an these results show that c-Jun can cooperate with ETV1 to anti-ETV1 antibody that has been successfully used before transactivate ETV1-regulated promoters independent of c-Jun (15), copurified c-Jun was detected by Western blotting transactivation function. (Fig. 5A). The negative control immunoprecipitation did not Next, we studied the importance of c-Jun on endogenous copurify c-Jun (Fig. 5A). Immunoprecipitated ETV1 could not MMP gene expression. As shown in Fig. 4A, transfecting be monitored in this immunoprecipitation experiment because LNCaP cells with either ETV1 or c-Jun(Ala63/73) significantly the antibody heavy chain comigrates on the SDS-PAGE gel induced expression of MMP-1, consistent with the promoter with ETV1.2 This demonstration of a physical association assays in Fig. 3B, and MMP-9. MMP-1 and MMP-9 were between ETV1 and c-Jun supports the c-Jun cooperative expressed significantly in M37 cells and undetectably in C14 function. cells, whereas MMP-7 and MMP-13 expression was only Colocalization between c-Jun and ETV1 was measured weakly increased in M37 cells as compared with C14 cells by immunocytochemistry in LNCaP cells lines, just as it was (Fig. 4B). To examine if the elevated expression of MMP-1 and done for c-Jun and AR (see Fig. 2). This experiment showed a MMP-9 in M37 cells was due to c-Jun cooperation with colocalization of c-Jun and ETV1 in the nuclei of LNCaP cells endogenous ETV1, we used siRNA to specifically reduce (Fig. 5B). The magnitude of colocalization depends on c-Jun ETV1 or c-Jun protein levels. As expected, transfection with levels because c-Jun/ETV1 colocalization is greatly enhanced

FIGURE 4. c-Jun stimulates ETV1 transactivation of MMP genes. Semiquantitative RT-PCR was used to measure the expression of MMP genes in LNCaP cells transfected with either c-Jun(Ala63/73)orETV1(A), C14 or M37 cells (B), M37 cells transfected with ETV1 siRNA (C) or c-Jun siRNA (D), or C14 and AJ81 cells (F). E. Western blotting was used to show reduced expression of ETV1 in M37 cells transfected with ETV1 siRNA.

FIGURE 5. Endogenous c-Jun associates and colocalizes with ETV1 in LNCaP cells. A. Whole-cell extracts prepared from LNCaP cells were subjected to immunoprecipitation with an anti-ETV1 antibody. Western blotting was used to detect coprecipitated c-Jun. The negative control immunoprecipitation was done with an immunoglobulin G antibody. ‘‘Input’’ represents 20% of extract used in the immunoprecipitation experiments. B. Immunocytochemistry was used to measure nuclear colocalization of ETV1 and c-Jun in LNCaP stable cell lines treated with 100 nmol/L DHT. Magnification, Â400. in M37 and inhibited in AJ81 cells, when compared with C14 tested by semiquantitative RT-PCR of prostate tissues control cells (Fig. 5B). Together, these data suggest that c-Jun purchased from the Cooperative Human Tissue Network. can mediate ETV1 transactivation by associating with ETV1 in Cooperative Human Tissue Network provides tissues that are the nuclei of prostate cancer cells. normal, benign prostatic hyperplasia, or malignant prostate cancer. To confirm that these tissues represent different stages c-Jun Mediates Invasion of Prostate Cancer Cells of prostate cancer, the expression of several other genes was Our data above (see Fig. 4) show that exogenous expression measured. This included PSA, which is highly expressed in of c-Jun or c-Jun(Ala63/73) led to elevated MMP gene both malignant prostate cancer and several benign prostatic expression. These data, together with our recent finding that hyperplasia tissues; EZH2, which is only expressed in the ETV1 mediates LNCaP cell invasion,1 suggest that c-Jun may two malignant prostate cancer samples; and E-cadherin, mediate cell invasion. This hypothesis was tested by comparing whose expression is significantly reduced in the benign prostatic the invasive capacity of M37 cells with that of C14 cells. As hyperplasia and malignant prostate cancer tissues as compared shown in Fig. 6A, DHT has a small but reproducible positive with normal (21). As we have previously reported,1 ETV1 effect on C14 cell invasion, as we have previously determined.1 mRNA is highly expressed in the two malignant prostate cancer Interestingly, this invasive ability is significantly elevated in samples and one benign prostatic hyperplasia (Fig. 6B); no M37 cells, both in the absence and presence of DHT (Fig. 6A), ETV1 was detected in the one normal tissue (Fig. 6B). showing that expression of c-Jun(Ala63/73) enhances the Interestingly, c-Jun expression closely parallels the expression invasive capacity of LNCaP cells. of ETV1, with the highest levels in the two malignant prostate cancer tissues and B2 (Fig. 6B). A similar expression pattern c-Jun Overexpression in Prostate Cancer Tumors Paral- was observed for AR, which, as expected (8), is expressed lels ETV1 Expression higher in C1 and C2 tissues than in the other tissues (Fig. 6B). Our previous data show that ETV1 mRNA and protein are Together, these limited tissue expression data show a similar overexpressed in prostate cancer tissues.1 Further, we expression pattern for ETV1, c-Jun, and AR. determined that this overexpression occurs independent of the translocation with the TMPRSS2 gene1 that was recently reported (16). If c-Jun is important in ETV1 activity, as our Discussion data suggest in this study, then c-Jun expression in prostate Several earlier studies from our lab have characterized the cancer tissues may parallel the expression of ETV1. This was coactivation function of c-Jun on AR transcriptional activity.

FIGURE 6. c-Jun enhances cell invasion capacity and is overexpressed in prostate tumors. A. Cell invasion assay was done on C14 or M37 cells in the presence of either ethanol (À) or 100 nmol/L DHT (+), as indicated. Invasive cell numbers given are relative to numbers of cells without DHT, and this activity was set to 1. Columns, average of three independent experiments; bars, SD. *, P < 0.05, statistical significance in the increased invasiveness of M37 as compared with C14 cells with DHT. B. Different prostate tissues were measured for expression of ETV1, c-Jun, and AR by RT-PCR. N, normal; B, benign prostatic hyperplasia; C, malignant prostate cancer. C. This model represents the involvement of AR, c-Jun, and ETV1 in prostate cancer cell invasion and proliferation. Solid thick arrow, enhancement of expression; thin arrow, positive effect; thin line without an arrowhead, negative effect; dashed arrow or line, weak positive or negative effect. Question mark, an unknown ETV1-regulated gene(s) that can induce cell invasion. ‘‘c-Jun’’ represents the coactivation or cooperative function of this protein and ‘‘AP-1’’ represents its transactivation function.

These studies identified regions of c-Jun necessary for AR wild-type c-Jun exhibited strong activity on the ETV1 coactivation (6) and regions of AR that respond to this c-Jun promoter (see Fig. 1F), showing that c-Jun can transactivate activity (27); showed that heterdimerization with c-Fos the ETV1 gene independent of ligand-activated AR. Whereas antagonizes c-Jun coactivation (5); functionally separated overexpression of c-Jun led to elevated ETV1 expression, c-Jun coactivation from transactivation (6); and importantly siRNA-mediated diminution of c-Jun resulted in substantially suggested that c-Jun coactivation supports prostate cancer cell reduced ETV1 expression (see Fig. 1E), strongly suggesting proliferation whereas c-Jun transactivation is antiproliferative that endogenous c-Jun and/or stably expressed c-Jun(Ala63/73) (19). In contrast to these two opposite effects of c-Jun on is involved in ETV1 expression in M37 cells. However, proliferation, our study here provides evidence that both the because siRNA inhibits the expression of both wild-type and c-Jun coactivation and transactivation functions have a mutant c-Jun, it is impossible to determine from this experi- positive effect on prostate cancer cell invasion. In addition, ment which of the two c-Jun activities, transactivation or this proinvasive effect of c-Jun in prostate cancer seems to coactivation, is more important for ETV1 expression in prostate depend on multiple activities by this proto-oncoprotein in a cancer cells. novel pathway that links the AR to ETV1 and is described in Not only is c-Jun able to elevate ETV1 expression but our another article.1 In that article, we reported that ETV1 is data show that it can also mediate ETV1 transcriptional activity, transcriptionally induced by ligand-activated AR in prostate as measured by both reporter gene assays and endogenous gene cancer cells.1 In the current study, we describe c-Jun activities expression. With respect to the reporter gene assays, both the in the entire pathway that begins with AR transactivation of Fes and MMP-1 promoters are about equally transactivated by ETV1 and terminates with prostate cancer cell invasion (see either exogenous ETV1 or wild-type c-Jun (see Fig. 3A and B). Fig. 6A). Thus, the role of c-Jun in prostate cancer cell Interestingly, transfected c-Jun(Ala63/73) had weaker but invasion seems to be mediated by multiple activities of this significant positive activity on both promoters (see Fig. 3A proto-oncoprotein. and B), probably due to its ability to cooperate with endogenous c-Jun acts in the beginning of the AR-ETV1 invasion ETV1 to activate transcription. This possibility is supported by pathway by enhancing ETV1 expression in prostate cancer our finding that the Fes promoter had higher activity in M37 cells. Our data suggest that this enhancement is due to (a) cells than in LNCaP cells both in the presence and absence c-Jun coactivation of AR and (a) c-Jun direct transactivation of of DHT (see Fig. 3F); the increased activity without the ETV1 promoter. The coactivation function is shown by our DHT probably reflects the ability of the c-Jun(Ala63/73) finding that either transient (see Fig. 1F) or stable expression, as protein expressed in M37 cells to enhance endogenous in M37 cells (see Fig. 1G), of the transactivation-deficient ETV1 activity on the Fes promoter. The effect of coexpressed mutant c-Jun(Ala63/73) led to elevated expression of ETV1. ETV1 and c-Jun was strongly synergistic on the Fes promoter This c-Jun mutant has been shown to be unable to activate AP- and weakly on the MMP-1 promoter (see Fig. 3A and B), 1–regulated promoters (6) and, in this study, the cloned ETV1 perhaps reflecting a promoter-specific difference in cooperation promoter that responds to ligand-activated AR.1 By contrast, magnitude between ETV1 and c-Jun. The cooperative activity

of c-Jun on ETV1 transactivation does not depend on c-Jun antagonistic to proliferation (19). These multiple effects of DNA binding, as shown by the equal ability of a DNA binding- c-Jun on prostate cancer behavior are undoubtedly due to deficient mutant as wild-type c-Jun to support ETV1 activation differential gene expression that is induced by c-Jun coactiva- of the MMP-1 promoter (see Fig. 3D). tion of AR and ETV1 and direct transactivation as a part of Interestingly, c-Jun enhancing activity has previously been AP-1. As an initial analysis of the c-Jun coactivation function reported on several Ets transcription factors, including PU.1 on androgen-regulated genes, a gene microarray system was (26) and all three PEA3 group proteins (12, 22). Here, we used to measure expression of androgen-regulated genes in have provided evidence that this c-Jun enhancing activity also LNCaP cells expressing c-Jun(Ala63/73), a mutant deficient in occurs on MMP gene expression. Our earlier work suggested transactivation but fully competent in AR coactivation (6). This that MMP-9 and MMP-13, like MMP-1 (27) and MMP-7 analysis shows that some androgen-regulated genes, including (12), are ETV1 target genes.1 In this study, we show that ETV1, exhibit elevated expression induced by c-Jun coactiva- MMP gene expression is higher in M37 cells than in control tion, whereas other genes are not changed in their expression LNCaP cells (see Fig. 4B). Similar results to those from M37 (see Table 1). This finding clearly shows that c-Jun coactivation cells were obtained when c-Jun(Ala63/73) was transiently of AR is not general but target gene specific. Perhaps c-Jun also expressed in LNCaP cells (see Fig. 4A), showing that has gene-specific effects on ETV1 transcriptional activity, a overexpression of the transactivation-deficient mutant was hypothesis that will form the basis of future studies aimed at sufficient to enhance MMP gene expression. All these results identifying genes mediating the ETV1 effect on prostate cancer show that the transactivation function of c-Jun is not cell invasion. necessary for enhanced ETV1 transcriptional activity on MMP gene expression and thus suggest that c-Jun may act as a coactivator for ETV1, as it does for AR. This coactivation Materials and Methods function is likely mediated by a physical interaction between Cell Culture and siRNA Transfection c-Jun and AR or ETV1 in LNCaP cells, as our immunopre- The culturing of LNCaP, C33, C81, C14, M37, AJ81, and cipitation and immunocytochemistry data here suggest (see A103 cells has been described (19). Note that the expression Fig. 5A and B). Interestingly, the immunocytochemistry data of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and also suggest that the c-Jun/AR nuclear colocalization sites h-actin was similar among these different cell lines, showing may be different from the c-Jun/ETV1 sites,1 and thus suggest that housekeeping gene expression is not altered in these cells. that c-Jun cooperates with AR or ETV1 by associating with Commercial siRNAs were obtained for ETV1 (Ambion), c-Jun, these two proteins at distinct nuclear sites. To obtain evidence and negative control (both from Santa CruzBiotechnology). for this, we did chromatin immunoprecipitation assays Note that we used two different siRNAs targeting different showing that c-Jun can be recruited in an androgen-dependent regions of ETV1 or c-Jun, and these had similar individual 2 manner to the ETV1 promoter region (see Fig. 2E) that effects. siRNA was transfected into cells using X-tremeGENE also recruits ligand-activated AR.1 Importantly, c-Jun is siRNA transfection following the manufacturer’s protocol also recruited to the PSA promoter but not the sGCa1 (Roche). Plasmid DNA was transfected using Lipofectamine promoter (see Fig. 2E). In view of our data showing that 2000 following the manufacturer’s protocol (Invitrogen). c-Jun(Ala63/73) can enhance the expression of ETV1 (see Fig. Cells were grown to 60% to 70% confluency in 10% fetal 1G) and PSA (19) but not sGCa12 (see Table 1), our bovine serum (FBS)–containing medium and then changed chromatin immunoprecipitation data here strongly suggest that to serum-free medium. After 48 h of incubation, ethanol or c-Jun is recruited in an androgen-induced manner only to 100 nmol/L DHT was added to the cells. After an additional 48 h those promoters that respond to c-Jun coactivation. Addition- of incubation, the cells were subjected to either semiquantitative ally, these data suggest that the c-Jun physical association with RT-PCR or Western blotting. AR (see Fig. 2A-D) is responsible for a promoter-specific recruitment of c-Jun, providing the first evidence that ligand- Plasmid and Reporter Gene Assay activated AR can bring c-Jun to AR-regulated promoters. The cloning of the human ETV1 promoter to make the Just as c-Jun enhancing activity on ETV1 expression seems reporter plasmid ETV1-Luc has previously been described.1 to be mediated by its coactivation and transactivation functions, This promoter exhibits androgen-induced activation and so too is the effect of c-Jun on MMP gene expression. It was contains a near consensus androgen-responsive element that is reported several years ago that c-Jun directly activates MMP able to recruit liganded AR in a chromatin immunoprecipitation genes through its transactivation function as a component of assay.1 The other reporter plasmids used, MMP1-Luc (23), AP-1 (28). Thus, taking into account our data presented here MMP1(mAP1)-Luc (25, 29), and Fes3ÂWT-Luc (pBfes.Luc; together with published results produces a complex and ref. 20), have previously been described. interesting model for c-Jun action in prostate cancer. c-Jun For reporter gene assays, cells were transfected and has multiple transcriptional functions as a coactivator and luciferase assays done as previously described (19). For all transactivator, which can lead to increased expression and transfections, empty vector (empty expression plasmid or activity of ETV1, resulting in elevated expression of MMPs or promoter-less reporter plasmid) was used to ensure equal other functionally related gene(s) that can then induce prostate amounts of each kind of vector, and pCH110, which expresses cancer cell invasion (Fig. 6C). In addition, our previous data h-galactosidase, was used to standardize transfection efficiency suggest that c-Jun coactivation can also enhance prostate cancer (16). Note that the h-galactosidase activity was very similar in cell proliferation, whereas c-Jun transactivation, surprisingly, is all the prostate cancer cell lines used,2 showing that the

constitutive promoter driving the expression of h-galactosidase NTD for immunoprecipitation and AD2 for Western blotting was equally active among the different cells lines. All luciferase (directed against the ETV1 acidic domain amino acids 42-73; values represent the average of three independent transfections ref. 15), anti-AR antibody PA1110 (Affinity Bioreagents), anti– plus standard deviatons. c-Jun antibody sc-45 (Santa CruzBiotechnology), or anti– h- actin antibody (Abcam). h-Actin was used as a control for Semiquantitative RT-PCR and Real-time Quantitative protein amount. PCR Analyses RNA was isolated using the TRIzol reagent and subjected to Immunocytochemistry either semiquantitative RT-PCR as previously described (19) or LNCaP cells were grown on glass coverslips and fixed with real-time quantitative-PCR using SYBR Green (iSCRIPT from formaldehyde. Then, the cells were incubated with Image-iT Bio-Rad). The upstream and downstream primers, respectively, FX signal enhancer (Molecular Probes). After rinsing with used for each gene were PSA, 5¶-GCAGCATTGAACCAGAG- PBS, cells were incubated with the anti-AR antibody sc-815 GAG-3¶ and 5¶-CCCATGACGTGATACCTTGA-3¶; c-Jun, 5¶- and anti–c-Jun antibody sc-45 (both from Santa Cruz TGACTGGAAAGATGGAAACG-3¶ and 5¶-CCGTTGCT- Biotechnology) or anti-ETV1 antibody AD2 and anti–c-Jun GGACTGGATTAT-3¶;AR,5¶-CAATGAGTACCGCATG- antibody sc-1694 (Santa CruzBiotechnology). This was CAC-3¶ and 5¶-GCCCATCCACTGGAATAATG-3; ETV1, followed by staining with secondary antibodies (Molecular 5¶-TACCCCATGGACCACAGATT-3¶ and 5¶-CACTGGG- Probes) against the respective primary antibodies, except for TCGTGGTACTCCT-3¶; MMP-1 5¶-ATGCTGAAACCCT- sc-1694, which is already conjugated to rhodamine. After GAAGGTG-3¶ and 5¶-CTGCTTGACCCTCAGAGACC-3¶; rinsing with PBS, the antifade reagent proLong Gold MMP-7, 5¶-GAGTGCCAGATGTTGCAGAA-3¶ and 5¶- (Molecular Probes) was used to mount the specimen for AAATGCAGGGGGATCTCTTT-3¶;MMP-9,5¶-GCCATT- fluorescence microscopy. Note that all micrographs were taken CACGTCGTCCTTAT-3¶ and 5¶-TTGACAGCGACAA- at the same microscope settings. GAAGTGG-3¶;MMP-13,5¶-GGAGCCTCTCAGTCAT- GGAG-3¶ and 5¶-TTGAGCTGGACTCATTGTCG-3¶;and Cell Invasion Assay GAPDH, 5¶-CGACCACTTTGTCAAGCTCA-3¶ and 5¶-AGG- Cell invasion was measured using the Cell Invasive Assay GGAGATTCAGTGTGGTG-3¶. GAPDH was used as a control Kit from Chemicon experiments following the manufacturer’s for mRNA amount. protocol. Briefly, cell suspensions containing 800,000 cells/mL (in serum-free medium) were used to monitor cell invasion into Immunoprecipitation a lower chamber containing RPMI 1640 with 10% charcoal- Whole-cell extracts from LNCaP cells grown to 90% stripped serum. After 72 h of incubation at 37jC, cells were confluence in 10% FBS–containing RPMI 1640 were subjected stained and quantified. to immunoprecipitation following the protocol from Upstate, using protein A-Sepharose (Amersham). The anti-ETV1 antibody NTD (15) and anti–c-Jun antibody sc-45 (Santa Cruz Affymetrix Gene Chip Assay Biotechnology) were used to immunoprecipitate ETV1 or C14 and M37 cells were grown to 60% to 70% confluency c-Jun, respectively. The anti-AR antibody PA1-110 (ABR) and in 10% FBS and then changed to FBS-free medium. After anti–c-Jun antibody sc-45 were used for Western blot analysis. 24 h of incubation, cells were treated with either ethanol or 100 nmol/L DHT. After 48 h of incubation, total mRNA was Chromatin Immunoprecipitation isolated and subjected to gene chip analysis using chips LNCaP cells were grown to 70% confluence in RPMI 1640 purchased from Affymetrix (GeneChip Human Genome containing 10% FBS. The medium was changed to RPMI 1640 U95Av2 Array) according to the manufacturer’s protocol. containing 2% dextran-coated charcoal/FBS and cells were incubated for 2 days. Then, the cells were treated with either ethanol (À) or 100 nmol/L DHT (+) for 8 or 24 h of incubation, Acknowledgments We thank Dr. R. Janknecht (Department of Biochemistry and Molecular Biology, after which cells were collected and subjected to chromatin Mayo Clinic, Rochester, MN) for providing the MMP1-Luc and MMP1(mAP1)- immunoprecipitation assay as described (21). The anti–c-Jun Luc plasmids and Drs. J-L. Baert (Institut de Biologie de Lille, Institut Pasteur, Lille Cedex, France) and Y. de Launoit (Laboratory of Molecular Virology, antibody sc-44 (Santa CruzBiotechnology) and protein A- Faculty of Medicine, The Free University of Brussels, Brussels, Belgium) for Sepharose (Amersham) were used for immunoprecipitation. providing an ETV1 expresssion plasmid, antibodies, and the FES3ÂWT-Luc The primers used to detect ETV1 promoter (À1 kb) are reporter plasmid. 5¶-TTTTGTGAATGGGACTGTCG-3¶ (upstream) and 5¶-AGGG- GAACAAGATGGCTTTT-3¶ (downstream). The primers used to detect the PSA promoter are 5¶-GCCTGGATCTGAGAGAGA- References 1. Bos TJ, Bohmann D, Tsuchie H, Tjian R, Vogt PK. v-jun encodes a nuclear TATCATC-3¶ (upstream) and 5¶-ACACCTTTTTTTTTCTG- protein with enhancer binding properties of AP-1. Cell 1988;52:705 – 12. GATTGTTG-3¶ (downstream). The primers used to detect the 2. Angel P, Karin M. The role of Jun, Fos and the AP-1 complex in sGCa1 promoter are previously described (24). cell-proliferation and transformation. Biochim Biophys Acta 1991;1072: 129 – 57. 3. Verma IM, Sassone-Corsi P. Proto-oncogene fos: complex but versatile SDS-PAGE and Western Blot regulation. Cell 1987;51:513 – 4. SDS-PAGE and Western blotting were carried out as 4. Gottlicher M, Heck S, Herrlich P. Transcriptional cross-talk, the second mode described (19). The antibodies used were anti-ETV1 antibody of steroid hormone receptor action. J Mol Med 1988;76:480 – 9.

5. Tillman K, Oberfield JL, Shen XQ, Bubulya A, Shemshedini L. c-Fos potentiates the functional interaction between the amino and carboxyl termini of dimerization with c-Jun represses c-Jun enhancement of androgen receptor the androgen receptor. J Biol Chem 2001;276:44704 – 11. transactivation. Endocrine 1988;9:193 – 200. 19. Chen SY, Cai C, Fisher CJ, et al. c-Jun enhancement of androgen receptor 6. Wise SC, Burmeister LA, Zhou XF, et al. Identification of domains of c-Jun transactivation is associated with prostate cancer cell proliferation. Oncogene mediating androgen receptor transactivation. Oncogene 1998;16:2001 – 9. 2006;25:7212 – 23. 7. Chang C, Kokontis J, Liao S. Molecular cloning of human and rat 20. de Launoit Y, Audette M, Pelczar H, Plaza S, Baert JL. The transcription of complementary DNA encoding androgen receptors. Science 1988;240:324 – 6. the intercellular adhesion molecule-1 is regulated by Ets transcription factors. 8. Jenster G. The role of the androgen receptor in the development and Oncogene 1998;16:2065 – 73. progression of prostate cancer. Semin Oncol 1999;26:407 – 21. 21. Cai C, Chen SY, Zheng Z, et al. Androgen regulation of soluble guanylyl a 9. Chen CD, Welsbie DS, Tran C, et al. Molecular determinants of resistance to cyclase 1 mediates prostate cancer cell proliferation. Oncogene 2007;26: antiandrogen therapy. Nat Med 2004;10:33 – 9. 1606 – 15. 10. Trapman J, Cleutjens KB. Androgen-regulated gene expression in prostate 22. Kemppainen JA, Lane MV, Sar M, Wilson EM. Androgen re- cancer. Semin Cancer Biol 1997;8:29 – 36. ceptor phosphorylation, turnover, nuclear transport, and transcriptional activation. Specificity for steroids and antihormones. J Biol Chem 11. Hsu T, Trojanowska M, Watson DK. Ets proteins in biological control and 1992;267:968 – 74. cancer. J Cell Biochem 2004;91:896 – 903. 23. Nakae K, Nakajima K, Iinazawa J, Kitaoka T, Hirano T. ERM, a PEA3 12. Crawford HC, Fingleton B, Gustavson MD, et al. The PEA3 subfamily of Ets subfamily of Ets transcription factors, can cooperate with c-Jun. J Biol Chem h transcription factors synergizes with -catenin-LEF-1 to activate matrilysin 1995;270:23795 – 800. transcription in intestinal tumors. Mol Cell Biol 2001;21:1370 – 83. 24. Fuchs B, Inwards CY, Janknecht R. Up-regulation of the matrix metal- 13. Yamamoto H, Horiuchi S, Adachi Y, et al. Expression of ets-related loproteinase-1 gene by the Ewing’s sarcoma associated EWS-ER81 and EWS-Fli- transcriptional factor E1AF is associated with tumor progression and 1 oncoproteins, c-Jun, and p300. FEBS Lett 2003;553:104 – 8. overexpression of matrilysin in human gastric cancer. Carcinogenesis 2004;25: 325 – 32. 25. Bosc DG, Goueli BS, Janknecht R. HER2/Neu-mediated activation of the ETS transcription factor ER81 and its target gene MMP-1. Oncogene 2001;20: 14. de Launoit Y, Chotteau-Lelievre A, Beaudoin C, et al. The PEA3 group of 6215 – 24. ETS-related transcription factors. Role in breast cancer metastasis. Adv Exp Med Biol 2000;480:107 – 16. 26. Smeal T, Angel P, Meek J, Karin M. Different requirements for formation of 15. Baert JL, Monte D, Musgrove EA, Albagli O, Sutherland RL, de Launoit Y. Jun: Jun and Jun: Fos complexes. Genes Dev 1989;3:2091 – 100. Expression of the PEA3 group of ETS-related transcription factors in human 27. Bubulya A, Zhou XF, Shen XQ, Fisher CJ, Shemshedini L. c-Jun targets breast-cancer cells. Int J Cancer 1997;70:590 – 7. amino terminus of androgen receptor in regulating androgen-responsive 16. Tomlins SA, Rhodes DR, Perner S, et al. Recurrent fusion of TMPRSS2 and transcription. Endocrine 2000;13:55 – 62. ETS transcription factor genes in prostate cancer. Science 2005;310:644 – 8. 28. Behre G, Whitmarsh AJ, Coghlan MP, et al. c-Jun is a JNK- 17. Bubulya A, Wise SC, Shen XQ, Burmeister LA, Shemshedini L. c-Jun can independent coactivator of the PU.1 transcription factor. J Biol Chem 1999; mediate androgen receptor-induced transactivation. J Biol Chem 1996;271: 274:4939 – 46. 24583 – 9. 29. Benbow U, Brinckerhoff CE. The AP-1 site and MMP gene regulation: what 18. Bubulya A, Chen Z-Y, Fisher CJ, Zheng Z, Shen X-Q, Shemshedini L. c-Jun is all the fuss about? Matrix Biol 1997;15:519 – 26.

Mol Cancer Res 2007;5(7). July 2007 Downloaded from mcr.aacrjournals.org on September 29, 2021. © 2007 American Association for Cancer Research. c-Jun Has Multiple Enhancing Activities in the Novel Cross Talk between the Androgen Receptor and Ets Variant Gene 1 in Prostate Cancer

Changmeng Cai, Chen-Lin Hsieh and Lirim Shemshedini

Mol Cancer Res 2007;5:725-735.

Updated version Access the most recent version of this article at: http://mcr.aacrjournals.org/content/5/7/725

Cited articles This article cites 29 articles, 9 of which you can access for free at: http://mcr.aacrjournals.org/content/5/7/725.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/5/7/725.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/5/7/725. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.