MicroRNA-19 (miR-19) Regulates Tissue Factor Expression in Breast Cancer Cells*

Tissue factor has been recognized as a regulator of tumor angiogenesis and metastasis. The tissue factor gene is selectively expressed in highly invasive breast cancer cells, and the mechanisms regulating tissue factor expression in these cells remain unclear. This study demonstrates that microRNA-19 (miR-19) regulates tissue factor expression in breast cancer cells, providing a molecular basis for the selective expression of the tissue factor gene. Tissue factor protein was barely detectable in MCF-7, T47D, and ZR-75-1 cells (less invasive breast lines) but was expressed at a significantly higher level in MDA-MB-231 and BT-20 cells (invasive breast lines) as assayed by Western blot. The tissue factor gene promoter was activated, and forced expression of tissue factor cDNA was achieved in MCF-7 cells, implying that the 3′-UTR of the tissue factor transcript is responsible for the suppression of tissue factor expression. Bioinformatics analysis predicted microRNA-binding sites for miR-19, miR-20, and miR-106b in the 3′-UTR of the tissue factor transcript. Reporter gene assay using the TF-3′-UTR luciferase reporter construct confirmed that the 3′-UTR negatively regulates gene expression in MCF-7 cells, an effect reversed by deletion of the miR-19-binding site. Application of the miR-19 inhibitor induces endogenous tissue factor expression in MCF-7 cells, and overexpression of miR-19 down-regulates tissue factor expression in MDA-MB-231 cells. RT-PCR analysis using cDNA made from Ago2-immunoprecipitated RNA samples confirmed that Ago2 binds preferentially to tissue factor 3′-UTR in MCF-7 cells, as compared with MDA-MB-231 cells, consistent with the observation that miR-19 levels are higher in MCF-7 cells.

Tissue factor, an initiator of the extrinsic coagulation cascade, was recently recognized as an important regulator of tumor angiogenesis and metastasis, independent of its function in the blood coagulation pathway (1). This is particularly true during breast cancer progression (2). This effect of tissue factor is mainly achieved through the tissue factor-factor VIIa complex that activates the protease-activated receptor 2 signaling pathway (3,4), a G protein-coupled receptor signaling event that generates second messages and alters transcription of the target genes (5). Tissue factor is highly overexpressed in many types of tumors, including breast cancer, especially at the late stages of disease progression (6 -10). A recent report indicates that tissue factor is expressed only in breast carcinoma cells having the highest invasive potential (11). Furthermore, formation of the tissue factor-factor VIIa complex promotes cellular signaling and migration of human breast cancer cells (12). Convincing evidence obtained over the last 5 years suggests that tissue factor may stimulate vascular epithelial growth factor production possibly through the protease-activated receptor 2 pathway, resulting in enhanced tumor angiogenesis (13). Although targeting of the tissue factor signaling pathway has been explored as a strategy to inhibit tumor angiogenesis (1) and prevent breast cancer progression (14), the mechanisms responsible for selective tissue factor expression remain to be defined.
MicroRNAs are a family of ϳ22-nucleotide-long RNA molecules that are highly conserved and expressed in many organisms (15). They are transcribed from various genetic regions and processed in the nuclei through cytoplasm, becoming mature single strand microRNA molecules. It is well recognized that microRNAs regulate gene expression in eukaryotic cells (15). The principal mechanism of microRNA regulation of gene expression is through the formation of a microRNA-RNA-induced silencing complex in the cytoplasm, which binds to the 3Ј-untranslated region (3Ј-UTR) of target transcripts, thereby leading to repression of protein translation or destabilization of the mRNAs (16). Altered expression of microRNA species has been well established in human cancer cells (17,18). Importantly, certain microRNA species have been demonstrated to regulate tumor invasion and metastasis of breast cancer (19,20) and are involved in mediating tumor angiogenesis (21).
The 3Ј-UTR of the tissue factor transcript is around 1 kb long, indicating a strong possibility that this 3Ј-UTR would contribute to the stability of the transcript and the regulation of the protein translation process. MicroRNA regulation of tissue factor expression has never been examined in any model system. This study investigated the potential regulation of tissue factor expression by microRNAs in breast cancer cells. We report here that tissue factor expression is regulated by miR-19 4 in our model systems, which may explain overexpression of the tissue factor gene in invasive breast cancer cells.

EXPERIMENTAL PROCEDURES
Materials-The PGL3 promoter vector and the Dual-Luciferase assay kit were purchased from Promega (Madison, WI). MicroRNA inhibitors and mimics were obtained from Dharmacon, Inc. (Lafayette, CO). The antibody for tissue factor was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) from ProMab Biotechnologies, Inc. (Richmond, CA). Ago2 antibody was ChIP grade and from Abcam Inc. (Cambridge, MA). The human tissue factor promoter reporter construct was obtained from Dr. Nigel Mackman (22). All other reagents were analytic grade and purchased from Sigma.
Confocal Microscopy-The YFP-TF chimera (sYT-pcDNA3.1(ϩ)) was generated by in-frame insertion of the monomeric form of yellow fluorescent protein (kind gift of Dr. Roger Tsien, University of California, San Diego (23)) between the signal and mature peptide sequences of the fulllength tissue factor cDNA clone (MGC-13630 from ATCC) by splicing and overlapping extension PCR. The expressed protein was analyzed in transient and stable transfected endothelial cells by Western blot, immunofluorescence, and activity assays (two-stage Xa generation assay) and was similar to the endogenous counterpart. MCF-7 and MDA-MB-231 cells, in RPMI 1640 and DMEM with supplements, respectively, were plated at 2.5 ϫ 10 5 cells per well in a 6-well plate and incubated overnight. The cells were then transfected with 3 g of the sYT-pcDNA3.1(ϩ) construct in 2 ml of serum-free medium containing 8 l of FuGENE HD transfection reagent (Roche Applied Science). Five hours after incubation, the medium was replaced with fresh medium supplemented with 10% fetal bovine serum. 48 h after transfection, the cells were washed three times with Hanks' balanced salt solution and viewed under a Leica TCS SP2 confocal microscope with LCS Lite imaging software (0.9 ϫ 63 objective, with excitation at 517 nm and emission at 527 nm). The images were quantified using the same software.
Cloning of the Tissue Factor 3Ј-UTR-Total RNA was isolated from MCF-7 cells with TRIzol reagent and reverse-transcribed using the SuperScript II kit (Invitrogen). The cDNA was then subjected to PCR amplification with the primers covering the 3Ј-UTR of the tissue factor transcript (1013-bp fragment, NM_001993, 1013-2026). The primers used were as follows: forward, 5Ј-TGCTCTAGAGGAAGCACTGTTG-GAGCTAC-3Ј, and reverse, 5Ј-TGCTCTAGACAATGT-CAACCATAGAAGC-3Ј. The PCR products were separated on 1% agarose gel containing ethidium bromide, and the DNA fragment was visualized under ultraviolet light. The DNA fragment was purified from the gel using QIAquick gel extraction kit (Qiagen Sciences, Valencia, CA) and cloned into the PGL3 promoter luciferase reporter construct (Promega, Madison, WI) downstream of the luciferase cDNA in both sense and antisense orientations. The cloned constructs were named TF-3Ј-UTR-S (sense orientation) and TF-3Ј-UTR-A (antisense orientation) and were confirmed by DNA sequencing. The TF-3Ј-UTR-A construct was used as a control wherever the TF-3Ј-UTR-S was used. The QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) was used to delete the miR-19-and miR-106b-binding sites from the TF-3Ј-UTR-S construct. The primers used for the deletions were as follows: miR-19 deletion forward, 5Ј-CACCATGGCAC-CCATGCTTTAGATTATATATTCCGC-3Ј, and miR-19 deletion reverse, 5Ј-GCGGAATATATAATCTAAAGCAT-GGGTGCCATGGTG-3Ј; miR-106b deletion forward, 5Ј-GGGCACCTTCCTAATATGCTGACTTAAGTGGC-3Ј, and miR-106b deletion reverse, 5Ј-GCCACTTAAGTCAG-CATATTAGGAAGGTGCCC-3Ј. The deletions were confirmed by DNA sequencing.
Dual-Luciferase Activity Assay-MCF-7 or MDA-MB-231 cells, growing in RPMI 1640 or DMEM with supplements, were plated at a density of 2.5 ϫ 10 5 cells per well in a 6-well plate. After 24 h, the medium was removed, and the transfection complex containing 2 g of the reporter constructs (TF-3Ј-UTR-S, TF-3Ј-UTR-A, or hTF-P2kb), 200 ng of the pRL-TK Renilla luciferase reporter vector (Promega, Madison, WI), and 10 l of FuGENE HD transfection reagent (Roche Applied Science) in 2 ml of serum-free medium was added. After 5 h of incubation, the medium was replaced with 2 ml of RPMI or DMEM with 10% fetal bovine serum, and cells were incubated overnight. The cells then were lifted and plated into a 24-well plate at a density of 250,000 per well. Luminescence was measured 72 h after transfection, using a Dual-Luciferase reporter system (Promega, Madison, WI) following the manufacturer's protocol. The firefly luciferase activity was normalized by the Renilla luciferase activity.
Application of MicroRNA Inhibitors or Mimics-MicroRNA inhibitors and mimics for hsa-miR-19a and hsa-miR-106b were purchased from Dharmacon, Inc. (Lafayette, CO). The day before transfection, MCF-7 or MDA-MB-231 cells were plated at 2.5 ϫ 10 5 cells per well in a 6-well plate. On the day of transfection, the cells were washed with PBS and incubated with 100 nM of the inhibitors or mimics for miR-19a or miR-106b in 2 ml of serum-free medium containing 8 l of FuGENE HD transfection reagent. After 5 h of incubation, the medium was replaced with 2 ml of RPMI 1640 or DMEM with supplements. For some of the experiments, cells were cotransfected with the microRNA inhibitors or mimics with 2 g of the TF-3Ј-UTR-S construct. After 72 h of incubation with the inhibitors or mimics, cells were harvested either for luciferase assay or for Western blot analysis.
Western Blot Analysis-Western blot was performed as we described previously (24). In brief, cells were lysed with the lysis buffer, sonicated on ice, and centrifuged at 15,000 ϫ g for 15 min to remove insoluble material. 40 g of cell lysate from each sample was separated in 10% SDS-polyacrylamide gel, transferred to PVDF membrane, and blotted with antibodies against human tissue factor, Ago2, and GAPDH.
MicroRNA Detection by Real Time PCR-miR-19 levels in MCF-7 and MDA-MB-231 cells were determined by real time PCR analysis. Total RNA was isolated using TRIzol reagent (Invitrogen), and microRNAs were enriched using the mirVana TM miRNA isolation kit (Ambion, Austin, TX). The microRNAs were reverse-transcribed using the RT 2 miRNA first strand kit from SABiosciences (Frederick, MD). Expression of miR-19 was assayed by real time PCR using the ABPrism 7700 sequence detection system (Applied Biosystems, Foster City, CA). The U6 small nuclear RNA served as an internal control for microRNA amplification (26). Primers for miR-19 and U6 were obtained from SABiosciences (Frederick, MD). Relative quantification of expression level was performed following the manufacturer's protocol.
Co-immunoprecipitation-Co-immunoprecipitation using Ago2 antibody was performed as described previously (24). In short, cells growing in 100-mm dishes were washed with PBS and harvested by adding 150 l of IP buffer (10 mM Tris-HCl, pH 7.4, 50 mM NaCl, 0.5 mM EDTA, 1 mM PMSF, and 1% Triton X-100). After sonicating the cells for 1-3 min on ice, insoluble material was removed by centrifugation. Supernatants were collected and protein concentrations determined. Supernatants were pre-cleared by agarose-coupled protein A, and the agarose beads were removed by centrifugation. Ago2 antibodies were added to the supernatants (1:100 ratio), and the reaction was incubated at 4°C overnight with gentle rotation. 50 l of agarose-coupled protein A was added to capture the antibody-protein complexes by rotating for 2 h at 4°C. The supernatants were then removed by centrifuging at 2,000 ϫ g for 5 min. Total RNA was isolated from the pellets using TRIzol reagent. Reverse transcription-PCR was performed as described earlier. A parallel immunoprecipitation with rabbit IgG was performed as a control.

Expression of Tissue Factor Gene in Breast Cancer Cells-
Previous studies reported that tissue factor is highly expressed in breast cancer cells having high invasive potential (11,27). To understand whether tissue factor is expressed differently among breast cancer cell lines, we examined its expression in five human breast cancer cell lines representing highly invasive and less invasive phenotypes. Reverse transcription-PCR assay revealed that tissue factor mRNA is detectable in all cell lines (Fig. 1A) tested, although it is expressed at relatively higher levels in MDA-MB-231 and BT-20 cells (highly invasive lines (28)). Western blot analysis demonstrated that tissue factor is barely detectable in MCF-7, T47D, and ZR-75-1 cells (less invasive cell lines (28)). However, tissue factor is highly expressed in more invasive breast cancer lines, MDA-MB-231 and BT-20 (Fig. 1B). These observations are consistent with previous reports (11,27). A reporter gene assay using a tissue factor gene promoter construct, hTF-P2kb (22), indicated that tissue factor transcription is indeed activated in MCF-7 cells by 1 M phorbol myristate acetate (PMA, Fig.  1C), suggesting that transcription regulation of the tissue factor gene is achievable in this cell line. A cDNA construct was established in which the intact tissue factor coding sequence (the 3Ј-UTR was not included) was fused to the C terminus of the YFP cDNA. Interestingly, when this construct was tran- siently introduced into the cells, tissue factor was found to be equally expressed in MCF-7 and MDA-MB-231 cells as observed by reverse transcription-PCR and confocal microscopy (Fig. 2). These observations suggest that the different expression patterns of the tissue factor gene in breast cancer cells may be due to post-transcriptional regulation involving the 3Ј-UTR of the tissue factor transcript.

3Ј-UTR of Tissue Factor Transcript Regulates Tissue Factor Expression in Breast Cancer
Cells-It has been established that microRNAs regulate gene expression by targeting the 3Ј-UTR of the transcripts, leading to suppression of protein translation or rapid degradation of mRNAs (15). The potential interaction of microRNAs with the 3Ј-UTR of a particular gene transcript can be predicted through established web programs (29). We used a well established microRNA-target predicting program developed at the Memorial Sloan-Kettering Cancer Center to predict microRNAs that may interact with the 3Ј-UTR of tissue factor transcript. As shown in Table 1, two regions of the tissue factor 3Ј-UTR were found to match the seed sequences (30) of miR-19, miR-106b, and miR-20 (Table 1). To experimentally confirm our prediction, the 3Ј-UTR of the tissue factor transcript was cloned into the PGL3 luciferase promoter reporter construct both in a sense and antisense orientation downstream of the luciferase cDNA sequence (TF-3Ј-UTR-S and TF-3Ј-UTR-A) (Fig. 3A). The TF-3Ј-UTR-A construct served as a control. As shown in Fig.  3B, inclusion of the 3Ј-UTR of the tissue factor transcript led to a significant reduction of luciferase activity in MCF-7 cells, but not in MDA-MB-231 cells, indicating that tissue factor expression is regulated by its own 3Ј-UTR in MCF-7 cells (Fig. 3B).
miR-19 Regulates Tissue Factor Expression-To further confirm our prediction, we generated reporter constructs with the deletion of the two microRNA-binding sites, respectively, from the TF-3Ј-UTR-S construct (Fig. 3C). These constructs, along with the TF-3Ј-UTR-S, were transfected into MCF-7 and MDA-MB-231 cells, and luciferase activity was analyzed. As shown in Fig. 3D, when the miR-19-binding site was deleted from the 3Ј-UTR, the luciferase activity increased to a level comparable with that of the control in MCF-7 cells. However, disruption of the binding site for miR-106b and miR-20 only had marginal effects on the luciferase activity, indicating that miR-19 is primarily responsible for the reduc-

TABLE 1 Potential candidates of microRNA species that might interact with the 3-UTR of human tissue factor mRNA
The sequence for tissue factor mRNA (TF) was obtained from the GenBank TM (NM_001993), and microRNA sequences were obtained from the website of Memorial Sloan-Kettering Cancer Center.

Sequence match MicroRNAs
hsa-miR-20 tion of luciferase activity observed in this model system. In contrast to MCF-7 cells, the luciferase activity was not significantly different in MDA-MB-231 cells transfected with various TF-3Ј-UTR-S reporter constructs (Fig. 3D), consistent with the observation that tissue factor is highly expressed in this cell line. Well established microRNA inhibitors were used to further confirm the regulation of tissue factor expression by miR-19 in MCF-7 cells. Addition of the miR-106b inhibitor slightly altered the luciferase activity in cells transfected with TF-3Ј-UTR-S. However, addition of the miR-19 inhibitor completely reversed the luciferase activity (Fig. 4A), strongly supporting the conclusion that miR-19 primarily regulates tissue factor expression in less invasive breast cancer cells. This was further confirmed by the observation that endogenous tissue factor protein expression was induced by the miR-19 inhibitor but not the miR-106b inhibitor (Fig. 4B) as measured by Western blot. Interestingly, tissue factor mRNA content was not significantly altered by the miR-19 inhibitor (Fig. 4B), suggesting that miR-19 mainly suppresses protein translation of the tissue factor gene. Using a commercially available microRNA mimic, we found that overexpression of miR-19 significantly down-regulates tissue factor expression in MDA-MB-231 cells, as analyzed by reporter gene assay and Western blot analysis (Fig. 5). Again, the miR-19 mimics did not alter tissue factor mRNA levels (Fig. 5C), consistent with the results obtained using the miR-19 inhibitor. miR-19 has been shown to be expressed at a higher level in MCF-7 cells (31,32). Consistent with previous reports, we also found that miR-19 is highly expressed in MCF-7 cells versus MDA-MB-231 cells as analyzed by real time PCR technique (Fig. 6A). Because Ago2 has been established as an essential mediator of microRNA binding to the RNA-induced silencing complex and the 3Ј-UTR of relevant transcripts (33,34), we immunoprecipitated Ago2 from cell lysates prepared from MCF-7 and MDA-MB-231 cells and analyzed tissue factor 3Ј-UTR expression in the precipitates. In contrast to the tissue factor mRNA level in nonprecipitated samples, a higher level of tissue factor mRNA, relative to GAPDH, was detected   in MCF-7 versus MDA-MB-231 cells, indicating that Ago2 targets tissue factor 3Ј-UTR preferentially in MCF-7 cells, as compared with MDA-MB-231 cells. Ago2 protein is expressed in both cell lines (Fig. 6B).

DISCUSSION
Proper regulation of tissue factor expression is critical for homeostasis. The tissue factor gene is classified as an immediate early response gene that is a target of many regulatory influences. Transcription regulation of the tissue factor gene has been well characterized, which is associated with DNA consensus binding sites that have been found within the tissue factor gene promoter (35). We have confirmed previous observations that expression of the tissue factor gene is barely detectable at the protein level in less invasive breast cancer cells yet is highly expressed in invasive breast cancer cells (11,27). The lower expression of the tissue factor gene in MCF-7 cells does not seem to be related to abnormal transcription suppression of the gene, as our results from the tissue factor gene promoter studies showed that the promoter is active and can be regulated through the MAPK pathway in this cell line, observations similar to a previous report (27). Furthermore, tissue factor mRNA is detectable in MCF-7 cells, and the genome hypomethylation pattern is similar between MCF-7 and MDA-MB-231 cells (36), implying that the tissue factor gene is not transcriptionally silenced in MCF-7 cells. The fact that YFP-TF was equally expressed in both MCF-7 and MDA-MB-231 cells indicates that there is no repression mechanism present in these cells targeting the coding sequence of the tissue factor gene. These observations led to the hypothesis that post-transcriptional regulation of tissue factor expression in less invasive breast cancer cells involves the 3Ј-UTR of the gene transcript. This hypothesis was confirmed by the experiments presented in this study.
Tissue factor gene expression has been shown to be regulated at a post-transcriptional level in cultured fibroblasts and monocytic cells (37,38), yet the detailed mechanisms remain unclear. Several lines of experimental evidence from this study support the conclusion that miR-19 regulates tissue factor expression at a post-transcriptional level in breast cancer cells. First, bioinformatics analysis using a well established program indicated that several microRNAs have a perfectly matched seed sequence with two regions of the 3Ј-UTR of the tissue factor transcript. This provides a rationale (29) to look into the potential regulation of tissue factor expression by these microRNA species. Second, we demonstrated that the 3Ј-UTR of the tissue factor transcript down-regulates this gene expression in MCF-7 but not MDA-MB-231 cells as assayed with the reporter gene technique. The down-regulation of the reporter gene activity was completely reversed when the miR-19-binding site in the 3Ј-UTR was removed, which provides strong evidence indicating that miR-19 primarily regulates tissue factor expression by targeting the 3Ј-UTR. Third, when the miR-19 inhibitor was applied, endogenous tissue factor expression was enhanced and the down-regulation of the reporter activity was reversed in MCF-7 cells. On the other hand, the miR-19 mimics reduced tissue factor expression in MDA-MB-231 cells, a cell line constantly ex-pressing a high level of tissue factor protein. Finally, the experiments with Ago2 immunoprecipitation indicated that the tissue factor 3Ј-UTR is favored by Ago2 protein binding in MCF-7 cells versus MDA-MB-231 cells. This is consistent with the result showing that miR-19 is expressed at a higher level in MCF-7 cells. Thus, the role of miR-19 in regulating tissue factor expression in our model systems is clearly demonstrated.
Previous observations also indicated that the miR-17-92 cluster, of which miR-19 is a member, is expressed at higher levels in MCF-7 cells compared with MDA-MB-231 cells (32), and miR-19 is expressed at higher levels among the members of this cluster in MCF-7 cells (31). Because the tissue factor mRNA levels were not significantly affected by miR-19 inhibitor in MCF-7 cells nor by miR-19 mimics in MDA-MB-231 cells, the main action of miR-19 seems to inhibit protein translation of the tissue factor gene in less invasive breast cancer cells. Repression of protein translation is one of the two well established mechanisms of microRNA action in suppressing gene expression (16). This mechanism of action, the post-transcriptional regulation of tissue factor expression by miR-19, could well explain the selective expression of the tissue factor gene in highly invasive breast cancer cells.
The up-regulation of the miR-17-92 cluster has been observed in human cancer cells, including breast cancer (39,40). These microRNAs are considered as oncogenes (41)(42)(43), which promote proliferation, inhibit apoptosis, and induce tumor angiogenesis (42,44). However, the contribution of each individual microRNA in the cluster to tumor progression remains to be characterized. As tissue factor is involved in tumor angiogenesis and metastasis (1), our finding that miR-19 regulates tissue factor expression in breast cancer cells provides novel information on the mechanisms of microRNA-induced tumor progression. The potential effect of miR-19 on breast cancer invasiveness and angiogenesis could be directly tested in the relevant model systems. In a more broad sense, regulation of tissue factor expression by microRNAs could be relevant to the advancement of homeostatic biology in which cell-specific expression of tissue factor plays a pivotal role (35). Model systems such as monocytes or endothelial cells may be used to further investigate the regulation of tissue factor expression by microRNAs.