c-MYC-regulated miR-23a/24-2/27a Cluster Promotes Mammary Carcinoma Cell Invasion and Hepatic Metastasis by Targeting Sprouty2*

Background: The regulation of the miR-23a/24-2/27a cluster is largely unknown. Results: EGF induced c-MYC expression to promote the expression of the miR-23a/24-2/27a cluster, resulting in decreased expression of Sprouty2 and increased activation of p44/42 MAPK to stimulate mammary carcinoma cell invasion and subsequent hepatic metastases. Conclusion: EGF promoted mammary carcinoma cell invasion and hepatic metastasis. Significance: The miR-23a/24-2/27a cluster might be used as a biomarker for breast cancer metastasis. Emerging evidence indicates that the miR-23a/24-2/27a cluster may possess a causal role in mammary tumorigenesis and function as a novel class of oncogenes. However, the regulatory mechanism of the miR-23a/24-2/27a cluster in mammary carcinoma cell invasion and migration is still largely unknown. We observed that the expression levels of miR-23a, miR-24-2 and miR-27a were significantly higher in breast cancer with lymph node metastasis, compared with that from patients without lymph node metastasis or normal tissue. Forced expression of the miR-23a/24-2/27a cluster promoted mammary carcinoma cell migration, invasion, and hepatic metastasis, through targeting Sprouty2 (SPRY2) and consequent activation of p44/42 MAPK. Epidermal growth factor induced the expression of the transcription factor c-MYC, which promoted the expression of mature miR-23a, miR-24-2, and miR-27a and subsequently decreased expression of SPRY2 and activated p44/42 MAPK to promote mammary carcinoma cell migration and invasion. We therefore suggest a novel link between epidermal growth factor and the miR-23a/24-2/27a cluster via the regulation of c-MYC, providing the potential for the miR-23a/24-2/27a cluster to be used as biomarker in the diagnosis and/or treatment of breast cancer.

MicroRNAs (miRNAs) 4 are small noncoding RNAs that regulate the translation of protein-coding genes by repressing translation of protein-coding mRNA or enhancing mRNA degradation (1-7, 9 -15). They are predicted to modulate the expression levels of at least one-third of all human proteincoding genes (16,17). Current target prediction computer programs (18,19) predict that one specific miRNA may target tens to hundreds of genes. Thus, one can expect that miRNAs play important roles in coordinating many cellular processes (1, 2, 20 -22). Deregulation of miRNAs has been reported to modulate normal cell growth and differentiation, potentially leading to a variety of disorders, including cancer (23)(24)(25)(26). Thus, the identification of miRNAs that are associated with pathology provides new approaches for understanding disease processes (10,27).
The miR-23a/24-2/27a cluster is an miRNA cluster, existing intergenically in the vertebrate genome (28). Members of the cluster are involved in cell cycle control and differentiation, in various cell types (29). The cluster has also been suggested to play a role in promoting apoptosis by both caspase-dependent and caspase-independent pathways (30). However, the mechanisms of miR-23a/24-2/27a cluster regulation in cancer progression remain poorly understood. Only a few target genes for the miR-23a/24-2/27a cluster have thus far been identified. Sprouty2 (SPRY2), an inhibitor of the Ras/MAPK pathway, is one of four highly conserved family members of Sprouty signal modulatory proteins. SPRY2 is recognized to be deregulated in various types of cancers, such as breast, liver, and prostate cancer, among others (31,32). SPRY2 might be an important modulator of pathways central to cancer progression, including cell growth, migration, and invasion (33)(34)(35).
In this report, we observed that the expression levels of members of the miR-23a/24-2/27a cluster were significantly higher in mammary carcinoma with lymph node metastasis compared with that from patients without lymph node metastasis or normal tissue. We further described the mechanism by which the miR-23a/24-2/27a cluster contributed to mammary carcinoma cell migration and invasion. We demonstrated that epidermal growth factor (EGF) induced the expression of c-MYC, which increased the expression of mature miR-23a, miR-24-2, and miR-27a; subsequently decreased the expression of the target gene, SPRY2; and promoted cell migration and invasion through activation of p44/42 MAPK.

EXPERIMENTAL PROCEDURES
Cell Culture-All human breast cancer cell lines used in this study were obtained from the American Type Culture Collection (Manassas, VA) and cultured in conditions as recommended. All cells were maintained in a humidified incubator at 37°C and 5% CO 2 .
Reagents-Media, sera, and antibiotics for cell culture were from Invitrogen. Protein electrophoresis reagents were from Bio-Rad. Iressa (Gefitinib) was purchased from Tocris Bioscience (Ellisville, MO). U0126 was purchased from Calbiochem. All other chemicals were from Sigma-Aldrich.
Patients and Specimens-The female patient population consisted of 14 consecutive non-invasive mammary ductal carcinoma patients, 14 consecutive invasive mammary ductal carcinoma patients, and 13 consecutive normal mammary (benign) tissue patients who underwent surgery at the First Affiliated Hospital of Anhui Medical University (Hefei, Anhui, China) between 2009 and 2010. The pathohistological diagnosis of the specimens was consistent with breast neoplasm in accordance with World Health Organization guidelines (36). Histology grade was based on the Scarff-Bloom-Richardson system (37). The protocol for the use of tissue samples from patients and follow-up study was approved by the Institutional Review Board, and patients signed informed consent. Our study procedures were in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
Transwell Migration and Invasion Assay-Assays were performed in BioCoat Matrigel invasion chambers (Corning Costar, Acton, MA) as described previously (38). Values for cell migration or invasion were expressed as the average number of cells per microscopic field.
ChIP Assay-Chromatin immunoprecipitation was performed using the EZ ChIP kit (Upstate Biotechnology, Inc., Lake Placid, NY) as described previously (40). Immunoprecipitations were performed using 2 g each of anti-c-MYC (sc-56505, Santa Cruz Biotechnology, Inc.), and anti-mouse IgG (EZ-ChIP TM ) overnight at 4°C with rotation.
Plasmid Constructs and Mutagenesis-The miR-23a/24-2/ 27a gene promoter expression plasmids were a kind gift from Dr. V. Narry Kim (Seoul National University). miR-23a, -24-2, and -27a precursors, including up-and downstream 80 -100-nt sequences, were amplified from human genomic DNA. The miRNA expression vector was created by inserting the PCR products downstream of the U6 promoter of the pll3.7 plasmid followed by 5T. To construct the luciferase reporter vector, the psiCHECK-2 plasmid (Promega) was modified by introducing a new Eco72I site at the multiple cloning regions downstream of the XhoI site. The 3Ј-untranslated region (UTR) sequence of SPRY2 was amplified with primers 5Ј-CCGCTCGAGGAA-AGGTGGGATAGTCTTTGCTGT-3Ј (forward) and 5Ј-TAT-GCATCTGTAACCCCTCATTTG-3Ј (reverse), and it was cloned into the XhoI and Eco72I sites downstream of the Renilla luciferase reporter gene of the modified psiCHECK-2 plasmid. Site-directed mutagenesis was performed by using the QuikChange II XL site-directed mutagenesis kit (Stratagene, La Jolla, CA). For the 3Ј-UTR of SPRY2, the seed sequence of miR-23a was changed from AATGTGA to TTACACT, miR-24-2 TGTGCCG was substituted by ACGCGGT, and miR-27a was changed from ACTGTGAA to TGACACTT.
Generation of Lentiviruses Expressing MicroRNA-Generation of lentivirus vectors was performed by cotransfecting pLL3.7 carrying the miRNA expression cassette with helper plasmid pCMV-VSV-G (envelope) and pHRЈ 8.9⌬VPR (core protein) in 293T cells using Effectene Transfection Reagent (Qiagen). The viral supernatant was collected 48 h after transfection.
Tumor Xenograft Studies-All studies were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. An institutional animal care and use committee approved all animal protocols. The procedure was carried out as described previously (39). The 5-week-old female BALB/c nude mice (Hunan SJA Laboratory Animal Co., Ltd.) were used for studies. The 5-week-old female BALB/c nude mice (Hunan SJA Laboratory Animal Co., Ltd.) were injected with 4.5 ϫ 10 6 cells via tail vein. When animals miR-23a/24-2/27a Promotes Breast Cancer Hepatic Metastasis were sacrificed, lungs and livers were harvested and fixed overnight in 4% paraformaldehyde, embedded in paraffin, and cut into 10-m-thick sections for histological studies and real-time PCR analysis for hHPRT to evaluate metastasis (41).
Statistical Analysis-All experiments were performed at least three times. All statistical analyses were performed using the SPSS software system for Windows (version 13.0; SPSS, Chicago, IL). The 2 test was used to analyze the difference in miRNA expression levels among breast cancer and normal samples. Numerical data are expressed as mean Ϯ S.E. from a representative experiment performed in triplicate. The statistical significance of each correlation was determined using the 2 test, and p Ͻ 0.05 was considered statistically significant.

miR-23a/24-2/27a Cluster Expression Is Increased in Mammary Carcinoma-
The biological roles of the miRNAs encoded by the miR-23a/24-2/27a cluster in cancer progression are yet to be fully delineated. To define the potential role of the miR-23a/24-2/27a cluster in the progression of mammary carcinoma, we performed quantitative real-time PCR analysis for expression of the members of the miR-23a/24-2/27a cluster on a panel of frozen mammary carcinoma samples. We observed that the expression levels of miR-23a, miR-24-2, and miR-27a were significantly higher in breast cancer with lymph node metastasis, compared with that from patients without lymph node metastasis or normal tissue (Fig. 1, A-C). Additionally, we observed a high correlation coefficient for expression of the individual members of the miR-23a/24-2/27a cluster (Fig. 1D), indicative of cluster co-expression in breast cancer.

miR-23a/24-2/27a Cluster Promotes Mammary Carcinoma
Cell Migration and Invasion in Vitro-Given the positive correlation between the level of each individual miRNA in the cluster and metastasis in patients with mammary carcinoma, we next assessed the potential impact of the miR-23a/24-2/27a cluster on mammary carcinoma cell migration and invasion, through manipulation of the expression level of members of the miR-23a/24-2/27a cluster by transfection of either ASO or synthetic miRNA mimic.
miR-23a/24-2/27a Cluster Initiates Distant Metastasis in Vivo-We further determined whether expression of the members of the miR-23a/24-2/27a cluster would promote metasta- injected with MCF-7 cells with lentivirus expressing miR-27a by H&E staining (data not shown) 8 weeks post-tail vein injection, whereas no metastases were detected in the livers of control mice. Furthermore, the burden of micrometastases was determined by quantitative real-time PCR analysis for the expression of human HPRT (hHPRT). We observed that the expression levels of hHPRT were significantly higher in the livers of all mice injected with MCF-7 cells with lentivirus expressing miR-23a, miR-24-2, or miR-27a, compared with the livers of control mice (Fig. 2H). In comparison, there is no significantly higher expression of hHPRT in the lungs of mice injected with MCF-7 cells with forced expression of miR-23a, miR-24-2, or miR-27a, compared with the levels in the lungs of the control mice (Fig. 2I), indicative that the miR-23a/24-2/27a cluster could specifically promote breast cancer metastasis to liver. SPRY2 Is Directly Targeted by the miR-23a/24-2/27a Cluster via Its 3Ј-UTR-To determine the mechanism by which these miRs promote migration and invasion, we sought to determine specific targets for this cluster of miRs. Four bioinformatics software programs, PicTar, TargetScan, Rna22, and miRanda (1,2), revealed that all three miRs have putative binding sites in the 3ЈUTR of SPRY2 that are evolutionarily conserved among vertebrate species (42).
The predominant mechanism of miRNA action is thought to be degradation of mRNA or inhibition of translation (1). We therefore investigated the effect of a control oligonucleotide and members of the miR-23a/24-2/27a cluster on repression of endogenous SPRY2 mRNA and protein. We observed that both SPRY2 mRNA (Fig. 4, C and G) and protein (Fig. 4, E, F, I, and J) were selectively decreased in the presence of miR-23a, miR-24-2, or miR-27a. In contrast, both SPRY2 mRNA (Fig. 4, D and  H) and protein (Fig. 4, E, F, I, and J) were increased after transfection of ASO for miR-23a, miR-24-2, or miR-27a, respectively, indicating SPRY2 is a bona fide target of the miR-23a/24-2/27a cluster. Thus, SPRY2 expression is directly controlled through 3Ј-UTR regulation by each miRNA of the miR-23a/24-2/27a cluster.
c-MYC Transcriptionally Up-regulates the miR-23a/24-2/ 27a Cluster-We next sought to delineate regulation of the expression of the miR-23a/24-2/27a cluster. Bioinformatics analysis revealed that one putative c-MYC regulatory element is present in the promoter of the miR-23a/24-2/27a cluster. Hence, we investigated whether c-MYC drove the expression of the miR-23a/24-2/27a cluster. To determine this possibility, we used an adenovirus to force the expression of c-MYC in MCF-7 cells. Real-time PCR analysis showed that the mature miR-23a, -24-2, -27a, and primary miR cluster transcripts were significantly increased above control levels, respectively (Fig. 5A). Together, these findings demonstrate that c-MYC increased the expression of the miR-23a/24-2/27a cluster in mammary carcinoma cells.

c-MYC Acts on an E-Box to Drive miR-23a/24-2/27a
Cluster Expression-Having demonstrated that c-MYC increased the expression of the miR-23a/24-2/27a cluster, we next searched for response elements (the recognition site of c-MYC) in the miRNA cluster promoter and observed that a region containing an E-box (a promoter element known to be bound by c-MYC) was present. This region was isolated from genomic DNA by PCR and cloned upstream of a luciferase reporter gene. Reporter assays demonstrated that luciferase activity was increased by c-MYC in cells transfected with the E-box-luciferase construct (PL842, PL639, and PL439); however, no increase in luciferase activity was observed in cells transfected with the pGL3-Basic plasmid lacking the E-box (Fig. 5B).
To demonstrate whether c-MYC occupied the promoter of the miR-23a/24-2/27a cluster, we performed chromatin immunoprecipitation assays for c-MYC binding sites. We observed a significant enrichment of the miR-23a/24-2/27a cluster promoter amplicon in c-MYC ChIP samples as compared with ChIP samples generated with control IgG (Fig. 5C).
A previous report (44) has demonstrated that c-MYC suppresses invasion and metastasis of mammary carcinoma cells, including MCF-7 cells. We further determined the possible role of c-MYC in breast cancer cell migration and invasion. Surprisingly, forced expression of c-MYC resulted in significantly increased migration and invasion in a number of breast cancer cell lines, including MCF-7 cells, compared with the control (Fig. 5, D and E) (data not shown). In contrast, depletion of c-MYC expression by siRNA resulted in a significant reduction in cell migration (Fig. 5F) and invasion (Fig. 5G). It is therefore apparent that forced expression of c-MYC enhances MCF-7 mammary carcinoma cell migration and invasion.

cells (E)/T47D cells (I) were grown and transfected with the individual miR-23a/24-2/27a
cluster mimics or miR-23a/24-2/27a cluster ASOs or scrambled sequence oligonucleotides as negative control. Total cellular protein was isolated and subjected to Western blot analysis for SPRY2 expression. ␤-Tubulin (E)/␤-actin (I) was used as loading control. F and J, quantitative result of Western blot. K, genomic location and sequence analysis of the miR-23a/24-2/27a cluster interactive sites with SPRY2. The cluster is present on chromosome 19. MRE, miRNA regulatory element. L-N, luciferase reporter assays. Relative luciferase activity was analyzed after luciferase reporter plasmids with SPRY2 3Ј-UTR constructs (either wild-type or mutant) or control reporter plasmid were cotransfected in MCF-7 cells with miR-23a (L), miR-24-2 (M), or miR-27a (N) mimic or control mimic, respectively. O, luciferase reporter assay. Cells were transfected with a reporter vector psiCHECK2-SPRY2 3Ј-UTR or psiCHECK2 plus either mimics of the members of the miR-23a/24-2/27a cluster or scrambled sequence oligonucleotides as negative control. Luciferase reporter assays were performed, and Renilla luciferase activity was normalized according to the internally controlled firefly luciferase activity. P, verification of the efficacy of SPRY2 siRNA by qPCR. Q and R, transwell assays. MCF-7 cells were grown and transfected with SPRY2 siRNA, individual miR-23a/24-2/27a cluster ASO, or individual miR-23a/24-2/27a cluster ASO plus SPRY2 siRNA or control scrambled siRNA duplex. These cells were subsequently subjected to migration (Q) and invasion (R) assays. S, Western blot. MCF-7 cells were grown and transfected with the individual miR-23a/24-2/27a cluster mimics/miR-23a/24-2/27a cluster ASOs or scrambled sequence oligonucleotides as negative control. Total cellular protein was isolated and subjected to Western blot analysis for phospho-p44/42 MAPK. Total MAPK was used as an internal control. T, quantitative result of Western blot. U and V, MCF-7 cells were grown and transiently transfected with the individual miR-23a/24-2/27a mimics or scrambled sequence oligonucleotide as negative control for 2 days and treated with 100 ng/ml UO126 or vehicle for 45 min. Cells were subsequently subjected to migration (U) and invasion (V) assays. *, p Ͻ 0.01. Error bars, S.E.
Additionally, reporter assays demonstrated that luciferase activity from the miR-23a/24-2/27a cluster promoter was significantly increased by EGF; however, no EGF-dependent increase in luciferase activity was observed in cells transfected with the pGL3-Basic plasmid lacking the promoter sequences (Fig. 6J). Thus, the miR-23a/24-2/27a cluster is modulated by EGF in vitro.

DISCUSSION
There is a paucity of reports on the mechanisms of miR-23a/ 24-2/27a cluster regulation in cancer progression. In this study, the expression of miR-23a/24-2/27a cluster members was examined in clinical specimens. Strikingly, we observed that the expression levels of the members of the miR-23a/24-2/27a cluster were significantly higher in breast cancer with lymph node metastasis, compared with that from patients without lymph node metastasis or normal tissue, suggesting that the miR-23a/24-2/27a cluster may be linked to metastasis in mammary carcinoma. We further demonstrated a role of the miR-23a/24-2/27a cluster members in promoting mammary carcinoma cell migration and invasion. The mechanistic insight provided by the clinical correlation was enhanced further by in vitro and in vivo studies. The finding herein, therefore, assists in understanding the functional role of the miR-23a/24-2/27a cluster in breast cancer progression and may provide new understanding of the mechanism of neoplastic progression. The miR-23a/24-2/27a cluster is expected to possess other functional roles in addition to our finding here. For example, functional activity of the miR-23a/24-2/27a cluster during osteoblast differentiation is indicative of a central role for the miR-23a/24-2/27a cluster in both progression and maintenance of the osteocyte phenotype (49). Also, miR-23a/24-2/27a cluster members are enriched in endothelial cells and highly vascularized tissues, suggesting that increased miR-23a/24-2/ 27a cluster expression may possess important function in neovascular age-related macular degeneration or tumor-related angiogenesis (28).
A prior report has indicated that SPRY2 was down-regulated in hepatocellular carcinoma (32). Interestingly, we also observed that the miR-23a/24-2/27a cluster members pro-moted micrometastases to murine liver but not to lung (Fig. 2,  H and I), indicative that this cluster of miRNAs might be specifically modulated in hepatic specific metastases.
In this study, we demonstrated that EGF, a major prometastatic growth factor in breast cancer (51)(52)(53)(54), induced the expression of c-MYC (55-59) that transcriptionally up-regulated the miR-23a/24-2/27a cluster, subsequently decreasing the expression of SPRY2. Decreased SPRY2 in turn increased the activity of p44/42 MAPK, promoting mammary carcinoma cell migration and invasion. However, it has previously been reported that c-MYC decreases the transcription of several miRNAs, including all members of the miR-23a/24-2/27a cluster, in the human p493 B cell line, bearing a tetracycline-repressible MYC construct, wherein tetracycline withdrawal results in rapid induction of c-MYC and mitochondrial biogenesis (60,61). We have, however, been able to repeat these results with the same p493 cell lines herein (data not shown). Interestingly, a recent report has demonstrated that c-MYC suppresses invasion and metastasis of mammary carcinoma cells (44). However, we could not repeat these results with the strains of MDA-MB-231 and BT549 cells in our possession (data not shown). Our data suggest that c-MYC exerts an alternate functional role in a highly context-dependent manner in the cell lines of different origin. We propose that EGF stimulates the expression of c-MYC to promote gene transcription of miR-23a/24-2/27a cluster in mammary carcinoma cells and further triggers the activation of p44/42 MAPK via repressed expression of SPRY2 to promote mammary carcinoma cell invasion and metastasis (Fig. 7).
Prior studies have reported that miR-23a and/or miR-27a can target SPRY2 in pancreatic cancer cells (62) or endothelial cells (28) without an overall examination of the functionality of miR-23a/24-2/27a cluster. In this study, we have defined the mechanism of regulation of the whole miR-23a/24-2/27a cluster, affecting both migration and invasion by specifically targeting different sites of SRPY2 3Ј-UTR in a mammary carcinoma cell. Furthermore, we have now also defined a specific and coordinated disease association (metastatic mammary carcinoma) for this miRNA cluster.
In this study, we demonstrated that the miR-23a/24-2/27a cluster enhanced MCF-7 cell migration and invasion but not proliferation. This might be due to distinct functions of the MAPK pathway during tumor progression (63). From another FIGURE 6. EGF regulates the expression of miR-23a/24-2/27a cluster partly via c-MYC in vitro. A, migration assay; B, invasion assay. MCF-7 cells were grown and transiently transfected with individual miR-23a/24-2/27a cluster ASO or scrambled sequence oligonucleotide as negative control (N.C.) for 2 days. Serumdeprived cells were subsequently treated with 20 ng/ml EGF or vehicle for 4 h and subjected to migration and invasion assays. C, Western blot. MCF-7 cells were grown and transiently transfected with the individual miR-23a/24-2/27a cluster ASOs or scrambled sequence oligonucleotides as negative control for 2 days. Serum-deprived cells were subsequently treated with 20 ng/ml EGF or vehicle for 4 h. Total cellular protein was isolated and subjected to Western blot analysis for phospho-p44/42 MAPK expression. Total MAPK was used as an internal control. D, the quantitative result of Western blot is shown. E, real-time PCR validated the expression of the miR-23a, miR-24-2, and miR-27a transcripts, respectively. F-H, Iressa repressed expression of the miR-23a/24-2/27a cluster. Serumdeprived MCF-7 cells were treated with Iressa for 30 min, and the expression of the miR-23a (F), miR-24-2 (G), and miR-27a (H) transcripts was analyzed by qPCR. U6 snRNA was used as loading control. I, real-time PCR validated that the expression of the miR-23a/24-2/27a cluster precursors was increased at all time intervals after treatment with EGF. J, EGF regulated the promoter activity of the miR-23a/24-2/27a cluster. For the luciferase reporter assay, MCF-7 cells were transfected with the luciferase reporter plasmid carrying various length promoter constructs (i.e. empty vector, 2046 bp, 1050 bp, 639 bp, and 239 bp) as indicated. Luciferase reporter gene activity normalized by Renilla/luciferase activity was analyzed 24 h after treatment with EGF. Basal luciferase activity (unstimulated sample with empty vector) of the reporter construct was defined as 100%. The data are presented as means Ϯ S.E. from three independent experiments, each performed in triplicate. K, verification of the efficacy of c-MYC siRNA by qPCR. L-O, real-time PCR demonstrated the expression of the miR-23a, miR-24-2, and miR-27a transcripts and primary miR (Pri-miR) cluster, respectively. MCF-7 cells were transfected with c-MYC siRNA for 24 h and subsequently treated with EGF for 4 h. *, p Ͻ 0.01. Error bars, S.E. angle, multiple target proteins of this cluster of miRs, which could be either growth-promoting or growth-suppressive, might simultaneously affect cell growth, depending on the particular microenvironment, to exert differential cellular events (64,65). Furthermore, the proliferation and migration/invasion may be functionally uncoupled, even mutually exclusive in certain cellular models as suggested (8,66).
It should be noted that the miR-23a/24-2/27a cluster could be expected to possess a wide range of functionalities due to its pleiotrophic regulation of additional genes in the progression of breast cancer. The cooperative expression and functionality of the members of the miR-23a/24-2/27a cluster has yet to be fully defined (50). Further studies are therefore required to fully understand these functional roles and interactions of this miRNA cluster in breast cancer.