MicroRNA-155 Regulates Cell Survival, Growth, and Chemosensitivity by Targeting FOXO3a in Breast Cancer*

Breast cancer is the second leading cause of cancer death in women. Despite improvement in treatment over the past few decades, there is an urgent need for development of targeted therapies. miR-155 (microRNA-155) is frequently up-regulated in breast cancer. In this study, we demonstrate the critical role of miR-155 in regulation of cell survival and chemosensitivity through down-regulation of FOXO3a in breast cancer. Ectopic expression of miR-155 induces cell survival and chemoresistance to multiple agents, whereas knockdown of miR-155 renders cells to apoptosis and enhances chemosensitivity. Further, we identified FOXO3a as a direct target of miR-155. Sustained overexpression of miR-155 resulted in repression of FOXO3a protein without changing mRNA levels, and knockdown of miR-155 increases FOXO3a. Introduction of FOXO3a cDNA lacking the 3′-untranslated region abrogates miR-155-induced cell survival and chemoresistance. Finally, inverse correlation between miR-155 and FOXO3a levels were observed in a panel of breast cancer cell lines and tumors. In conclusion, our study reveals a molecular link between miR-155 and FOXO3a and presents evidence that miR-155 is a critical therapeutic target in breast cancer.


MicroRNAs (miRs)
are short single-stranded RNAs that have become known as important regulators of various cellular processes by controlling gene expression at the post-transcriptional level (1)(2)(3)(4). Deregulated microRNAs in cancer function as either tumor suppressors (5)(6)(7) or oncogenes (8 -11) and play a central role in carcinogenesis. Accumulating evidence shows that miR-155 is an oncogenic microRNA. First, microRNA profiling studies indicated frequent increase of miR-155 in various types of human malignancy, including different forms of B cell lymphoma and carcinoma of breast, lung, colon, head/neck, and kidney (9,(11)(12)(13)(14)(15)(16). Recent studies have demonstrated association of elevated miR-155 with late stage and poor overall survival in several types of malignancy (17)(18)(19). We have previously shown that miR-155 is induced by transforming growth factor-␤ and plays an important role in epithelial-to-mesenchymal transition and demonstrated frequent overexpression of miR-155 in invasive breast cancer (20). Further, a possible link between miR-155 and inflammation in cancer has been reported (21). Moreover, miR-155 transgenic mice develop B-cell lymphoma (22), and miR-155-knock-out mice exhibit impaired immune function (23). FKHRL1 (FOXO3a) is a major member of the forkhead transcriptional factor family (24). Members of this family are characterized by a distinctive forkhead DNA binding domain, which is negatively regulated by protein kinases Akt, SGK, and IKK␤ (25)(26)(27)(28). The phosphorylation of FOXO3a by these kinases leads to its translocation from the nucleus to the cytoplasm and loss of the proapoptotic function. In the unphosphorylated active form, FOXO3a resides in the nucleus and induces cell death by up-regulation of apoptotic proteins, such as BIM, p27, BNIP3, and 24p3 (29 -32) and repression of antiapoptotic molecule FLIP and BCL-XL (33,34).
To date, the majority of validated miR-155 functional biology and protein targets define the importance of miR-155 in immunology (23,(35)(36)(37)(38)(39) and various forms of lymphoma (9,11,40,41); however, the role of miR-155 in human breast cancer remains elusive. Here, we report that miR-155 induces cell survival and plays an important role in chemoresistance in breast cancer. Its anti-apoptotic function is mediated by direct inhibition of FOXO3a. Thus, our findings not only demonstrate regulation of FOXO3a at post-transcriptional levels but also identify miR-155 as a critical therapeutic target in breast cancer.

EXPERIMENTAL PROCEDURES
Cell Lines and Breast Tumor Specimens-Breast cancer cell lines were obtained from ATCC and grown according to ATCC recommended culture conditions. All 77 primary and 38 recurrent (e.g. tumors reoccur in breast, bone, lung, and liver after surgery and radiation) human breast cancers and 11 normal breast specimens were obtained from patients who underwent surgery at H. Lee Moffitt Cancer Center and approved by the institutional review board. The stage and grade of primary breast tumors were as follows: 9 cases with stage I disease, 38 stage II, 22 stage III, and 8 stage IV as well as 28 low grade (e.g. I and II) and 49 high grade (e.g. III and IV) diseases. The histology of the primary breast cancer included 67 infiltrating ductal carcinomas, 5 ductal and lobular carcinomas, 2 mucinous adenocarcinomas, and 3 lobular carcinomas. Of primary and recurrent breast cancers, 71 were ER␣-positive, 17 were triple (ER␣/ PR/Her2)-negative, 11 were PR/Her2-positive, and 5 were PR-positive/Her2-negative, and 11 patients presented with unknown PR and Her2 status. Each cancer specimen contained at least 80% tumor cells, as confirmed by microscopic examination. Tissues were either preserved by snap-freeze and stored at Ϫ80°C or embedded in paraffin blocks for subsequent RNA and/or protein extraction.
Plasmids-Lenti-miR-155 and control was purchased from System Biosciences and packaged according to protocol. Hsa-miR-155 and control mimic precursor molecules were purchased from Ambion. miR-155 and control 2-OЈ-methyl antisense oligonucleotides were custom-synthesized from IDT-DNA with sequence as previously described (20). FOXO3a expression vector was kindly provided by Dr. B. M. T. Burgering.
Northern Blot, Locked Nucleic Acid in Situ Hybridization (LNA-ISH), and Immunohistochemical Staining-Total RNA from cell lines and breast cancer and normal tissue was isolated using TRIzol reagent (Invitrogen) according to the manufacturer's protocol. Northern analysis on breast cancer cell lines and tissue RNA were performed using the Starfire TM oligonucleotide-labeling kit (IDT-DNA). The expression of miR-155 was determined by comparing the miR155/U6 ratio between normal mammary and tumor tissues (i.e. the ratio miR155/U6 in normal mammary gland was considered as 1.0, and the cutoff value for overexpression of miR-155 is Ն2-fold in breast tumors). LNA-ISH and immunohistochemistry were performed as described previously (20). Briefly, LNA probes were synthesized complementary to human mature miR-155 (5Ј-CCCCTATCACGATTAGCATTAA-3Ј) and scrambled negative control (5Ј-TTCACAATGCGTTATCGGATGT-3Ј) and digoxigenin-labeled at the 5Ј-end (Exiqon). Immunohistochemical staining was hybridized with anti-FOXO3a antibody. The positive reaction of miR-155 and FOXO3a in LNA-ISH and immunohistochemistry was scored into four grades, according to the intensity of the staining: 0, 1ϩ, 2ϩ, and 3ϩ. The percentages of miR-155-and FOXO3a-positive cells were also scored into four categories: 0 (0%), 1 (1-33%), 2 (34 -66%), and 3 (67-100%). The product of the intensity by percentage scores was used as the final score. The final scores were classified as follows: 0 -4, negative; 5-9, positive. microRNA qRT-PCR Detection and Quantification-Hsa-miR-155 and U6 microRNA levels were detected using the TaqMan microRNA reverse transcription kit (Applied Biosys- tems). Briefly, 200 ng of total RNA from each cell line and tumor RNA were used for primer-specific reverse transcription (RT) in both Hsa-miR-155 and U6, and then 2 l of the RT product was used for subsequent quantitative PCR. The quantitative PCR was performed on an Applied Biosystems 7900HT fast real-time PCR system, and data were collected and analyzed using ABI SDS version 2.3. To calculate relative concentration, miR-155 and U6 C T values for all samples were obtained. A normalized expression for each sample was obtained by dividing the C T value of miR-155 by the same sample's U6 C T and designated as ⌬C T. This value was then transformed by performing 2 Ϫ(⌬CT) . Furthermore, the (⌬⌬C T ) method was used in comparing miR-155 expression in immortalized cells with cancer cells or comparing normal breast with cancer tissues according to the manufacturer's protocol.
Cell Viability and Apoptosis Assays-Optimal drug concentration for induction of apoptosis in BT-474 and Hs578T cell lines was titrated by measuring cleaved caspase activity with Caspase-Glo 3/7 assays (Promega). The following concentrations were used for all subsequent experiments: doxorubicin (2.5 M), paclitaxel (5 nM), and VP-16 (400 M). After infection of miR-155-lentivius (BT-474) or transfection of 2Ј-O-Me anti-miR-155 (Hs578T) and treatment with individual drugs, cell viability was examined with a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay as described previously (42). Apoptosis was detected with the Cell Death Detection ELISAPLUS kit according the manufacturer's protocol (Roche Applied Science), terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling, and Caspase-Glo 3/7 assays (Promega). Each experiment was repeated at least three times in triplicate. The results are expressed as the enrichment factor relative to the untreated controls.
Target in Vitro Luciferase Report Assay-The pMIR-REPORT plasmids for the miR-155 target FOXO3a 3Ј-UTR were constructed as wild type (WT) pmiR-FOXO3a containing two tandem repeats of miR-155 response element from FOXO3a 3Ј-UTR and as mutant (MUT) pmiR-FOXO3a by replacing 2 nucleotides within the "seed sequence" (e.g. 2-8 nucleotides; Fig. 2A). The sequences used to create the pmiR-FOXO3a are as follows: forward, 5Ј-ctagATGAACTTACAG-GTGAGCATTAAATGAACTTACAGGTGAGCATTAA-3Ј; reverse, 5Ј-agctTTAATGCTCACCTGTAAGTTCATTTAA-TGCTCACCTGTAAGTTCAT-3Ј. The oligonucleotides were annealed and inserted into the pMIR-REPORT vector (Ambion). The empty vector (pMIR-REPORT) was used as a negative control. Cells were transfected with 0.2 g of the reporter plasmids, 0.1 g of pCMV-␤-gal, and, where applicable, 5 nM miR-155 precursor or control or 50 nM miR-155 ASO or control per well on 96-well plates. Following 24 h of incubation, cells were subjected to a luciferase reporter assay using the Luciferase Assay System (Promega). Luciferase activities were normalized by ␤-galactosidase activities. Each experiment was repeated at least three times in triplicate.
Statistical Analysis-Statistical significance was analyzed by unpaired Student's t test, and p Յ 0.05 was considered to be statistically significant.  Table 1). To examine the effect of miR-155 on chemosensitivity, we ectopically expressed miR-155 in BT-474 (e.g. low miR-155) and knocked down miR-155 in HS578T (e.g. high miR-155) cells. Briefly, BT-474 and HS578T cells were infected and transfected with lentivirus expressing miR-155 and 2-OЈmethyl antisense-miR-155 oligonucleotides, respectively. The cells infected with lentiviral vector and transfected with scrambled oligonucleotides were used as controls. After 48 h of incubation, expression of miR-155 was confirmed by Northern blot analysis (Fig. 1, B and C). Notably, modulation of miR-155 alone is sufficient to significantly affect cell growth and the programmed cell death (i.e. ectopic expression of miR-155 in BT-474 promotes cell proliferation and cell survival (Fig. 1, D-F; p Ͻ 0.05), whereas depletion of miR-155 in HS578T induces cell growth arrest and apoptosis (Fig. 1, G-I; p Ͻ 0.01)). Moreover, expression of miR-155 renders BT-474 cells resistant to doxorubicin, VP16, and paclitaxel (Fig. 1, E and F). On the other hand, knockdown of miR-155 sensitizes HS578T cells to apoptosis induced by these chemotherapeutic agents (Fig. 1,  H and I). These data indicate that miR-155 is a determinant of chemosensitivity in breast cancer cells.
3Ј-UTR of FOXO3a Interacts with miR-155-To investigate whether miR-155 repression of FOXO3a is mediated by direct interaction of miR-155 with FOXO3a 3Ј-UTR, we cloned two tandem repeats of WT miR-155-FOXO3a response element or MUT into pMIR-REPORT plasmid downstream of luciferase (Fig. 2, A and D). Basal levels of pMIR-FOXO3a reporter activ-
Downstream Targets of FOXO3a Were Inhibited by miR-155-Because FOXO3a is an important transcriptional factor that regulates proapoptotic and growth inhibition genes, we next examined if expression of Bim and p27, two major downstream targets of FOXO3a, is affected by modulation of miR-155 in breast cancer. Immunoblotting analysis with anti-Bim and -p27 antibodies revealed that basal levels of Bim and p27 were reduced by stable expression of miR-155 in BT-474 cells (Fig.  3A, lanes 1 and 4) but increased by knockdown of miR-155 in HS578T (Fig. 3C, lanes 1 and 4). It has been demonstrated that doxorubicin induces/activates FOXO3a, resulting in apoptosis (45). Thus, we further determined the effect of miR-155 on doxorubicin-induced FOXO3a and apoptosis. Fig. 3, A and B, shows that expression of miR-155 considerably reduced the effects of doxorubicin on FOXO3a, Bim, and p27 expression and PARP cleavage and apoptosis. In contrast, depletion of miR-155 enhanced doxorubicin-stimulated FOXO3a, Bim, and p27 expression as well as PARP cleavage and apoptosis (Fig. 3, C  and D). These results further support the findings of induction of cell survival and growth by miR-155 through targeting FOXO3a.
Introduction of FOXO3a cDNA Lacking 3Ј-UTR Largely Abrogates miR-155 Cellular Function-Because miR-155 directly targets FOXO3a through interaction between FOXO3a 3Ј-UTR and miR-155, we reasoned that ectopic expression of FOXO3a by transfection of the cDNA that only contains the open reading frame of FOXO3a (FOXO3a-ORF) should escape the regulation by miR-155 and thus attenuate or decrease miR-155 function. To this end, we transfected pcDNA-FOXO3a-ORF and/or miR-155 into BT-474 cells and treated them with and without doxorubicin for 24 h. Immunoblotting analysis revealed that expression of FOXO3a alone moderately induced expression of Bim and p27 as well as PARP cleavage (Fig. 4A, lane 3), but combination of FOXO3a and doxorubicin considerably enhanced these effects (Fig. 4A, lane  7). In agreement with Fig. 3, expression of miR-155 alone reduced expression of FOXO3a, Bim, and p27 as well as PARP cleavage (Fig. 4A, lanes 2 and 6) compared with the cells transfected with pcDNA3 vector (Fig. 4A, lanes 1 and 5) in the absence and the presence of doxorubicin. However, co-expression of FOXO3a with miR-155 largely reduced the miR-155inhibitory effect on Bim and p27 expression and PARP cleavage (Fig. 4A, lanes 4 and 8 versus lanes 2 and 6). Further, miR-155inhibited apoptosis and caspase 3/7 activity induced by doxorubicin were also significantly reduced by ectopic expression of FOXO3a (Fig. 4, B and C). Based on these results, we conclude that the FOXO3a is a major target of miR-155 and largely mediates miR-155 antiapoptotic function.

Inverse Correlation of Expression of miR-155 and FOXO3a in Breast
Cancer-Having demonstrated FOXO3a as a major target of miR-155, we next investigated the correlation between miR-155 and FOXO3a expression in breast cancer cell lines and breast tumors. Of 12 cell lines examined, three expressing high levels of miR-155 exhibited undetectable or low level FOXO3a. Of nine cell lines with low levels of miR-155, eight expressed high levels of FOXO3a (Fig. 5A). Moreover, we examined 77 human breast cancer specimens and 11 normal breast tissues with Western blots, Northern blots, immunohistochemical staining, and LNA-microRNA in situ hybridization. Up-regulation of miR-155 was detected in 55 breast cancers and one normal breast tissue (Fig. 5, B-D). Of the 55 tumors with elevated miR-155, 41 (75%) had low levels of FOXO3a (p Ͻ 0.001), whereas 16 of 22 (73%) specimens with down-regulated miR-155 presented high levels of FOXO3a (Fig. 5E). However, we did not observe a significant relationship of miR-155 levels with tumor stage, grade, and status of ER␣, PR, and Her2, which could be due to the limited number of tumor samples examined. In addition, we examined miR-155 and FOXO3a levels in 38 recurrent breast cancers due to chemo-and/or radioresistance after surgical removal. The qRT-PCR and/or immunoblotting analyses show that 31 recurrent tumors express elevated miR-155 and low FOXO3a (p Ͻ 0.001; Fig. 5D and data not shown). These findings suggest that miR-155 regulation of FOXO3a in vivo and that elevated levels of miR-155 are associated with chemo-and/or radio-resistance in breast cancer.

DISCUSSION
miR-155 has emerged as an essential regulator of cellular physiology, particularly important in the mammalian immune system (23,36,37). For instance, miR-155 is detected during an immune response in activated mature B and T lymphocytes (46), germinal centers B cells (23), and monocytes (21). BIC/ miR-155 knock-out mice resulted in impaired immune response and cytokine production (23), further supporting the vital role of miR-155 in immunology. In addition, Down syndrome or trisomy 21 was recently linked with high levels of miR-155 and thus provides further insights into the resulting cognitive impairment and congenital heart defects seen in patients (47). In cancer, deregulation of miR-155 is implicated in a wide range of malignancies, including various forms of lymphoma and carcinomas of breast, lung, pancreas, head and neck, and kidney (9,11,(13)(14)(15). In immunology and lymphoma, miR-155 has been extensively investigated; however, it is only evident that miR-155 expression is elevated in breast cancer (13,15), and detailed function remains elusive. We demonstrate in this study that miR-155 is a determinant of chemosensitivity by targeting FOXO3a in breast cancer. We show that miR-155 directly interacts with 3Ј-UTR of FOXO3a and blocks FOXO3a translation. As a result, Bim and p27, major downstream targets of FOXO3a, are inhibited by miR-155. Ectopic expression of FOXO3a-ORF largely abrogated miR-155-induced chemoresistance. In addition, inverse correlation between miR-155 and FOXO3a expression was detected in breast cancer cell lines and tumors. These findings demonstrate for the first time that deregulation of miR-155 in breast cancer is associated with chemosensitivity and that FOXO3a is a bona fide target of miR-155.
FOXO3a is a well studied transcriptional factor that contains a forkhead DNA binding domain and plays a crucial role in apoptosis and cell growth by transcriptional regulation of a number of apoptosis/cell growth-associated genes (24,32,43). Overexpression of FOXO3a inhibits tumor cell growth in vitro and tumor size in vivo in breast cancer cells (28,48). In addition, genetic deletion of three FOXO (FOXO1, FOXO3, and FOXO4) alleles generates progressive cancerous phenotypes, such as thymic lymphoma and hemangioma (49). These data elucidated FOXOs as bona fide tumor suppressor genes. Recent studies also reveal the importance of FOXOs in preserving the self-renewal capacity of hematopoietic stem cells (50,51). Moreover, FOXO3a has been reported to be deregulated in breast cancer, and loss of FOXO3a is often linked to a decline in apoptotic activity and increased chemoresistance in cancer cells (43,(52)(53)(54)(55)(56)(57)(58). In addition, post-translational regulation of FOXOs has been wildly studied. Our finding of miR-155 negative regulation of FOXO3a not only provides an underlying mechanism for aberrant expression of FOXO3a in breast cancer but also reveals regulation of FOXO3a at the post-transcriptional level.
In an earlier work, we demonstrated that miR-155 is induced by transforming growth factor-␤ in mouse mammary epithelial cell, NMuMG. Expression of miR-155 promotes whereas knockdown of miR-155 reduces transforming growth factor-␤induced tight junction dissolution, cell migration, and invasion by targeting RhoA (20). Previous studies have shown that miR-155 regulates a number of genes that are involved in immune response, inflammation, and cell growth/survival. The transcription factors Pu1 and inositol phosphatase SHIP1 have been validated previously as direct targets of the miR-155-mediated immunoresponse (40,59). BACH1 and ZIC3 are targeted by miR-155 and mediate miR-155 function in viral infection (60,61). Moreover, miR-155 represses tumor protein 53-induced nuclear protein 1 (TP53NP1), leading to pancreatic tumor development (62). In addition, two studies using gene expression microarray analyses showed that miR-155 and its viral orthologue Kaposi's sarcoma-associated herpesvirus miR-K12-11 negatively regulate more than 180 mRNAs, some of which encode proteins involved in cell growth and survival, including PCSK5 and Rho GTPase-activating protein 21 (60,61). Thus, FOXO3a is a major but not necessarily the only target that mediates miR-155 function in the control of cell growth, survival, and chemosensitivity.
Finally, we noticed that miR-155 is more frequently up-regulated in breast tumors than cancer cell lines. A possible reason is that miR-155 in tumor cells could be induced by cytokines that are released from the tumor microenvironment. In agreement with this notion, miR-155 has been shown to be transcriptionally regulated by NFB, AP1, and Foxp3 in response to cytokines during immune cell maturation and development (36,(63)(64)(65). We also demonstrated transcriptional regulation of miR-155 via transforming growth factor-␤/Smad pathway (20). In addition, accumulated studies show that gene expression, biology, and clinical outcome of cancer are significantly influenced by the microenvironment (66,67). A three-dimensional culture model has been established to recapitulate the in vivo functions, interactions, and architecture of the mammary gland and breast tumor (66,67), which more closely resembles the tumor microenvironment than traditional tissue culture. In order to address if miR-155 expression is influenced by microenvironment, further investigation is required using a threedimensional culture system and animal models.
In summary, we demonstrated that miR-155 contributes to chemoresistance in breast cancer. FOXO3a is negatively regulated by miR-155 and mediates miR-155 function in the control of breast cancer cell survival and growth. In combination with our previous findings of miR-155 induction of epithelial-mesenchymal transition, cell migration, and invasion in mammary epithelial cells, miR-155, therefore, is a critical therapeutic target for breast cancer intervention.