MicroRNA-661, a c/EBPα Target, Inhibits Metastatic Tumor Antigen 1 and Regulates Its Functions

MicroRNAs (miRs) have been identified as posttranscriptional modifiers of target gene regulation (1). miRs are 20 to 22 bases long and are derived from longer noncoding primary transcripts by the actions of the Drosha and Dicer RNA cleaving enzymes (2, 3). Recent spurt of research has revealed that miRs have important roles in development and disease including cancer (4–6). In general, miRs pair up with the 3′ untranslated region (UTR) of target mRNAs in a sequence-specific manner with the help of RNA-induced silencing complex. These pairings lead to the down-regulation of the target mRNA by translational inhibition or degradation of mRNA (7). Recent research suggests that miRs suppress or activate cancer and metastasis. miR-10B, for example, has been shown to promote metastasis in breast cancer (8), whereas miR-7 has been shown to suppress breast and brain cancers (9, 10).

A large body of work (11–13) has linked the up-regulation of metastatic tumor antigen 1 (MTA1) to the maintenance and progression of more invasive phenotypes of many human cancers. MTA1, the founding member of the MTA family, was initially identified as a differentially expressed gene in metastatic rat mammary adenocarcinoma (14). MTA1 is a component of the chromatin remodeling complex and modulates transcription of its target gene chromatin, by recruiting HDACs or RNA PolII. MTA1 is widely up-regulated in many carcinomas, including breast, colorectal, gastric, esophageal, pancreatic, ovarian, non–small cell lung, hepatocellular and renal carcinomas; thymoma; and hematopoietic malignancies (12). Forced overexpression of MTA1 in the mouse mammary gland epithelium leads to hyper proliferation in glands of virgin mice and mammary gland adenocarcinomas (15). Although the role of MTA1 in human cancer has been established, it remains unclear whether MTA1 is targeted by miRs.

Cell lines, culture conditions, and transfections. Human cell lines were cultured as described in supplementary methods. Transfections for miRNA mimics and plasmid constructs were done with Oligofectamine or Fugene as described in the Supplementary Methods.

Plasmid constructs, luciferase assay, and Western blotting. MTA1 3′ UTR and miR-661 promoter region were cloned into pGL3 control vector and pGL3 basic promoter less vector, respectively, as described in Supplementary Methods. Luciferase assays and Western blotting were done as described in the Supplementary Methods.

Quantitative real-time PCR analysis of miRs. RNA was isolated using mirVana miRNA isolation kit. For quantitative analysis of miRNAs, two-step TaqMan real-time PCR analysis was performed using primers and probes obtained from Applied Biosystems.

Migration, soft-agar, and confocal studies. Migration, Invasion, Soft agar assays, and Confocal studies were done as described in the Supplementary methods.

Chromatin immunoprecipitation. Chromatin immunoprecipitation and Electrophoretic Mobility Shift Assay was performed as described in the Supplementary methods.

To search for miRs that might regulate MTA1 expression, we screened the 3′UTR region of MTA1 mRNA using the public database for possible complementation of a minimum of 8-bp to the seed region of miRs (16). This exercise resulted in identification of human miR-661 and miR-559 as candidate miRs that might regulate MTA1 (Fig. 1A). To assess the effects of these miRs on MTA1 expression, we transfected human cancer cell lines with miR-559 or miR-661 or a negative control (designated con-miR) and MTA1 protein levels were determined by Western blot analysis. Both miRs inhibited the expression level of MTA1 but not MTA2 or vinculin. (Fig. 1B). To investigate whether the 3′UTR region of MTA1 was directly targeted by the miRs, we cloned the 3′UTR region of MTA1 that is complementary to miR-559 or miR-661 into the pGL3-luciferase reporter (MTA1 3′UTR-luc). Transfection of the miRs and MTA1 3′UTR-luc into cells inhibited luciferase reporter activity, whereas no such inhibitory effect was observed in the control, con-miR (Fig. 1C). These findings suggested that miR-559 or miR-661 targeted the 3′UTR region of MTA1 and inhibited MTA1 expression. Because the expression of only miR-661 was observed in human cells, we focused on the regulation of MTA1 expression and functions by miR-661 in subsequent experiments.

We analyzed the 2-kb region directly upstream of the miR-661 to identify the mechanisms by which miR-661 might be regulated in physiologic settings. Using transcriptional factor prediction program Alibaba2 (17), we determined whether binding motifs for various transcription factors were present. This analysis revealed the presence of consensus motifs for c/EBPα within the −1426 to −1417, −1322 to −1311, −885 to −875, −740 to −730, and −331 to −320 in the upstream of the miR-661 (Fig. 2A). Because expression of c/EBPα has been shown to reduce as cancer progresses (18, 19) and MTA1 expression is up-regulated during the progression of cancer cells to more invasive phenotypes (12), we reasoned that c/EBPα has a role in miR-661 expression and in turn MTA1 regulation. To evaluate this hypothesis, we used chromatin immunoprecipitation analysis to determine that c/EBPα is recruited at its binding sites, −331 to −320 (R4), on the miR-661 promoter region but not to the other three regions of the miR-661 chromatin (Fig. 2B). To establish a direct binding of c/EBPα to the miR-661 promoter DNA, we performed an electrophoretic mobility shift assay analysis using a miR-661 DNA fragment containing the consensus site for c/EBPα. We found that incubating the labeled DNA fragment with nuclear extract from MDA-MB-231 cells transiently transfected with c/EBPα resulted in the formation of discrete protein-DNA complex. These complexes could be effectively supershifted by inclusion of c/EBPα antibodies and their formation blocked by the unlabeled DNA fragment (Fig. 2C). We next cloned the putative miR-661 promoter into a TATA-less basic pGL3-luc reporter (pGLmiR-661) to study the functionality of the interaction of c/EBPα with miR-661 DNA. We found that transient expression of c/EBPα efficiently stimulated the transcription of miR-661 from the pGLmiR-661 reporter in HeLa and MDA-231 cells (Fig. 2D). Similarly, transient expression of c/EBPα led to the decreased expression of MTA1 and elevated expression of miR-661 in MDA-231 cells (Fig. 2E). However, the same was not observed for the control vector. Together, these findings suggest that c/EBPα directly interacted with the putative miR-661 promoter and positively regulated miR-661 expression.

Because miR-661 targeted MTA1, because c/EBPα positively regulated the expression of miR-661 (this study), and because loss of c/EBPα is commonly observed during cancer progression, we hypothesized that a dynamic relationship exists between MTA1, miR-661, and c/EBPα and specifically that c/EBPα is involved in the action of miR-661 in cancer cells. We tested these hypotheses by analyzing the levels of MTA1, miR-661, and c/EBPα in exponentially breast cancer progressive isogenic model MCF-10A cells (nonmalignant), MCF-10AT cells (weakly tumorigenic cells), MCF-10CA1D cells (undifferentiated metastatic cells), and MCF-10DCIS cells (highly proliferative, aggressive, and invasive cells; ref. 20). The levels of MTA1 protein were progressively up-regulated, whereas those of miR-661 and c/EBPα were down-regulated from non invasive MCF-10A to highly invasive MCF-10DCIS cells (Fig. 3A). To confirm these findings, we showed that c/EBPα expression led to up-regulation of miR-661 expression and down-regulation of MTA1 (Fig. 3A). To independently validate these findings, we showed that c/EBPα down-regulated the activity of MTA1 3′UTR-luc in MDA-231, ZR-75 breast cancer cells, and PC-3 prostate cancer cells (Fig. 3B). These results suggested that c/EBPα up-regulates the level of miR-661 and consequently affects the functions of the miR-661-MTA1 pathway in human cancer cells.

To evaluate the effect of miR-661mediated MTA1 down-regulation on the biology of breast cancer cells, we next examined the effects of miR-661 on of MDA-231 cells. We found that miR-661 expression in MDA-231 cells was accompanied by a substantial inhibition of cell motility, reduced cell invasiveness and suppression of anchorage-independent growth in soft agar, and a reduced ability of the cells to form tumors in a xenograft model (Fig. 4A–D). As expected from the data presented here, introduction of miR-661 in tumors was accompanied by a reduced MTA1 expression. Collectively, these findings allow us to propose a model wherein the expression of miR-661 and its activator, c/EBPα, are progressively reduced during cancer progression and the loss of miR-661, in turn allows MTA1 levels to be sustained. These results raise the possibility that miR-661 may be developed as a therapeutic agent for down-regulating the expression of MTA1, which is widely up-regulated in human cancers (12).

No potential conflicts of interest were disclosed.

Grant support: NIH grant CA98823 (R. Kumar).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.