Distinct Expression Pattern and Post-Transcriptional Regulation of Cell Cycle Genes in the Glandular Epithelia of Avian Ovarian Carcinomas

The cell cycle system is controlled in a timely manner by three groups of cyclins, cyclin dependent kinases and cyclin dependent kinase inhibitors. Abnormal alterations of cell cycle regulatory mechanisms are a common feature of many diseases including numerous tumor types such as ovarian cancer. Although a variety of cell cycle regulatory genes are well known in mammalian species including human and mice, they are not well studied in avian species, especially in laying hens which are recognized as an excellent animal model for research relevant to human ovarian carcinogenesis. Therefore, in the present study, we focused on comparative expression and regulation of expression of candidate genes which might be involved in the cell cycle program in surface epithelial ovarian cancer in laying hens. Our current results indicate that expression levels of cell cycle gene transcripts are greater in cancerous as compared to normal ovaries. In particular, cyclin A2 (CCNA2), CCND1, CCND2, CCND3, CCNE2, cyclin dependent kinase 1 (CDK1), CDK3, CDK5, cyclin dependent kinases inhibitor 1A (CDKN1A) and CDKN1B were upregulated predominantly in the glandular epithelia of cancerous ovaries from laying hens. Further, several microRNAs (miRs), specifically miR-1798, miR-1699, miR-223 and miR-1744 were discovered to influence expression of CCND1, CCNE2, CDK1, and CDK3 mRNAs, respectively, via their 3′-UTR which suggests that post-transcriptional regulation of gene expression influences their expression in laying hens. Moreover, miR-1626 influenced CDKN1A expression and miR-222, miR-1787 and miR-1812 regulated CDKN1B expression via their 3′-UTR regions. Collectively, results of the present study demonstrate increased expression of cell cycle-related genes in cancerous ovaries of laying hens and indicate that expression of these genes is post-transcriptionally regulated by specific microRNAs.


Introduction
Epithelial ovarian cancer (EOC), the most lethal gynecological malignancy, claims the lives of over 15,000 women and 22,000 are diagnosed with the disease in the US each year [1]. However, over 75% of woman diagnosed are at an advanced stage of EOC, because it is generally asymptomatic and there is no specific biomarker(s) for early detection [2]. Therefore, to prevent and cure this lethal disease and to improve the long-term survival of patients with EOC, the most promising approach is to identify markers for early diagnosis. To overcome the problem that EOC is rarely detected at an early stage, many animal models have been developed, but they have not proven to be successful. For instance, genetically manipulated rodent models have been used to elucidate some aspects of the pathogenesis and etiologies of EOC; however, the non-spontaneous nature of their ovarian cancer limits their clinical relevance [3,4,5]. In fact, the laying hen is the only animal that spontaneously develops ovarian cancer of the surface epithelium of the ovaries at a high rate, as also occurs in women [6]. Thus, the laying hen is a unique animal model for human EOC research aimed at development of a biomarker(s) for detection and early diagnosis, as well as for discovery of anticancer drugs/biomaterials for prevention and treatment of this deadly disease.
The cell cycle in most eukaryotic cells includes a series of coordinated events consisting of cell growth, replication of genetic material, segregation of the duplicated chromosomes and cell division [7]. In general, the cell division cycle in mammals is precisely and harmoniously regulated in a timely manner by different active heterodimeric complexes that include cyclin dependent kinases (CDKs) and their cognate cyclin partners, as well as CDK inhibitors (CDKIs) [8]. Thus, tumor development frequently results when there is deregulation of the cell cycle control system including abnormal regulation of expression of cell cycle genes [1]. In human cancerous tissues, such as neoplasms, different families of cell cycle genes and regulators are frequently mutated and dysfunctional [8]. Although expression and functional roles of many CDKs, cyclins and CDKIs are well studied in mammalian species, including humans and mice, little is known about their expression and regulation in avian species, especially laying hens.
MicroRNAs (miRNAs) are endogenous non-coding short RNAs involved in various biological processes that regulate gene expression via degradation or inhibition of expression of target mRNAs. The involvement of miRNA-mediated regulatory mechanisms affecting gene transcription and translation in human cell cycle progression has been reported [9,10,11]. Indeed, miRNA-based fine tuning of expression of cell cycle genes is very important because improper cell cycle control is likely to lead to initiation and development of proliferative diseases, such as cancer. Although numerous miRNAs have been indentified in chickens, the functional aspects of most chicken miRNAs are not known and reports of miRNA-mediated post-transcriptional mechanism regulating cell cycle progression in chickens are not available. Understanding the target spectrum of cell cycle-related miRNAs and their functional interactions is expected to help elucidate the molecular and epigenetic regulatory mechanisms affecting transcriptional and translational events critical to control of the cell cycle and progression into carcinogenesis. Therefore, the objectives of this study with laying hens were to determine: 1) the expression of cyclins, CDKs and CDKIs in normal and cancerous ovaries; and 2) whether cyclins, CDKs and CDKIs are regulated by post-transcriptional actions of specific microRNAs (miRs) using a miR target validation assay. Our results confirm that the laying hen is a unique model for the research on human ovarian cancer and cell cycle-related genes and that regulatory factors for cell cycle-related genes play a key role in ovarian carcinomas. These cell cycle-related genes may be important targets for discovery of a biomarker(s) for diagnosis and evaluation of therapeutics designed to treat EOC in women.

Comparative Expression of Cyclin Genes in Normal and Cancerous Ovaries from Laying Hens
To determine tissue-specific expression patterns of the cell cyclerelated genes in normal (n = 5) and cancerous (n = 10) ovaries from laying hens, we performed RT-PCR, and quantitative PCR analyses. As illustrated in Figure 1, expression of cyclin A2 (CCNA2), CCND1, CCND2, CCND3 and CCNE2 mRNAs was 3.42-(P,0.01), 1.32-(P,0.05), 2.41-(P,0.01), 3.31-(P,0.05) and 2.36-fold (P,0.001) greater in cancerous ovaries from hens. Next, cell-specific localization of these genes in the normal and cancerous ovaries was determined using in situ hybridization analysis. The mRNAs for CCNA2, CCND1, CCND2, CCND3 and CCNE2 were localized predominantly to the glandular epithelium (GE) in cancerous ovaries, but there was very weak or no detectable expression of these genes in the luminal epithelium (LE), stromal cells or blood vessels in normal and cancerous ovaries.

Comparative Expression of Cyclin Dependent Kinase and Cyclin Dependent Kinase Inhibitor Genes in Normal and Cancerous Ovaries of Laying Hens
As shown in Figure 2, the results from RT-PCR and quantitative PCR analyses showed that expression of mRNAs for cyclin dependent kinase 1 (CDK1), CDK3 and CDK5 were 2.87-(P,0.01), 5.18-(P,0.01) and 3.66-fold (P,0.01) greater in cancerous ovaries from hens, respectively. Interestingly, cyclin dependent kinase inhibitor 1A (CDKN1A) and CDKN1B mRNAs were 5.62-(P,0.01) and 2.31-fold (P,0.05) more abundant in cancerous as compared with normal ovaries. In situ hybridization analyses demonstrated that expression of CDK1, CDK3, CDK5, CDKN1A and CDKN1B mRNAs was abundant in GE and to a much lesser extent in stromal cells of cancerous ovaries, whereas there was very little or no expression of these genes in normal ovaries.

Post-transcriptional Regulation of Specific Cell Cycle Regulatory Genes by Chicken microRNAs
A microRNA (miR) target validation assay was used to test the hypothesis that expression of cell cycle genes is regulated at the post-transcriptional level by specific miRNAs. Analysis of potential miRNA binding sites within the 39-UTR of the six cell cycle regulatory genes was performed using the miRNA target prediction database (miRDB; http://mirdb.org/miRDB/). This analysis revealed putative binding sites for several chicken miRNAs (miR-1798 for CCND1; miR-1699 for CCNE2; miR-223 for CDK1; miR-1744 for CDK3; miR-1626 for CDKN1A; and miR-222, miR-1787 and miR-1812 for CDKN1B), but not for the other four genes of interest. Therefore, we determined whether these miRNAs influenced expressions of cell cycle regulatory genes via the 39-UTR. As illustrated in Figures 3 and 4, in the presence of miR-1798 and miR-1699, the intensity and percentage of GFP-CCND1-expressing cells (12.7% in control vs. 4.2% in miR-1798) and GFP-CCNE2-expressing cells (96.4% in control vs. 71.4% in miR-1699) decreased (P,0.01). In addition, as shown in Figures 5  and 6, in the presence of miR-223 and miR-1744, the intensity and percentage of GFP-CDK1-expressing cells (17.2% in control vs. 1.3% in miR-223) and GFP-CDK3-expressing cells (16.1% in control vs. 6.8% in miR-1744) were decreased (P,0.01). Moreover, in the presence of miR-1626, the intensity and percentage of GFP-CDKN1A-expressing cells (54.6% in control vs. 34.7% in miR-1626) were decreased (P,0.01) (Figure 7). In addition, for CDKN1B, in the presence of miR-222, miR-1787 and miR-1812, the intensity and percentage of GFP-CDKN1Bexpressing cells (29.0% in control vs. 15.6% in miR-1787, 12.6% in miR-1812, 9.8% in miR-222) were decreased (P,0.01) ( Figure 8). These results indicate that at least one to three miRNAs bind directly to the cell cycle-related gene transcripts to regulate expression.

Discussion
Results of the present study provide the first evidence of significant differences in expression of CCNA2, CCND1, CCND2, CCND3, CCNE2, CDK1, CDK 3, CDK5, CDKN1A and CDKN1B genes in cancerous as compared to normal ovaries of laying hens. In addition, our results indicate that several microRNAs (miRs), specifically miR-222, miR-223, miR-1626, miR-1699, miR-1744, miR-1787, miR-1798 and miR-1812 interact with sites in the 39-UTR of the cell cycle genes and regulatory factors affecting cell cycle genes including CCND1, CCNE2, CDK1, CDK3, CDKN1A and CDKN1B to influence post-transcriptional regulation of its expression in laying hens. These results support our hypothesis that cell cycle genes are critical regulators for growth and developmental aspects of epithelial cells of the ovaries of hens and that there is dysregulation of their level of expression as ovaries of laying hens transition from a normal to a cancerous state.
In the United States, ovarian cancer is the most common malignancy in the female genital tract and the fifth leading cause of cancer-related deaths among women. The surface epithelialderived ovarian cancer (EOC) accounts for 90% of all ovarian cancers [12]. The idea that the repeated rupture of the ovarian surface epithelium during the monthly ovulation event in women may contribute to or accelerate the incidence of the EOC was proposed by Fathalla about four decades ago [13]; however, the etiology and pathology of the EOC is complicated and not fully understood. Results of a number of epidemiological and histological studies strongly support the idea that there is an increased incidence of EOC dependent on the frequency of ovulation and other factors associated with the reproductive tract [14]. However, there is evidence suggesting that serous-type carcinomas do not begin as precursor lesions in the fimbria of the fallopian tube in women and then spread to the surface epithelium of the ovary or shed into the peritoneal cavity [15,16,17]. In addition, there is a report indicating that EOC is caused from instability of the copy number of a certain gene, but not from the mutation of the gene(s) [18]. At present, the laying hen is an established animal model for study of EOC to elucidate the pathogenesis and etiologies of EOC because they spontaneously develop surface epithelium-derived ovarian cancer at a high rate as occurs in women [6]. There are no reports that ovarian carcinoma has its origin in the oviduct (fallopian tube) of laying hens. Further, except for serous carcinoma, there are no reports of the other three types of ovarian cancer including endometrioid, mucinous and clear cell carcinomas in humans. Therefore, future studies will investigate the origins of the different types of ovarian carcinoma are regulatory mechanism(s) that govern the initiation and development of ovarian carcinogenesis in women and laying hen.
Cyclins are a family of proteins that control the cell cycle by binding and activating cyclin-dependent kinases. As illustrated in Figures 1, D-type cyclins, which are G 1 phase regulators of the cell cycle [1,8], such as CCND1, CCND2 and CCND3 are predominantly found in cancerous ovaries, but there was weak or little expression in normal ovaries of laying hens in the present study. In humans, CCND1 is frequently overexpressed in a variety of tumor types and is associated with carcinogenesis and metastasis [19]. Dhar and colleagues reported that expression of CCND1 was upregulated in about 90% of patients with EOC and expressed mainly in both borderline and invasive tumours without any association between immunoreactive protein overexpression and stage of tumor differentiation or grade of tumor [20]. In the present study, we found CCND1, CCND2 and CCND3 mRNAs in the nucleus and the cytoplasm of epithelial cells of normal ovaries, but exclusively in the cytoplamic compartment of the epithelial cells in cancerous ovaries of laying hens. This result is consistent with deregulation of CCND1 expression leading to localization of the protein in both the cytoplasmic and nuclear compartments of cells from cancerous ovaries [20]. Interestingly, although CCND2 and CCND3 are not overexpressed in human ovarian cancer [21], messenger RNA expression levels for both CCND2 and CCND3 were significantly upregulated in cancerous ovaries of laying hens. The difference in expression of these genes between humans and laying hens should be elucidated. Expression of CCNE2 mRNA, a G 1 -S phase regulator [7], in cancerous ovaries from hens was 2.36-fold (P,0.001) greater than in normal ovaries, and mainly detected in the glandular epithelium (GE). This result supports the idea that CCNE protein is valuable prognostic factors for EOC patients because amplification and over-expression of the CCNE1 gene occurs in many cases with a gradual increase from benign to borderline to malignant tumors [21,22]. In addition, CCNA2 mRNA, an S phase regulators [1,7], was found predominantly in GE and its expression was 3.42-fold (P,0.01) greater in cancerous ovaries of laying hens. In humans, CCNA expression increased in  the ovarian carcinoma cell line compared with normal cells [23] and CCNA protein was detected mainly in serous and endometrioid carcinomas, but not in mucinous and clear cell carcinomas [24].
Cyclin dependent kinases (CDKs) are the catalytic subunits of a large family of heterodimeric serine/threonine protein kinases that have essential roles in controlling progression of the cell cycle [25]. In the present study, we found that CDK1, CDK3 and CDK5 mRNAs were up-regulated predominantly in GE of cancerous ovaries, but there was weak or no expression of these mRNAs in normal ovaries of laying hens. In general, CDK1 has an essential role in the transition of cells into mitosis under normal circumstances as a component of the cell cycle control system and it is also involved in rapid arrest in G 2 phase in response to DNA damage [26]. Overexpression of CDK1 is detected in 79% of EOC patients, but not in benign epithelial tumors or normal epithelial tissues in women [27]. In addition, overexpression of CDK2 was found in only 6% of EOC patients, but its level of expression was positively correlated with CCNE abundance, suggesting that overexpression of both CDK2 and CCNE is significantly associated with development of malignant ovarian tumors [22,28]. Furthermore, CDK4 is overexpressed in 14% to  15% of ovarian tumors [29] and its activity in malignant ovarian tumors is significantly greater than in benign tumors [22]. These results suggest that CDK4 activity play important roles in ovarian carcinogenesis. Collectively, results of the present study strongly  indicate that CDK activity is regulated by cyclin synthesis and degradation, and that orderly progression through the cell cycle requires coordinated activation of the CDK proteins by binding to the cyclin partner [8]. Furthermore, the results confirm that the laying hen is an appropriate animal model for identifying and developing biomarkers for early diagnosis and evaluation of therapeutics for treatment of ovarian cancer [5,6,30,31,32,33,34,35,36].
MicroRNAs, short and noncoding RNAs of 18 to 23 nucleotides in length, regulate complex patterns of gene expression posttranscriptionally and are capable of defining and altering cell fate by silencing translational of gene transcripts through cleavage of their target mRNAs through base pairing at partially or fully complementary sites [37]. As shown in Figures 3, 4, 5, 6, miR-1798 and miR-1699 influence the expression of CCND1 and CCNE2, respectively, while miR-223 and miR-1744 regulate expression of CDK1 and CDK3, respectively. By regulating post-transcriptional events, miRs affect function of a number of cellular processes in development, differentiation and oncogenesis [38,39,40]. Results of the present study demonstrated that miR-1798 inhibits expression of CCND1 in laying hens. In human ovarian cancer, deregulation of CCND1 expression mainly occurs without any gene amplification [20]. Thus, we suggest dysfunction of miR-1787 leads to the overexpression of CCND1 in cancerous ovaries of laying hens. In fact, deregulation of miRs is generally considered to be a prerequisite for initiation and progression of carcinogenesis in humans. For instance, functional overexpression of miR-31 inhibits proliferation and induces apoptosis in a variety of seroustype ovarian cancer cell lines, such as SKOV3, with a dysfunctional p53 pathway [41]. In addition, miR expression of transcriptional targets of p53 (i.e. miR-34b and miR-34c) is markedly down-regulated in human EOC tissues [42]. These results indicated that miRs may be useful in predicting outcomes of many diverse carcinomas, including EOC.
As shown in Figure 2, there is overexpression of cyclin dependent kinase inhibitor 1A (CDKN1A) and CDKN1B mRNAs in cancerous ovaries of laying hens. CDKN1A and CDKN1B (also known as p21/WAF1 and p27, respectively) are potent CDK inhibitors and act as regulators in the G 1 phase of the cell cycle. In humans, aberrant expression of CDKN1A and/or CDKN1B in various carcinomas of lung, colorectum, cervix, head and neck leads to carcinogenesis by blocking DNA synthesis and inhibiting cell growth [43,44,45,46]. In other words, loss of both genes may contribute to tumor progression. For instance, loss of CDKN1B protein was significantly associated with a relatively shorter time to cell cycle progression and decreased overall survival rates in patients with advanced ovarian cancer [47]. These results indicate that both CDKN1A and CDKN1B are potential prognostic markers to predict progression of EOC and survival in EOC patients. However, it has been reported that CDKN1B has dual functional roles during tumorigenesis. In mice, Cdkn1b acts as a tumor suppressor due to its cyclin-CDK regulatory function and it acts as an oncogene through a cyclin-CDK-independent function [48]. These results may explain why CDKN1A and CDKN1B genes have controversial patterns of expression pattern in human and mouse tumors. Therefore, we suggest that synchronous up-regulation of CDKN1A and CDKN1B genes with other CDK genes in the present study may be caused by a CDKindependent-oncogenic-function of both CDKN1A and CDKN1B instead of an inhibitory function of these CDKs [48]. Future research is required to gain a better understanding of the CDKindependent-oncogenic-function of both genes in ovarian carcinogenesis in laying hens. In addition, our miR target validation assay demonstrated that miR-1626 regulates CDKN1A expression and miR-222, miR-1787 and miR-1812 influence post-transcriptional modification of transcripts of the CDKN1B gene. These results suggest that down-regulation of these miRs might contribute to the overexpression of cell cycle genes and regulatory factors in chicken ovarian cancer and to transcriptional deregulation of many genes in the genome, that may lead to uncontrolled carcinogenesis.
Collectively, results of the present study indicate that overexpression of cell cycle-related genes (i.e. cyclins, their associated kinases and inhibitors) may be involved in uncontrolled cell proliferation, growth and loss of function in cells that leads to ovarian tumorigenesis in laying hens. Furthermore, post-transcriptional regulation of the specific miRs that influence expression of cell cycle genes likely leads to an alternative mechanism(s) for regulation of their expression. Although results of this study indicate that various miRs might be involved in many different oncogenic/carcinogenic pathways, details of altered expression patterns and their relevance to EOC remain to be elucidated. Thus, further research is clearly required to unravel the mechanism(s) for post-transcriptional regulation of cell cycledependent gene expression and different oncogenic pathways leading to ovarian carcinogenesis in women and in laying hens.

Experimental Animals and Animal Care
The experimental use of chickens for this study was approved by the Institute of Laboratory Animal Resources, Seoul National University (SNU-070823-5). White Leghorn (WL) laying hens were subjected to standard management practices at the Seoul National University Animal Farm, Seoul, Republic of Korea. All chickens were exposed to a light regimen of 15 h light and 9 h dark with ad libitum access to feed and water, as well as standard management practices for feeding and husbandry.

Tissue Samples
In this study, a total of 136 laying hens (88 over 36 months of age and 48 over 24 months of age), which had completely stopped laying eggs, were euthanized for biopsy and cancerous (n = 10) ovaries were collected. As a control, normal (n = 5) ovaries were also collected from laying hens. We examined tumor stage and the degree of metastasis in 10 hens with cancerous ovaries according to characteristic features of chicken epithelial ovarian cancers [5] (see the Table 1 of the reference). In three hens, ovarian tumor cells were classified as Stage III as they had metastasized to the gastrointestinal tract and superficial surface of the liver with profuse ascites in the abdominal cavity. In five hens, the tumors had metastasized to distant organs such as liver parenchyma, lung, gastrointestinal tract and oviduct with profuse ascites, so these were classified at Stage IV tumors. The other two hens did not have tumors in any other organs; therefore, their ovarian tumors were classified as Stage I. The collected cancerous and normal ovaries containing follicles, glands, stromal cells and blood vessels were frozen in liquid nitrogen or fixed in 4% paraformaldehyde for further analyses. Frozen tissue samples were cut into 5-to 7mm pieces before being frozen in liquid nitrogen. The other samples were cut into 10 mm pieces and fixed in 4% paraformaldehyde in PBS (pH 7.4). After 24 h, fixed tissues were changed to 70% ethanol for 24 h and then dehydrated and embedded in Paraplast-Plus (Leica Microsystems, Wetzlar, Germany). Paraffin-embedded tissues were sectioned at 5 mm and stained with hematoxylin and eosin. Epithelial ovarian cancers in laying hens were classified based on their cellular subtypes and patterns of cellular differentiation with reference to ovarian malignant tumor types in humans.

RNA Isolation
Total cellular RNA was isolated from frozen tissues using Trizol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's recommendations. The quantity and quality of total RNA was determined by spectrometry and denaturing agarose gel electrophoresis, respectively.

Semiquantitative RT-PCR Analysis
The expression of mRNAs for cell cycle genes in normal and cancerous ovaries of laying hens was assessed using semiquantitative RT-PCR as described previously [49]. Information on the primer sets is provided in Table 1. The cDNA was synthesized from total cellular RNA (2 ug) using random hexamer (Invitrogen, Carlsbad, CA) and oligo (dT) primers and AccuPo-werH RT PreMix (Bioneer, Daejeon, Korea). The cDNA was diluted (1:10) in sterile water before use in PCR. The primers, PCR amplification and verification of their sequences were conducted as described previously [49]. After PCR, equal amounts of reaction product were analyzed using a 1% agarose gel, and PCR products were visualized using ethidium bromide staining. The amount of DNA present was quantified by measuring the intensity of light emitted from correctly sized bands under ultraviolet light using a Gel Doc TM XR+ system with Image Lab TM software (Bio-Rad).

Quantitative RT-PCR Analysis
Total RNA was extracted from normal and cancerous ovarian tissue using TRIzol (Invitrogen) and purified using an RNeasy Mini Kit (Qiagen). Complementary DNA was synthesized using a SuperscriptH III First-Strand Synthesis System (Invitrogen). Gene expression levels were measured using SYBRH Green (Biotium, Hayward, CA, USA) and a StepOnePlus TM Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The glyceraldehydes 3-phosphate dehydrogenase (GAPDH) gene was analyzed simultaneously as a control and used for normalization of data. Each target gene and GAPDH were analyzed in triplicate. Using the standard curve method, we determined the expression quantities of the examined genes using the standard curves and C t values, and normalized them to GAPDH expression values. ROX dye (Invitrogen) was used as a negative control for the fluorescence measurements. Sequence-specific products were identified by generating a melting curve in which the C t value represented the cycle number at which a fluorescent signal rose statistically above background, and relative gene expression was quantified using the 2 -DD C t method [50]. For the control, the relative quantification of gene expression was normalized to the C t of the control ovaries. Information on the primer sets is provided in Table 2.

In Situ Hybridization Analysis
For hybridization probes, PCR products were generated from cDNA with the primers used for RT-PCR analysis. The products were gel-extracted and cloned into pGEM-T vector (Promega). After verification of the sequences, plasmids containing gene sequences were amplified with T7-and SP6-specific primers (T7:59-TGT AAT ACG ACT CAC TAT AGG G-39; SP6:59-CTA TTT AGG TGA CAC TAT AGA AT-39) then digoxigenin (DIG)-labeled RNA probes were transcribed using a DIG RNA labeling kit (Roche Applied Science, Indianapolis, IN). Tissues were collected and fixed in 4% paraformaldehyde. The tissues

MicroRNA Target Validation Assay
The 39-UTR of selected genes was cloned and confirmed by sequencing. The 39-UTR was subcloned between the eGFP gene and the bovine growth hormone (bGH) poly-A tail in pcDNA3eGFP (Clontech, Mountain View, CA) to generate the eGFP-miRNA target 39-UTR (pcDNA-eGFP-39UTR) fusion constructs. For the dual fluorescence reporter assay, the fusion constructs containing the DsRed gene and each microRNA were designed to be co-expressed under control of the CMV promoter (pcDNA-DsRed-miRNA). The pcDNA-eGFP-39UTR and pcDNA-DsRed-miRNA (4 mg) were co-transfected into 293FT cells using the calcium phosphate method. When the DsRed-miRNA is expressed and binds to the target site of the 39-UTR downstream of the GFP transcript, green fluorescence intensity decreases due to degradation of the GFP transcript. At 48 h posttransfection, dual fluorescence was detected by fluorescence microscopy and calculated by FACSCalibur flow cytometry (BD Biosciences). For flow cytometry, the cells were fixed in 4% paraformaldehyde and analyzed using FlowJo software (Tree Star Inc., Ashland, OR). We captured fluorescence images using a confocal laser scanning microscope and zen 2009 microscopy software (Carl zeiss, Germany) with the following settings: Lenses, 206; Frame size, 128(X) and 128(Y); Laser settings, 3% FITC at 488 nm) and 2.8% DsRed at 555 nm; Scan time, 491 msec; Pinhole size; 1Airy unit.

Statistical Analyses
Data for quantitative PCR were subjected to analysis of variance (ANOVA) according to the general linear model (PROC-GLM) of the SAS program (SAS Institute, Cary, NC). Data are presented as mean 6 SEM unless otherwise stated. Differences in the variance between normal and each classification of cancerous ovary group were analyzed using the F test, and differences in the means were subjected to Student's t test. Differences were considered significant at P,0.05.