MiR-20a Is Upregulated in Anaplastic Thyroid Cancer and Targets LIMK1

Background There have been conflicting reports regarding the function of miR-20a in a variety of cancer types and we previously found it to be dysregulated in sporadic versus familial papillary thyroid cancer. In this study, we studied the expression of miR-20a in normal, benign and malignant thyroid samples, and its effect on thyroid cancer cells in vitro and in vivo. Methodology/Principal Findings The expression of miR-20a in normal, benign and malignant thyroid tissue was determined by quantitative RT-PCR. Thyroid cancer cells were transfected with miR-20a and the effect on cellular proliferation, tumor spheroid formation, and invasion was evaluated. Target genes of miR-20 were determined by genome-wide mRNA expression analysis with miR-20a overexpression in thyroid cancer cells and target prediction database. Target genes were validated by quantitative PCR and immunoblotting, and luciferase assays. MiR-20a expression was significantly higher in anaplastic thyroid cancer than in differentiated thyroid cancer, and benign and normal thyroid tissues. MiR-20a significantly inhibited thyroid cancer cell proliferation in vitro (p<0.01) and in vivo (p<0.01), tumor spheroid formation (p<0.05) and invasion (p<0.05) in multiple thyroid cancer cell lines. We found that LIMK1 was a target of miR-20a in thyroid cancer cell lines and direct knockdown of LIMK1 recapitulated the effect of miR-20a in thyroid cancer cells. Conclusions/Significance To our knowledge, this is the first study to demonstrate that miR-20a plays a role as a tumor suppressor in thyroid cancer cells and targets LIMK1. Our findings suggest the upregulated expression of miR-20a in anaplastic thyroid cancer counteracts thyroid cancer progression and may have therapeutic potential.


Introduction
Thyroid cancer is the most common endocrine cancer and one of the fastest growing cancer diagnoses in the United States [1,2]. Thyroid cancers originate from follicular cells and parafollicular cells [3,4]. Thyroid cancers originating from follicular cells account for over 95% of all thyroid cancer cases and are classified into four major histologic groups (papillary thyroid cancer (PTC), follicular thyroid cancer (FTC), Hürthle cell carcinoma (HCC), and anaplastic thyroid carcinoma (ATC)). MicroRNAs (miRNAs) have been shown to be dysregulated in thyroid cancers originating from follicular cells [5][6][7]. MiRNAs are small, noncoding RNAs, which are approximately 21 nucleotides long and regulate gene expression [8,9]. Generally, miRNAs bind to the 39-untranslated region (39-UTR) of the target gene, leading to repressed translation or degradation of mRNA [9,10].
MiR-20a is a member of the miR-17-92 cluster located on chromosome 13. Previous studies have shown that miR-20 may function to promote or inhibit the hallmarks of malignant cell phenotype in a cell type specific manner [11][12][13]. For example, miR-20a overexpression inhibits cellular proliferation, invasion, and tumor metastasis in breast cancer cell lines [11,12]. On the other hand, miR-20a suppresses E2F1 expression in human B cell line P-493-6, a transcription factor that promotes G1-S phase progression in mammalian cells [14]. This finding suggests that miR-20a function may be different depending on cell type. We previously found miR-20a to be upregulated in familial PTC as compared to sporadic cases, which are thought to be more aggressive [3,15]. Takakura et al. [16] also found that miRNAs of the miR-17-92 cluster (miR-17-3p, -17-5p, -18a, -19a, -20a, -19b, and -92-1) were overexpressed in ATC cell lines.
In this study, we characterize the expression of miR-20a in normal, benign and malignant thyroid samples, and studied its effect on thyroid cancer cells in vitro and in vivo. We also performed analysis of miR-20a target genes using target prediction database and genome-wide expression with miR-20a overexpression, and validated the target genes with luciferase assay. Lastly, we show that LIMK1 recapitulates the effects of miR-20a in thyroid cancer cells.

Human thyroid tissue samples and animal experiments
Thyroid tissue samples were snap frozen at the time of thyroidectomy under a protocol approved by the Office of Human Subject Research at the National Institutes of Health Clinical Center, after written informed consent. All tissue samples underwent secondary additional histology review by an endocrine pathologist to confirm the diagnosis and identify samples with greater than 80% tumor cells. Tissue samples were classified as normal, benign (multinodular goiter, follicular adenoma, Hürthle cell adenoma), differentiated thyroid cancer [DTC] (classic PTC, follicular variant of PTC, FTC), and ATC. Normal thyroid tissue was obtained from patients undergoing thyroidectomy for benign or malignant disease from the contralateral thyroid lobe. In all, 8 ATC, 22 DTC, 24 benign, and 11 normal thyroid tissue samples were analyzed.
The National Cancer Institute Animal Care and Use Committee approved the protocols for animal care and handling in the present study. Any mouse experiencing significantly abnormal neurological signs, bleeding from any orifice, impaired mobility, rapid weight loss, debilitating diarrhea, rough hair coat, hunched posture, labored breathing, lethargy, persistent recumbence, jaundice, anemia, self-induced trauma, becomes moribund or otherwise becomes unable to obtain food or water, or with a tumor 2 cm or greater in diameter has been immediately euthanized by CO 2 chamber.

MiRNA transfection
Mature miRNA precursor (pre-miR-20a; Applied Biosystems, Foster City, CA) was transfected into cells at a concentration of 25 nM using Lipofectamine RNAiMAX (Invitrogen, Carlsbad, CA), following the manufacturer's protocol. An oligonucleotide not representing any known miRNA (Pre-miR miRNA Precursor Molecules-Negative Control #1; Applied Biosystems, Foster City, CA) was used as a negative control.

RNA isolation and quantitative real-time RT-PCR
Total RNA was isolated from the cell lines using the TRIzol reagent (Invitrogen, Carlsbad, CA). The TaqMan MiRNA Assay (Applied Biosystems, Carlsbad, CA) was used to measure the miRNA expression level. Total RNA was reverse transcribed with a miRNA-specific primer, followed by real-time PCR with TaqMan probes. U6 was used as an endogenous control. The relative amount of mRNAs in LIM kinase 1 (LIMK1) was determined using the TaqMan Assay (Applied Biosystems, Carlsbad, CA) on an ABI 7900 HT system, and human GAPDH was used as an endogenous control. The DD Ct method was used to calculate expression levels.

Proliferation assay
Cell proliferation was determined using the CyQUANT Cell Proliferation Assay (Invitrogen, Carlsbad, CA), according to the manufacturer's protocol. The fluorescence intensity was measured using a fluorescence microplate reader (Molecular Devices, Sunnyvale, CA), with excitation at 485 nm and emission detection at 538 nm.

Invasion assay
Cellular invasion was measured using the BD BioCoat Matrigel Invasion Chamber (BD Biosciences, Bedford, MA), according to the manufacturer's instructions. Cell culture medium with 10% FBS was used as a chemoattractant in the lower well of the Boyden chamber. After rehydration of the basement membrane, thyroid cancer cells were seeded in the upper compartment of the chamber in serum-free medium (4610 4 cells per well). After incubation at 37uC in 5% CO 2 for 22 hours, the non-invading cells were removed from the upper surface, and the cells that had invaded the membrane to the lower surface were stained with Diff-Quik Stain Set (Siemens Healthcare Diagnostics, Inc., Newark, DE). Images were taken from the membrane of each insert under a microscope (506 magnification) using a digital camera. The images were viewed on the computer screen and the cells in individual fields of each insert were manually counted. The percent of cells invading was determined by counting the number of cells invading through the Matrigel matrix and membrane relative to the number of cells migrating through the membrane of the control inserts without the Matrigel matrix. An invasion index was calculated based on the ratio of the percent of invading cells divided by the percent of invading cells of control cells.

Spheroid culture
Two days after miRNA transfection, FTC-133 cells were trypsinized, counted, re-suspended in culture media, and plated in an Ultra Low Cluster plate (Costar, Corning, NY) at 3.5610 4 per well. The plates were cultured at 37uC in 5% CO 2 , and the medium was changed every 2 to 3 days. After 2 weeks of culture, cells were stained with Crystal Violet and photographed under a microscope. The total area occupied by spheroids within an image was measured by circumscribing the perimeter of each spheroid, marking the entire area, and calculating the pixel numbers using ImageJ software (Maryland, USA).

Tumor xenograft studies
FTC-133 cells transfected with miR-20a or miR-NC were inoculated subcutaneously (10 5 viable cells) in the left and right flanks of athymic nude mice. Tumors were measured two times a week with calipers, and volumes were calculated as length 6width 6 height. Autopsy tumor samples were photographed to document gross morphology, and then samples were weighed.

Migration assay
Thyroid cancer cell migration was assessed using a scratchwound assay. 150,000 cells were transfected with miRNAs (25 nM) or siRNAs (60 nM) and were plated in six-well plates and allowed to attach and grow for 44 hours (miRNAs) or 72 hours (siRNAs). Thereafter, three vertical wounds were made with a sterile 10-ml pipette tip and a horizontal line was made across the three lines so that cells could be observed at the same point. The cells were inspected every 12 hours and measurements taken up to 24 hours.
Genome-wide mRNA expression array FTC-133 cells were transfected with miR-20a and miR-NC. Three days post-transfection, cells were harvested. Total RNA was extracted from cells using Trizol (Invitrogen, USA). RNA quality was ensured using the Agilent RNA 6000 Nano kit and the Bioanalyzer 2100. One-hundred fifty nanograms of total RNA was used to perform cDNA reverse transcription, synthesis, amplification, fragmentation, and terminal labeling with the GeneChip WT Sense Target Labeling and Control Reagents (Affymetrix, Santa Clara, CA). Approximately 25 ng/mL of cDNA was hybridized to the Affymetrix Human Gene 1.0 ST Array GeneChip. The arrays were washed and stained using the fluidics protocol FS450_0007 procedure on an Affymetrix Fluidics Station 450. The probe intensities were scanned with the GeneChip Scanner 3000. The raw data was normalized and analyzed using Partek Genomic Suite (Partek, Inc., St. Louis, MO). Variance analysis was used to determine the probe sets that were significantly different between the two groups. The gene list was filtered with a fold-change cutoff of 1.5, resulting in an output of significant differential expression at p#0.05 and 1.5-fold or more differences.

Luciferase reporter assay
The 1223 base pair 39-UTR of human LIMK1 was cloned into an empty luciferase reporter vector pEZX-MT01 (GeneCopoeia, Rockville, MD), generating a wild-type LIMK1 UTR luciferase reporter construct (pEZX-LIMK1-UTR). For the dual luciferase assay, FTC-133 cells were plated in triplicate into 12-well plates and co-transfected with 0.25 mg of the reporter construct and 15 pmol of miR-20a or miR-NC by using Lipofectamine 2000 (Invitrogen). At 24 hours, the cells were lysed and assayed for both firefly and renilla luciferase using Luc-Pair miR Luciferase Assay Kit (GeneCopoeia, Rockville, MD) on a SpectraMax M5e microplate reader (Molecular Device, Sunnyvale, CA), according to the manufacturers' instructions.

Data analysis
Data is presented as mean 6 standard error of the mean. To determine statistical significance, variance analysis and t test were used, as appropriate. A p value of less than 0.05 was considered statistically significant.

Results
MiR-20a is overexpressed in anaplastic thyroid cancer (ATC) We found the expression level of miR-20a was significantly higher in ATC than in DTC, benign and normal thyroid tissues (Fig. 1). There was no significant difference in miR-20a expression level by BRAF mutation status (p = 0.62) or extent of disease (p = 0.70 for tumor size; p = 0.12 for lymph node metastasis) in DTC or PTC.

MiR-20a regulates thyroid cancer cell proliferation, spheroid formation, and invasion
We overexpressed miR-20a in four thyroid cancer cell lines (TPC-1, XTC-1, FTC-133, and C643) using miR-NC as a negative control to determine its effect on cell proliferation. MiR-20a overexpression significantly inhibited cell proliferation by 24% in TPC-1 cells at 144 hours (p,0.001), 34% in XTC-1 cells at 144 hours (p,0.001), 22% in FTC-133 cells at 144 hours (p, 0.001), and 22% in C643 cells at 216 hours ( Fig. 2A-D). We evaluated the effect of miR-20a on tumor growth in vivo. We found that tumor xenografts derived from FTC-133 cells transfected with miR-20a were significantly smaller than tumor xenografts from the miR-NC group (p,0.01) (Fig. 2E), and the tumor weights derived from FTC-133 cells transfected with miR-20a were also significantly less than the tumor weights in the miR-NC group (p,0.05) (Fig. 2F).
We also studied the effect of miR-20a on thyroid cancer cell tumor spheroid formation. The FTC-133 cell line forms spheroids when cultured in ultra-low adherent culture flask and with miR-20a transfection, the number and size of spheroids were significantly decreased (Fig. 3).

MiR-20a regulates LIMK1 expression in thyroid cancer cells
Given that miR-20a had an effect on cell proliferation and invasion in vitro and in vivo, we were interested in determining the target gene(s) of miR-20a. We used two approaches to determine miR-20a targets: (1) a target prediction database and (2) genomewide expression analysis with miR-20a overexpression. We found 3635 predicted target genes for miR-20a using the TargetScan 5.0 software. We found 58 genes with altered expression upon miR-20a overexpression using genome-wide expression analysis. Among the 45 genes whose expression level was down-regulated, 37 genes were predicted target genes of miR-20a ( Table 1). LIMK1 was the most downregulated gene from this analysis and has been previously reported to have a role in tumor cell invasion and metastasis [19][20][21][22]. Thus, we were interested in determining whether LIMK1 was a direct target of miR-20a. We found that LIMK1 protein expression in thyroid cancer cell lines (C643, XTC-1, FTC-133, and TPC-1) was decreased with miR-20a overexpression (Fig. 5A). The decrease in LIMK1 protein expression was observed for up to 14 days after transfection (Fig. 5B).
To determine if whether LIMK1 was a direct target of miR-20a, we used a luciferase reporter vector pEZX-MT01 with the 39- UTR of human LIMK1 cloned into it, generating a LIMK1 39-UTR luciferase reporter construct (pEZX-LIMK1-UTR). We performed luciferase assays with the pEZX-LIMK1-UTR (vector with 39-UTR of LIMK1) co-transfected into the FTC-133 cell line with miR-20a or miR-NC. We found significantly decreased luciferase activity with miR-20a overexpression as compared to negative control (Fig. 6A), suggesting that miR-20a directly downregulates LIMK1 expression. Given that miR-20a overexpression downregulated LIMK1 in thyroid cancer cell lines and the most prominent effect of miR-20a on thyroid cancer cells was the  Table 1. Genes identified to be regulated by miR-20a using both microarray analysis and target scan analysis.* *Genes listed were common to genome-wide gene expression analysis and target scan database, and based on change in gene expression of 2-fold or greater with adjusted p value of 0.05. In Table S1 is the entire gene list with 1.5-fold or greater change in expression, with adjusted p value of 0.05, and genes which are predicted targets by target scan analysis. doi:10.1371/journal.pone.0096103.t001 inhibition of cellular invasion, we explored whether LIMK1 has an effect on cellular invasion and migration. We found that knockdown of LIMK1 resulted in decreased cellular invasion but not migration ( Fig. 6B and C).

Discussion
In this study, we found miR-20a was overexpressed in ATC as compared to DTC, benign and normal thyroid tissue. Ectopic overexpression of miR-20a significantly inhibited thyroid cancer cell proliferation in vitro and in vivo, and significantly inhibited tumor spheroid formation and invasion in multiple thyroid cancer cell lines. This suggests that miR-20a has a tumor suppressive function when it is upregulated in thyroid cancer. We also found that miR-20a regulates LIMK1 expression, suggesting that LIMK1 is a target gene that may mediate the suppressive effects of miR-20a on growth and invasion of thyroid cancer cells. However, one limitation of our study is the small number of ATC tumor samples analyzed but it is a rare malignancy.
Takakura and associates reported that the miR-17-92 cluster (miR-17-3p, -17-5p, -18a, -19a, -20a, -19b, and -92-1) was overexpressed in ATC cell lines [16]. Using quantitative RT-PCR, they showed that miR-17-3p and miR-17-5p were overexpressed in three of six ATC tissue samples compared to normal tissue samples. They reported that transfection of inhibitors of miR-17-5p suppressed the expression level of the miR-17 family (miR-17-5p, miR-20a, and miR-106a and b) in ARO cells, resulting in cell growth reduction. However, our study of the function of miR-20a in thyroid cancer was different than the study conducted by Takakura and colleagues. First, we specifically overexpressed miR-20a to understand its effect on tumor cell biology in both undifferentiated and differentiated thyroid cancer cell lines, and we did not use inhibitors of multiple members of the miR-17-92 cluster with possible off target effects, which cannot be only selective to miR-20a. Second, we noticed that Takakura et al. used cells treated only with the transfection reagent as the negative control instead of using scrambled oligonucleotides, and we used scrambled oligonucleotides as the negative control. Additionally, the cell lines we used (C643, TPC-1, FTC-133 and XTC-1) were different than the cell lines used (ARO and FRO) by Takakura and colleagues. Lastly, the ARO cell lines used in the study by Takakura and associates may not have been authenticated thyroid cancer cell lines [18].
We found that miR-20a regulates LIMK1 expression in thyroid cancer cell lines. LIMK1 is regulated by the Rho signaling pathway, and it modulates actin dynamics by regulating the activity of the cofilin family proteins [30,31]. Previous studies have shown that LIMK1 plays a central and important role in tumor cell invasion and metastasis [19][20][21][22]. LIMK1 overexpression increases the invasiveness of breast and prostate cancer cells in vitro and in vivo, and knocking down of LIMK1 suppresses breast and prostate cancer cell invasion in vitro and in vivo [19][20][21][22]32]. Based on the results from our genome-wide gene expression and target prediction analyses, we asked whether miR-20a overexpression affects LIMK1 expression in thyroid cancer cell lines. Indeed, we found that LIMK1 in all thyroid cancer cell lines (C643, XTC-1, FTC-133, and TPC-1) was inhibited with miR-20a overexpres-sion, which suggests that the suppressive effect of miR-20a on cellular proliferation and invasion may be mediated by its effect on LIMK1. Indeed, direct knockdown of LIMK1 had the same effects on cellular invasion and migration as observed with the overexpression of miR-20a.
Given the tumor suppressive effect of miR-20a in thyroid cancer cells we observed in vitro and in vivo, it is possible that successful delivery of miR-20a could result in tumor suppression/regression regardless of the cell type and or basal miR-20a levels [33]. The tumor suppressive effects of miR-20a could also be mediated by other genes than LIMK1. We validated LIMK1 as a target because it had the lowest expression with miR-20a overexpression but as listed in Table 1 many candidate target genes were altered with miR-20a overexpression and thus could also mediate its tumor suppressive effects.
To our knowledge, this is the first study to characterize the effect of miR-20a on thyroid cancer cell phenotypes and to show that miR-20a regulates LIMK1 expression. Our findings suggest the upregulated expression of miR-20a in anaplastic thyroid cancer counteracts thyroid cancer progression and may have therapeutic potential [34].