Tel: +81-43-226-2971, Fax: +81-43-226-2975
Cancer Cell Biology
miR-145, miR-133a and miR-133b: Tumor-suppressive miRNAs target FSCN1 in esophageal squamous cell carcinoma
Article first published online: 2 MAR 2010
Copyright © 2010 UICC
International Journal of Cancer
Volume 127, Issue 12, pages 2804–2814, 15 December 2010
How to Cite
Kano, M., Seki, N., Kikkawa, N., Fujimura, L., Hoshino, I., Akutsu, Y., Chiyomaru, T., Enokida, H., Nakagawa, M. and Matsubara, H. (2010), miR-145, miR-133a and miR-133b: Tumor-suppressive miRNAs target FSCN1 in esophageal squamous cell carcinoma. Int. J. Cancer, 127: 2804–2814. doi: 10.1002/ijc.25284
- Issue published online: 2 MAR 2010
- Article first published online: 2 MAR 2010
- Manuscript Accepted: 15 FEB 2010
- Manuscript Received: 29 SEP 2009
- Ministry of Education, Science, Sports and Culture Grants-in-Aid for Scientific Research (C). Grant Number: 21592187, 2009
- esophageal squamous cell carcinoma;
MicroRNAs (miRNAs), noncoding RNAs 21–25 nucleotides in length, regulate gene expression primarily at the posttranscriptional level. Growing evidence suggests that miRNAs are aberrantly expressed in many human cancers, and that they play significant roles in carcinogenesis and cancer progression. A search for miRNAs with a tumor-suppressive function in esophageal squamous cell carcinoma (ESCC) was performed using the miRNA expression signatures obtained from ESCC clinical specimens. A subset of 15 miRNAs was significantly downregulated in ESCC. A comparison of miRNA signatures from ESCC and our previous report identified 4 miRNAs that are downregulated in common (miR-145, miR-30a-3p, miR-133a and miR-133b), suggesting that these miRNAs are candidate tumor suppressors. Gain-of-function analysis revealed that 3 transfectants (miR-145, miR-133a and miR-133b) inhibit cell proliferation and cell invasion in ESCC cells. These miRNAs (miR-145, miR-133a and miR-133b), which have conserved sequences in the 3′UTR of FSCN1 (actin-binding protein, Fascin homolog 1), inhibited FSCN1 expression. The signal from a luciferase reporter assay was significantly decreased at 2 miR-145 target sites and 1 miR-133a/b site, suggesting both miRNAs directly regulate FSCN1. An FSCN1 loss-of-function assay found significant cell growth and invasion inhibition, implying an FSCN1 is associated with ESCC carcinogenesis. The identification of tumor-suppressive miRNAs, miR-145, miR-133a and miR-133b, directly control oncogenic FSCN1 gene. These signal pathways of ESCC could provide new insights into potential mechanisms of ESCC carcinogenesis.
Human esophageal cancer occurs worldwide with a variable geographic distribution. The disease ranks eighth in order of occurrence and sixth as the leading cause of cancer mortality, affecting men more than women.1 There are 2 main forms, each with distinct etiologic and pathologic characteristics: esophageal squamous cell carcinoma (ESCC) and adenocarcinoma. ESCC is the most frequent subtype of esophageal cancer, although the incidence of adenocarcinoma in the western world is increasing faster than other malignancies. ESCC is one of the most aggressive and lethal malignancies in eastern Asia. Because most cases of ESCC are not diagnosed until the disease is at an advanced stage, the overall 5-year survival rate is very poor.2–4 Recently, the combination of chemotherapy and radiotherapy, alone or as an adjunct to surgery, has improved the prognosis of ESCC patients.5, 6 Research over the last 20 years has identified a number of oncogenic and tumor-suppressor proteins that are associated with the induction of ESCC.7, 8 However, molecular indicators of the origin of cellular deregulation in ESCC have not been identified. Elucidation of the molecular pathways involved in ESCC carcinogenesis could lead to improvements in disease diagnosis and therapy.
MicroRNAs (miRNAs) are a class of naturally occurring small (21–25 nucleotides) noncoding RNAs. Mature miRNAs play important regulatory roles in cell growth, proliferation, differentiation and cell death.9–12 miRNAs bind through partial sequence homology to the 3′ untranslated region of target messenger RNAs and either block translation or promote mRNA degradation. miRNAs can function either as oncogenes or tumor suppressors and are aberrantly expressed in several types of human cancer. Upregulated miRNAs in cancer may function as oncogenes by negatively regulating tumor suppressor genes. In contrast, downregulated miRNAs may normally function as tumor suppressor genes and inhibit cancer by regulating oncogenes. Growing evidence has indicated that unique miRNA expression profiles for each cancer type would be a useful biomarker for cancer diagnosis and prognosis.13–16 Genome-wide miRNA expression studies in esophageal cancer have been published17–19; however, further analysis is needed to clarify the role of miRNAs in esophageal cancer.
In our study, we identified a subset of 15 miRNAs that were significantly downregulated in ESCC. Three miRNAs (miR-145, miR-133a and miR-133b) have tumor-suppressive function and directly control the oncogenic actin-binding protein, Fascin homolog 1 gene (FSCN1). Furthermore, FSCN1 expression is inversely correlated with miR-133a and miR-133b expression levels in clinical ESCC specimens. Insight into the association between miRNAs and their target gene networks could enhance our understanding of the molecular mechanism of ESCC carcinogenesis.
Material and Methods
Clinical ESCC specimens and ESCC cell culture
Ten pairs of primary esophageal squamous cell carcinoma and corresponding normal esophageal epitheliums were obtained from patients in Chiba University Hospital from 2008 to 2009. All tissue specimens were obtained from untreated patients undergoing primary surgical treatment. Normal tissues were obtained far from the center of the cancer in surgical specimens. No cancer cells were detected in neighboring formalin-fixed paraffin-embedded specimens. Written consent of tissue donation for research purposes was obtained from each patient before tissue collection. The protocol was approved by the Institutional Review Board of Chiba University. The specimens were snap frozen in liquid nitrogen and stored at −80°C.
Human ESCC cell lines (TE2, TE3, TE12 and TE13) were provided by the Cell Resource Center for Biomedical Research Institute of Development, Aging and Center, Tohoku University, Japan. All cells were grown in Dulbecco's Modified Eagle's Medium/Nutrient Mixture F-12 Ham (DMEM/F-12) supplemented with 10% fetal bovine serum in a humidified atmosphere containing 5% CO2 at 37°C.
Tissues and cells were treated with TRIzol reagent (Invitrogen, Carlsbad, CA), according to the manufacturer's instructions, for total RNA extraction. RNA concentration was determined spectrophotometrically, and integrity was checked by gel electrophoresis. RNA quality was confirmed in an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA).
miRNA expression signature and data normalization
miRNA expression patterns were evaluated using the TaqMan Low Density Array Human MicroRNA Panel v1.0 (Applied Biosystems, Foster City, CA). The assay is composed of 2 steps: generation of cDNA by reverse transcription and a TaqMan real-time PCR assay. In brief, miRNAs in the samples were converted into cDNA using 365 specific stem-loop reverse transcription primers. After cDNA conversion, the quantity of mature miRNAs was evaluated using specific TaqMan real-time PCR primers and probes. Real-time PCR was performed in duplicate using GeneAmp Fast PCR Master Mix (Applied Biosystems) and the ABI 7900HT Real-Time PCR System. The comparative CT method was used to determine the expression levels of mature miRNAs. Real-time PCR theory and the list of human miRNAs can be found on the company's website.
Cell proliferation and invasion assays in ESCC cell lines
Mature miRNA molecules, Pre-miR™ miRNA precursors (hsa-miR-145, hsa-miR-30a-3p, hsa-miR-133a, hsa-miR-133b and negative control miRNA) (Applied Biosystems) and si-FSCN1 (LU-019576-00 and J-019576-08) and negative control siRNA (D-001810-10) (Thermo Fisher Scientific, Waltham, MA) were incubated with Lipofectamine™ RNAiMAX (Invitrogen, Tokyo, Japan) before transfection. As recommended by the manufacturer, we first tested the transfection efficacy of miRNA in ESCC cell lines based on the downregulation of protein tyrosine kinase 9 (PTK9) mRNA after transfection with miR-1 (This method was recommended by the manufacture).
Cells were reverse transfected with 10 nM miRNA or si-FSCN1 and plated in 96-well plates at 3 × 103 cells per well. After 72 hr, cell viability was determined by XTT assay using the Cell Proliferation Kit II (Roche Molecular Biochemicals, Mannheim, Germany) as previously described.20 In each treatment group, 3 wells were assayed for cell viability.
A cell invasion assay was carried out using modified Boyden Chambers containing transwell-precoated matrigel membrane filter inserts with 8-μm pores in 24-well tissue culture plates (BD Biosciences, Bedford, MA). All experiments were performed in triplicate.
Screening of miR-145 target genes by microarray analysis
A genome-wide screen for target genes silenced by miR-145 in cancer cells compared the expression profiles of miR-145-transfected cancer cells with miRNA-negative control transfectants. Oligo-microarray Human 44K (Agilent) was used for expression profiling in miR-145 transfectants in ESCC cell lines TE2 and TE13, as described previously.21 The predicted target genes and their conserved sites, where the seed region of each miRNA binds, were investigated using TargetScan (release 5.1, http://www.targetscan.org/). The sequences of the predicted mature miRNAs were confirmed using miRBase (release 13.0, http://microrna.sanger.ac.uk/).
Detection of FSCN1 mRNA and protein expression levels
The first strand of cDNA was synthesized using 1 μg of total RNA and random primers from the Reverse Transcription (RT) System (Promega, Tokyo, Japan). Gene-specific PCR products were assayed continuously using a 7900 Real-Time PCR System (Applied Biosystems) according to the manufacturer's protocol. The initial PCR step was a 10-min hold at 95°C; the cycles (n = 40) consisted of a 15-sec denaturation step at 95°C followed by 1 min of annealing/extension at 63°C. TaqMan® probes, and primers for glyceraldehydes-3 phosphate dehydrogenase (GAPDH) (P/N: Hs99999905_m1) and FSCN1 (P/N: Hs00818825_m1) were Assay-On-Demand Gene Expression Products (Applied Biosystems). All reactions were performed in triplicate and included a negative control lacking cDNA.
Two days after transfection, 50 μg of protein lysate was separated by NuPAGE on a 4–12% bis-tris gel (Invitrogen) and transferred to PVDF membrane. Immunoblotting was performed with diluted (1:100) monoclonal FSCN1 antibody (ab49815, abcam, UK). The membrane was washed before the membrane was incubated with goat anti-mouse IgG (H+L)-HRP conjugate (Bio-Rad, Hercules, CA). Specific complexes were visualized with an echochemiluminescence detection system (GE Healthcare Bio-Sciences, Princeton, NJ).
Cloning of the 3′UTR of FSCN1 and luciferase reporter assay
The dual-luciferase plasmid, psiCHECK-2 vector was obtained from Promega (Promega, Madison, WI). miRNA target sequences were inserted between the XhoI-PmeI restriction sites in the multiple cloning region in the 3′UTR of the hRluc gene. The 3′UTR region of FSCN1 contains 4 putative miR-145 recognition sites and 1 133a/b recognition site amplified from the genomic DNA of cancer cells. The amplified fragment was cloned into the psiCHECK-2 vector and confirmed by sequencing using the following primers:
116-123F, GATCGCTCGAGACTCCTGTGGACTTCTTC TTCG
116-123R, CTCTAGGTTTAAACGCGGAGAAGGGGTT AGCA
377-384F, CGATCGCTCGAGTCTGGCACCTCTTTCTT CTGA
377-384R, GGCCGCTCTAGGTTTAAACGACATGTGCC CAGCTCTCCT
729-735R, GCTCTAGGTTTAAACGGACGCCTCCAGCA ATAATA
1140-1147F, GATCGCTCGAGCTCTGGGTGTCTTGGT CTT
1140-1147R, CGCTCTAGGTTTAAACGGGGCTGCAGA CTGAGTTAT.
TE2 cells were transfected with 5 ng of vector, 10 nM of miRNAs and 1 μl of Lipofectamine 2000 (Invitrogen) in 100 μl of Opti-MEM I Reduced-Serum Medium (Invitrogen). The activities of firefly and Renilla luciferases in cell lysates were determined with a dual-luciferase assay system (Promega, Tokyo, Japan). Normalized data were calculated as the quotient of Renilla activity/firefly luciferase activity.
Western blot analysis
Primary mouse monoclonal antibodies against FSCN1 (abcam) were diluted to 1:200. The slides were treated with Biotinylated Anti-Mouse IgG (H+L), made in horse (Vector Laboratories, Burlingame, CA). The chromagen was diaminobenzidine-hydrogen peroxide (DAB; Sigma-Aldrich, St. Louis, MO), and the counterstain agent was 0.5% hematoxylin. Staining of the endothelia and myofibroblasts served as internal positive controls.
The relationship between the 2 variables and numerical values obtained by real-time RT-PCR was analyzed using the Mann-Whitney U test. The relationships between miRNAs expression and FSCN1 expression were analyzed using the Spearman rank correlation. Statistical significance was defined as p < 0.05. All statistical analyses were performed using Expert StatView (version 4, SAS Institute, Cary, NC).
Identification of downregulated miRNAs in ESCC specimens
The expression of 365 mature miRNAs was evaluated in 10 matched pairs of cancer and noncancerous epithelium from ESCC patients (Table 1, patient numbers 1–10). After global normalization of the raw data, we identified 18 downregulated miRNAs between cancer and normal tissues when we used a cutoff p value of <0.05 and value of tumor/normal fold change of <0.2. Use of a different normalization technique using internal references RNU44 and RNU48 revealed 15 commonly downregulated miRNAs (Table 2). Recently, we identified a subset of 7 downregulated miRNAs based on miRNA expression signatures in bladder cancer.20 A comparison of the ESCC and bladder cancer miRNA signatures revealed 4 miRNAs (miR-145, miR-30a-3p, miR-133a and miR-133b) that were downregulated in both cancers (Table 2, underlined).
Effect of miRNA transfection in ESCC cell lines
The expression levels of 4 miRNAs (miR-145, miR-30a-3p, miR-133a and miR-133b) were extremely low in the ESCC cell lines (data not shown), suggesting that the corresponding endogenous miRNAs do not affect viability in these cell lines. In a gain-of-function analysis, these mature miRNA molecules were separately transfected into 4 ESCC cell lines (TE2, TE3, TE12 and TE13).
The XTT proliferation assay showed significant cell growth inhibition of miR-145 transfectants in ESCC cell lines: 79% in TE2, 75% in TE3, 71% in TE12 and 63% in TE13 (p < 0.05) (Fig. 1a). Transfection of miR-133a and miR-133b also inhibited cell growth in ESCC cell lines (miR-133a83% in TE2, 80% in TE3, 79% in TE12 and 63% in TE13, p < 0.05; miR-133b: 82% in TE2, 80% in TE3, 79% in TE12 and 70% in TE13, p < 0.05) (Fig. 1a). However, no significant reduction in growth was observed in miR-30a-3p transfectants in ESCC cell lines (97% in TE2, 96% in TE3 and 102% in TE12). Only in TE13 cells did miR-30a-3p transfection inhibit cell growth (83%, p < 0.05).
The matrigel invasion assay demonstrated significant inhibition of cell invasion in the miR-145 (34% in TE2, 42% in TE3, 46% in TE12 and 22% in TE13), miR-133a (58% in TE2, 50% in TE3, 38% in TE12 and 33% in TE13) and miR-133b (70% in TE2, 45% in TE3, 42% in TE12 and 52% in TE13) transfectants (Fig. 1b). No significant difference was observed in miR-30a-3p transfectants in any of the ESCC cell lines (92% in TE2, 88% in TE3, 91% in TE12 and 88% in TE13).
FSCN1 as a target of miR-145, miR-133a and miR-133b
Genome-wide gene expression of target genes silenced by miR-145 was examined using oligo microarray analysis of miR-145-transfected ESCC cell lines (TE2 and TE13). A total of 51 genes were downregulated less than 0.5-fold (log2 ratio > −0.1) in both miR-145-transfectant cell lines compared to the control. FSCN1 was listed as the top candidate gene in the expression profiles (Table 3) and was, therefore, the focus of further experiments. Entries from the microarray data were approved by the Gene Expression Omnibus (GEO) and were assigned GEO accession number, GSE20028.
TE2 cells were used to determine whether the suppression of miRNAs (miR-145, miR-133a and miR-133b) affected FSCN1 expression. First, the 3′UTR of FSCN1 was examined to determine if the entire region is a functional target of miR-145 and miR-133a. The miR-133b miRNA was not examined because of its functional sequence similarity to miR-133a (miR-133a: UUGGUCCCCUUCAACCAGCUGU, miR-133b: UUGGUCCCCUUCAACCAGCUA). The luciferase reporter assay confirmed the 3′UTR of FSCN1 as the actual target of miR-145 and miR-133a (Fig. 2, lower). Next, 5 conserved sites (4 in miR-145 and 1 in miR-133a/b) in the 3′UTR of FSCN1 were confirmed as actual functional sites. Luciferase activity was significantly decreased in 2 miR-145 target sites (position 377–383 and 1140–1146 in 3′UTR of FSCN1) and the miR-133a/b site (position 745–751 in 3′UTR of FSCN1), suggesting that these 3 specific sites are targeted by the 2 miRNAs (Fig. 2, lower).
The expression level of FSCN1 mRNA was significantly decreased in the 2 ESCC cell lines (TE2 and TE13) transfected with miR-145, miR-133a and miR-133b (18, 41 and 45% of control in TE2 cells; 39, 42 and 45% in TE13 cells) (Fig. 3a). The protein expression level was also markedly reduced in ESCC cell lines transfected with miR-145, miR-133a and miR-133b transfectants (Fig. 3b).
Effect of FSCN1 loss-of-function on cell proliferation and invasion in ESCC cell lines
A loss-of-function assay using si-RNA analysis was performed to examine the oncogenic function of FSCN1 that is directly targeted by miR-145, miR-133a and miR-133b. We determined whether si-FSCN1 was reduced both mRNA and protein expression levels of si-FSCN1 transfectants in ESCC cell lines, TE2 and TE13. After 72-hr transfection, mRNA and protein of FSCN1 were successfully reduced by the si-FSCN-1 and si-FSCN1-2 (Fig. 4a). The XTT assay revealed significant cell growth inhibition in the si-FSCN1 transfectants compared to the si-RNA control transfectant (63 and 60% of control in TE2 cells; 54 and 58% in TE13 cells) (Fig. 4b). The matrigel invasion assay showed that the number of invading cells was significantly decreased in si-FSCN1 transfectants compared to control cells (28 and 22% of control in TE2 cells; 18 and 19% in TE13 cells) (Fig. 4c).
FSCN1 mRNA expression in ESCC clinical specimens
The clinicopathological findings of 20 ESCC patients using miRNA and FSCN1 expression are shown in Table 1. Total RNA was isolated from matched noncancerous esophageal epithelium and ESCC tissues, from which miRNAS and FSCN1 mRNA expression levels were determined by TaqMan real-time PCR (Fig. 5a). Of 20 matched normal and cancer tissues, the expression levels of miR-145, miR-133a and miR-133b were significantly lower in cancer tissues compared to noncancerous tissues. In contrast, FSCN1 expression was significantly higher in ESCC tissues than in noncancerous tissues.
The expression level correlations of FSCN1 and the 3 miRNAs (miR-145, miR-133a and miR-133b) were tested using the Spearman rank correlation. There was a significant inverse correlation between FSCN1 and miR-133a or miR-133b expression levels, p = 0.022 and p = 0.033, respectively (Fig. 5b). However, a significant inverse correlation was not observed between FSCN1 and miR-145 expression levels, p = 0.099.
miRNAs can control the expression levels of target genes. Thus, dysregulation of miRNAs is expected in human diseases, such as cancer, which are attributed to dysregulated gene expression of tumor suppressors and oncogenes. miRNA alteration could contribute to human carcinogenesis. Investigation of the differentially expressed miRNAs in cancer specimens has yielded important information on carcinogenesis. ESCC is one of the most lethal malignancies in the world, with a 5-year survival rate of 5–25% after curative surgery.22, 23 An understanding of the molecular pathways involved in ESCC carcinogenesis would help improve diagnosis and therapy of the disease.
We screened 365 miRNAs, from ESCC and normal epithelial specimens, using stem-loop RT-PCR, and identified 15 downregulated miRNAs (miR-375, let-7c, miR-145, miR-143, miR-100, miR-133a, miR-99a, miR-133b, miR-1, miR-30a-3p, miR-504, miR-139-5p, miR-204, miR-203 and miR-326) in cancer tissues. We previously identified 27 miRNAs that were up- or downregulated in bladder cancer.20 Seven of these miRNAs (miR-145, miR-30a-3p, miR-133a, miR-133b, miR-195, miR-125b and miR-199*) were reduced in bladder cancer. Stem-loop RT-PCR confirmed that these miRNAS were significantly downregulated in 104 bladder cancer specimens when compared to 31 normal epithelial samples. Expression profiles have also shown a reduction of miR-145 in colorectal and prostate cancers.24–26 In lung adenocarcinoma, inhibition of cell growth was observed in EGFR mutant cancer cells after transfection of miR-145.27 Examinations of these expression profiles have uncovered downregulated miRNAs, such as miR-145, miR-133a and miR-133b, that are common to various cancers and have led to the suggestion that miRNAs function as tumor suppressors. Our results support this hypothesis, in that significant decreases in cell growth and invasive activity were observed in ESCC cell lines after transfection with miR-145, miR-133a and miR-133b. Similar results were observed with the bladder cancer cell lines (data not shown).
miRNAs regulate a variety of cellular pathways by regulating the expression of multiple target genes.28 Therefore, we carried out genome-wide gene expression analysis to find miR-145 targets. The microarray data obtained after miR-145 transfection of cancer cells led us to focus on FSCN1. Interestingly, TargetScan miRNA database found 4 conserved sequence sites for miR-145 and 1 for miR-133a/b in the 3′UTR of FSCN1.
FSCN1 is a 55-kDa globular protein that organizes F-actin into well-ordered, tightly packed parallel bundles in cells. Vertebrate genomes encode 3 forms of FSCN: FSCN1, which is widely expressed by mesenchymal tissues and in the nervous system; FSCN2, which is expressed by retinal photoreceptor cells; and FSCN3, which is testis specific.29 FSCN1 contributes to the organization of 2 major forms of actin-based structures: cortical cell protrusions that mediate cell interactions and migration, and cytoplasmic microfilament bundles that contribute to cell architecture and intracellular movements. Many of the studies described above demonstrate how closely the formation of FSCN1-containing protrusions is linked with the activation of cell migration by physiological stimuli. Two new biological contexts where FSCN1 participates in motility and invasion have recently emerged: the movement of intracellular pathogenic bacteria and the clinical aggressiveness of human carcinomas. With the exception of the basal layer of the epidermis, FSCN1 has not been reported in normal adult epithelia. During mouse development, the equivalent epithelia are also FSCN1 negative.29 A recent burst of publications have indicated that upregulated FSCN1 protein was observed in human carcinomas from many body sites. In studies of NSCLC and gastric adenocarcinoma, high FSCN1 expression within the tumor correlated with poor survival.30, 31 Increased FSCN1 content has also been documented in other invasive tumor types such as high-grade astrocytoma.32 In ESCC, FSCN1 protein was usually increased in the tumor when compared to normal epithelium, and FSCN1 overexpression was significantly associated with poor prognosis.33 These data implicate FSCN1 in the progression to an invasive phenotype and raise the possibility that FSCN1 could be of value as a novel prognostic marker for early identification of aggressive carcinomas.
The control of FSCN1 expression in epithelial neoplasmas has not been described. This is the first study to show that miR-145 and miR-133a/b directly regulate FSCN1 and contribute to cellular proliferation and invasion in ESCC. The miRNA expression signature and identification of candidate target genes may provide an understanding of potential carcinogenic mechanisms in ESCC. These findings have therapeutic implications and may be exploited for future treatment of ESCC.