YM
Yin‐Yuan Mo
Author with expertise in MicroRNA Regulation in Cancer and Development
Achievements
Cited Author
Open Access Advocate
Key Stats
Upvotes received:
0
Publications:
15
(93% Open Access)
Cited by:
5,201
h-index:
44
/
i10-index:
70
Reputation
Biology
< 1%
Chemistry
< 1%
Economics
< 1%
Show more
How is this calculated?
Publications
0

MicroRNA-21 Targets the Tumor Suppressor Gene Tropomyosin 1 (TPM1)

Shuomin Zhu et al.Mar 15, 2007
MicroRNAs are small noncoding RNA molecules that control expression of target genes. Our previous studies show that mir-21 is overexpressed in tumor tissues compared with the matched normal tissues. Moreover, suppression of mir-21 by antisense oligonucleotides inhibits tumor cell growth both in vitro and in vivo. However, it remains largely unclear as to how mir-21 affects tumor growth, because our understanding of mir-21 targets is limited. In this study, we performed two-dimensional differentiation in-gel electrophoresis of tumors treated with anti-mir-21 and identified the tumor suppressor tropomyosin 1 (TPM1) as a potential mir-21 target. In agreement with this, there is a putative mir-21 binding site at the 3′-untranslated region (3′-UTR) of TPM1 variants V1 and V5. Thus, we cloned the 3′-UTR of TPM1 into a luciferase reporter and found that although mir-21 down-regulated the luciferase activity, anti-mir-21 up-regulated it. Moreover, deletion of the mir-21 binding site abolished the effect of mir-21 on the luciferase activity, suggesting that this mir-21 binding site is critical. Western blot with the cloned TPM1-V1 plus the 3′-UTR indicated that TPM1 protein level was also regulated by mir-21, whereas real-time quantitative reverse transcription-PCR revealed no difference at the mRNA level, suggesting translational regulation. Finally, overexpression of TPM1 in breast cancer MCF-7 cells suppressed anchorage-independent growth. Thus, down-regulation of TPM1 by mir-21 may explain, at least in part, why suppression of mir-21 can inhibit tumor growth, further supporting the notion that mir-21 functions as an oncogene. MicroRNAs are small noncoding RNA molecules that control expression of target genes. Our previous studies show that mir-21 is overexpressed in tumor tissues compared with the matched normal tissues. Moreover, suppression of mir-21 by antisense oligonucleotides inhibits tumor cell growth both in vitro and in vivo. However, it remains largely unclear as to how mir-21 affects tumor growth, because our understanding of mir-21 targets is limited. In this study, we performed two-dimensional differentiation in-gel electrophoresis of tumors treated with anti-mir-21 and identified the tumor suppressor tropomyosin 1 (TPM1) as a potential mir-21 target. In agreement with this, there is a putative mir-21 binding site at the 3′-untranslated region (3′-UTR) of TPM1 variants V1 and V5. Thus, we cloned the 3′-UTR of TPM1 into a luciferase reporter and found that although mir-21 down-regulated the luciferase activity, anti-mir-21 up-regulated it. Moreover, deletion of the mir-21 binding site abolished the effect of mir-21 on the luciferase activity, suggesting that this mir-21 binding site is critical. Western blot with the cloned TPM1-V1 plus the 3′-UTR indicated that TPM1 protein level was also regulated by mir-21, whereas real-time quantitative reverse transcription-PCR revealed no difference at the mRNA level, suggesting translational regulation. Finally, overexpression of TPM1 in breast cancer MCF-7 cells suppressed anchorage-independent growth. Thus, down-regulation of TPM1 by mir-21 may explain, at least in part, why suppression of mir-21 can inhibit tumor growth, further supporting the notion that mir-21 functions as an oncogene. MicroRNAs (miRNAs) 2The abbreviations used are: miRNA, microRNA; siRNA, short interfering RNA; RT, reverse transcription; qRT, quantitative reverse transcription; GFP, green fluorescent protein; EGFP, enhanced green fluorescent protein; TPM, tropomyosin; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazoliumbromide; 5-aza-dC, 5-aza-2′-deoxycytidine; UTR, untranslated region; 2-DIGE, two-dimensional differentiation in-gel. are a class of naturally occurring small noncoding RNAs that regulate gene expression by targeting mRNAs for translational repression or cleavage (1Pillai R.S. RNA. 2005; 11: 1753-1761Crossref PubMed Scopus (604) Google Scholar, 2Zamore P.D. Haley B. Science. 2005; 309: 1519-1524Crossref PubMed Scopus (1132) Google Scholar). Like protein-coding mRNAs, miRNAs are transcribed as long primary transcripts in the nucleus. However, unlike protein-coding mRNAs, miRNAs are subsequently cleaved to produce stem-loop-structured precursor molecules of ∼70 nucleotides in length (pre-miRNAs) by the nuclear RNase III enzyme Drosha (3Kim V.N. Nat. Rev. Mol. Cell. Biol. 2005; 6: 376-385Crossref PubMed Scopus (2013) Google Scholar). The pre-miRNAs are then exported to the cytoplasm, where the RNase III enzyme Dicer further processes them into mature miRNAs (∼22 nucleotides). Thus, miRNAs are related to, but distinct from, short inferring RNAs (siRNAs) (4Bartel D.P. Cell. 2004; 116: 281-297Abstract Full Text Full Text PDF PubMed Scopus (29863) Google Scholar, 5Fitzgerald K. Curr. Opin. Drug Discovery Dev. 2005; 8: 557-566PubMed Google Scholar). A key difference between siRNAs and miRNAs is that siRNAs require almost identical sequences to targets to exert their silencing function, whereas miRNAs bind through partial sequence homology to the 3′-untranslated region (3′-UTR) of target genes. Because of this unique feature, a single miRNA has multiple targets. Thus, miRNAs could regulate a large fraction of protein-coding genes, and as high as 30% of all genes could be miRNA targets (6Lewis B.P. Burge C.B. Bartel D.P. Cell. 2005; 120: 15-20Abstract Full Text Full Text PDF PubMed Scopus (9936) Google Scholar). As a new layer of gene regulation mechanism, miRNAs have diverse functions, including the regulation of cellular differentiation, proliferation, and apoptosis (7Croce C.M. Calin G.A. Cell. 2005; 122: 6-7Abstract Full Text Full Text PDF PubMed Scopus (1221) Google Scholar, 8Chen C.Z. Li L. Lodish H.F. Bartel D.P. Science. 2004; 303: 83-86Crossref PubMed Scopus (2804) Google Scholar). Hence, deregulation of miRNA expression may lead to a variety of disorders. Aberrant expression of miRNAs in cancer has been well documented (7Croce C.M. Calin G.A. Cell. 2005; 122: 6-7Abstract Full Text Full Text PDF PubMed Scopus (1221) Google Scholar). Apparently, miRNAs may function as tumor suppressors or oncogenes by targeting oncogenes or tumor suppressor genes (9Chen C.Z. N. Engl. J. Med. 2005; 353: 1768-1771Crossref PubMed Scopus (693) Google Scholar). In this regard, tumor-suppressive miRNAs are usually underexpressed in tumors. For instance, let-7 is down-regulated in lung cancer (10Takamizawa J. Konishi H. Yanagisawa K. Tomida S. Osada H. Endoh H. Harano T. Yatabe Y. Nagino M. Nimura Y. Mitsudomi T. Takahashi T. Cancer Res. 2004; 64: 3753-3756Crossref PubMed Scopus (2164) Google Scholar, 11Johnson S.M. Grosshans H. Shingara J. Byrom M. Jarvis R. Cheng A. Labourier E. Reinert K.L. Brown D. Slack F.J. Cell. 2005; 120: 635-647Abstract Full Text Full Text PDF PubMed Scopus (3119) Google Scholar). Furthermore, more than 60% of investigated patients suffering from B-cell chronic lymphocytic leukemia (B-CLL) have been reported to show a deletion at chromosome 13q14 where the mir-15 and mir-16 genes are located; these genes are under-represented in many B-CLL patients (12Calin G.A. Dumitru C.D. Shimizu M. Bichi R. Zupo S. Noch E. Aldler H. Rattan S. Keating M. Rai K. Rassenti L. Kipps T. Negrini M. Bullrich F. Croce C.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 15524-15529Crossref PubMed Scopus (4274) Google Scholar). Deregulation of miRNAs has also been reported in many other types of cancers. However, although miRNAs have been the subject of extensive research in recent years, the molecular basis of miRNA-mediated gene regulation and the effect of these genes on tumor growth remain largely unknown because of our limited understanding of miRNA target genes. Identification of miRNA target genes has been a great challenge. Computational algorithms have been the major driving force in predicting miRNA targets (13Stark A. Brennecke J. Russell R.B. Cohen S.M. PLoS Biol. 2003; 1: e60Crossref PubMed Scopus (628) Google Scholar, 14Lewis B.P. Shih I.H. Jones-Rhoades M.W. Bartel D.P. Burge C.B. Cell. 2003; 115: 787-798Abstract Full Text Full Text PDF PubMed Scopus (4250) Google Scholar, 15Kiriakidou M. Nelson P.T. Kouranov A. Fitziev P. Bouyioukos C. Mourelatos Z. Hatzigeorgiou A. Genes Dev. 2004; 18: 1165-1178Crossref PubMed Scopus (644) Google Scholar). The approaches are mainly based on base pairing of miRNA and target gene 3′-UTR, emphasizing the location of miRNA complementary elements in 3′-UTR of target mRNAs, the concentration in the seed (6-8 bp) of continuous Watson-Crick base pairing in the 5′ proximal half of the miRNA, and the phylogenetic conservation of the complementary sequences in 3′-UTRs of orthologous genes. However, evidence suggests that perfect seed pairing may not necessarily be a reliable predictor for miRNA-target interactions (16Didiano D. Hobert O. Nat. Struct. Mol. Biol. 2006; 13: 849-851Crossref PubMed Scopus (354) Google Scholar), which may explain why many predicted target sites are nonfunctional. A recent study also suggests that there may be at least three types of miRNA-mRNA interactions in mammals (17Smalheiser N.R. Torvik V.I. Methods Mol. Biol. 2006; 342: 115-127PubMed Google Scholar). Hence, with few exceptions, large portion of the physiologic targets for miRNAs remain to be identified or verified experimentally. In this study, we analyzed tumors derived from breast cancer MCF-7 cells treated with antisense mir-21 oligonucleotide (anti-mir-21) or the negative control by two-dimensional differentiation in-gel (2-DIGE) and identified the tumor suppressor tropomyosin 1 (TPM1) as a putative mir-21 target. Subsequent experiments confirmed that mir-21 down-regulated expression of TPM1, whereas anti-mir-21 up-regulated its expression through the mir-21 binding site at the 3′-UTR region. Furthermore, ectopic expression of TPM1 suppressed anchorage-independent growth. Cell Culture—MCF-7 cells (obtained from American Type Cell Collection, Manassas, VA) were grown in RPMI 1640 (Cambrex, Walkersville, MD) supplemented with 10% fetal bovine serum (Sigma-Aldrich), 2 mm glutamine, 100 units of penicillin/ml, and 100 μg of streptomycin/ml (Cambrex). MCF10A cells (ATCC) were grown in serum-free mammary epithelial growth medium (from Cambrex) supplemented with 100 ng/ml cholera toxin (EMD Biosciences, San Diego, CA). 293T cells (ATCC) were grown in Dulbecco’s modified Eagle’s medium (Cambrex) supplemented with 10% fetal bovine serum. All cells were incubated at 37 °C in a humidified chamber supplemented with 5% CO2. Reagents—Anti-mir-21 (AM17000, ID No. AM10206) and the negative control (AM17010) were purchased from Ambion (Austin, TX). Anti-TPM1 antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Transfection—Transfection of MCF-7 cells was performed with Optifect reagent (Invitrogen) following the manufacturer’s protocol. Briefly, the cells were seeded in 6-well plates at 30% confluence on the day before transfection. Three μg of TPM1-expressing plasmid or control vector was used for each transfection in antibiotic free Opti-MEM medium (Invitrogen). Transfection of 293T cells was performed using the calcium phosphate method as described previously (18Mo Y.Y. Beck W.T. Exp. Cell Res. 1999; 252: 50-62Crossref PubMed Scopus (34) Google Scholar). The negative control oligonucleotide or anti-mir-21 oligonucleotide (both from Ambion) at 50 nm or 3 μg of appropriate plasmid (otherwise indicated) was used for each transfection. Transfection efficiency was monitored by spiking GFP-expressing vector or β-galactosidase-expressing vector when necessary. Detection of Mature mir-21 by TaqMan Real-time PCR— TaqMan miRNA assays (ABI, Forest City, CA) used the stem-loop method (19Chen C. Ridzon D.A. Broomer A.J. Zhou Z. Lee D.H. Nguyen J.T. Barbisin M. Xu N.L. Mahuvakar V.R. Andersen M.R. Lao K.Q. Livak K.J. Guegler K.J. Nucleic Acids Res. 2005; 33: e179Crossref PubMed Scopus (4082) Google Scholar, 20Lao K. Xu N.L. Yeung V. Chen C. Livak K.J. Straus N.A. Biochem. Biophys. Res. Commun. 2006; 343: 85-89Crossref PubMed Scopus (119) Google Scholar) to detect the expression level of mature mir-21. For RT reactions, 10 ng total RNA was used in each reaction (15 μl) and mixed with the RT primer (3 μl). The RT reaction was carried out under the following conditions: 16 °C for 30 min; 42 °C for 30 min; 85 °C for 5 min; and then held on 4 °C. After the RT reaction, the cDNA products were diluted at 150×, and 1.33 μl of the diluted cDNA was used for PCR reaction along with TaqMan primers (2 μl). The PCR reaction was conducted at 95 °C for 10 min followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s in the ABI 7500 real-time PCR system. The real-time PCR results were analyzed and expressed as relative miRNA expression of CT (threshold cycle) value, which was then converted to fold changes (19Chen C. Ridzon D.A. Broomer A.J. Zhou Z. Lee D.H. Nguyen J.T. Barbisin M. Xu N.L. Mahuvakar V.R. Andersen M.R. Lao K.Q. Livak K.J. Guegler K.J. Nucleic Acids Res. 2005; 33: e179Crossref PubMed Scopus (4082) Google Scholar). The RT primer, PCR primers, and TaqMan probe for mir-21 (19Chen C. Ridzon D.A. Broomer A.J. Zhou Z. Lee D.H. Nguyen J.T. Barbisin M. Xu N.L. Mahuvakar V.R. Andersen M.R. Lao K.Q. Livak K.J. Guegler K.J. Nucleic Acids Res. 2005; 33: e179Crossref PubMed Scopus (4082) Google Scholar) were purchased from ABI. U6 or 5S RNA was used for normalization. Detection of TPM1 mRNA—To detect relative levels of TPM1 transcription, qRT-PCR was performed using the Cyber Green method under the following conditions: 94 °C for 3 min followed by 30 cycles of 94 °C for 0.5 min, 54 °C for 1 min, and 72 °C for 0.5 min. PCR primers were TPM1-5.1, sense, 5′-CTCTCAACGATATGACTTCCA-3′, and TPM1-3.1, antisense, 5′-TTTTTTTAGCTTACACAGTGTT-3′. Both were purchased from Sigma-Genosys (Woodland, TX). Constructs—To construct a plasmid expressing mir-21, we first modified pCMV-Myc (Clontech, Mountain View, CA) by deleting the Myc tag by PCR. We then amplified a 500-bp DNA fragment carrying pre-mir-21 from MCF10A genomic DNA using PCR primers mir-21-5.1, 5′-GAATTCTGATTGAACTTGTTCATTTT-3′ where the EcoRI site is underlined, and mir-21-3.1, 5′-GGTACCAATTAAGACTATCCCCATTTCTCCA-3′, where the KpnI site is underlined. The amplified fragment was first cloned into pCR8 (Invitrogen) and was subsequently cloned into this modified pCMV vector at the EcoRI and KpnI sites. Full-length TPM1 plus 3′-UTR was amplified from MCF-7 cells using primers TPM1-EcoRI-5.1, 5′-GAATTCTGGACGCCATCAAGAAGAAGA-3′, and TPM1-UTR-NotI-3.1, 5′-GCGGCCGCCCTACAATGTGCATTTTATTCC-3′, then cloned into pCR8, and finally subcloned into the original pCMV-Myc. The 250-bp 3′-UTR region of TPM1 was also amplified from MCF-7 cells using primers TPM1-UTR-XbaI-5.1, 5′-TCTAGACTCTCAACGATATGACTTCCA-3′, and TPM1-UTR-XbaI-3.1, 5′-TCTAGATTTTTTTAGCTTACACAGTGTT-3′, using the same approach described above, and was finally cloned into pGL3 control vector (Promega, Madison, WI) at the XbaI site. To construct a plasmid expressing the GFP-TPM1 fusion protein (pEGFP-TPM1+UTR), we also used primers TPM1-R1-5.1 and TPM1-UTR-NotI-3.1 as indicated above. This fragment was finally cloned into pEGFP-C3 (Clontech) at the EcoRI and NotI sites in-frame with the GFP coding region. To clone the 3′-UTR of TPM1 into a GFP reporter, which was different from the GFP fusion construct, we first modified the pEGFP-C3 by introducing a stop codon in the front of the multiple cloning sites by PCR and then cloning the TPM1-UTR fragment into EcoRI site of this modified vector. All PCR products were verified by DNA sequencing before cloning into the final destination vectors. Luciferase Assay—293T cells were seeded in 6-well plates and transfected with luciferase reporters using the calcium phosphate method as described above. After transfection, the cells were split into 12-well plates (in duplicates) and harvested for luciferase assays 24 h later using a luciferase assay kit (Promega) according to the manufacturer’s protocol. β-Galactosidase was used for normalization. Cell Growth Assay—After transfection with vector control or TPM1-expressing vector, the cells were seeded into 96-well plates at 2500 cell/well. The MTT assay was used to determine relative cell growth as described previously (21Si M.L. Zhu S. Wu H. Lu Z. Wu F. Mo Y.Y. Oncogene, in press. 2006; Google Scholar). Anchorage-independent Assay—To determine anchorage-independent growth of transfected cells, the cells were grown in soft agar according to a published method (22Finlay T.H. Tamir S. Kadner S.S. Cruz M.R. Yavelow J. Levitz M. Endocrinology. 1993; 133: 996-1002Crossref PubMed Google Scholar). Briefly, 1 day after transfection with TPM1, cells were harvested and mixed with tissue culture medium containing 0.7% agar to result in a final agar concentration of 0.35%. Then, 1-ml samples of this cell suspension were immediately plated in 12-well plates covered with 0.6% agar in tissue culture medium and cultured at 37 °C with 5% CO2. To assess cell viability before plating in soft agar, cell number was determined by trypan blue staining in Vi-Cell XR (Beckman Coulter, Fullerton, CA). Western Blot—Total protein was isolated from tumor samples or 293T cells transfected with an appropriate plasmid in cell lysis buffer (20 mm Tris, pH 8.0, 100 mm NaCl, 1 mm EDTA, 0.5% Nonidet P-40). Protein concentration was measured using the Bio-Rad protein assay kit. The membrane was first probed with antibodies against Myc (Applied Biomaterials) or GFP (Clontech), and then with anti-β-actin antibody (Sigma-Aldrich). Secondary antibodies were labeled with either Alexa Fluor 680 (Invitrogen) or IRDye800 (Rockland Immunochemicals, Gilbertsville, PA). Signals were visualized using the Odyssey infrared imaging system (Li-Cor Biosciences, Lincoln, NE). Animal Work—Female nude (nu/nu) mice (4-5 weeks old) were purchased from Harlan (Indianapolis, IN) and were maintained in the Southern Illinois University School of Medicine’s accredited animal facility. All animal studies were conducted in accordance with National Institutes of Health animal use guidelines and a protocol approved by the Southern Illinois University Animal Care Committee. Exponentially growing MCF-7 cells were harvested, mixed with 50% Matrigel (BD Biosciences) at 15 million cells/ml, and injected (1.5 million cells/spot) into mammary pads of female nude mice. To facilitate tumor growth, a 0.72-mg 17β-estradiol pellet (Innovative Research of America, Sarasota, FL) was implanted beneath the back skin. Tumors usually appeared 1 week after inoculation when anti-mir-21 or negative control oligonucleotide was delivered to tumor sites by injecting 50 μl (50 nm) of the oligonucleotide carrying 12 μl of Optifect. One week later another injection of the same amount was performed. Tumor size was monitored every other day; 4 weeks after inoculation of MCF-7 cells, tumors were harvested, weighed, and frozen immediately in a -80 °C freezer. Proteomic Analysis of Tumor Samples—Tumor samples that were harvested and stored at -80 °C were sent directly for 2-DIGE and mass spectrometry analysis, a service provided by Applied Biomics (Hayward, CA). Total protein was extracted and labeled with either Cy3 or Cy5. Isoelectric focusing in the first dimension was carried out at pH 3-10, and in the second dimension was carried out in 8-14% gradient SDS-PAGE. Differentially expressed proteins were cut out and subjected to trypsin digestion before mass spectrometry analysis. Statistical Analysis—Data are expressed as means ±S.E., and p < 0.01 is considered as statistically significant by Student’s t test. Suppression of Tumor Growth by Anti-mir-21—We have previously shown that transient transfection of MCF-7 cells with anti-mir-21 causes tumor growth inhibition in a xenograft carcinoma mouse model (21Si M.L. Zhu S. Wu H. Lu Z. Wu F. Mo Y.Y. Oncogene, in press. 2006; Google Scholar). Thus, we asked here whether intratumoral delivery of anti-mir-21 has the same effect on tumor growth. Tumors treated with anti-mir-21 grew substantially smaller in size than those treated with the negative control; tumors treated with anti-mir-21 revealed a lower level of Ki-67 staining compared with the vector control (not shown). This is consistent with the previous finding (21Si M.L. Zhu S. Wu H. Lu Z. Wu F. Mo Y.Y. Oncogene, in press. 2006; Google Scholar), suggesting that suppression of tumor growth likely occurs because of reduced cell proliferation, increased apoptosis, or both as suggested previously (21Si M.L. Zhu S. Wu H. Lu Z. Wu F. Mo Y.Y. Oncogene, in press. 2006; Google Scholar, 23Chan J.A. Krichevsky A.M. Kosik K.S. Cancer Res. 2005; 65: 6029-6033Crossref PubMed Scopus (2240) Google Scholar). Therefore, these results not only support the notion that mir-21 is an oncogenic miRNA but also imply that anti-mir-21 has a therapeutic potential. TPM1 Is Up-regulated in Anti-mir-21-treated Tumor Samples as Detected by 2-DIGE Analysis—Although mir-21 is overexpressed in many types of tumors, suggesting its role in cancer development, the underlying mechanism of mir-21-mediated tumorigenesis is still unclear largely because of limited knowledge about mir-21 targets. Although various computer-aided algorithms have predicted many putative mir-21 targets, these targets have not been validated experimentally. Because miRNAs are believed to regulate gene expression mainly through translational repression in mammalian cells, we thought to determine the differential expression of proteins from the tumor samples after treatment with anti-mir-21. Protein was extracted from tumors derived from MCF-7 cells treated with either the negative control (labeled with Cy3) or anti-mir-21 (labeled with Cy5). Unlike conventional two-dimensional gels in which two samples are run in separate gels, this method separates two samples labeled with different fluorescent dyes in a single gel, thus eliminating gel-to-gel variation and allowing for easy comparison of relative expression levels. After separating the proteins by isoelectric focusing and SDS-PAGE, we found that several proteins were either up-regulated or down-regulated as shown by either red or green color, respectively (Fig. 1). This result, in fact, is in agreement with the finding that mir-122 also causes up-regulation or down-regulation of many proteins (24Krutzfeldt J. Rajewsky N. Braich R. Rajeev K.G. Tuschl T. Manoharan M. Stoffel M. Nature. 2005; 438: 685-689Crossref PubMed Scopus (3403) Google Scholar), because some of these differentially expressed proteins may be due to the secondary effect of miRNA regulation. Analysis of another pair of tumor samples harvested from different mice revealed an almost identical pattern to that of Fig. 1, suggesting the reproducibility of this method. We are particularly interested in those proteins up-regulated by anti-mir-21 because they are potential direct targets for mir-21. We picked 10 protein spots that were up-regulated more than 2-fold in the tumor samples treated with anti-mir-21 compared with the negative control; these are circled in Fig. 1B. Mass spectrometry analysis identified seven of them with a good score. Among them, three proteins have been implicated in tumorigenesis: TPM1 (25Cooper H.L. Feuerstein N. Noda M. Bassin R.H. Mol. Cell. Biol. 1985; 5: 972-983Crossref PubMed Scopus (107) Google Scholar), integrin-β4-binding protein (ITGB4BP), (26Sanvito F. Vivoli F. Gambini S. Santambrogio G. Catena M. Viale E. Veglia F. Donadini A. Biffo S. Marchisio P.C. Cancer Res. 2000; 60: 510-516PubMed Google Scholar) and selenium-binding protein-1 (SELENBP1) (27Chen G. Wang H. Miller C.T. Thomas D.G. Gharib T.G. Misek D.E. Giordano T.J. Orringer M.B. Hanash S.M. Beer D.G. J. Pathol. 2004; 202: 321-329Crossref PubMed Scopus (109) Google Scholar). Therefore, we tested these three genes by cloning their UTRs into a luciferase reporter. Interestingly, Western blot analysis of the tumor samples also indicated that the endogenous TPM1 was increased in the anti-mir-21-treated tumors by almost 2-fold (Fig. 1, C and D). Further characterization identified TPM1 as a direct target for mir-21 as described below. TPM1 Carries a Putative mir-21 Binding Site, Which Is Responsible for Regulation by mir-21—The tropomyosins (TMs or TPMs) are a group of proteins that bind to the sides of actin filaments; there are at least four separate proteins, TPM1, -2, -3, and -4, encoded by different genes (28Lees-Miller J.P. Helfman D.M. BioEssays. 1991; 13: 429-437Crossref PubMed Scopus (299) Google Scholar). TPM1 has seven variants through alternative splicing. Coincidentally, TPM1 variants 1 and 5 carry a putative mir-21 binding site, as predicted by the Sanger miRNA data base target search program (Fig. 2A). Variant 1 differs from variant 5 in a sequence coding for 24 amino acids and also by lacking an additional 48 nucleotides upstream of the 3′-UTR (Fig. 2A). The potential base pairing between mir-21 and TPM1 3′-UTR is shown in Fig. 2B. Thus, we tried to amplify this UTR region of both variants from MCF-7 cells, which, however, appeared to express only TPM1 variant 1. Hence, we cloned this variant 1 3′-UTR into pGL3 control vector. As shown in Fig. 2C, the luciferase activity in 293T cells for Luc-TPM1-V1-UTR was about 20% less than that of pGL3 control vector, suggesting that TPM1 3′-UTR carries a regulatory element(s). To confirm that this regulatory region is mir-21 specific, we transfected 293T cells with the same Luc-TPM1-UTR plasmid along with either the pCMV vector or the mir-21-expressing plasmid. The ectopic expression of mir-21 was confirmed by TaqMan real-time PCR, which revealed about a 4-fold higher mir-21 expression in the mir-21-transfected cells than in vector control (Fig. 3A). In contrast, anti-mir-21 reduced mir-21 by almost 50%, as determined by the same method (Fig. 3B). We then transfected the 293T cells with various amounts of mir-21-expressing vector. As shown in Fig. 3C, reduction of luciferase activity by mir-21 was dose-dependent, suggesting that this regulation is specifically responsive to mir-21. In contrast, mir-21 had no effect on Luc-TPM1-V4-UTR, which is derived from variant 4 and lacks the mir-21 binding site (Fig. 3D). In addition, we tested the effect of anti-mir-21 on the luciferase activity of Luc-TPM1-V1-UTR. As expected, mir-21 suppressed the luciferase activity, whereas anti-mir-21 increased the luciferase activity (Fig. 4E), further suggesting that expression of TPM1 is specifically regulated by mir-21. To determine the role of the mir-21 binding site in regulating its expression, we deleted the mir-21 binding site in variant 1 (Luc-TPM1-V1-UTR-d). As shown in Fig. 3F, neither mir-21 nor anti-mir-21 had any effect on the luciferase activity, highlighting the importance of this mir-21 binding site.FIGURE 4Regulation of TPM1 expression by mir-21 at the translational level. A, expression of TPM1-Myc in 293T cells as determined by Western blot. The same membrane was first probed with anti Myc-antibody and then with anti β-actin as described under “Experimental Procedures.” The effect of mir-21 or anti-mir-21 on the protein levels of TPM1-Myc (B and D) or TPM1-d (C and E), in which the mir-21 binding site was deleted, is shown. B and C are representative of at least three separate experiments; D and E are the means of three separate experiments ± S.E. F, mir-21 or anti-mir-21 has no effect on the mRNA levels of TPM1-Myc, as determined by real-time qRT-PCR. V, vector (pCMV); 21, pCMV-mir-21; N, negative control oligonucleotide; A, anti-mir-21. **, p < 0.01.View Large Image Figure ViewerDownload Hi-res image Download (PPT) To ensure that down-regulation of luciferase activity by mir-21 was not due to the reporter we used, we made similar reporter constructs in EGFP vector. In this case, we cloned TPM1-V1-UTR into the EcoRI site of the modified pEGFP-C3 (see “Experimental Procedures”). Consistent with the luciferase data, the level of EGFP-TPM1-V1-UTR was reduced by mir-21 but was increased by anti-mir-21, as measured either by Western blot or fluorescence microscopy (not shown). mir-21 Regulates TPM1 at the Translational Level—Translational repression is a major mechanism of miRNAs to regulate gene expression (29Engels B.M. Hutvagner G. Oncogene. 2006; 25: 6163-6169Crossref PubMed Scopus (377) Google Scholar). To determine whether mir-21 also suppresses TPM1 through translational repression, we cloned the full-length TPM1 plus the 3′-UTR into pCMV-Myc. Expression of Myc-tagged TPM1 was confirmed by anti-Myc antibody (Fig. 4A). Importantly, although ectopic expression of mir-21 significantly reduced TPM1 protein, anti-mir-21 enhanced TPM1 protein (Fig. 4, B and D). To further determine the importance of the mir-21 binding site, we did similar experiments with pCMV-Myc-TPM1+UTR-d in which the mir-21-binding site was deleted. Deletion of this site abolished the effect of mir-21 or anti-mir-21 on TPM1 expression at the protein level (Fig. 4, C and E). However, despite the effect of mir-21 or anti-mir-21 on TPM1 at the protein level, no effect on the TPM1 mRNA level was detected by real-time qRT-PCR for pCMV-Myc-TPM1+UTR (Fig. 4F). Therefore, these results suggest that the mir-21 binding site present in the TPM1-UTR region is critical for mir-21-mediated regulation at the translational level. In addition, we made a GFP fusion construct with the full-length TPM1 plus the 3′-UTR (GPF-TPM1+UTR). We first confirmed its expression of its fusion protein by Western blot (not shown). Fluorescence microscopy clearly showed a distinguished subcellular localization of GFP-TPM1. Although we detected the green protein all over the cell for GFP alone, GFP-TPM1 fusion protein was localized exclusive
0
Citation1,065
0
Save
0

p53 represses c-Myc through induction of the tumor suppressor miR-145

Mohit Sachdeva et al.Feb 7, 2009
The tumor suppressor p53 negatively regulates a number of genes, including the proto-oncogene c-Myc, in addition to activating many other genes. One mechanism of the p53-mediated c-Myc repression may involve transcriptional regulation. However, it is not clear whether microRNAs (miRNAs) play a role in the p53-mediated posttranscriptional regulation of c-Myc. In this study, we show that a putative tumor suppressor, miR-145 , is expressed through the phosphoinositide-3 kinase (PI-3K)/Akt and p53 pathways. Importantly, p53 transcriptionally induces the expression of miR-145 by interacting with a potential p53 response element (p53RE) in the miR-145 promoter. We further show that c-Myc is a direct target for miR-145 . Although miR-145 silences the expression of c-Myc, anti- miR-145 enhances its expression. This specific silencing of c-Myc by miR-145 accounts at least in part for the miR-145 -mediated inhibition of tumor cell growth both in vitro and in vivo. Finally, the blockade of miR-145 by anti- miR-145 is able to reverse the p53-mediated c-Myc repression. Together, these results define the role of miR-145 in the posttranscriptional regulation of c-Myc by p53 and suggest that, as a new member of the p53 regulatory network, miR-145 provides a direct link between p53 and c-Myc in this gene regulatory network.
0
Citation804
0
Save
0

LncRNA loc285194 is a p53-regulated tumor suppressor

Qian Liu et al.Apr 3, 2013
Protein-coding genes account for only a small part of the human genome, whereas the vast majority of transcripts make up the non-coding RNAs including long non-coding RNAs (lncRNAs). Accumulating evidence indicates that lncRNAs could play a critical role in regulation of cellular processes such as cell growth and apoptosis as well as cancer progression and metastasis. LncRNA loc285194 was previously shown to be within a tumor suppressor unit in osteosarcoma and to suppress tumor cell growth. However, it is unknown regarding the regulation of loc285194. Moreover, the underlying mechanism by which loc285194 functions as a potential tumor suppressor is elusive. In this study, we show that loc285194 is a p53 transcription target; ectopic expression of loc285194 inhibits tumor cell growth both in vitro and in vivo . Through deletion analysis, we identify an active region responsible for tumor cell growth inhibition within exon 4, which harbors two miR-211 binding sites. Importantly, this loc285194-mediated growth inhibition is in part due to specific suppression of miR-211. We further demonstrate a reciprocal repression between loc285194 and miR-211; in contrast to loc285194, miR-211 promotes cell growth. Finally, we detect downregulation of loc285194 in colon cancer specimens by quantitative PCR arrays and in situ hybridization of tissue microarrays. Together, these results suggest that loc285194 is a p53-regulated tumor suppressor, which acts in part through repression of miR-211.
0
Citation371
0
Save
0

MicroRNA-145 Suppresses Cell Invasion and Metastasis by Directly Targeting Mucin 1

Mohit Sachdeva et al.Dec 9, 2009
Abstract MicroRNAs are important gene regulators that could play a profound role in tumorigenesis. Our previous studies indicate that miR-145 is a tumor suppressor capable of inhibiting tumor cell growth both in vitro and in vivo. In this study, we show that miR-145 exerts its function in a cell-specific manner. Although miR-145 inhibits cell growth in MCF-7 and HCT-116 cells, it has no significant effect on cell growth in metastatic breast cancer cell lines. However, miR-145 significantly suppresses cell invasion in these cells; in contrast, the antisense oligo against miR-145 increases cell invasion. miR-145 is also able to suppress lung metastasis in an experimental metastasis animal model. This miR-145–mediated suppression of cell invasion is in part due to the silencing of the metastasis gene mucin 1 (MUC1). Using luciferase reporters carrying the 3′-untranslated region of MUC1 combined with Western blot and immunofluorescence staining, we identify MUC1 as a direct target of miR-145. Moreover, ectopic expression of MUC1 enhances cell invasion, which can be blocked by miR-145. Of interest, suppression of MUC1 by miR-145 causes a reduction of β-catenin as well as the oncogenic cadherin 11. Finally, suppression of MUC1 by RNAi mimics the miR-145 action in suppression of invasion, which is associated with downregulation of β-catenin and cadherin 11. Taken together, these results suggest that as a tumor suppressor, miR-145 inhibits not only tumor growth but also cell invasion and metastasis. Cancer Res; 70(1); 378–87
0
Citation366
0
Save
0

Fatty Acid Synthase Gene Is Up-regulated by Hypoxia via Activation of Akt and Sterol Regulatory Element Binding Protein-1

Eiji Furuta et al.Feb 15, 2008
Abstract The fatty acid synthase (FAS) gene is significantly up-regulated in various types of cancers, and blocking the FAS expression results in apoptosis of tumor cells. Therefore, FAS is considered to be an attractive target for anticancer therapy. However, the molecular mechanism by which the FAS gene is up-regulated in tumor cells is poorly understood. We found that FAS was significantly up-regulated by hypoxia, which was also accompanied by reactive oxygen species (ROS) generation in human breast cancer cell lines. The FAS expression was also activated by H2O2, whereas N-acetyl-l-cystein, a ROS inhibitor, suppressed the expression. We also found that the hypoxia significantly up-regulated sterol regulatory–element binding protein (SREBP)-1, the major transcriptional regulator of the FAS gene, via phosphorylation of Akt followed by activation of hypoxia-inducible factor 1 (HIF1). Moreover, our results of reporter assay and chromatin immunoprecipitation analysis indicate that SREBP-1 strongly bound to the SREBP binding site/E-box sequence on the FAS promoter under hypoxia. In our xenograft mouse model, FAS was strongly expressed in the hypoxic regions of the tumor. In addition, our results of immunohistochemical analysis for human breast tumor specimens indicate that the expressions of both FAS and SREBP-1 were colocalized with hypoxic regions in the tumors. Furthermore, we found that hypoxia-induced chemoresistance to cyclophosphamide was partially blocked by a combination of FAS inhibitor and cyclophosphamide. Taken together, our results indicate that FAS gene is up-regulated by hypoxia via activation of the Akt and HIF1 followed by the induction of the SREBP-1 gene, and that hypoxia-induced chemoresistance is partly due to the up-regulation of FAS. [Cancer Res 2008;68(4):1003–11]
0

Exosome-mediated transfer of miR-10b promotes cell invasion in breast cancer

Ramesh Singh et al.Nov 26, 2014
Exosomes are 30-100 nm membrane vesicles of endocytic origin, mediating diverse biological functions including tumor cell invasion, cell-cell communication and antigen presentation through transfer of proteins, mRNAs and microRNAs. Recent evidence suggests that microRNAs can be released through ceramide-dependent secretory machinery regulated by neutral sphingomyelinase 2 (nSMase2) enzyme encoded by the smpd3 gene that triggers exosome secretion. However, whether exosome-mediated microRNA transfer plays any role in cell invasion remains poorly understood. Thus, the aim of this study was to identify the exosomal microRNAs involved in breast cancer invasion.The expression level of endogenous and exosomal miRNAs were examined by real time PCR and the expression level of target proteins were detected by western blot. Scanning electron and confocal microscopy were used to characterize exosomes and to study its uptake and transfer. Luciferase reporter plasmids and its mutant were used to confirm direct targeting. Furthermore, the functional significance of exosomal miR-10b was estimated by invasion assay.In this study, we demonstrate that microRNA carrying exosomes can be transferred among different cell lines through direct uptake. miR-10b is highly expressed in metastatic breast cancer MDA-MB-231 cells as compared to non-metastatic breast cancer cells or non-malignant breast cells; it is actively secreted into medium via exosomes. In particular, nSMase2 or ceramide promotes the exosome-mediated miR-10b secretion whereas ceramide inhibitor suppresses this secretion. Moreover, upon uptake, miR-10b can suppress the protein level of its target genes such as HOXD10 and KLF4, indicating its functional significance. Finally, treatment with exosomes derived from MDA-MB-231 cells could induce the invasion ability of non-malignant HMLE cells.Together, our results suggest that a set of specific microRNAs may play an important role in modulating tumor microenvironment through exosomes. Thus, a better understanding of this process may aid in the development of novel therapeutic agents.
0

IGF2BP2 regulates DANCR by serving as an N6-methyladenosine reader

Xiaoge Hu et al.Dec 5, 2019
Abstract The major function of Insulin-like growth factor 2 mRNA-binding protein 2 (IGF2BP2) is to regulate cell metabolism. However, emerging evidence indicates that IGF2BP2 plays a role in cancer, but the underlying mechanism is largely unknown. Here we showed that upregulation of IGF2BP2 is associated with poor outcomes of pancreatic cancer patients and suppression of IGF2BP2 inhibits cell proliferation. We further showed that IGF2BP2 regulates lncRNA DANCR. Ectopic expression IGF2BP2 enhances, whereas knockdown (KD) or knockout (KO) of IGF2BP2 suppresses DANCR expression. Moreover, in vivo RNA precipitation and reciprocal RNA immunoprecipitation revealed that IGF2BP2 interacts with DANCR. DANCR promotes cell proliferation and stemness-like properties. Experiments with xenograft models revealed that while ectopic expression of DANCR promotes, DANCR KO suppresses tumor growth. Mechanistically, DANCR is modified at N6-methyladenosine (m6A) and mutagenesis assay identified that adenosine at 664 of DANCR is critical to the interaction between IGF2BP2 and DANCR where IGF2BP2 serves a reader for m6A modified DANCR and stabilizes DANCR RNA. Together, these results suggest that DANCR is a novel target for IGF2BP2 through m6A modification, and IGF2BP2 and DANCR work together to promote cancer stemness-like properties and pancreatic cancer pathogenesis.
0
Citation270
0
Save
0

miR-7 Suppresses Brain Metastasis of Breast Cancer Stem-Like Cells By Modulating KLF4

Hiroshi Okuda et al.Feb 6, 2013
Abstract Despite significant improvement in survival rates of patients with breast cancer, prognosis of metastatic disease is still dismal. Cancer stem-like cells (CSC) are considered to play a role in metastatic progression of breast cancer; however, the exact pathologic role of CSCs is yet to be elucidated. In this report, we found that CSCs (CD24−/CD44+/ESA+) isolated from metastatic breast cell lines are significantly more metastatic than non-CSC populations in an organ-specific manner. The results of our microRNA (miRNA) profile analysis for these cells revealed that CSCs that are highly metastatic to bone and brain expressed significantly lower level of miR-7 and that this miRNA was capable of modulating one of the essential genes for induced pluripotent stem cell, KLF4. Interestingly, high expression of KLF4 was significantly and inversely correlated to brain but not bone metastasis-free survival of patients with breast cancer, and we indeed found that the expression of miR-7 significantly suppressed the ability of CSCs to metastasize to brain but not to bone in our animal model. We also examined the expression of miR-7 and KLF4 in brain-metastatic lesions and found that these genes were significantly down- or upregulated, respectively, in the tumor cells in brain. Furthermore, the results of our in vitro experiments indicate that miR-7 attenuates the abilities of invasion and self-renewal of CSCs by modulating KLF4 expression. These results suggest that miR-7 and KLF4 may serve as biomarkers or therapeutic targets for brain metastasis of breast cancer. Cancer Res; 73(4); 1434–44. ©2012 AACR.
0
Citation270
0
Save
Load More