Cooperation meets competition inmicroRNA-mediated DMPK transcript regulation

Edyta Koscianska; Tomasz M. Witkos; Emilia Kozlowska; Marzena Wojciechowska; Wlodzimierz J. Krzyzosiak

doi:10.1093/nar/gkv849

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

DeepDyve requires Javascript to function. Please enable Javascript on your browser to continue.

Cooperation meets competition inmicroRNA-mediated DMPK transcript regulation

Koscianska, Edyta; Witkos, Tomasz M.; Kozlowska, Emilia; Wojciechowska, Marzena; Krzyzosiak, Wlodzimierz J. 2015-10-30 00:00:00 9500–9518 Nucleic Acids Research, 2015, Vol. 43, No. 19 Published online 24 August 2015 doi: 10.1093/nar/gkv849 Cooperation meets competition in microRNA-mediated DMPK transcript regulation Edyta Koscianska , Tomasz M. Witkos, Emilia Kozlowska, Marzena Wojciechowska and Wlodzimierz J. Krzyzosiak Department of Molecular Biomedicine, Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14, 61-704 Poznan, Poland Received February 06, 2015; Revised July 30, 2015; Accepted August 10, 2015 ABSTRACT nucleotide-long non-coding RNAs are potent regulators of gene expression. They act post-transcriptionally and The fundamental role of microRNAs (miRNAs) in exert their regulatory effects mainly by binding to the the regulation of gene expression has been well- 3 -untranslated region (3 UTR) of target mRNAs, which established, but many miRNA-driven regulatory results in mRNA deadenylation and decay, translational mechanisms remain elusive. In the present study, we suppression or, rarely, mRNA cleavage (1,2). By target- demonstrate that miRNAs regulate the expression ing multiple transcripts and affecting the expression of of DMPK, the gene mutated in myotonic dystrophy numerous proteins, miRNAs engage in various biologi- cal pathways in cells (i.e., proliferation, differentiation, type 1 (DM1), and we provide insight regarding the development, apoptosis, metabolism and neurodegener- concerted effect of the miRNAs on the DMPK target. ation). The interaction between miRNAs and mRNAs Speciﬁcally, we examined the binding of several miR- is influenced by many factors; however, nucleotides 2–8 NAs to the DMPK 3 UTR using luciferase assays. We of the miRNA, termed the ‘seed’ sequence, are essential validated the interactions between the DMPK tran- for target recognition and binding (1). The number and script and the conserved miR-206 and miR-148a. We distribution of miRNA binding sites as well as plausible suggest a possible cooperativity between these two miRNA cooperation are particularly important. More miRNAs and discuss gene targeting by miRNA pairs than a decade ago, the insertion of multiple binding sites that vary in distance between their binding sites and in reporter constructs used to validate miRNA–mRNA expression proﬁles. In the same luciferase reporter interactions has been suggested to ensure a higher effi- ciency of such constructs (3,4). It was later demonstrated system, we showed miR-15b/16 binding to the non- that two sites in the same or different miRNAs could act conserved CUG repeat tract present in the DMPK synergistically and that the distance between neighboring transcript and that the CUG-repeat-binding miRNAs miRNA binding sites affects the strength of the target might also act cooperatively. Moreover, we detected down-regulation. Specifically, an optimal down-regulation miR-16 in cytoplasmic foci formed by exogenously was observed when the distance between the 3 end of the expressed RNAs with expanded CUG repeats. There- first miRNA site and the 5 end of the subsequent one fore, we propose that the expanded CUGs may serve was > 7and < 40 nt (5), and when the 5 ends of both as a target for concerted regulation by miRNAs and miRNA seeds were separated by between 13 and 35 nt may also act as molecular sponges for natural miR- (6). More recently, the concept of miRNA synergy was NAs with CAG repeats in their seed regions, thereby reexamined and addressed in a more detailed way, showing affecting their physiological functions. that this mechanism of miRNA-mediated regulation can affect thousands of human genes. In a transcriptome-wide approach, it was demonstrated that miRNA sites spaced INTRODUCTION by a maximum of 26 nt may act cooperatively and that The involvement of microRNAs (miRNAs) in the the human transcriptome is enriched for miRNA-binding pathogenic mechanisms of many human diseases has sites located at a cooperativity-permitting distance (7). become increasingly apparent. These endogenous ∼22- To whom correspondence should be addressed. Tel: +48 61 8528503; Fax: +48 61 8520532; Email: edytak@ibch.poznan.pl Correspondence may also be addressed to Wlodzimierz J. Krzyzosiak. Tel: +48 61 8528503; Fax: +48 61 8520532; Email: wlodkrzy@ibch.poznan.pl Present addresses: Tomasz M. Witkos, Faculty of Life Sciences, University of Manchester, Manchester M13 9PL, UK. Marzena Wojciechowska, Department of Molecular and Systems Biology, European Centre for Bioinformatics and Genomics, Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14, 61-704, Poznan, Poland. C The Author(s) 2015. Published by Oxford University Press on behalf of Nucleic Acids Research. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Nucleic Acids Research, 2015, Vol. 43, No. 19 9501 Moreover, a workflow for the identification and analysis of Taken together, the abovementioned reports indicate the RNA triplexes composed of two cooperating miRNAs and pathological potential of miRNA dysregulation in DM1. a target mRNA has been proposed, and all of the predicted However, regarding the possible treatment of DM1, of par- human target genes of synergistic miRNA regulation have ticular importance is a report showing that some miRNAs been collected in the TriplexRNA database (8). The syn- are predicted to preferentially bind and repress toxic tran- ergistic activity of co-expressed miRNAs as well as those scripts with longer CUG repeats (25). that exhibit differential expression across tissues could be In this study, we focus on the miRNA-mediated regu- employed in therapeutic interventions using miRNA mim- lation of the DMPK transcript, which provides a unique ics and/or miRNA inhibitors. Interestingly, the natural model for the investigation of miRNA binding in the con- antagonizing activity of two miRNAs with overlapping text of potential miRNA cooperativity. Using a luciferase binding sites has been observed. More specifically, miR-184 reporter system, we validated the regulation of the DMPK was found to interfere with miR-205 in the suppression transcript by conserved miRNAs, miRs 206 and 148a, as of the SHIP2 gene. The binding of miR-184 to its seed well as miR-15b/16 binding to a non-conserved CUG tract. sequence prevented the inhibitory effect of miR-205 on We demonstrated a possible cooperativity between the miR- SHIP2 mRNA (9). 206/148a pair and the potential for cooperative targeting of Myotonic dystrophy type 1 (DM1) is an incurable the CUG tract by CUG-repeat-binding miRNAs. In addi- neuromuscular disorder that is caused by an expanded tion, we demonstrated the enrichment of miR16 in RNA CTG*CAG repeat in the 3 UTR of the dystrophia foci composed of exogenously expressed CUG-repeat tran- myotonica-protein kinase (DMPK)gene(10). The normal scripts by RNA fluorescence in situ hybridization (RNA human DMPK gene harbors 5–37 copies of the trinu- FISH), supporting the possibility of miRNA sequestration cleotide motif, but a dynamic mutation may increase this by the CUG repeats present in the DMPK 3 UTR. number to over 5000 repeat copies. The expanded CUGs in DM1 result in the nuclear retention of mutant DMPK MATERIALS AND METHODS mRNA and reduced DMPK protein levels (11). Mutant Computational prediction of the miRNAs binding to the transcripts sequester the muscleblind-like 1 (MBNL1) splic- DMPK 3 UTR ing factor, leading to the abnormal alternative splicing of a multitude of other transcripts and the expression of fe- The pipeline for the computational prediction of miRNA tal forms of their protein products in DM1 adults (12,13). binding to the DMPK transcript is presented in Figure 1. Spliceopathy is therefore thought to be the major factor The following steps were taken: (i) finding the conserved underlying the pathogenesis of DM1. However, alternative sites of the miRNA families conserved among vertebrates mechanisms such as additional changes in gene expression, predicted by any of three of the most commonly used antisense transcripts, translation efficiency, misregulated al- miRNA prediction programs (TargetScan Release 6.2, DI- ternative polyadenylation and miRNA deregulation may ANAmicroT v. 5.0 and miRanda August 2010 Release) (26– contribute to the pathogenesis of DM1 (14,15). 28); (ii) adding the poorly conserved miRNA–mRNA sites A few reports detailing a close connection between miR- that were predicted by all of these programs; (iii) predicting NAs and DM1 have been published (reviewed in (16)). The additional miRNAs that would bind to the CUG repeats deregulation of specific miRNAs has been linked with mus- using an in-house method. cular dystrophies and cardiomyopathies (17–19)and with myotonic dystrophy type 2 (DM2) (20). In DM1, alterations Identification of miRNAs that potentially bind to CUG re- in the miRNA expression patterns have been observed in peats muscle-specific miRNAs (myomiRs). More specifically, in DM1 skeletal muscle, miR-1 and miR-335 are up-regulated Mature human and murine miRNA sequences were re- whereas miRs 29b, 29c and 33 are down-regulated, com- trieved from miRBase (ver. 20) (29). In-house scripts written pared with the control muscles (21). In addition, miR-1 in the Python programming language were used to detect is down-regulated in cardiac muscle (22), whereas miR- miRNAs with at least six complementary matches to CUG 206 is up-regulated in the skeletal muscle of DM1 patients repeats in the same reading frame within the miRNA seed (23). The deregulation of DM1-associated miRNAs has regions (positions 2–7 of the miRNA sequences). also been linked to alterations in their putative target ex- pression, indicating that miRNA misregulation in DM1 Examination of the frequency distribution of binding sites for is functionally relevant and may contribute to the disease miRNA pairs pathology (21,22). Importantly, the decreased expression of mature miR-1 and increased levels of its targets in the The sequences of human miRNA families with shared hearts of individuals with myotonic dystrophy are medi- seed regions were obtained from TargetScan, Release ated by the functional depletion of MBNL1, a sequestered 6.2 (30). The mature miRNA expression profiles (qPCR splicing factor, which affects the processing of pre-miR-1 miRNA profiling) were obtained from miRNAMap 2.0 (22). Recently, a study investigating a transgenic yfl model (31). Broadly expressed miRNAs were defined as miRNAs of DM1 (i(CTG)480 Drosophila line carrying 480 CTG re- that were expressed in at least 10 of the 18 examined nor- peats) revealed that miRNA alterations were caused directly mal adult tissues at considerable levels (at least 2% of the by CTG expansions (24). Specifically, the expression of 20 total miRNA expression). miRNAs with at least 75% of the miRNAs was changed in DM1 flies compared with con- reads found in the tissue with the highest expression levels trol flies; 19 were down-regulated and one was up-regulated. were classified as tissue-specific miRNAs. Next, miRNAs 9502 Nucleic Acids Research, 2015, Vol. 43, No. 19 were grouped into miRNA families (miRNAs with com- For miRNA overexpression, commercial plasmid con- mon seed regions), and only the miRNA families with all structs expressing miRNA precursors (pri-miR-1, pri-miR- of the members nominated as broadly expressed or tissue- 206, pri-miR-214 (System Biosciences), pri-miR-148a and specific were considered to calculate the distances between pri-miR-29a (Cell Biolabs)) were used. Moreover, the con- putative miRNA binding sites (all of the distances were struct expressing the miR-15b/16 precursor was generated counted between the 5 ends of the binding sites). Lists of based on an empty plasmid (System Biosciences) by cloning miRNA families with broadly expressed and tissue-specific the appropriate sequence at the EcoRI and NotI sites. The miRNAs can be found in the supplementary data (Supple- mir-15b/16–2 cluster that was flanked by a ∼250-nt se- mentary Table S1). The distances between all of the distinc- quence was amplified from the genomic DNA using PCR tive pairs of miRNAs were used as reference controls. A and the following primers: forward 5 -GACGCGGAATTC Perl script provided by TargetScan was used to predict the CGAAGCCATGGAATTGACTT and reverse 5 -GAAG miRNA target sites for the longest variants of the human CGGCCGCAAGAACAAAAACAAAGGAAAAGGA. 3 UTR transcripts, and in-house scripts written in Python were used to process the data. Cell transfection HEK 293T and HeLa cells were transfected using Lipofec- Cell culture tamine 2000 (Invitrogen) according to the manufacturer’s HEK 293T and HeLa cells were obtained from the protocols. For the luciferase assays, the cells were trans- American Type Culture Collection (ATCC) and grown in fected in 12-well plates at ∼80% confluence. For each trans- Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma- fection experiment, 200 ng of the appropriate reporter con- Aldrich) supplemented with 8% fetal bovine serum (FBS) struct and 400 ng of the appropriate miRNA-coding vec- (Sigma-Aldrich), 2 mM L-glutamine and an antibiotic– tor were used. For miRNA cooperativity studies, 200 ng of antimycotic solution (Sigma-Aldrich) at 37 C in a humidi- the appropriate reporters and a total of 400 ng of miRNA- fied atmosphere containing 5% CO .At24hbefore trans- coding plasmids (200 ng of either vector) were used. The fection, the cells were plated in 12-well or 6-well dishes in cells were harvested 24 h after transfection and assayed for DMEM medium and harvested 24, 48, and 72 or 96 h post- luciferase activity. transfection for the luciferase assay, real-time polymerase For the miRNA overexpression required for real-time chain reaction (PCR), and western blot analyses, respec- PCR and western blot analyses, the HEK 293T cells were tively. grown to 80% and 60% confluence, respectively, transfected in 12- and 6-well plates with 1 g/ml pri-miRNA plasmid vectors, and harvested at 48 and 72/96 h, respectively. Plasmid constructs and synthetic miRNA oligonucleotides For the fluorescence in situ hybridization (FISH) experi- To generate reporter constructs bearing miRNA-binding ments, the HEK 293T cells were grown to 80% confluence sites, the pmirGLO Dual-Luciferase miRNA Target Ex- on microscope slides in 12-well plates, transfected with 0.5 pression Vector (Promega) was used. Specific oligonu- g/ml plasmid vectors expressing DMPK exons 11–15 con- cleotides of different lengths with DraIand XbaI restric- taining either 960 CUG or CAG repeats in exon 15 or with tion sites and containing single binding sites (b.s.) for the a control plasmid lacking the repeats (36). For miR-214 de- analyzed miRNAs (DMPK b.s. for miRs 1, 148a, 206 and tection, the cells were additionally co-transfected with 30 combined 148a/206) as well as the oligonucleotides com- nM of miR-214 mimic (synthetic oligonucleotides were pur- posed of 5 CUG, 20 CUG and 20 CAA repeats were syn- chased from IDT and prepared as described in (33)). thesized (IBB Warsaw). The appropriate oligos were an- nealed and cloned into the pmirGLO vector, which was Luciferase reporter assay previously digested with DraI(Fermentas) and XbaI(Fer- mentas) restriction enzymes downstream of the luc2 gene. After harvesting, the cells were lysed in a passive lysis buffer For all of the miRNAs, three types of constructs were pre- (Promega). The enzymatic activities of fireyfl and Renilla lu- pared, namely wild-type (WT) constructs, constructs carry- ciferases were measured using a Centro LB 960 luminome- ing mutations (MUT) predicted to disrupt binding and per- ter (Berthold Technologies) and the substrates and pro- fect match (PM) constructs, as previously described (32,33). cedures provided with a Dual-Luciferase Reporter Assay The reporter constructs carrying 44, 53 and 72 CUG re- System (Promega). The values for firefly luciferase activity peats were generated based on the reporter containing 20 for every reporter construct were normalized to the corre- CUGs via the SLIP (synthesis of long iterative polynu- sponding values of Renilla luciferase activity to account for cleotide) method (34,35). Briefly, the pmirGLO plasmid varying transfection efficiency. The relative expression val- with the cloned sequence corresponding to 20 CUG repeats ues for all of the constructs were obtained by comparing was independently digested with DraIand SalI restriction their normalized luciferase activities with those of the con- enzymes and subjected to one thermal cycle of 95 C for 5 trol plasmid. ◦ ◦ min, 50 C for 10 min and 72 C for 3 min with 2.5 units of Pfu polymerase, reaction buffer, and 250 M each dNTP. RNA isolation and real-time PCR The product was directly transformed into Escherichia coli and selected on an LB plate containing ampicillin. The se- The total RNA from HEK 293T cells was isolated using quences of all of the constructs are presented in the supple- TRI Reagent (MRC, Inc., BioShop) according to the man- mentary data (Supplementary Table S2). ufacturer’s instructions. cDNA was obtained from 500 ng Nucleic Acids Research, 2015, Vol. 43, No. 19 9503 of total RNA using Superscript III (Life Technologies) and yeast tRNA, 10% dextran sulfate, 2 mM vanadyl ribonucle- random hexamer primers (Promega). For subsequent quan- oside complex and 2 ng/l appropriate fluorescently labeled titative real-time analyses, 50 ng of cDNA was used. Real- RNA or DNA/LNA probes. Specifically, the miR-16–5p time PCR was performed with a LightCycler 480 II system probe labeled at the 5 -end with TYE665 (Cy5) and mod- (Roche) using TaqMan Gene Expression Assays and Taq- ified at positions 5, 10, 14 and 19 with locked nucleic acids Man Universal Master Mix II (Applied Biosystems). The (LNA) (Exiqon) was used in combination with the (CAG) results obtained for the assessment of DMPK mRNA lev- probe labeled at the 5 -end with FAM and modified with els were normalized to the levels of actin mRNA. 2 -O-methyl at positions 1 and 2 (IDT). In parallel experi- ments, the miR-214 probe labeled at the 5 -end with FAM and modified with 2 -O-methyls (IDT) was used together Northern blot analysis with the (CAG) probe labeled at the 5 -end with TYE665 High-resolution northern blot analysis was performed as (Cy5) and modified at positions 6, 12 and 17 with LNAs previously described (37,38). Briefly, 25 g of total RNA (Exiqon). The (CTG) probe, which was used additionally was extracted from HEK 293T cells and resolved in a 12% as a control to detect CAG RNAs, was labeled at the 5 - denaturing polyacrylamide gel in 0.5× TBE. The RNA end with TYE563 (Cy3) and modified as described for the (CAG) probe. Post-hybridization washing was performed was transferred to a GeneScreen Plus hybridization mem- brane (PerkinElmer) using semi-dry electroblotting (Sigma- in 30% formamide and 2× SSC at 45 C for 30 min followed Aldrich), immobilized by subsequent UV irradiation (120 by 1× SSC at 37 C for 30 min. The slides were mounted in 2 ◦ mJ/cm ) (UVP) and baking at 80 C for 30 min. The mem- SlowFade Gold Antifade reagent with DAPI (Invitrogen) branes were probed with specific DNA oligonucleotides for further microscopy. (Supplementary Table S3) that were complementary to the To determine the spatial interactions between the ex- annotated human miRNAs (miRBase). The probes were la- panded repeat RNA foci and specific miRNAs, the same beled with [ P] ATP (5000 Ci/mmol; Hartmann Analyt- exposure was established for all of the images from a single ics) using OptiKinase (USB). The hybridizations were per- experiment, and a z-stack that was sufficient to cover all of formed at 37 C overnight in a PerfectHyb buffer (Sigma- the foci was acquired for each cell with a slice thickness of Aldrich). The marker lanes contained a mixture of radiola- 1 musingaPL-Apo63× oil objective in conjunction with beled RNA oligonucleotides (17-, 19-, 21-, 23- and 25-nt in a cooled AxioCam HRc camera; the images in one stack length). Hybridizations to U6 RNA provided loading con- were overlaid and saved as TIFF files. Approximately 100 trols. Radioactive signals were detected by phosphorimag- cells were randomly selected to characterize the mutant re- ing (Multi Gauge v3.0; Fujifilm). peat RNA and miRNA interactions in transfected HEK293 cells. The FISH images were processed with LSM 510 soft- ware (Zeiss). Western blot analysis After transfection with the appropriate miRNA-coding Statistical analysis plasmids, HEK 293T cells were lysed with 1× PBS sup- The experiments were repeated at least three times. The plemented with protease inhibitor cocktail (Roche). A to- graphs were generated using GraphPad Prism 5 (GraphPad tal of 25–30 g of protein lysate was separated by 12% Software). The figures for the luciferase assays were gen- SDS-PAGE. After electrophoresis, the proteins were elec- erated after averaging the results from the repeated exper- trotransferred onto a nitrocellulose membrane (Sigma). iments for a particular construct. The values for the error All of the immunodetection steps were performed on a bars (means with SEM) and the statistical significance were SNAP id (Millipore) in PBS buffer containing 0.25% non- calculated using GraphPad Prism 5. The statistical signifi- fat milk and 0.1% Tween 20, and the membranes were cance of the luciferase reduction in the case of transfection washed in PBS/Tween. For DMPK and GAPDH de- with constructs carrying miRNA b.s. was assessed using a tection, the blots were probed with the primary mouse one-sample t-test after checking that the data followed a anti-DMPK (1:500, Millipore) and mouse anti-GAPDH normal distribution, with a hypothetical value of 100% as- (1:5000, Millipore) antibodies, respectively, and they were signed to the cells that were transfected with an empty vec- subsequently probed with biotinylated secondary antibod- tor control. P-values < 0.05 (two-tailed) were considered ies (1:500 Sigma). The membranes were incubated with a significant. streptavidin–AP conjugate (1:2000, Millipore), and the im- munoreactive bands were visualized using the Sigma Fast RESULTS BCIP/NBT kit (Sigma). In our previous study, we performed an in-depth computa- tional analysis of the miRNA interactions using all of the RNA FISH and microscopy mRNAs derived from genes that trigger hereditary neuro- RNA FISH was performed in transiently transected HEK logical disorders, known as trinucleotide repeat expansion 293T cultured cells as previously described (39). Briefly, 48 diseases (TREDs), and we showed that DMPK, the gene h post-transfection, the cells were fixed in 4% PFA /PBS at that is mutated in DM1, may be subject to miRNA regu- 4 C and washed three times in PBS for 5 min each. Pre- lation (40). In the present study, following our guidelines hybridization was performed in 30% formamide and 2× for the efficient prediction of miRNA–mRNA binding sites SSC buffer for 10 min followed by hybridization in a buffer (miRNA b.s.), we selected both highly and poorly conserved containing 30% formamide, 2× SSC, 0.02% BSA, 66 g/ml sites for experimental validation (Figure 1). 9504 Nucleic Acids Research, 2015, Vol. 43, No. 19 Figure 1. Analyzed miRNAs targeting the 3 UTR of the DMPK transcript. A schematic presentation of the selected miRNA target site distribution in the DMPK 3 UTR is shown. Additionally, a list of the miRNAs with CAG motifs in their seed regions that are complementary to the CUG repeats in the DMPK 3 UTR and a pipeline for the computational prediction of the miRNAs that bind to the DMPK 3 UTR are presented. Regulation of the DMPK gene by the conserved miR-206 and Supplementary Table S1). The binding parameters for these miR-148a candidate miRNAs meet the recommended bioinformatics criteria, and their experimental validation was of particular The top candidate miRNAs (miR-148a/152 and miR- interest in the context of the pathogenesis and therapy of 1/206) that are predicted regulators of the DMPK tran- DM1 (22). script (40) were ranked highly by the most commonly used Experiments employing a set of reporter constructs and algorithms to rank miRNA b.s., based on sequence conser- luciferase assays were performed to experimentally verify vation criteria. Both of the putative sites within the DMPK the predicted binding of miR-206 and miR-148a to their 3 UTR were the only two highly conserved sites for miRNA target sites within the 3 UTR of the DMPK, as previ- families that are broadly conserved among vertebrates. The ously described (32,33). Briefly, the following constructs miRs 148a/152 and 1/206 exhibit disparate expression pat- were tested in parallel: wild-type reporters (WT) bearing a terns in human tissues. Both miR-1 and miR-206 are my- single native b.s. for either miRNA, constructs with muta- omiRs, which are abundantly expressed in smooth, skeletal, tions (MUT) that disrupt the 5 seed site (negative controls) and/or cardiac muscles, whereas miR-148a and miR-152 and constructs showing perfect complementarity (PM) to are broadly expressed across various tissue types (31)(see miRNA b.s. (positive controls). Having performed a north- Nucleic Acids Research, 2015, Vol. 43, No. 19 9505 ern blot analysis, which revealed that miR-206 was not ex- script, but differ in four nucleotides at their 3 ends (Fig- pressed in HEK 293T cells and miR-148a was expressed at ure 3A). Both miR-1 and miR-206 are widely studied and alow level(Figure 2A), we validated the predicted miRNA– well-defined myomiRs; they enhance skeletal muscle dif- mRNA interactions using a miRNA overexpression system. ferentiation, regulate numerous targets and are involved Specifically, HEK 293T cells were co-transfected with a re- in a wide array of specific muscular pathways (reviewed porter construct (carrying a potential b.s. for the studied in (44,45)). Moreover, these two miRNAs exhibit differ- miRNA) and the appropriate miRNA-coding plasmid (Sys- ent expression patterns, depending on the muscle type, tem Biosciences, Cell Biolabs). The transient transfection of and differentially regulate transcripts that bear their b.s. the cells was followed by measuring the reporter activity. (46). Although canonical miRNA-target specificity is trig- We obtained considerable repression of luciferase ex- gered primarily by complementarity within the seed re- pression following the transfection of both of the WT re- gion, non-canonical interactions also depend on 3 compen- porters, namely the WT construct carrying the b.s. for miR- satory sites (1,47). To determine whether miR-1 and miR- 206 (WT 206) and the WT construct carrying the b.s. for 206 differ in their regulatory potential, we performed a rel- miR-148a (WT 148a) (Figure 2B). This result indicates that evant luciferase experiment. We co-transfected HEK 293T both miR-206 and miR-148a are functional and may down- cells with the reporter construct carrying the potential b.s. regulate the DMPK transcript. The effect exerted by these for miR-1 (WT 1) and a miR-1-encoding plasmid (System two miRNAs was comparable, with miR-148a being slightly Biosciences) together with the adequate controls. In con- more effective. Specifically, the reduction of luciferase ac- trast to the repressed luciferase expression that followed the tivity was reproducible and statistically significant for both transfection of the WT 206 construct, after transfection of WT constructs, with suppression to 81% and 71% for the WT 1, we observed only a slight decrease in luciferase ac- WT construct carrying the b.s. for miR-206 and miR-148a, tivity (suppression to 92%) (Figure 3B). There was no vis- respectively. Both of the PM constructs repressed the lu- ible decrease in the protein level of DMPK and no statis- ciferase activity to very low levels (13–16%), whereas the tically significant reduction of the mRNA level following luciferase activity for both of the MUT constructs exhib- miR-1 overexpression (Figure 3C, D, E). These results sug- ited efficient de-repression ( ≥90%). To validate our experi- gest that miR-1 is a weaker regulator of DMPK expression mental approach, we performed additional luciferase tests than miR-206; however, it cannot be ruled out that, under in HeLa cells and also HEK 293T cells but without the ad- certain conditions, this miRNA may affect the expression dition of the appropriate miRNA-coding vectors (Supple- of DMPK to variable degrees. mentary Text 1, Supplementary Figure S1A, S1B). The ob- tained results verified the reliability of the experimental sys- Potential cooperativity between miR-206 and miR-148a tem used. Thus, the presented study provides the first evi- dence of the direct binding of miRs 206 and 148a with the Having validated the individual interactions of miR-206 DMPK 3 UTR and positively validates these miRNAs as and miR-148a with the DMPK 3 UTR, we proceeded negative regulators of the DMPK gene. with the same type of experiments to determine the spac- Next, we evaluated the expression of the DMPK protein ing requirements for cooperative miRNA target site inter- and mRNA following the transfection of HEK 293T cells actions. The distance between the miR-206 and 148a sites with plasmids encoding either miR-206 or miR-148a. Real- in the intact DMPK 3 UTR measured between the 5 ends time PCR performed after transfection of the studied miR- of both seeds is 15 nt, which is considered to be the op- NAs did not reveal a considerable decrease in the DMPK timal spacing for efficient target repression ( 5–7). Addi- mRNA (Figure 2C). Although miRNA binding frequently tionally, this miRNA pair was predicted in silico to have leads to the reduction of the cellular levels of targeted tran- the potential to act cooperatively, as measured by the en- scripts (41,42), our observation is consistent with the find- ergy gain achieved through RNA triplex formation in com- ings that no or minimal changes in the respective mRNA parison to two separate miRNA–mRNA duplexes using levels were observed or that these changes were only re- the TriplexRNA database (8). To study the possible co- ported for certain targets (43). By contrast, western blot regulation of DMPK expression by the miR-206/148a pair, analyses performed with either miR-206 or miR-148a re- new sets of reporter constructs were prepared that differed vealed a considerable decrease in the DMPK protein level, in the distance between the b.s. of miR-206 and miR-148a. which was slightly more evident following the overexpres- Three reporters were constructed: (i) a wild-type reporter sion of miR-206 (Figure 2D). In muscle cells, which are the bearing the native sequence of the DMPK 3 UTR encom- primary cell type affected in DM1, miR-206 is highly ex- passing the b.s. of both miRNAs (WT 148/206); (ii) a re- pressed. Therefore, to confirm the DMPK down-regulation porter containing the sequence between miRNA seeds that observed in HEK 293T cells after miR-206 overexpression, were artificially extended to 31 nt (Ext 148/206), a nt length we also addressed the regulation of DMPK expression in that is considered to promote miRNA cooperativity (6)and human muscle cells including DM1 myoblasts (Supplemen- (iii) a reporter in which the sequence between miRNA seeds tary Text 2 and Figure S2). The observed increase of miR- was extended to 85 nt (Long 148/206), a nt length that 206 expression, together with the decrease of DMPK pro- is considered to possess very limited miRNA cooperativ- tein levels, is in agreement with the hypothesis that miR-206 ity (Figure 4A).Inthe case of theExt 148/206 construct, act as a negative regulator of DMPK expression. additional nucleotides were introduced to maintain a na- Finally, we aimed to analyze the binding specificity of tive sequence context, which is required for a proper inter- miR-1 and miR-206, which share the seed sequence and action with both miR-206 and miR-148a. In addition, an are predicted to bind to the same site in the DMPK tran- EcoRI restriction site was introduced, permitting the sub- 9506 Nucleic Acids Research, 2015, Vol. 43, No. 19 Figure 2. Regulation of the DMPK gene by miR-206 and miR-148a. (A) Northern blot detection of miRs 206 and 148a in untreated HEK 293T cells and cells that were transfected with miRNA-coding plasmids (System Biosciences, Cell Biolabs). M denotes the size marker: end-labeled 17, 19, 21, 23 and 25-nt oligoribonucleotides. En and Ex indicate the endogenous and vector-expressed miRNA levels, respectively. Hybridization to U6 RNA served as a loading control. (B) Relative repression of luciferase expression. Reporter constructs carrying a single b.s. for miR-206 and miR-148a were tested. For each luciferase experiment, the miRNA activity on four constructs was measured in parallel: an empty pmirGLO vector (Control), a wild-type potential b.s. for the appropriate miRNA (WT), a mutated b.s. (MUT), and a site with full complementarity (PM). The fireyfl luciferase activity was normalized against that of Renilla luciferase. The standard errors were calculated from nine independent experiments. The asterisks indicate statistical significance at P < 0.001. (C) Relative DMPK mRNA levels. Real-time PCR was performed after the HEK 293T cells were transfected with miR-206 and miR-148a. The bar graphs show the quantification of the DMPK mRNA levels normalized to the actin mRNA level, based on the data from vfi e independent experiments. (D) Western blot analysis of the DMPK protein levels 72 and 96 h after the HEK 293T cells were transfected with miR-206 and miR-148a, as indicated. The GAPDH protein served as a loading control. sequent extension of the spacing between these miRNAs significantly suppressed luciferase expression, demonstrat- b.s. In the case of the longest reporter construct, namely ing their ability to bind miRNAs, but their efficiency varied Long 148/206, an appropriate DNA fragment comprising (Figure 4C). Specifically, after the overexpression of both an irrelevant sequence that further separated the sites for miRNAs, the WT construct carrying the native sequences of miR-206 and miR-148a was cloned into the EcoRI site. the miR-148a and miR-206 sites (WT 148/206) reduced the The secondary structures that can be formed by transcripts luciferase activity to 66% compared with the control, which from every reporter construct, as predicted by the RNA is a better score compared with that observed for the down- metaserver available on the GeneSilico website, showed regulation resulting from the activity of either individual a high degree of secondary structure preservation in the miRNA (compare with Figure 2B and Supplementary Fig- miRNA b.s. (Figure 4B). ure S1A). However, in the case of the Ext 148/206 construct Luciferase assays were performed as described to vali- with the miRNA site separation extended to 31 nt, the re- date the individual miRNA–mRNA interactions. The re- duction of luciferase activity was much stronger (suppres- porters bearing two miRNA sites with increasing separa- sion to 42%). By contrast, in the case of the Long 148/206 tions, i.e., WT, Ext and Long, were tested in parallel follow- construct with miRNA site separation extended to 85 nt, the ing the transfection of HEK 293T cells and the simultane- reduction of luciferase activity was again at a level similar ous overexpression of both miRNAs (an equal amount of to that produced by the WT 148/206 construct (suppres- the plasmids encoding miRs 148a and 206 was transfected sion to 65%). The obtained results indicate that miRNAs into the cells). Importantly, all of the analyzed constructs 148a and 206 may act cooperatively and that their optimal Nucleic Acids Research, 2015, Vol. 43, No. 19 9507 Figure 3. Regulation of the DMPK gene by miR-1. (A) Schematic presentation of the base pairing of miR-1 and miR-206 with their target DMPK sequences. The nucleotides in the seed regions of these miRNAs are marked in green, whereas disparate nucleotides are shown in orange. (B)Relative repression of luciferase expression. Reporter constructs carrying a single b.s. for miR-1 were tested; the miRNA activity on four constructs was measured in parallel (Control, WT, MUT and PM) as described in Figure 2. The standard errors were calculated from eight independent experiments. The asterisk indicates statistical significance ( P < 0.05). (C) Western blot analysis of the DMPK protein levels after the HEK 293T cells were transfected with miR- 1. The GAPDH protein served as a loading control. (D) Relative DMPK mRNA levels. Real-time PCR was performed after the HEK 293T cells were transfected with miR-1. The bar graph shows the quantification of the DMPK mRNA level normalized to the actin mRNA level, based on the data from four experiment repeats. (E) Northern blot detection of miR-1 in the untreated HEK 293T cells and cells that were transfected with the miR-1-coding plasmid (System Biosciences). M denotes the size marker: end-labeled 17, 19, 21, 23 and 25-nt oligoribonucleotides. En and Ex indicate the endogenous and vector-expressed miRNA levels, respectively. Hybridization to U6 RNA served as a loading control. distance between b.s. is in agreement with that previously to structural consequences. Specifically, nucleotides in the studied in other human miRNAs (5–8). region of the miRNA b.s. were predicted to be more tightly The cooperativity between miR-206 and miR-148a in the paired, and a stem rather than a loop was formed at the regulation of DMPK could be further confirmed by the ob- miR-206 b.s., resulting in the lower efficiency of this con- servation that the expression of both the DMPK protein struct compared with Ext 148/206. After the transfection and mRNA was affected following the simultaneous over- of impExt 148/206, the luciferase expression decreased to expression of the cooperating miRNAs in HEK 293T cells 85% (Supplementary Figure S4), which was rather similar (Figure 4D, E). Interestingly, the real-time PCR results re- to the reduction achieved by the overexpression of only one vealed a minor decrease in the DMPK mRNA level after miRNA, suggesting a loss of miRNA cooperativity. In the transfection of the miR-148a/206 pair, which was statisti- same experimental setup, the Ext 148/206 reporter could cally significant and greater than the effect exerted by a sin- suppress luciferase up to 42% (Figure 4C). gle miR-148a or miR-206 (Figure 2C). It is worth noting that the effect of miRNA cooperativity was achieved when Global analysis of the miRNA–mRNA binding sites a half concentration of each miRNA was used (see Mate- Although a few significant advances in the field of miRNA rials and Methods). The DMPK protein was considerably cooperativity have been made (5–8), neither its mecha- repressed when miR-206 and miR-148a were mildly but si- nism nor its biological relevance is yet fully understood. multaneously overexpressed in HEK 293T cells, whereas the Therefore, after discovering the co-regulation of the DMPK protein level was not decreased when a lower concentration transcript via the ubiquitously expressed miR-148a and of either of the single miRNAs was used (Supplementary tissue-specific miR-206, we investigated whether this type Figure S3). of miRNA cooperativity in the regulation of gene expres- The RNA structure itself does not seem to be the main sion could be a common phenomenon that is dependent factor influencing the potential miRNA binding; however, on the miRNA expression profiles, i.e., whether broadly ex- spatial separation and a rather loose structure are prefer- pressed miRNAs and tissue-specific miRNAs could work in able (48–50). To assess the impact of the secondary struc- tandem. ture of the target site on the cooperative binding of the Using the qPCR expression profiles of human miRNAs miRNAs, an impaired reporter construct was also tested, in different tissues (miRNAMap 2.0, (31)), we selected namely the impaired Ext 148/206 (hereafter referred to as miRNA families in which every member of the family was impExt 148/206). The impairment of this reporter con- ubiquitously expressed (94 miRNA families) as well as a cerned its sequence; however, this change in sequence led set of tissue-specific miRNA families (25 miRNA fami- 9508 Nucleic Acids Research, 2015, Vol. 43, No. 19 Figure 4. Potential cooperativity between miR-206 and miR-148a. (A) A schematic presentation of the reporter constructs that were used to study the role of the distance between the b.s. of miR-206 and miR-148a. Three types of reporters are shown with the 5 ends of the miRNA b.s. separated by 15, 31 and 85 nt. The miR-148a and miR-206 seed regions are marked in light red and light green, respectively. (B) Graphical representation of the secondary structures formed by the reporter constructs of different lengths as predicted by RNAmetaserver. WT 148/206, Ext 148/206 and Long 148/206 correspond to constructs with miRNA b.s. spaced by 15, 31 and 85 nt, respectively. (C) Relative repression of luciferase expression by constructs that differ in the distance between the b.s. of miRs 148a and 206. The miRNA activities on the four constructs were assessed in parallel (Control, WT 148/206, Ext 148/206 and Long 148/206) after the simultaneous overexpression of both miRNAs. The standard errors were calculated from six independent experiments. The asterisks indicate statistical significance; a double asterisk denotes P < 0.01, and a triple asterisk denotes P < 0.001. (D)RelativeDMPKmRNAlevels. Real-time PCR was performed after the simultaneous overexpression of miR-148a and miR-206 in HEK 293T cells. The bar graph shows the quantification of the DMPK mRNA level normalized to the actin mRNA level, based on the data from vfi e independent experiments. ( E) Western blot analysis of the DMPK protein levels after the simultaneous overexpression of miR-148a and miR-206 in HEK 293T cells. A representative blot is shown. The GAPDH protein served as a loading control. lies) (see Materials and Methods and Supplementary Ta- three types of miRNA pairs examined (Figure 5). For the ble S1). Next, for all of the human transcripts in the Tar- pairs of miRNAs with one broadly expressed miRNA and getScan database Release 6.2 (see Materials and Methods), one tissue-specific miRNA, we observed an enrichment for we calculated the distances between the putative b.s. for overlapping sites (spacing of less than 7-nt, which causes three types of miRNA pairs: (i) pairs with one broadly ex- the sequence of the ‘seed’ region to be shared by two miR- pressed miRNA and one tissue-specific miRNA; (ii) pairs NAs), but no enrichment of pairs with 7–28-nt spacing, the with two different tissue-specific miRNAs and (iii) pairs distance that enabled cooperativity of miR-148a and miR- with two different broadly expressed miRNAs. Our analysis 206 in the DMPK transcript. This phenomenon suggests revealed different preferences for the distances between the that there may be a tendency for tissue-specific miRNAs Nucleic Acids Research, 2015, Vol. 43, No. 19 9509 Figure 5. Distribution of the frequencies of the pairwise distance between the predicted binding sites of miRNA pairs. The graph shows the distribution of the b.s. distance frequencies between pairs with two broadly expressed miRNAs (blue line), two tissue-specific miRNAs (green line) or one broadly expressed miRNA and one tissue-specific miRNA (red line) in relation to the distribution of the miRNA b.s. distances for all of the miRNA pairs. The dotted lines indicate the 7- and 28-nt pairwise distances. In the upper panel, the schematic demonstrates how the different miRNA pairs were classifie d. to regulate gene expression by competing with broadly ex- found 33 miRNAs with at least six complementary matches pressed miRNAs for binding sites rather than via cooper- to CUG repeats in their seed regions (Figure 1 and Sup- ative regulation by binding to the same transcripts in close plementary Table S4). All of these miRNAs were predicted vicinity. This competition for binding sites could be an ad- to have additional pairing sites to a CUG sequence. It is ditional mechanism for providing and maintaining tissue worth noting that the members of one miRNA family (miR- specificity. By contrast, we observed an enrichment of pairs 15ab/-16/-195/-424/-497/-6838) bear a common seed re- of miRNAs with two tissue-specific miRNAs with 7–28-nt gion that is fully complementary to the CUG repeats. The spacing. This result could be explained by the observation observations that miRNAs from the miR-15/107 group that that tissue-specific miRNAs, which are often expressed in share a 5 AGCAGC sequence (52) play key roles in gene similar types of tissues (brain-, muscle-, liver-, testes- and regulation and have overlapping targets (53,54)are also placenta-specific miRNA families), orchestrate the same of relevance to this issue. These miRNAs are expressed at developmental/physiological processes and miRNA coop- medium-to-high levels across many tissues, and members erativity increases the effectiveness of miRNA regulatory with the 7-nt common sequence AGCAGCA exhibit the mechanisms ((7,51) and references therein). highest expression levels (55). In the last set of experiments, we examined whether the miRNAs that were identified in silico as bearing CAG re- Interaction of miR-15b/16 with CUG repeats in the DMPK peats in their seed regions bind to the CUG expanded re- transcript peats in the DMPK transcript. We used the same exper- imental approach with a luciferase reporter system. The In our prediction of miRNA-mediated regulation of the candidate miRNAs were chosen based on the number of DMPK gene, we also sought to explore the possibility of matches with the CUG repeats in their seeds (Supple- miRNA binding to CUG repeats, which are the underlying mentary Table S4). Specifically, we selected miR-15b and cause of DM1 when they are extensively expanded. Previ- miR-16 (seven matches) and miR-214 and miR-29a (six ous studies have postulated the existence of several miR- matches). First, we co-transfected HEK 293T cells with NAs that can potentially bind to the CUG repeats present reporter constructs carrying either 20 CUG repeats (re- in the DMPK 3 UTR (25). If these interactions occur, the ferred to as WT CUG because there are 20 CUGs in CUG repeats could act as multiple overlapping b.s. for sev- the human DMPK gene sequence deposited in GenBank eral miRNAs and therefore have unique miRNA coopera- (NM 001081560.2)) or 20 CAA repeats (referred to as tive potential. We searched in silico for novel human miR- MUT CUG) and an appropriate miRNA-expressing vec- NAs that can interact with repeated CUG sequences. We 9510 Nucleic Acids Research, 2015, Vol. 43, No. 19 tor. In the luciferase assays, we obtained considerable and DMPK gene, rather than the 20 CTGs anticipated based on statistically significant repression of the luciferase expres- the reference sequence deposited in GenBank. Therefore, sion only after transfection of the reporters and miR- this observation is in agreement with the results of the lu- 15b/16 (luciferase repression equal to 62%). In the case of ciferase assay showing that the CUG-repeat-binding miR- the other tested miRNAs, the luciferase activity was not re- NAs cannot affect the control reporter bearing the 5 CUGs pressed (Figure 6A). (see Figure 6B). To assess whether the presence of abnor- Next, we aimed to determine whether the length of the mally expanded CUG tracts may affect DMPK expression, CUG sequence correlated with the level of transcript down- we performed western blot analyses using fibroblast cells de- regulation. Therefore, we prepared and tested reporter con- rived from DM1 patients (Supplementary Text 3 and Fig- structs that differed with respect to the number of CUG re- ure S6). We did not detect a significant change in the protein peats: the reporters carried 5, 20 (WT), 44, 53 and 72 CUG level in untreated cells (Supplementary Figure S6A), but the triplets, the last two mimicking the expansion of CUGs that DMPK protein level was reduced after overexpression of is observed in DM1 patients. Similarly, in the luciferase as- miR-16 in fibroblasts with a CUG expansion (Supplemen- says, we obtained significant repression of the luciferase ex- tary Figure S6B), which may further confirm the possibility pression after the transfection of miR-15b/16 with most that miRNAs with CAG repeats in their seed regions are of the reporters, with the exception of the reporter with 5 able to bind to the CUG repeats in the DMPK transcript. CUGs (Figure 6B). Moreover, a trend toward stronger re- pression was observed when constructs with a longer repeat Sequestration of miRNAs with a CAG sequence in their seeds length were used; the luciferase repression was equal to 64% by the expanded CUG repeats and 44% of the expression levels for the constructs with 20 and 72 CUGs, respectively. By contrast, the luciferase Having examined the binding of miR-15b/16 and miR-214 activity was not noticeably reduced when other miRNAs to the reporter constructs bearing CUG triplet repeats, we were co-transfected along with the tested reporters, with studied the plausible sequestration of these miRNAs to the the exception of miR-214. The latter considerably repressed expanded CUG repeats exogenously expressed in human luciferase activity, but only in the presence of the longest cells. We hypothesized that the presence of elongated CUG reporter (with 72 CUGs) (repression to 73%). This find- repeats would promote the binding of these miRNAs to ing suggests that 7-mer CAG miRNAs may preferentially the mutant DMPK transcript and that the sequestration of bind to and suppress the expression of transcripts contain- the CUG-repeat-binding miRNAs by abnormally length- ing CUG repeats. Overall, these results indicate that cer- ened mutant mRNAs could impair the miRNA-mediated tain miRNAs can target DMPK mRNA with expanded silencing machinery. To determine whether miRs 16 and triplets, and the degree of miRNA-mediated repression de- 214, which were previously tested in a reporter system, may pends on the number of repeats. The interaction between be arrested by the CUG repeats present in the 3 UTR of the DMPK 3 UTR and miR-15b/16 with a CUG binding the truncated DMPK transcripts, we used fluorescence in motif in their seeds containing seven matching nucleotides situ hybridization (FISH) in HEK 293T cells as a model (AGCAGCA), is a notable example. Interestingly, this in- system in this study. We transfected HEK 293T cells with teraction is even stronger in terms of the DMPK repression plasmid vectors expressing DMPK exons 11–15 contain- achieved by appropriate reporter constructs carrying 20 or ing either 960 CUG or 960 CAG repeats (36), or with the more CUG repeats than the interaction between the ana- control plasmid lacking the repeats. The cells were also co- lyzed conserved miRNAs (non-CUG-repeat-binding miR- transfected with the miR-214 mimic; co-transfection with NAs), namely miRs 148a and 206, when expressed individ- miR-16 was not necessary because it is expressed in HEK ually (compare with Figure 2B and Supplementary Figure 293T cells at a high level (Figure 6C). This experimental ap- S1A). proach permitted the simultaneous detection of both the Some of the CUG-repeat-binding miRNAs are endoge- selected miRNAs and the foci of either CUG- or CAG- nously expressed in cells at high levels (55)(Figure 6C). In triplet repeat mRNAs formed in the transfected cells. Using our experimental luciferase-based system, we overexpressed appropriate complementary probes that were fluorescently miRNAs from plasmids and ensured nearly equal concen- labeled with TYE665 (Cy5) or FAM at their 5 -ends (for trations of all of the tested miRNAs. However, the luciferase details, see Materials and Methods), in the FISH experi- tests used to study miRNA binding to CUG repeats were ment we detected that overexpression of 960 CUG repeat also positive in the absence of miRNA overexpression (Fig- transcripts caused the formation of miR-16-enriched CUG ure 6D). RNA foci, in contrast to the uniform distribution of miR-16 Next, we attempted to evaluate the expression of the in non-transfected cells (Figure 7A and Supplementary Fig- DMPK protein and mRNA following the transfection of ure S5). Interestingly, miR-16-containing CAG RNA foci HEK 293T cells with plasmids encoding miRs 15b/16, 214 were almost completely absent in cells overexpressing 960 and 29a, as well as the positive control transfection with CAG repeat transcripts (Figure 7B). This strongly suggests siRNAs specific to the DMPK sequence ( 56). The west- that miR-16 sequestration was specific for long CUG re- ern blot and real-time PCR analyses did not show a con- peats. The vast majority of the miR-16-enriched RNA foci siderable decrease in the DMPK protein and mRNA lev- were found in the cytoplasm, where 68% were co-localized els after transfection with the miRNAs (Figure 6E, F). Se- with the CUG RNA foci (Figure 7D). The relevant co- quencing of the endogenous DMPK 3 UTR fragment de- localization with CUG RNA inclusions was not observed rived from the HEK 293T cells used here revealed that these for miR-214, the miRNA bearing six matches with CUG cells harbored only 5 CTG repeats in the 3 UTR of the repeats in its seed (Figure 7C), or for the appropriate con- Nucleic Acids Research, 2015, Vol. 43, No. 19 9511 Figure 6. Interaction between miRNAs with CAG sequences in their seed regions and the CUG repeats in the DMPK transcript. (A) Relative repression of luciferase expression after transfection of reporter constructs carrying 20 CUG or 20 CAA repeats (denoted as WT CUG and MUT CUG) and either 9512 Nucleic Acids Research, 2015, Vol. 43, No. 19 trols. Therefore, the miR-16 co-localization with CUG foci cally significant enrichment for the former group, with 22% detected by RNA FISH confirmed our results from the lu- of transcripts with at least one b.s. for CUG-repeat-binding ciferase assays with the use of the same HEK 293T cells. miRNAs versus 13% for all transcripts (P = 0.0018; Yates’ Expanded CUG repeat tracts form both nuclear and cy- continuity corrected chi-squared test) (Figure 8A and Sup- toplasmic aggregates (57,58); however, the pathogenicity as- plementary Table S6). This trend was even stronger for tran- sociated with DM1 has been linked to the fact that mutant scripts that were uniquely deregulated in the CUG trans- DMPK transcripts are retained in the cell nucleus (11,59– genic mice, but not in the Mbnl1 or Clcn1 knockout mice, 60). In our experimental system, in rapidly proliferating which are alternative DM1 mouse models. Additionally, the −/− HEK 293T cells, we observed both nuclear and cytoplasmic transcripts that were uniquely deregulated in the Mbnl1 LR CUG foci, but the co-localization of such foci with miR-16 mice, but not in the HSA mice, did not include any bind- occurred to a greater extent in the cytoplasm (most likely ing sites for the CUG-binding miRNAs. This observation due to a cytoplasmic abundance of miR-16). Nevertheless, supports the hypothesis that the miRNA deregulation ob- because miRNAs function in both the cytoplasm and cell served in animal models of DM1 and DM1 patients may nucleus (reviewed in (61)), we believe that our findings are have a significant impact on changes in the transcriptome relevant and further support the possibility of miRNA se- and may contribute to the pathogenesis of DM1. questration by the abnormally expanded CUG repeats in the mutant DMPK transcripts. More evidence for miRNA DISCUSSION sequestration was indirectly provided by luciferase assays that were performed after the co-expression of appropriate The importance and role of miRNAs in human physiol- reporter constructs and CUG-expanded repeats; an experi- ogy and disease have been increasingly appreciated; how- ment with a reporter carrying a target sequence for miR-16 ever, many miRNA-driven regulatory mechanisms remain revealed that the presence of CUG repeats diminishes the unclear. This study reports that miRNAs regulate the ex- miR-16 effect (Supplementary Text 4 and Figure S7). pression of the DMPK gene and examines how miRNAs Having experimentally demonstrated the binding of the function on a specific target sequence and on a repeated miRNAs to CUG tracts using luciferase assays and having CUG sequence present in the DMPK transcript. observed miR-16-enriched CUG RNA foci by RNA FISH, we sought to examine whether the enhanced interactions miR-206 and miR-148a regulate the DMPK transcript and between the miRNAs and the aberrantly extended CUG re- may functionally cooperate peats in the DMPK 3 UTR could explain the changes in the transcriptome of DM1 mice reported by Thornton and We experimentally assessed the validity of three predicted colleagues (62). We computationally analyzed the miRNA interactions with miR-1, miR-206 and miR-148a. Using regulation of genes with significantly altered expression in luciferase reporter assays and sets of reporter constructs, LR CUG transgenic mice (HSA ) in comparison with the we provide the first evidence of DMPK down-regulation parental line. First, analogous to the human miRNAs, we by miR-206 and miR-148a. No changes were detected in predicted the presence of murine miRNAs with CAG re- the expression levels of the endogenous DMPK mRNA, peats in their seed sequences (Supplementary Table S5). but there was a significant decrease in the DMPK protein Next, we compared the prevalence of the predicted con- levels for both miR-148a and miR-206, which was even served b.s. for those miRNAs in the 3 UTRs of transcripts slightly higher for miR-206. Interestingly, in our experimen- that are altered in CUG transgenic mice, and in the 3 UTRs tal setup, we could not detect any regulation of the DMPK of all of the murine transcripts predicted to have at least one transcript by miR-1, which targets the same DMPK se- b.s. for any miRNAs using TargetScan Mouse Release 6.2. quence and has the same seed as miR-206, but differs some- We observed an overall enrichment for transcripts bearing what at the 3 end. miR-206 has two additional Watson- conserved miRNA b.s. in the CUG transgenic mice (78% of Crick base pairs to the DMPK b.s., which are absent in transcripts with at least one conserved b.s. versus 64% for the case of miR-1 binding. The variety in the 3 -end be- all murine transcripts; P = 0.0003; Yates’ continuity cor- tween miRs 1 and 206 seems to be important for their in- rected chi-squared test). Moreover, we discovered a statisti- hibitory activities against their targets. The impact of the ← −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− − the miR-15b/16-, miR-214- or miR-29a-expressing vectors, as indicated. The standard errors were calculated from twelve, eight and vfi e independent experiments that were performed after overexpression of miRs 15b/16, 214 and 29a, respectively. The asterisks indicate statistical significance ( P < 0.001). (B) Relative repression of luciferase expression by the reporters differing in the number of CUG repeats. Constructs with 5, 20 (WT), 44, 53 and 72 CUGs were tested in parallel. The luminescence was measured after the overexpression of miRs 15b/16 and 214, as indicated. The standard errors were calculated from ten and four independent experiments, respectively. The asterisks indicate statistical significance: a single asterisk denotes P < 0.05, a double asterisk denotes P < 0.01 and a triple asterisk denotes P < 0.001. (C) Northern blot detection of miRs 15b, 16 and 214 and 29a in untreated HEK 293T cells and cells transfected with the miRNA-encoding plasmids (System Biosciences). M denotes the size marker: end-labeled 17, 19, 21, 23 and 25-nt oligoribonucleotides. En and Ex indicate the endogenous and vector-expressed miRNA levels, respectively. The hybridization to U6 RNA serves as a loading control. (D)The relative repression of luciferase expression in HEK 293T cells measured after testing the reporter constructs carrying 20 CUG or 20 CAA repeats (denoted as WT CUG and MUT CUG) and 72 CUGs without miRNA overexpression. The standard errors were calculated from seven independent experiments. The asterisks indicate statistical significance: a single asterisk denotes P < 0.05, a double asterisk denotes P < 0.01 and a triple asterisk denotes P < 0.001. (E) Relative DMPK mRNA levels. Real-time PCR was performed after the HEK 293T cells were transfected with miRs 15b/16, 214, 29 and the appropriate siRNAs as a positive control (ssiDM10 (56)). The bar graphs show the quantification of the DMPK mRNA levels normalized to the actin mRNA level, based on the data from vfi e independent experiments. ( F) Western blot analysis of the DMPK protein levels after transfection of HEK 293T cells with miRs 15b/16, 214 and 29a, as indicated. The GAPDH protein served as a loading control. Nucleic Acids Research, 2015, Vol. 43, No. 19 9513 Figure 7. Visualization of miRNAs and cytoplasmic foci of CUG-triplet repeat mRNAs. (A) Representative RNA-FISH confocal images of cultured HEK 293T cells with endogenous miR-16 expression that were transfected with plasmid vectors expressing DMPK exons 11–15 containing 960 CUG repeats. (B) Representative RNA-FISH confocal images of cultured HEK 293T cells with endogenous miR-16 expression that were transfected with plasmid vectors expressing DMPK exons 11–15 containing 960 CAG repeats. (C) Representative RNA-FISH confocal images of cultured HEK 293T cells that were transfected with plasmid vectors expressing DMPK exons 11–15 containing 960 CUG repeats and the miR-214 mimic. The appropriate miRNAs and CUG repeat-containing RNAs were detected by differentially labeled fluorescent probes as depicted in the figure legend. The nuclei were stained wi th DAPI (blue). Scale bar: 20 m. (D) Graphs showing the statistics for the estimation of the foci found in the cytoplasm (C) and nucleus (N), as well as information about the co-localization of the miRNA and CUG foci. The number of miR-16 foci in cells expressing 960 CUG and 960 CAG repeats were compared using the Mann Whitney test; ***−P-value < 0.001. that two sites of the same or different miRNAs could act 3 -end variability of the miRNAs on target recognition and synergistically; the optimal distance between two miRNA binding strength has been previously reported. For instance, the functionality of the isomiR-214 variants, as opposed to sites was determined to be approximately 7–40 nt between the other tested miRNA isoforms, has been proposed to be neighboring miRNA b.s. (5–7). Potential cooperativity be- due to compensatory base pairing at the 3 end of the miR- tween more distant miRNA sites has also been addressed 214 isomiRs (33). (25,63–64); however, the optimal spacing requirements for The phenomenon of miRNA cooperativity and/or syn- neighboring miRNAs have not been precisely defined. In ergy has not been fully elucidated. It was demonstrated the present study, the distance between the miR-206 and 9514 Nucleic Acids Research, 2015, Vol. 43, No. 19 Figure 8. Potential sequestration of CUG-repeat binding miRNAs. (A) Deregulated genes in the CUG transgenic mice that may be subject to regulation by CUG-repeat-binding miRNAs (62). The left graph shows the percentage of genes bearing putative b.s. for the conserved miRNAs in all murine transcripts LR and the transcripts that are deregulated in the CUG transgenic mice (HSA ). The right graph compares the percentage of genes with the predicted miRNA b.s. for miRNAs with CAG sequences in their seed regions (Supplementary Table S5) in the following groups: all murine transcripts (all mouse genes), LR −/− all transcripts deregulated in the CUG transgenic mice (all HSA ), the transcripts deregulated in CUG transgenic mice, but not in the Mbnl1 or −/− LR −/− −/− Clcn1 mice (unique to HSA ), and the transcripts deregulated only in the Mbnl1 mice, but not in the CUG transgenic mice (unique to Mbnl1 ). All comparisons were made using Yates’ continuity corrected chi-squared test. **−P-value < 0.01, ***−P-value < 0.001. (B) Effects of the extended CUG repeats on miRNA activity. The mutant DMPK transcripts may sequester miRNAs with CAG repeats in their seed sequence. The sequestration of miRNAs could prevent their normal activity, leading to the reduction of accessible miRNA molecules and, consequently, improper repression of their targets. gle miRNAs. Interestingly, Ext 148/206, the reporter con- 148a sites in the intact DMPK 3 UTR measured between the 5 ends of both seeds was 15 nt. Based on the knowl- struct with slightly extended miRNA sites (to 31 nt between edge that these two miRs may act on the DMPK mRNA, the 5 ends of the seed) showed stronger suppression (by we analyzed the binding mechanism of the miR-206/148a approximately 20%). Of the three reporter constructs that pair. We found that this pair may cooperate because the were tested in our study, Ext 148/206 was the most effec- joint binding of miRs 206 and 148a increased the down- tively targeted construct by the simultaneously expressed regulation of DMPK expression. The reporter construct miRs 148a and 206. The observed effect was consistent with carrying the sequence corresponding to the native b.s. of previous findings showing that cooperativity is facilitated miR-148a and miR-206 (WT 148/206) with the 5 seed ends when the miRNA b.s. are directly adjoining (3,65)orare spaced by 15 nt (the lower limit of the cooperative distance) separated by a few additional nucleotides (66). In the case was efficiently suppressed by these miRNAs; and the sup- of the Long 148/206 construct with miRNA seed ends sep- pression was greater than that achieved by either of the sin- arated by 85 nt, the measured repression was restored to the Nucleic Acids Research, 2015, Vol. 43, No. 19 9515 level of the WT reporter. A similar decrease in cooperative that certain miRNAs directly target the expanded CUG re- repression has been reported previously (5,65). It has also peats in the DMPK mRNA and that the degree of miRNA- been assumed that the availability of miRNA sites that were mediated repression increases with the length of the re- separated is reduced due to the lower concentration of b.s. peated sequence. Because human miRNAs are able to de- (25), and that the loss of the directly adjacent b.s. may lead crease mRNA levels (42), mutant mRNAs with expanded to a shift toward independent activities (7). Together, our CUG triplets may serve as targets for therapeutic interven- results indicate that miR-148a and miR-206 are functional tion by reducing the amount of the toxic transcripts. Re- when their b.s. are located at a distance between 15 and 85 cently, several approaches aimed at targeting CUG repeats nt, and this observation may relate to the regulation of other with antisense oligonucleotides and short interfering RNAs genes targeted by these miRNAs. Here, we refer to miRNA (siRNAs) have been proposed (67–70); therefore, close at- cooperativity because the level of joint repression produced tention should also be paid to the potential use of natural by these two miRNA sites was stronger than the repression or artificial miRNA binding to such repeats. induced by only one of them; however, this repression was Having examined the miRNA targeting of the CUG re- still not greater than that expected from the independent peat tracts of different lengths present in the DMPK tran- contribution of two single sites. In particular, the resulting script using the luciferase system, we examined the poten- increase in the level of suppression (by approximately 20%) tial sequestration of these miRNAs by the exogenously ex- resembled the effect obtained using multiple (but the same) pressed mutant DMPK mRNAs in cells. We overexpressed rather than single binding sites in the reporter constructs either CUG or CAG expanded repeats in the context of (32). DMPK exons 11–15 (36) in HEK 293T cells and found Of particular importance is the fact that the studied miR- that miR-16, with the 7-nt seed sequence AGCAGCA, was NAs exhibit different expression profiles; miR-148a is ubiq- located in cytoplasmic foci formed by the mRNAs with uitously expressed in many tissue and cell types, whereas CUG repeats. We propose that the mutant DMPK tran- miR-206 is almost exclusively expressed in muscles. The scripts may serve as molecular sponges for natural miR- synergistic and/or antagonizing activity of miRNAs with NAs with CAG repeats in their seed sequences, sequester- different expression levels could be employed in therapeu- ing them and thereby preventing their physiological activity. tic interventions using miRNA mimics and/or miRNA in- We speculate that in addition to the sequestration of specific hibitors. The natural selective activity of two miRNAs with RNA-binding proteins, such as the MBNL family proteins overlapping binding sites has been previously reported (9). (12,13), the CUG repeats may also bind certain miRNAs Our genome-wide search for pairs of miRNA sites based on and RNA-induced silencing complexes (RISCs). their expression profiles suggests that both competition for The extension of the CUG repeats amplifies the naturally binding sites and miRNA cooperativity by binding to adja- existing overlapping miRNA b.s., and our luciferase assay cent sites are common miRNA regulatory mechanisms. experiments showed that the miRNA binding capacity in- In summary, our findings for miRNA cooperativity pro- creases with the length of the CUG tracts, which may pro- vide mechanistic insights into the miRNA-mediated regu- mote miRNA cooperativity. The mechanism of concerted lation of the DMPK transcript. Interference with miRNA miRNA regulation by two or more different miRNAs with function has known therapeutic potential; therefore, better overlapping or neighboring b.s. has not been fully eluci- knowledge of the miRNA-mediated gene regulation is es- dated, and we can only speculate how these miRNAs are sential because these potent regulatory RNAs can be used associated with the CUG tracts. The footprint of argonaute in various therapeutic strategies. (Ago), a central component of RISCs, was experimentally defined as 45–62 nt with the use of crosslinking immunopre- cipitation of Ago protein-RNA complexes (71), which indi- CUG repeats present in the DMPK 3 UTR may sequester cates that Ago bound to one b.s. covers the other sites po- CUG-repeat-binding miRNAs sitioned in close vicinity and prevents the binding of other The presence of abnormally long CUG repeats in the Ago molecules. Although long tracts of triplet repeats are DMPK 3 UTR can potentially affect the general physio- likely to form tight hairpin structures, the hairpin stems logical miRNA-mediated regulatory mechanisms of gene are easily accessible to RNA binding proteins (reviewed in expression because these expanded CUG repeats can be a (72)), and we think that they are likely to be the associa- target of the miRNA silencing machinery. Using bioinfor- tion sites for RISCs loaded with CUG binding miRNAs. matics, Hon et al. identified a group of 7 miRNAs that were However, we cannot excluded the possibility that miRNAs predicted to bind to CUG repeats via the CAG sequences containing CAG sequences in their seed regions can directly present in their seed regions. It was also calculated that the associate with CUG repeats without being constantly asso- miRNA binding affinity to CUG repeats may increase with ciated with RISCs, because the presence of stable Ago-free the number of CUG repeat sequences (25). miRNA–mRNA duplexes has been recently demonstrated In the present study, we performed a detailed analy- in mammalian cells (73). The increase in the local concen- sis of miRNA binding to CUG repeats in the DMPK tration of CUG binding miRNAs associated with CUG 3 UTR using both computational and experimental ap- tracts could then enhance the preference for Ago bind- proaches. Using luciferase assays, we examined three miR- ing, resulting in sequestration of the miRNA silencing ma- NAs and demonstrated that miR-15b/16 can significantly chinery. In this context, the sequestration of CUG-repeat- down-regulate its target. Importantly, the repression effi- binding miRNAs could cause them to become inactive or ciency of this miRNA increased with the length of the re- simply prevent them from regulating other transcripts that peated CUG sequence. This result corroborates the findings contain these miRNA b.s., leading to a more widespread 9516 Nucleic Acids Research, 2015, Vol. 43, No. 19 impairment of the miRNA-mediated regulation of gene ex- with a Zeiss Axiovert 200M microscope equipped with a pression (Figure 8B). cooled AxioCam HRc camera. It has been demonstrated that CTG expansions are di- rectly linked to alterations in miRNA regulation and RNA FUNDING toxicity in DM1 (24). An analysis of the muscle miRNA transcriptome in a Drosophila model of CUG toxicity re- Polish Ministry of Science and Higher Education vealed 20 alterations in the miRNA profiles triggered by the [N N301 523038 to E.K.]; National Science Center expression of repeats, and the misregulation of several miR- [2011/03/B/NZ1/03259 and 2012/06/A/NZ1/00094 to NAs has been shown to be independent of Mbnl1 seques- W.J.K.]. This publication was also supported by the Polish tration. In addition, the expansion of CUGs in transgenic Ministry of Science and Higher Education, under the Lead- mice has been shown to affect gene expression at the RNA ing National Research Centre (KNOW) program for the level, and the observed effects were mediated mainly, but years 2014-2019. Funding for open access charge: National not entirely, through the sequestration of Mbnl1 (62). Science Center [2012/06/A/NZ1/00094 to W.J.K.]. We consider that the sequestration of the CUG-repeat- Conflict of interest statement. None declared. binding miRNAs may have wider implications for the RNA-mediated regulation of gene expression. Long non- coding RNAs (lncRNAs) were recently shown to play a vital REFERENCES role in connecting many RNA regulatory pathways (74–78). 1. Bartel,D.P. (2009) MicroRNAs: target recognition and regulatory A growing number of direct links between the levels of miR- functions. Cell, 136, 215–233. NAs, mRNAs and lncRNAs are being discovered (recently 2. Eulalio,A., Huntzinger,E., Nishihara,T., Rehwinkel,J., Fauser,M. and Izaurralde,E. (2009) Deadenylation is a widespread effect of reviewed in (79), and lncRNAs are thought to act as com- miRNA regulation. RNA, 15, 21–32. peting endogenous RNAs (ceRNAs) and miRNA sponges 3. Doench,J.G., Petersen,C.P. and Sharp,P.A. (2003) siRNAs can (80)). Based on these reports, we speculate that the seques- function as miRNAs. Genes Dev., 17, 438–442. tration of CUG-repeat-binding miRNAs can affect the ex- 4. Pillai,R.S., Artus,C.G. and Filipowicz,W. (2004) Tethering of human pression levels of lncRNAs with binding sites for these miR- Ago proteins to mRNA mimics the miRNA-mediated repression of protein synthesis. RNA, 10, 1518–1525. NAs. Imbalances between miRNAs and interacting ceR- 5. Grimson,A., Farh,K.K., Johnston,W.K., Garrett-Engele,P., Lim,L.P. NAs due to altered expression levels or the cellular distribu- and Bartel,D.P. (2007) MicroRNA targeting specificity in mammals: tion and mutagenesis of interacting sites have been shown determinants beyond seed pairing. Mol. Cell, 27, 91–105. to contribute to disease pathogenesis (80). It was also pro- 6. Saetrom,P., Heale,B.S., Snove,O. Jr, Aagaard,L., Alluin,J. and Rossi,J.J. (2007) Distance constraints between microRNA target sites posed that competitor RNAs should be among the most dictate efficacy and cooperativity. Nucleic Acids Res., 35, 2333–2342. abundant RNAs in the cell, or that they should contain 7. Rinck,A., Preusse,M., Laggerbauer,B., Lickert,H., Engelhardt,S. dozens of binding sites for a single miRNA species (81,82). and Theis,F.J. (2013) The human transcriptome is enriched for In this scenario, the long CUG repeats extended in DM1 ap- miRNA-binding sites located in cooperativity-permitting distance. pear to be ideal candidates. Therefore, we believe that lncR- RNA Biol., 10, 1125–1135. 8. Schmitz,U., Lai,X., Winter,F., Wolkenhauer,O., Vera,J. and NAs may play a role in the pathogenesis of TREDs, and Gupta,S.K. (2014) Cooperative gene regulation by microRNA pairs future studies are needed to unravel their functions. and their identification using a computational workflow. Nucleic In conclusion, it is of utmost importance to understand Acids Res., 42, 7539–7552. the role of miRNAs in DM1 to identify potential novel ther- 9. Yu,J., Ryan,D.G., Getsios,S., Oliveira-Fernandes,M., Fatima,A. and Lavker,R.M. (2008) MicroRNA-184 antagonizes microRNA-205 to apies. Given the complexity and degree of interactions be- maintain SHIP2 levels in epithelia. Proc. Natl. Acad. Sci. U.S.A., tween miRNAs and their various targets, which may be ei- 105, 19300–19305. ther protein-coding or non-coding transcripts, a better un- 10. Mahadevan,M., Tsilfidis,C., Sabourin,L., Shutler,G., Amemiya,C., derstanding of the rules governing gene regulation by non- Jansen,G., Neville,C., Narang,M., Barcelo,J., O’Hoy,K. et al. (1992) coding RNAs (ncRNAs) and the intricate mechanisms that Myotonic dystrophy mutation: an unstable CTG repeat in the 3 untranslated region of the gene. Science, 255, 1253–1255. are involved in DM1 pathogenesis is necessary. 11. Davis,B.M., McCurrach,M.E., Taneja,K.L., Singer,R.H. and Housman,D.E. (1997) Expansion of a CUG trinucleotide repeat in SUPPLEMENTARY DATA the 3 untranslated region of myotonic dystrophy protein kinase transcripts results in nuclear retention of transcripts. Proc. Natl. Supplementary Data are available at NAR Online. Acad. Sci. U.S.A., 94, 7388–7393. 12. Lin,X., Miller,J.W., Mankodi,A., Kanadia,R.N., Yuan,Y., ACKNOWLEDGEMENT Moxley,R.T., Swanson,M.S. and Thornton,C.A. (2006) Failure of MBNL1-dependent post-natal splicing transitions in myotonic We thank Thomas A. Cooper for kindly providing the dystrophy. Hum. Mol. Genet., 15, 2087–2097. plasmid vectors expressing the highly expanded repeats 13. Mankodi,A., Takahashi,M.P., Jiang,H., Beck,C.L., Bowers,W.J., Moxley,R.T., Cannon,S.C. and Thornton,C.A. (2002) Expanded and Denis Furling for the DM1 myoblasts. We also thank CUG repeats trigger aberrant splicing of ClC-1 chloride channel Krzysztof Sobczak for sharing materials used for myoblast pre-mRNA and hyperexcitability of skeletal muscle in myotonic differentiation and Edyta Jaworska for assistance with the dystrophy. Mol. Cell, 10, 35–44. miRNA profiling. The real-time PCR analyses were per- 14. Sicot,G., Gourdon,G. and Gomes-Pereira,M. (2011) Myotonic dystrophy, when simple repeats reveal complex pathogenic entities: formed using a LightCycler 480 II system (Roche) in the new findings and future challenges. Hum. Mol. Genet., 20, Laboratory of Subcellular Structures Analysis at the Insti- R116–R123. tute of Bioorganic Chemistry PAS in Poznan. The FISH 15. Batra,R., Charizanis,K., Manchanda,M., Mohan,A., Li,M., images were obtained at the Electron and Confocal Mi- Finn,D.J., Goodwin,M., Zhang,C., Sobczak,K., Thornton,C.A. et al. croscopy Core at Adam Mickiewicz University in Poznan (2014) Loss of MBNL leads to disruption of developmentally Nucleic Acids Research, 2015, Vol. 43, No. 19 9517 regulated alternative polyadenylation in RNA-mediated disease. 35. Figura,G., Koscianska,E. and Krzyzosiak,W.J. (2015) In Vitro Mol. Cell, 56, 311–322. Expansion of CAG, CAA, and Mixed CAG/CAA Repeats. Int J Mol 16. Koscianska,E., Kozlowska,E., Jaworska,E. and Krzyzosiak,W. Sci., 16, 18741–18751. (2014) MicroRNA deregulation in trinucleotide repeat expansion 36. Ho,T.H., Savkur,R.S., Poulos,M.G., Mancini,M.A., Swanson,M.S. disorders. In: Arbuthnot,P and Weinberg,MS (eds). Applied RNAi: and Cooper,T.A. (2005) Colocalization of muscleblind with RNA From Fundamental Research to Therapeutic Applications. Caister foci is separable from mis-regulation of alternative splicing in Academic Press, pp. 227–246. myotonic dystrophy. J. Cell Sci., 118, 2923–2933. 17. Latronico,M.V. and Condorelli,G. (2009) MicroRNAs and cardiac 37. Koscianska,E., Starega-Roslan,J., Czubala,K. and Krzyzosiak,W.J. pathology. Nat. Rev. Cardiol., 6, 419–429. (2011) High-resolution northern blot for a reliable analysis of 18. van Rooij,E., Liu,N. and Olson,E.N. (2008) MicroRNAs flex their microRNAs and their precursors. ScientificWorldJournal , 11, muscles. Trends Genet., 24, 159–166. 102–117. 19. Eisenberg,I., Eran,A., Nishino,I., Moggio,M., Lamperti,C., 38. Koscianska,E., Starega-Roslan,J., Sznajder,L.J., Olejniczak,M., Amato,A.A., Lidov,H.G., Kang,P.B., North,K.N., Galka-Marciniak,P. and Krzyzosiak,W.J. (2011) Northern blotting Mitrani-Rosenbaum,S. et al. (2007) Distinctive patterns of analysis of microRNAs, their precursors and RNA interference microRNA expression in primary muscular disorders. Proc. Natl. triggers. BMC Mol. Biol., 12, 14. Acad. Sci. U.S.A., 104, 17016–17021. 39. Wojciechowska,M., Taylor,K., Sobczak,K., Napierala,M. and 20. Greco,S., Perfetti,A., Fasanaro,P., Cardani,R., Capogrossi,M.C., Krzyzosiak,W.J. (2014) Small molecule kinase inhibitors alleviate Meola,G. and Martelli,F. (2012) Deregulated microRNAs in different molecular features of myotonic dystrophy type 1. RNA myotonic dystrophy type 2. PLoS One, 7, e39732. Biol., 11, 742–754. 21. Perbellini,R., Greco,S., Sarra-Ferraris,G., Cardani,R., 40. Witkos,T.M., Koscianska,E. and Krzyzosiak,W.J. (2011) Practical Capogrossi,M.C., Meola,G. and Martelli,F. (2011) Dysregulation Aspects of microRNA Target Prediction. Curr.Mol.Med., 11, and cellular mislocalization of specific miRNAs in myotonic 93–109. dystrophy type 1. Neuromuscul. Disord., 21, 81–88. 41. Baek,D., Villen,J., Shin,C., Camargo,F.D., Gygi,S.P. and Bartel,D.P. 22. Rau,F., Freyermuth,F., Fugier,C., Villemin,J.P., Fischer,M.C., (2008) The impact of microRNAs on protein output. Nature, 455, Jost,B., Dembele,D., Gourdon,G., Nicole,A., Duboc,D. et al. (2011) 64–71. Misregulation of miR-1 processing is associated with heart defects in 42. Guo,H., Ingolia,N.T., Weissman,J.S. and Bartel,D.P. (2010) myotonic dystrophy. Nat. Struct. Mol. Biol., 18, 840–845. Mammalian microRNAs predominantly act to decrease target 23. Gambardella,S., Rinaldi,F., Lepore,S.M., Viola,A., Loro,E., mRNA levels. Nature, 466, 835–840. Angelini,C., Vergani,L., Novelli,G. and Botta,A. (2010) 43. Llorens,F., Banez-Coronel,M., Pantano,L., del Rio,J.A., Ferrer,I., Overexpression of microRNA-206 in the skeletal muscle from Estivill,X. and Marti,E. (2013) A highly expressed miR-101 isomiR myotonic dystrophy type 1 patients. J. Transl. Med., 8, 48. is a functional silencing small RNA. BMC Genom., 14, 104. 24. Fernandez-Costa,J.M., Garcia-Lopez,A., Zuniga,S., 44. Erriquez,D., Perini,G. and Ferlini,A. (2013) Non-coding RNAs in Fernandez-Pedrosa,V., Felipo-Benavent,A., Mata,M., Jaka,O., muscle dystrophies. Int. J. Mol. Sci., 14, 19681–19704. Aiastui,A., Hernandez-Torres,F., Aguado,B. et al. (2012) Expanded 45. Ge,Y. and Chen,J. (2011) MicroRNAs in skeletal myogenesis. Cell CTG repeats trigger miRNA alterations in Drosophila that are Cycle, 10, 441–448. conserved in myotonic dystrophy type 1 patients. Hum. Mol. Genet., 46. Lin,C.Y., Lee,H.C., Fu,C.Y., Ding,Y.Y., Chen,J.S., Lee,M.H., 22, 704–716. Huang,W.J. and Tsai,H.J. (2013) MiR-1 and miR-206 target different 25. Hon,L.S. and Zhang,Z. (2007) The roles of binding site arrangement genes to have opposing roles during angiogenesis in zebrafish and combinatorial targeting in microRNA repression of gene embryos. Nat. Commun., 4, 2829. expression. Genome Biol., 8, R166. 47. Brennecke,J., Stark,A., Russell,R.B. and Cohen,S.M. (2005) 26. Betel,D., Koppal,A., Agius,P., Sander,C. and Leslie,C. (2010) Principles of microRNA-target recognition. PLoS Biol., 3, e85. Comprehensive modeling of microRNA targets predicts functional 48. Kertesz,M., Iovino,N., Unnerstall,U., Gaul,U. and Segal,E. (2007) non-conserved and non-canonical sites. Genome Biol., 11, R90. The role of site accessibility in microRNA target recognition. Nat. 27. Garcia,D.M., Baek,D., Shin,C., Bell,G.W., Grimson,A. and Genet., 39, 1278–1284. Bartel,D.P. (2011) Weak seed-pairing stability and high target-site 49. Long,D., Lee,R., Williams,P., Chan,C.Y., Ambros,V. and Ding,Y. abundance decrease the proficiency of lsy-6 and other microRNAs. (2007) Potent effect of target structure on microRNA function. Nat. Nat. Struct. Mol. Biol., 18, 1139–1146. Struct. Mol. Biol., 14, 287–294. 28. Maragkakis,M., Reczko,M., Simossis,V.A., Alexiou,P., 50. Robins,H., Li,Y. and Padgett,R.W. (2005) Incorporating structure to Papadopoulos,G.L., Dalamagas,T., Giannopoulos,G., Goumas,G., predict microRNA targets. Proc. Natl. Acad. Sci. U.S.A., 102, Koukis,E., Kourtis,K. et al. (2009) DIANA-microT web server: 4006–4009. elucidating microRNA functions through target prediction. Nucleic 51. Guo,Z., Maki,M., Ding,R., Yang,Y., Zhang,B. and Xiong,L. (2014) Acids Res., 37, W273–W276. Genome-wide survey of tissue-specific microRNA and transcription 29. Kozomara,A. and Griffiths-Jones,S. (2014) miRBase: annotating factor regulatory networks in 12 tissues. Sci. Rep., 4, 5150. high confidence microRNAs using deep sequencing data. Nucleic 52. Finnerty,J.R., Wang,W.X., Hebert,S.S., Wilfred,B.R., Mao,G. and Acids Res., 42, D68–D73. Nelson,P.T. (2010) The miR-15/107 group of microRNA genes: 30. Lewis,B.P., Burge,C.B. and Bartel,D.P. (2005) Conserved seed evolutionary biology, cellular functions, and roles in human diseases. pairing, often flanked by adenosines, indicates that thousands of J. Mol. Biol., 402, 491–509. human genes are microRNA targets. Cell, 120, 15–20. 53. Nelson,P.T., Wang,W.X., Mao,G., Wilfred,B.R., Xie,K., 31. Hsu,S.D., Chu,C.H., Tsou,A.P., Chen,S.J., Chen,H.C., Hsu,P.W., Jennings,M.H., Gao,Z. and Wang,X. (2011) Specific sequence Wong,Y.H., Chen,Y.H., Chen,G.H. and Huang,H.D. (2008) determinants of miR-15/107 microRNA gene group targets. Nucleic miRNAMap 2.0: genomic maps of microRNAs in metazoan Acids Res., 39, 8163–8172. genomes. Nucleic Acids Res., 36, D165–D169. 54. Wang,W.X., Kyprianou,N., Wang,X. and Nelson,P.T. (2010) 32. Koscianska,E., Baev,V., Skreka,K., Oikonomaki,K., Rusinov,V., Dysregulation of the mitogen granulin in human cancer through the Tabler,M. and Kalantidis,K. (2007) Prediction and preliminary miR-15/107 microRNA gene group. Cancer Res., 70, 9137–9142. validation of oncogene regulation by miRNAs. BMC Mol. Biol., 8, 55. Wang,W.X., Danaher,R.J., Miller,C.S., Berger,J.R., Nubia,V.G., 79. Wilfred,B.S., Neltner,J.H., Norris,C.M. and Nelson,P.T. (2014) 33. Kozlowska,E., Krzyzosiak,W.J. and Koscianska,E. (2013) Regulation Expression of miR-15/107 family microRNAs in human tissues and of huntingtin gene expression by miRNA-137, -214, -148a, and their cultured rat brain cells. Genomics Proteomics Bioinformatics, 12, respective isomiRs. Int. J. Mol. Sci., 14, 16999–17016. 19–30. 34. Takahashi,N., Sasagawa,N., Suzuki,K. and Ishiura,S. (1999) 56. Langlois,M.A., Boniface,C., Wang,G., Alluin,J., Salvaterra,P.M., Synthesis of long trinucleotide repeats in vitro. Neurosci. Lett., 262, Puymirat,J., Rossi,J.J. and Lee,N.S. (2005) Cytoplasmic and nuclear 45–48. retained DMPK mRNAs are targets for RNA interference in myotonic dystrophy cells. J. Biol. Chem., 280, 16949–16954. 9518 Nucleic Acids Research, 2015, Vol. 43, No. 19 57. Amack,J.D. and Mahadevan,M.S. (2001) The myotonic dystrophy 71. Chi,S.W., Zang,J.B., Mele,A. and Darnell,R.B. (2009) Argonaute expanded CUG repeat tract is necessary but not sufficient to disrupt HITS-CLIP decodes microRNA-mRNA interaction maps. Nature, C2C12 myoblast differentiation. Hum. Mol. Genet., 10, 1879–1887. 460, 479–486. 58. Dansithong,W., Wolf,C.M., Sarkar,P., Paul,S., Chiang,A., Holt,I., 72. Krzyzosiak,W.J., Sobczak,K., Wojciechowska,M., Fiszer,A., Morris,G.E., Branco,D., Sherwood,M.C., Comai,L. et al. (2008) Mykowska,A. and Kozlowski,P. (2012) Triplet repeat RNA structure Cytoplasmic CUG RNA foci are insufficient to elicit key DM1 and its role as pathogenic agent and therapeutic target. Nucleic Acids features. PLoS One, 3, e3968. Res., 40, 11–26. 59. Taneja,K.L., McCurrach,M., Schalling,M., Housman,D. and 73. Janas,M.M., Wang,B., Harris,A.S., Aguiar,M., Shaffer,J.M., Singer,R.H. (1995) Foci of trinucleotide repeat transcripts in nuclei Subrahmanyam,Y.V., Behlke,M.A., Wucherpfennig,K.W., Gygi,S.P., of myotonic dystrophy cells and tissues. J. Cell Biol., 128, 995–1002. Gagnon,E. et al. (2012) Alternative RISC assembly: binding and 60. Timchenko,L.T., Timchenko,N.A., Caskey,C.T. and Roberts,R. repression of microRNA-mRNA duplexes by human Ago proteins. (1996) Novel proteins with binding specificity for DNA CTG repeats RNA, 18, 2041–2055. and RNA CUG repeats: implications for myotonic dystrophy. Hum. 74. Cesana,M., Cacchiarelli,D., Legnini,I., Santini,T., Sthandier,O., Mol. Genet., 5, 115–121. Chinappi,M., Tramontano,A. and Bozzoni,I. (2011) A long 61. Ohrt,T., Muetze,J., Svoboda,P. and Schwille,P. (2012) Intracellular noncoding RNA controls muscle differentiation by functioning as a localization and routing of miRNA and RNAi pathway components. competing endogenous RNA. Cell, 147, 358–369. Curr.Top.Med.Chem., 12, 79–88. 75. Karreth,F.A., Tay,Y., Perna,D., Ala,U., Tan,S.M., Rust,A.G., 62. Osborne,R.J., Lin,X., Welle,S., Sobczak,K., O’Rourke,J.R., DeNicola,G., Webster,K.A., Weiss,D., Perez-Mancera,P.A. et al. Swanson,M.S. and Thornton,C.A. (2009) Transcriptional and (2011) In vivo identification of tumor- suppressive PTEN ceRNAs in post-transcriptional impact of toxic RNA in myotonic dystrophy. an oncogenic BRAF-induced mouse model of melanoma. Cell, 147, Hum. Mol. Genet., 18, 1471–1481. 382–395. 63. Kloosterman,W.P., Wienholds,E., Ketting,R.F. and Plasterk,R.H. 76. Poliseno,L., Salmena,L., Zhang,J., Carver,B., Haveman,W.J. and (2004) Substrate requirements for let-7 function in the developing Pandolfi,P.P. (2010) A coding-independent function of gene and zebrafish embryo. Nucleic Acids Res., 32, 6284–6291. pseudogene mRNAs regulates tumour biology. Nature, 465, 64. Vella,M.C., Choi,E.Y., Lin,S.Y., Reinert,K. and Slack,F.J. (2004) The 1033–1038. C. elegans microRNA let-7 binds to imperfect let-7 complementary 77. Sumazin,P., Yang,X., Chiu,H.S., Chung,W.J., Iyer,A., sites from the lin-41 3 UTR. Genes Dev., 18, 132–137. Llobet-Navas,D., Rajbhandari,P., Bansal,M., Guarnieri,P., Silva,J. 65. Broderick,J.A., Salomon,W.E., Ryder,S.P., Aronin,N. and et al. (2011) An extensive microRNA-mediated network of Zamore,P.D. (2011) Argonaute protein identity and pairing RNA-RNA interactions regulates established oncogenic pathways in geometry determine cooperativity in mammalian RNA silencing. glioblastoma. Cell, 147, 370–381. RNA, 17, 1858–1869. 78. Tay,Y., Kats,L., Salmena,L., Weiss,D., Tan,S.M., Ala,U., Karreth,F., 66. Doench,J.G. and Sharp,P.A. (2004) Specificity of microRNA target Poliseno,L., Provero,P., Di Cunto,F. et al. (2011) selection in translational repression. Genes Dev., 18, 504–511. Coding-independent regulation of the tumor suppressor PTEN by 67. Evers,M.M., Pepers,B.A., van Deutekom,J.C., Mulders,S.A., den competing endogenous mRNAs. Cell, 147, 344–357. Dunnen,J.T., Aartsma-Rus,A., van Ommen,G.J. and van 79. Yoon,J.H., Abdelmohsen,K. and Gorospe,M. (2014) Functional Roon-Mom,W.M. (2011) Targeting several CAG expansion diseases interactions among microRNAs and long noncoding RNAs. Semin. by a single antisense oligonucleotide. PLoS One, 6, e24308. Cell Dev. Biol., 34, 9–14. 68. Sicot,G. and Gomes-Pereira,M. (2013) RNA toxicity in human 80. Tay,Y., Rinn,J. and Pandolfi,P.P. (2014) The multilayered complexity disease and animal models: from the uncovering of a new mechanism of ceRNA crosstalk and competition. Nature, 505, 344–352. to the development of promising therapies. Biochim. Biophys. Acta, 81. Broderick,J.A. and Zamore,P.D. (2014) Competitive endogenous 1832, 1390–1409. RNAs cannot alter microRNA function in vivo. Mol. Cell, 54, 69. Udd,B. and Krahe,R. (2012) The myotonic dystrophies: molecular, 711–713. clinical, and therapeutic challenges. Lancet Neurol., 11, 891–905. 82. Denzler,R., Agarwal,V., Stefano,J., Bartel,D.P. and Stoffel,M. (2014) 70. Sobczak,K., Wheeler,T.M., Wang,W. and Thornton,C.A. (2013) Assessing the ceRNA hypothesis with quantitative measurements of RNA interference targeting CUG repeats in a mouse model of miRNA and target abundance. Mol. Cell, 54, 766–776. myotonic dystrophy. Mol. Ther., 21, 380–387. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nucleic Acids Research Oxford University Press http://www.deepdyve.com/lp/oxford-university-press/cooperation-meets-competition-inmicrorna-mediated-dmpk-transcript-07WCWnFel4

Cooperation meets competition inmicroRNA-mediated DMPK transcript regulation

Koscianska, Edyta; Witkos, Tomasz M.; Kozlowska, Emilia; Wojciechowska, Marzena; Krzyzosiak, Wlodzimierz J.

Nucleic Acids Research , Volume 43 (19) – Oct 30, 2015

Download PDF

Share Full Text for Free

19 pages

Loading...

Page 2

Loading...

Page 3

Loading...

Page 4

Loading...

Page 5

Loading...

Page 6

Loading...

Page 7

Loading...

Page 8

Loading...

Page 9

Loading...

Page 10

Loading...

Page 11

Loading...

Page 12

Loading...

Page 13

Loading...

Page 14

Loading...

Page 15

Loading...

Page 16

Loading...

Page 17

Loading...

Page 18

Loading...

Page 19

References (86)

Wang-Xia Wang, N. Kyprianou, Xiaowei Wang, P. Nelson (2010)
Dysregulation of the mitogen granulin in human cancer through the miR-15/107 microRNA gene group.
Cancer research, 70 22
David Garcia, Daehyun Baek, C. Shin, G. Bell, Andrew Grimson, D. Bartel (2011)
Weak Seed-Pairing Stability and High Target-Site Abundance Decrease the Proficiency of lsy-6 and Other miRNAs
Nature structural & molecular biology, 18
M. Latronico, G. Condorelli (2009)
MicroRNAs and cardiac pathology
Nature Reviews Cardiology, 6
J. Finnerty, Wang-Xia Wang, Sébastien Hébert, Bernard Wilfred, Guogen Mao, P. Nelson (2010)
The miR-15/107 group of microRNA genes: evolutionary biology, cellular functions, and roles in human diseases.
Journal of molecular biology, 402 3
Ana Kozomara, S. Griffiths-Jones (2013)
miRBase: annotating high confidence microRNAs using deep sequencing data
Nucleic Acids Research, 42
M. Mahadevan, C. Tsilfidis, L. Sabourin, G. Shutler, C. Amemiya, G. Jansen, C. Neville, M. Narang, J. Barceló, K. O'Hoy (1992)
Myotonic dystrophy mutation: an unstable CTG repeat in the 3' untranslated region of the gene.
Science, 255 5049
Zhiyun Guo, M. Maki, R. Ding, Yalan Yang, Bao-Ting Zhang, Lili Xiong (2014)
Genome-wide survey of tissue-specific microRNA and transcription factor regulatory networks in 12 tissues
Scientific Reports, 4
R. Pillai, C. Artus, W. Filipowicz (2004)
Tethering of human Ago proteins to mRNA mimics the miRNA-mediated repression of protein synthesis.
RNA, 10 10
W. Kloosterman, Erno Wienholds, R. Ketting, R. Plasterk (2016)
Substrate requirements for let-7 function in the developing zebrafish embryo
Nucleic Acids Research, 44
Ranjan Batra, K. Charizanis, Mini Manchanda, Apoorva Mohan, Moyi Li, Dustin Finn, M. Goodwin, Chaolin Zhang, K. Sobczak, C. Thornton, M. Swanson (2014)
Loss of MBNL leads to disruption of developmentally regulated alternative polyadenylation in RNA-mediated disease.
Molecular cell, 56 2
I. Eisenberg, A. Eran, I. Nishino, M. Moggio, C. Lamperti, A. Amato, H. Lidov, P. Kang, K. North, S. Mitrani-Rosenbaum, K. Flanigan, L. Neely, D. Whitney, A. Beggs, I. Kohane, L. Kunkel (2008)
Correction for Eisenberg et al., Distinctive patterns of microRNA expression in primary muscular disorders
Proceedings of the National Academy of Sciences, 105
P. Nelson, Wang-Xia Wang, Guogen Mao, Bernard Wilfred, Kevin Xie, Mary Jennings, Zhen Gao, Xiaowei Wang (2011)
Specific sequence determinants of miR-15/107 microRNA gene group targets
Nucleic Acids Research, 39
M. Vella, E. Choi, Shin-Yi Lin, Kristy Reinert, F. Slack (2004)
The C. elegans microRNA let-7 binds to imperfect let-7 complementary sites from the lin-41 3'UTR.
Genes & development, 18 2
J. Broderick, P. Zamore (2014)
Competitive endogenous RNAs cannot alter microRNA function in vivo.
Molecular cell, 54 5
W. Krzyzosiak, K. Sobczak, M. Wojciechowska, Agnieszka Fiszer, A. Mykowska, P. Kozłowski (2011)
Triplet repeat RNA structure and its role as pathogenic agent and therapeutic target
Nucleic Acids Research, 40
S. Greco, A. Perfetti, P. Fasanaro, R. Cardani, M. Capogrossi, G. Meola, F. Martelli (2012)
Deregulated MicroRNAs in Myotonic Dystrophy Type 2
PLoS ONE, 7
Dang Long, Rosalind Lee, P. Williams, C. Chan, V. Ambros, Ye Ding (2007)
Potent effect of target structure on microRNA function
Nature Structural &Molecular Biology, 14
J. Brennecke, A. Stark, R. Russell, S. Cohen (2005)
Principles of MicroRNA–Target Recognition
PLoS Biology, 3
M. Wojciechowska, Katarzyna Taylor, K. Sobczak, Marek Napierala, Wlodzimierz Krzyzosiak (2014)
Small molecule kinase inhibitors alleviate different molecular features of myotonic dystrophy type 1
RNA Biology, 11
T. Ho, R. Savkur, Michael Poulos, M. Mancini, M. Swanson, T. Cooper (2005)
Colocalization of muscleblind with RNA foci is separable from mis-regulation of alternative splicing in myotonic dystrophy
Journal of Cell Science, 118
Michael Kertesz, Nicola Iovino, Ulrich Unnerstall, Ulrike Gaul, E. Segal (2007)
The role of site accessibility in microRNA target recognition
Nature Genetics, 39
E. Koscianska, Vesselin Baev, Konstantinia Skreka, Katerina Oikonomaki, V. Rusinov, Martin Tabler, K. Kalantidis (2007)
Prediction and preliminary validation of oncogene regulation by miRNAs
BMC Molecular Biology, 8
Emilia Kozłowska, W. Krzyzosiak, E. Koscianska (2013)
Regulation of Huntingtin Gene Expression by miRNA-137, -214, -148a, and Their Respective isomiRs
International Journal of Molecular Sciences, 14
Xiaoyan Lin, Jill Miller, A. Mankodi, R. Kanadia, Yuan Yuan, R. Moxley, M. Swanson, C. Thornton (2006)
Failure of MBNL1-dependent post-natal splicing transitions in myotonic dystrophy.
Human molecular genetics, 15 13
R. Perbellini, S. Greco, Gianluca Sarra-Ferraris, R. Cardani, M. Capogrossi, G. Meola, F. Martelli (2011)
Dysregulation and cellular mislocalization of specific miRNAs in myotonic dystrophy type 1
Neuromuscular Disorders, 21
Huili Guo, Nicholas Ingolia, J. Weissman, D. Bartel (2010)
Mammalian microRNAs predominantly act to decrease target mRNA levels
Nature, 466
P. Sætrom, Bret Heale, O. Snøve, L. Aagaard, Jessica Alluin, J. Rossi (2007)
Distance constraints between microRNA target sites dictate efficacy and cooperativity
Nucleic Acids Research, 35
E. Koscianska, Julia Staręga-Rosłan, Lukasz Sznajder, M. Olejniczak, P. Galka-Marciniak, W. Krzyzosiak (2011)
Northern blotting analysis of microRNAs, their precursors and RNA interference triggers
BMC Molecular Biology, 12
F. Rau, F. Freyermuth, C. Fugier, J. Villemin, Marie-Christine Fischer, B. Jost, D. Dembélé, G. Gourdon, A. Nicole, D. Duboc, K. Wahbi, J. Day, H. Fujimura, M. Takahashi, D. Auboeuf, Natacha Dreumont, D. Furling, N. Charlet-Berguerand (2011)
Misregulation of miR-1 processing is associated with heart defects in myotonic dystrophy
Nature Structural &Molecular Biology, 18
Jeffrey Amack, M. Mahadevan (2001)
The myotonic dystrophy expanded CUG repeat tract is necessary but not sufficient to disrupt C2C12 myoblast differentiation.
Human molecular genetics, 10 18
(2003)
siRNAs can function as miRNAs
(2009)
Argonaute HITS-CLIP decodes microRNA-mRNA interaction maps
Nature, 460
John Doench, P. Sharp (2004)
Specificity of microRNA target selection in translational repression.
Genes & development, 18 5
Manolis Maragkakis, M. Reczko, V. Simossis, P. Alexiou, Giorgos Papadopoulos, Theodore Dalamagas, G. Giannopoulos, G. Goumas, Evangelos Koukis, K. Kourtis, Thanasis Vergoulis, N. Koziris, T. Sellis, P. Tsanakas, A. Hatzigeorgiou (2009)
DIANA-microT web server: elucidating microRNA functions through target prediction
Nucleic Acids Research, 37
J. Fernández‐Costa, A. García-López, S. Zuniga, V. Fernández-Pedrosa, Amelia Felipo-Benavent, M. Mata, O. Jaka, A. Aiastui, F. Hernández-Torres, B. Aguado, M. Perez-Alonso, J. Vilchez, A. Munain, R. Artero (2013)
Expanded CTG repeats trigger miRNA alterations in Drosophila that are conserved in myotonic dystrophy type 1 patients.
Human molecular genetics, 22 4
L. Poliseno, L. Salmena, Jiangwen Zhang, B. Carver, W. Haveman, P. Pandolfi (2010)
A coding-independent function of gene and pseudogene mRNAs regulates tumour biology
Nature, 465
Deborah Hurt, K. Winestock, M. O'connor, M. Johnston (2015)
Nucleic Acids Research
Nucleic Acids Research, 43
Yejing Ge, Jie Chen (2011)
MicroRNAs in skeletal myogenesis
Cell Cycle, 10
T. Ohrt, Joerg Muetze, P. Svoboda, P. Schwille (2012)
Intracellular localization and routing of miRNA and RNAi pathway components.
Current topics in medicinal chemistry, 12 2
Maja Janas, B. Wang, Abigail Harris, M. Aguiar, J. Shaffer, Y. Subrahmanyam, M. Behlke, K. Wucherpfennig, S. Gygi, É. Gagnon, C. Novina (2012)
Alternative RISC assembly: binding and repression of microRNA-mRNA duplexes by human Ago proteins.
RNA, 18 11
Yvonne Tay, L. Kats, L. Salmena, Dror Weiss, S. Tan, U. Ala, F. Karreth, L. Poliseno, P. Provero, F. Cunto, J. Lieberman, I. Rigoutsos, P. Pandolfi (2011)
Coding-Independent Regulation of the Tumor Suppressor PTEN by Competing Endogenous mRNAs
Cell, 147
B. Lewis, C. Burge, D. Bartel (2005)
Conserved Seed Pairing, Often Flanked by Adenosines, Indicates that Thousands of Human Genes are MicroRNA Targets
Cell, 120
Grzegorz Figura, E. Koscianska, W. Krzyzosiak (2015)
In Vitro Expansion of CAG, CAA, and Mixed CAG/CAA Repeats
International Journal of Molecular Sciences, 16
Daehyun Baek, J. Villén, C. Shin, F. Camargo, S. Gygi, D. Bartel (2008)
The impact of microRNAs on protein output
Nature, 455
Tomasz Witkos, E. Koscianska, W. Krzyzosiak (2011)
Practical Aspects of microRNA Target Prediction
Current Molecular Medicine, 11
D. Bartel (2009)
MicroRNAs: Target Recognition and Regulatory Functions
Cell, 136
Rémy Denzler, Vikram Agarwal, Joanna Stefano, D. Bartel, M. Stoffel (2014)
Assessing the ceRNA hypothesis with quantitative measurements of miRNA and target abundance.
Molecular cell, 54 5
Marc-André Langlois, C. Boniface, G. Wang, Jessica Alluin, P. Salvaterra, J. Puymirat, J. Rossi, N. Lee (2005)
Cytoplasmic and Nuclear Retained DMPK mRNAs Are Targets for RNA Interference in Myotonic Dystrophy Cells*
Journal of Biological Chemistry, 280
Jia Yu, D. Ryan, S. Getsios, Michelle Oliveira-Fernandes, A. Fatima, R. Lavker (2008)
MicroRNA-184 antagonizes microRNA-205 to maintain SHIP2 levels in epithelia
Proceedings of the National Academy of Sciences, 105
K. Taneja, M. McCurrach, M. Schalling, D. Housman, R. Singer (1995)
Foci of trinucleotide repeat transcripts in nuclei of myotonic dystrophy cells and tissues
The Journal of Cell Biology, 128
Sheng-Da Hsu, Chia-Huei Chu, A. Tsou, Shu-Jen Chen, Hua-Chien Chen, P. Hsu, Y. Wong, Yi-Hsuan Chen, Gian-Hung Chen, Hsien-Da Huang (2007)
miRNAMap 2.0: genomic maps of microRNAs in metazoan genomes
Nucleic Acids Research, 36
W. Dansithong, C. Wolf, P. Sarkar, S. Paul, Andy Chiang, I. Holt, G. Morris, D. Branco, M. Sherwood, L. Comai, C. Berul, S. Reddy (2008)
Cytoplasmic CUG RNA Foci Are Insufficient to Elicit Key DM1 Features
PLoS ONE, 3
M. Evers, B. Pepers, J. Deutekom, S. Mulders, J. Dunnen, A. Aartsma-Rus, G. Ommen, W. Roon-Mom (2011)
Targeting Several CAG Expansion Diseases by a Single Antisense Oligonucleotide
PLoS ONE, 6
G. Sicot, G. Gourdon, M. Gomes-Pereira (2011)
Myotonic dystrophy, when simple repeats reveal complex pathogenic entities: new findings and future challenges.
Human molecular genetics, 20 R2
Pavel Sumazin, Xuerui Yang, Hua-Sheng Chiu, W. Chung, A. Iyer, D. Llobet-Navás, Presha Rajbhandari, M. Bansal, P. Guarnieri, Jose Silva, A. Califano (2011)
An Extensive MicroRNA-Mediated Network of RNA-RNA Interactions Regulates Established Oncogenic Pathways in Glioblastoma
Cell, 147
M. Kiebler, Jacki Heraud (2009)
Faculty Opinions recommendation of Argonaute HITS-CLIP decodes microRNA-mRNA interaction maps.
A. Mankodi, M. Takahashi, Hong Jiang, C. Beck, W. Bowers, R. Moxley, S. Cannon, C. Thornton (2002)
Expanded CUG repeats trigger aberrant splicing of ClC-1 chloride channel pre-mRNA and hyperexcitability of skeletal muscle in myotonic dystrophy.
Molecular cell, 10 1
G. Sicot, M. Gomes-Pereira (2013)
RNA toxicity in human disease and animal models: from the uncovering of a new mechanism to the development of promising therapies.
Biochimica et biophysica acta, 1832 9
Lawrence Hon, Zemin Zhang (2007)
The roles of binding site arrangement and combinatorial targeting in microRNA repression of gene expression
Genome Biology, 8
H. Robins, Y. Li, R. Padgett (2005)
Incorporating structure to predict microRNA targets.
Proceedings of the National Academy of Sciences of the United States of America, 102 11
U. Schmitz, Xin Lai, Felix Winter, O. Wolkenhauer, J. Vera, S. Gupta (2014)
Cooperative gene regulation by microRNA pairs and their identification using a computational workflow
Nucleic Acids Research, 42
B. Davis, M. McCurrach, M. McCurrach, K. Taneja, R. Singer, D. Housman (1997)
Expansion of a CUG trinucleotide repeat in the 3' untranslated region of myotonic dystrophy protein kinase transcripts results in nuclear retention of transcripts.
Proceedings of the National Academy of Sciences of the United States of America, 94 14
N. Takahashi, N. Sasagawa, Koichi Suzuki, S. Ishiura (1999)
Synthesis of long trinucleotide repeats in vitro
Neuroscience Letters, 262
E. Rooij, Ning Liu, E. Olson (2008)
MicroRNAs flex their muscles.
Trends in genetics : TIG, 24 4
D. Betel, Anjali Koppal, P. Agius, C. Sander, C. Leslie (2010)
Comprehensive modeling of microRNA targets predicts functional non-conserved and non-canonical sites
Genome Biology, 11
Andrea Rinck, M. Preusse, B. Laggerbauer, H. Lickert, S. Engelhardt, Fabian Theis (2013)
The human transcriptome is enriched for miRNA-binding sites located in cooperativity-permitting distance
RNA Biology, 10
Yvonne Tay, Johannes Rinn, P. Pandolfi (2014)
The multilayered complexity of ceRNA crosstalk and competition
Nature, 505
Je-Hyun Yoon, K. Abdelmohsen, M. Gorospe (2014)
Functional interactions among microRNAs and long noncoding RNAs.
Seminars in cell & developmental biology, 34
E. Koscianska, Julia Staręga-Rosłan, K. Czubała, W. Krzyzosiak (2011)
High-Resolution Northern Blot for a Reliable Analysis of MicroRNAs and Their Precursors
The Scientific World Journal, 11
Robert Osborne, Xiaoyan Lin, S. Welle, K. Sobczak, Jason O’Rourke, M. Swanson, C. Thornton (2009)
Transcriptional and post-transcriptional impact of toxic RNA in myotonic dystrophy.
Human molecular genetics, 18 8
F. Llorens, Mónica Báñez-Coronel, L. Pantano, J. Río, I. Ferrer, X. Estivill, E. Martí (2013)
A highly expressed miR-101 isomiR is a functional silencing small RNA
BMC Genomics, 14
F. Karreth, Yvonne Tay, D. Perna, U. Ala, S. Tan, A. Rust, G. DeNicola, Kaitlyn Webster, Dror Weiss, P. Pérez-Mancera, M. Krauthammer, R. Halaban, P. Provero, D. Adams, D. Tuveson, P. Pandolfi (2011)
In Vivo Identification of Tumor- Suppressive PTEN ceRNAs in an Oncogenic BRAF-Induced Mouse Model of Melanoma
Cell, 147
Wang-Xia Wang, R. Danaher, Craig Miller, J. Berger, Vega Nubia, Bernard Wilfred, J. Neltner, C. Norris, P. Nelson (2014)
Expression of miR-15/107 Family MicroRNAs in Human Tissues and Cultured Rat Brain Cells
Genomics, Proteomics & Bioinformatics, 12
M. Cesana, D. Cacchiarelli, I. Legnini, Tiziana Santini, O. Sthandier, M. Chinappi, A. Tramontano, I. Bozzoni (2011)
A Long Noncoding RNA Controls Muscle Differentiation by Functioning as a Competing Endogenous RNA
Cell, 147
L. Timchenko, Nikolaj Timchenko, C. Caskey, R. Roberts (1996)
Novel proteins with binding specificity for DNA CTG repeats and RNA CUG repeats: implications for myotonic dystrophy.
Human molecular genetics, 5 1
Cheng-Yung Lin, Hung‐Chieh Lee, C. Fu, Yu-yun Ding, Jie-Shin Chen, Ming-Hsuan Lee, Wei-Jhen Huang, H. Tsai (2013)
miR-1 and miR-206 target different genes to have opposing roles during angiogenesis in zebrafish embryos
Nature Communications, 4
K. Sobczak, T. Wheeler, Wenli Wang, C. Thornton (2013)
RNA interference targeting CUG repeats in a mouse model of myotonic dystrophy.
Molecular therapy : the journal of the American Society of Gene Therapy, 21 2
A. Eulalio, E. Huntzinger, Tadashi Nishihara, J. Rehwinkel, Maria Fauser, E. Izaurralde (2008)
Deadenylation is a widespread effect of miRNA regulation.
RNA, 15 1
Andrew Grimson, K. Farh, W. Johnston, P. Garrett-engele, Lee Lim, D. Bartel (2007)
MicroRNA targeting specificity in mammals: determinants beyond seed pairing.
Molecular cell, 27 1
(2004)
Substrate requirements for let-7 function in the developing zebrafish embryo
Nucleic Acids Res., 32
J. Broderick, William Salomon, S. Ryder, N. Aronin, P. Zamore (2011)
Argonaute protein identity and pairing geometry determine cooperativity in mammalian RNA silencing.
RNA, 17 10
B. Mari, B. Bardoni (2014)
Applied RNAi: from fundamental research to therapeutic applications
Frontiers in Genetics, 5
B. Udd, R. Krahe (2012)
The myotonic dystrophies: molecular, clinical, and therapeutic challenges
The Lancet Neurology, 11
S. Gambardella, F. Rinaldi, S. Lepore, A. Viola, E. Loro, C. Angelini, L. Vergani, G. Novelli, A. Botta (2010)
Overexpression of microRNA-206 in the skeletal muscle from myotonic dystrophy type 1 patients
Journal of Translational Medicine, 8
D. Erriquez, G. Perini, A. Ferlini (2013)
Non-Coding RNAs in Muscle Dystrophies
International Journal of Molecular Sciences, 14
(2014)
MicroRNA deregulation in trinucleotide repeat expansion disorders
Applied RNAi: From Fundamental Research to Therapeutic Applications

Publisher: Oxford University Press
Copyright: The Author(s) 2015. Published by Oxford University Press on behalf of Nucleic Acids Research.
ISSN: 0305-1048
eISSN: 1362-4962
DOI: 10.1093/nar/gkv849
pmid: 26304544
Publisher site: See Article on Publisher Site

Abstract

9500–9518 Nucleic Acids Research, 2015, Vol. 43, No. 19 Published online 24 August 2015 doi: 10.1093/nar/gkv849 Cooperation meets competition in microRNA-mediated DMPK transcript regulation Edyta Koscianska , Tomasz M. Witkos, Emilia Kozlowska, Marzena Wojciechowska and Wlodzimierz J. Krzyzosiak Department of Molecular Biomedicine, Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14, 61-704 Poznan, Poland Received February 06, 2015; Revised July 30, 2015; Accepted August 10, 2015 ABSTRACT nucleotide-long non-coding RNAs are potent regulators of gene expression. They act post-transcriptionally and The fundamental role of microRNAs (miRNAs) in exert their regulatory effects mainly by binding to the the regulation of gene expression has been well- 3 -untranslated region (3 UTR) of target mRNAs, which established, but many miRNA-driven regulatory results in mRNA deadenylation and decay, translational mechanisms remain elusive. In the present study, we suppression or, rarely, mRNA cleavage (1,2). By target- demonstrate that miRNAs regulate the expression ing multiple transcripts and affecting the expression of of DMPK, the gene mutated in myotonic dystrophy numerous proteins, miRNAs engage in various biologi- cal pathways in cells (i.e., proliferation, differentiation, type 1 (DM1), and we provide insight regarding the development, apoptosis, metabolism and neurodegener- concerted effect of the miRNAs on the DMPK target. ation). The interaction between miRNAs and mRNAs Speciﬁcally, we examined the binding of several miR- is influenced by many factors; however, nucleotides 2–8 NAs to the DMPK 3 UTR using luciferase assays. We of the miRNA, termed the ‘seed’ sequence, are essential validated the interactions between the DMPK tran- for target recognition and binding (1). The number and script and the conserved miR-206 and miR-148a. We distribution of miRNA binding sites as well as plausible suggest a possible cooperativity between these two miRNA cooperation are particularly important. More miRNAs and discuss gene targeting by miRNA pairs than a decade ago, the insertion of multiple binding sites that vary in distance between their binding sites and in reporter constructs used to validate miRNA–mRNA expression proﬁles. In the same luciferase reporter interactions has been suggested to ensure a higher effi- ciency of such constructs (3,4). It was later demonstrated system, we showed miR-15b/16 binding to the non- that two sites in the same or different miRNAs could act conserved CUG repeat tract present in the DMPK synergistically and that the distance between neighboring transcript and that the CUG-repeat-binding miRNAs miRNA binding sites affects the strength of the target might also act cooperatively. Moreover, we detected down-regulation. Specifically, an optimal down-regulation miR-16 in cytoplasmic foci formed by exogenously was observed when the distance between the 3 end of the expressed RNAs with expanded CUG repeats. There- first miRNA site and the 5 end of the subsequent one fore, we propose that the expanded CUGs may serve was > 7and < 40 nt (5), and when the 5 ends of both as a target for concerted regulation by miRNAs and miRNA seeds were separated by between 13 and 35 nt may also act as molecular sponges for natural miR- (6). More recently, the concept of miRNA synergy was NAs with CAG repeats in their seed regions, thereby reexamined and addressed in a more detailed way, showing affecting their physiological functions. that this mechanism of miRNA-mediated regulation can affect thousands of human genes. In a transcriptome-wide approach, it was demonstrated that miRNA sites spaced INTRODUCTION by a maximum of 26 nt may act cooperatively and that The involvement of microRNAs (miRNAs) in the the human transcriptome is enriched for miRNA-binding pathogenic mechanisms of many human diseases has sites located at a cooperativity-permitting distance (7). become increasingly apparent. These endogenous ∼22- To whom correspondence should be addressed. Tel: +48 61 8528503; Fax: +48 61 8520532; Email: edytak@ibch.poznan.pl Correspondence may also be addressed to Wlodzimierz J. Krzyzosiak. Tel: +48 61 8528503; Fax: +48 61 8520532; Email: wlodkrzy@ibch.poznan.pl Present addresses: Tomasz M. Witkos, Faculty of Life Sciences, University of Manchester, Manchester M13 9PL, UK. Marzena Wojciechowska, Department of Molecular and Systems Biology, European Centre for Bioinformatics and Genomics, Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14, 61-704, Poznan, Poland. C The Author(s) 2015. Published by Oxford University Press on behalf of Nucleic Acids Research. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Nucleic Acids Research, 2015, Vol. 43, No. 19 9501 Moreover, a workflow for the identification and analysis of Taken together, the abovementioned reports indicate the RNA triplexes composed of two cooperating miRNAs and pathological potential of miRNA dysregulation in DM1. a target mRNA has been proposed, and all of the predicted However, regarding the possible treatment of DM1, of par- human target genes of synergistic miRNA regulation have ticular importance is a report showing that some miRNAs been collected in the TriplexRNA database (8). The syn- are predicted to preferentially bind and repress toxic tran- ergistic activity of co-expressed miRNAs as well as those scripts with longer CUG repeats (25). that exhibit differential expression across tissues could be In this study, we focus on the miRNA-mediated regu- employed in therapeutic interventions using miRNA mim- lation of the DMPK transcript, which provides a unique ics and/or miRNA inhibitors. Interestingly, the natural model for the investigation of miRNA binding in the con- antagonizing activity of two miRNAs with overlapping text of potential miRNA cooperativity. Using a luciferase binding sites has been observed. More specifically, miR-184 reporter system, we validated the regulation of the DMPK was found to interfere with miR-205 in the suppression transcript by conserved miRNAs, miRs 206 and 148a, as of the SHIP2 gene. The binding of miR-184 to its seed well as miR-15b/16 binding to a non-conserved CUG tract. sequence prevented the inhibitory effect of miR-205 on We demonstrated a possible cooperativity between the miR- SHIP2 mRNA (9). 206/148a pair and the potential for cooperative targeting of Myotonic dystrophy type 1 (DM1) is an incurable the CUG tract by CUG-repeat-binding miRNAs. In addi- neuromuscular disorder that is caused by an expanded tion, we demonstrated the enrichment of miR16 in RNA CTG*CAG repeat in the 3 UTR of the dystrophia foci composed of exogenously expressed CUG-repeat tran- myotonica-protein kinase (DMPK)gene(10). The normal scripts by RNA fluorescence in situ hybridization (RNA human DMPK gene harbors 5–37 copies of the trinu- FISH), supporting the possibility of miRNA sequestration cleotide motif, but a dynamic mutation may increase this by the CUG repeats present in the DMPK 3 UTR. number to over 5000 repeat copies. The expanded CUGs in DM1 result in the nuclear retention of mutant DMPK MATERIALS AND METHODS mRNA and reduced DMPK protein levels (11). Mutant Computational prediction of the miRNAs binding to the transcripts sequester the muscleblind-like 1 (MBNL1) splic- DMPK 3 UTR ing factor, leading to the abnormal alternative splicing of a multitude of other transcripts and the expression of fe- The pipeline for the computational prediction of miRNA tal forms of their protein products in DM1 adults (12,13). binding to the DMPK transcript is presented in Figure 1. Spliceopathy is therefore thought to be the major factor The following steps were taken: (i) finding the conserved underlying the pathogenesis of DM1. However, alternative sites of the miRNA families conserved among vertebrates mechanisms such as additional changes in gene expression, predicted by any of three of the most commonly used antisense transcripts, translation efficiency, misregulated al- miRNA prediction programs (TargetScan Release 6.2, DI- ternative polyadenylation and miRNA deregulation may ANAmicroT v. 5.0 and miRanda August 2010 Release) (26– contribute to the pathogenesis of DM1 (14,15). 28); (ii) adding the poorly conserved miRNA–mRNA sites A few reports detailing a close connection between miR- that were predicted by all of these programs; (iii) predicting NAs and DM1 have been published (reviewed in (16)). The additional miRNAs that would bind to the CUG repeats deregulation of specific miRNAs has been linked with mus- using an in-house method. cular dystrophies and cardiomyopathies (17–19)and with myotonic dystrophy type 2 (DM2) (20). In DM1, alterations Identification of miRNAs that potentially bind to CUG re- in the miRNA expression patterns have been observed in peats muscle-specific miRNAs (myomiRs). More specifically, in DM1 skeletal muscle, miR-1 and miR-335 are up-regulated Mature human and murine miRNA sequences were re- whereas miRs 29b, 29c and 33 are down-regulated, com- trieved from miRBase (ver. 20) (29). In-house scripts written pared with the control muscles (21). In addition, miR-1 in the Python programming language were used to detect is down-regulated in cardiac muscle (22), whereas miR- miRNAs with at least six complementary matches to CUG 206 is up-regulated in the skeletal muscle of DM1 patients repeats in the same reading frame within the miRNA seed (23). The deregulation of DM1-associated miRNAs has regions (positions 2–7 of the miRNA sequences). also been linked to alterations in their putative target ex- pression, indicating that miRNA misregulation in DM1 Examination of the frequency distribution of binding sites for is functionally relevant and may contribute to the disease miRNA pairs pathology (21,22). Importantly, the decreased expression of mature miR-1 and increased levels of its targets in the The sequences of human miRNA families with shared hearts of individuals with myotonic dystrophy are medi- seed regions were obtained from TargetScan, Release ated by the functional depletion of MBNL1, a sequestered 6.2 (30). The mature miRNA expression profiles (qPCR splicing factor, which affects the processing of pre-miR-1 miRNA profiling) were obtained from miRNAMap 2.0 (22). Recently, a study investigating a transgenic yfl model (31). Broadly expressed miRNAs were defined as miRNAs of DM1 (i(CTG)480 Drosophila line carrying 480 CTG re- that were expressed in at least 10 of the 18 examined nor- peats) revealed that miRNA alterations were caused directly mal adult tissues at considerable levels (at least 2% of the by CTG expansions (24). Specifically, the expression of 20 total miRNA expression). miRNAs with at least 75% of the miRNAs was changed in DM1 flies compared with con- reads found in the tissue with the highest expression levels trol flies; 19 were down-regulated and one was up-regulated. were classified as tissue-specific miRNAs. Next, miRNAs 9502 Nucleic Acids Research, 2015, Vol. 43, No. 19 were grouped into miRNA families (miRNAs with com- For miRNA overexpression, commercial plasmid con- mon seed regions), and only the miRNA families with all structs expressing miRNA precursors (pri-miR-1, pri-miR- of the members nominated as broadly expressed or tissue- 206, pri-miR-214 (System Biosciences), pri-miR-148a and specific were considered to calculate the distances between pri-miR-29a (Cell Biolabs)) were used. Moreover, the con- putative miRNA binding sites (all of the distances were struct expressing the miR-15b/16 precursor was generated counted between the 5 ends of the binding sites). Lists of based on an empty plasmid (System Biosciences) by cloning miRNA families with broadly expressed and tissue-specific the appropriate sequence at the EcoRI and NotI sites. The miRNAs can be found in the supplementary data (Supple- mir-15b/16–2 cluster that was flanked by a ∼250-nt se- mentary Table S1). The distances between all of the distinc- quence was amplified from the genomic DNA using PCR tive pairs of miRNAs were used as reference controls. A and the following primers: forward 5 -GACGCGGAATTC Perl script provided by TargetScan was used to predict the CGAAGCCATGGAATTGACTT and reverse 5 -GAAG miRNA target sites for the longest variants of the human CGGCCGCAAGAACAAAAACAAAGGAAAAGGA. 3 UTR transcripts, and in-house scripts written in Python were used to process the data. Cell transfection HEK 293T and HeLa cells were transfected using Lipofec- Cell culture tamine 2000 (Invitrogen) according to the manufacturer’s HEK 293T and HeLa cells were obtained from the protocols. For the luciferase assays, the cells were trans- American Type Culture Collection (ATCC) and grown in fected in 12-well plates at ∼80% confluence. For each trans- Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma- fection experiment, 200 ng of the appropriate reporter con- Aldrich) supplemented with 8% fetal bovine serum (FBS) struct and 400 ng of the appropriate miRNA-coding vec- (Sigma-Aldrich), 2 mM L-glutamine and an antibiotic– tor were used. For miRNA cooperativity studies, 200 ng of antimycotic solution (Sigma-Aldrich) at 37 C in a humidi- the appropriate reporters and a total of 400 ng of miRNA- fied atmosphere containing 5% CO .At24hbefore trans- coding plasmids (200 ng of either vector) were used. The fection, the cells were plated in 12-well or 6-well dishes in cells were harvested 24 h after transfection and assayed for DMEM medium and harvested 24, 48, and 72 or 96 h post- luciferase activity. transfection for the luciferase assay, real-time polymerase For the miRNA overexpression required for real-time chain reaction (PCR), and western blot analyses, respec- PCR and western blot analyses, the HEK 293T cells were tively. grown to 80% and 60% confluence, respectively, transfected in 12- and 6-well plates with 1 g/ml pri-miRNA plasmid vectors, and harvested at 48 and 72/96 h, respectively. Plasmid constructs and synthetic miRNA oligonucleotides For the fluorescence in situ hybridization (FISH) experi- To generate reporter constructs bearing miRNA-binding ments, the HEK 293T cells were grown to 80% confluence sites, the pmirGLO Dual-Luciferase miRNA Target Ex- on microscope slides in 12-well plates, transfected with 0.5 pression Vector (Promega) was used. Specific oligonu- g/ml plasmid vectors expressing DMPK exons 11–15 con- cleotides of different lengths with DraIand XbaI restric- taining either 960 CUG or CAG repeats in exon 15 or with tion sites and containing single binding sites (b.s.) for the a control plasmid lacking the repeats (36). For miR-214 de- analyzed miRNAs (DMPK b.s. for miRs 1, 148a, 206 and tection, the cells were additionally co-transfected with 30 combined 148a/206) as well as the oligonucleotides com- nM of miR-214 mimic (synthetic oligonucleotides were pur- posed of 5 CUG, 20 CUG and 20 CAA repeats were syn- chased from IDT and prepared as described in (33)). thesized (IBB Warsaw). The appropriate oligos were an- nealed and cloned into the pmirGLO vector, which was Luciferase reporter assay previously digested with DraI(Fermentas) and XbaI(Fer- mentas) restriction enzymes downstream of the luc2 gene. After harvesting, the cells were lysed in a passive lysis buffer For all of the miRNAs, three types of constructs were pre- (Promega). The enzymatic activities of fireyfl and Renilla lu- pared, namely wild-type (WT) constructs, constructs carry- ciferases were measured using a Centro LB 960 luminome- ing mutations (MUT) predicted to disrupt binding and per- ter (Berthold Technologies) and the substrates and pro- fect match (PM) constructs, as previously described (32,33). cedures provided with a Dual-Luciferase Reporter Assay The reporter constructs carrying 44, 53 and 72 CUG re- System (Promega). The values for firefly luciferase activity peats were generated based on the reporter containing 20 for every reporter construct were normalized to the corre- CUGs via the SLIP (synthesis of long iterative polynu- sponding values of Renilla luciferase activity to account for cleotide) method (34,35). Briefly, the pmirGLO plasmid varying transfection efficiency. The relative expression val- with the cloned sequence corresponding to 20 CUG repeats ues for all of the constructs were obtained by comparing was independently digested with DraIand SalI restriction their normalized luciferase activities with those of the con- enzymes and subjected to one thermal cycle of 95 C for 5 trol plasmid. ◦ ◦ min, 50 C for 10 min and 72 C for 3 min with 2.5 units of Pfu polymerase, reaction buffer, and 250 M each dNTP. RNA isolation and real-time PCR The product was directly transformed into Escherichia coli and selected on an LB plate containing ampicillin. The se- The total RNA from HEK 293T cells was isolated using quences of all of the constructs are presented in the supple- TRI Reagent (MRC, Inc., BioShop) according to the man- mentary data (Supplementary Table S2). ufacturer’s instructions. cDNA was obtained from 500 ng Nucleic Acids Research, 2015, Vol. 43, No. 19 9503 of total RNA using Superscript III (Life Technologies) and yeast tRNA, 10% dextran sulfate, 2 mM vanadyl ribonucle- random hexamer primers (Promega). For subsequent quan- oside complex and 2 ng/l appropriate fluorescently labeled titative real-time analyses, 50 ng of cDNA was used. Real- RNA or DNA/LNA probes. Specifically, the miR-16–5p time PCR was performed with a LightCycler 480 II system probe labeled at the 5 -end with TYE665 (Cy5) and mod- (Roche) using TaqMan Gene Expression Assays and Taq- ified at positions 5, 10, 14 and 19 with locked nucleic acids Man Universal Master Mix II (Applied Biosystems). The (LNA) (Exiqon) was used in combination with the (CAG) results obtained for the assessment of DMPK mRNA lev- probe labeled at the 5 -end with FAM and modified with els were normalized to the levels of actin mRNA. 2 -O-methyl at positions 1 and 2 (IDT). In parallel experi- ments, the miR-214 probe labeled at the 5 -end with FAM and modified with 2 -O-methyls (IDT) was used together Northern blot analysis with the (CAG) probe labeled at the 5 -end with TYE665 High-resolution northern blot analysis was performed as (Cy5) and modified at positions 6, 12 and 17 with LNAs previously described (37,38). Briefly, 25 g of total RNA (Exiqon). The (CTG) probe, which was used additionally was extracted from HEK 293T cells and resolved in a 12% as a control to detect CAG RNAs, was labeled at the 5 - denaturing polyacrylamide gel in 0.5× TBE. The RNA end with TYE563 (Cy3) and modified as described for the (CAG) probe. Post-hybridization washing was performed was transferred to a GeneScreen Plus hybridization mem- brane (PerkinElmer) using semi-dry electroblotting (Sigma- in 30% formamide and 2× SSC at 45 C for 30 min followed Aldrich), immobilized by subsequent UV irradiation (120 by 1× SSC at 37 C for 30 min. The slides were mounted in 2 ◦ mJ/cm ) (UVP) and baking at 80 C for 30 min. The mem- SlowFade Gold Antifade reagent with DAPI (Invitrogen) branes were probed with specific DNA oligonucleotides for further microscopy. (Supplementary Table S3) that were complementary to the To determine the spatial interactions between the ex- annotated human miRNAs (miRBase). The probes were la- panded repeat RNA foci and specific miRNAs, the same beled with [ P] ATP (5000 Ci/mmol; Hartmann Analyt- exposure was established for all of the images from a single ics) using OptiKinase (USB). The hybridizations were per- experiment, and a z-stack that was sufficient to cover all of formed at 37 C overnight in a PerfectHyb buffer (Sigma- the foci was acquired for each cell with a slice thickness of Aldrich). The marker lanes contained a mixture of radiola- 1 musingaPL-Apo63× oil objective in conjunction with beled RNA oligonucleotides (17-, 19-, 21-, 23- and 25-nt in a cooled AxioCam HRc camera; the images in one stack length). Hybridizations to U6 RNA provided loading con- were overlaid and saved as TIFF files. Approximately 100 trols. Radioactive signals were detected by phosphorimag- cells were randomly selected to characterize the mutant re- ing (Multi Gauge v3.0; Fujifilm). peat RNA and miRNA interactions in transfected HEK293 cells. The FISH images were processed with LSM 510 soft- ware (Zeiss). Western blot analysis After transfection with the appropriate miRNA-coding Statistical analysis plasmids, HEK 293T cells were lysed with 1× PBS sup- The experiments were repeated at least three times. The plemented with protease inhibitor cocktail (Roche). A to- graphs were generated using GraphPad Prism 5 (GraphPad tal of 25–30 g of protein lysate was separated by 12% Software). The figures for the luciferase assays were gen- SDS-PAGE. After electrophoresis, the proteins were elec- erated after averaging the results from the repeated exper- trotransferred onto a nitrocellulose membrane (Sigma). iments for a particular construct. The values for the error All of the immunodetection steps were performed on a bars (means with SEM) and the statistical significance were SNAP id (Millipore) in PBS buffer containing 0.25% non- calculated using GraphPad Prism 5. The statistical signifi- fat milk and 0.1% Tween 20, and the membranes were cance of the luciferase reduction in the case of transfection washed in PBS/Tween. For DMPK and GAPDH de- with constructs carrying miRNA b.s. was assessed using a tection, the blots were probed with the primary mouse one-sample t-test after checking that the data followed a anti-DMPK (1:500, Millipore) and mouse anti-GAPDH normal distribution, with a hypothetical value of 100% as- (1:5000, Millipore) antibodies, respectively, and they were signed to the cells that were transfected with an empty vec- subsequently probed with biotinylated secondary antibod- tor control. P-values < 0.05 (two-tailed) were considered ies (1:500 Sigma). The membranes were incubated with a significant. streptavidin–AP conjugate (1:2000, Millipore), and the im- munoreactive bands were visualized using the Sigma Fast RESULTS BCIP/NBT kit (Sigma). In our previous study, we performed an in-depth computa- tional analysis of the miRNA interactions using all of the RNA FISH and microscopy mRNAs derived from genes that trigger hereditary neuro- RNA FISH was performed in transiently transected HEK logical disorders, known as trinucleotide repeat expansion 293T cultured cells as previously described (39). Briefly, 48 diseases (TREDs), and we showed that DMPK, the gene h post-transfection, the cells were fixed in 4% PFA /PBS at that is mutated in DM1, may be subject to miRNA regu- 4 C and washed three times in PBS for 5 min each. Pre- lation (40). In the present study, following our guidelines hybridization was performed in 30% formamide and 2× for the efficient prediction of miRNA–mRNA binding sites SSC buffer for 10 min followed by hybridization in a buffer (miRNA b.s.), we selected both highly and poorly conserved containing 30% formamide, 2× SSC, 0.02% BSA, 66 g/ml sites for experimental validation (Figure 1). 9504 Nucleic Acids Research, 2015, Vol. 43, No. 19 Figure 1. Analyzed miRNAs targeting the 3 UTR of the DMPK transcript. A schematic presentation of the selected miRNA target site distribution in the DMPK 3 UTR is shown. Additionally, a list of the miRNAs with CAG motifs in their seed regions that are complementary to the CUG repeats in the DMPK 3 UTR and a pipeline for the computational prediction of the miRNAs that bind to the DMPK 3 UTR are presented. Regulation of the DMPK gene by the conserved miR-206 and Supplementary Table S1). The binding parameters for these miR-148a candidate miRNAs meet the recommended bioinformatics criteria, and their experimental validation was of particular The top candidate miRNAs (miR-148a/152 and miR- interest in the context of the pathogenesis and therapy of 1/206) that are predicted regulators of the DMPK tran- DM1 (22). script (40) were ranked highly by the most commonly used Experiments employing a set of reporter constructs and algorithms to rank miRNA b.s., based on sequence conser- luciferase assays were performed to experimentally verify vation criteria. Both of the putative sites within the DMPK the predicted binding of miR-206 and miR-148a to their 3 UTR were the only two highly conserved sites for miRNA target sites within the 3 UTR of the DMPK, as previ- families that are broadly conserved among vertebrates. The ously described (32,33). Briefly, the following constructs miRs 148a/152 and 1/206 exhibit disparate expression pat- were tested in parallel: wild-type reporters (WT) bearing a terns in human tissues. Both miR-1 and miR-206 are my- single native b.s. for either miRNA, constructs with muta- omiRs, which are abundantly expressed in smooth, skeletal, tions (MUT) that disrupt the 5 seed site (negative controls) and/or cardiac muscles, whereas miR-148a and miR-152 and constructs showing perfect complementarity (PM) to are broadly expressed across various tissue types (31)(see miRNA b.s. (positive controls). Having performed a north- Nucleic Acids Research, 2015, Vol. 43, No. 19 9505 ern blot analysis, which revealed that miR-206 was not ex- script, but differ in four nucleotides at their 3 ends (Fig- pressed in HEK 293T cells and miR-148a was expressed at ure 3A). Both miR-1 and miR-206 are widely studied and alow level(Figure 2A), we validated the predicted miRNA– well-defined myomiRs; they enhance skeletal muscle dif- mRNA interactions using a miRNA overexpression system. ferentiation, regulate numerous targets and are involved Specifically, HEK 293T cells were co-transfected with a re- in a wide array of specific muscular pathways (reviewed porter construct (carrying a potential b.s. for the studied in (44,45)). Moreover, these two miRNAs exhibit differ- miRNA) and the appropriate miRNA-coding plasmid (Sys- ent expression patterns, depending on the muscle type, tem Biosciences, Cell Biolabs). The transient transfection of and differentially regulate transcripts that bear their b.s. the cells was followed by measuring the reporter activity. (46). Although canonical miRNA-target specificity is trig- We obtained considerable repression of luciferase ex- gered primarily by complementarity within the seed re- pression following the transfection of both of the WT re- gion, non-canonical interactions also depend on 3 compen- porters, namely the WT construct carrying the b.s. for miR- satory sites (1,47). To determine whether miR-1 and miR- 206 (WT 206) and the WT construct carrying the b.s. for 206 differ in their regulatory potential, we performed a rel- miR-148a (WT 148a) (Figure 2B). This result indicates that evant luciferase experiment. We co-transfected HEK 293T both miR-206 and miR-148a are functional and may down- cells with the reporter construct carrying the potential b.s. regulate the DMPK transcript. The effect exerted by these for miR-1 (WT 1) and a miR-1-encoding plasmid (System two miRNAs was comparable, with miR-148a being slightly Biosciences) together with the adequate controls. In con- more effective. Specifically, the reduction of luciferase ac- trast to the repressed luciferase expression that followed the tivity was reproducible and statistically significant for both transfection of the WT 206 construct, after transfection of WT constructs, with suppression to 81% and 71% for the WT 1, we observed only a slight decrease in luciferase ac- WT construct carrying the b.s. for miR-206 and miR-148a, tivity (suppression to 92%) (Figure 3B). There was no vis- respectively. Both of the PM constructs repressed the lu- ible decrease in the protein level of DMPK and no statis- ciferase activity to very low levels (13–16%), whereas the tically significant reduction of the mRNA level following luciferase activity for both of the MUT constructs exhib- miR-1 overexpression (Figure 3C, D, E). These results sug- ited efficient de-repression ( ≥90%). To validate our experi- gest that miR-1 is a weaker regulator of DMPK expression mental approach, we performed additional luciferase tests than miR-206; however, it cannot be ruled out that, under in HeLa cells and also HEK 293T cells but without the ad- certain conditions, this miRNA may affect the expression dition of the appropriate miRNA-coding vectors (Supple- of DMPK to variable degrees. mentary Text 1, Supplementary Figure S1A, S1B). The ob- tained results verified the reliability of the experimental sys- Potential cooperativity between miR-206 and miR-148a tem used. Thus, the presented study provides the first evi- dence of the direct binding of miRs 206 and 148a with the Having validated the individual interactions of miR-206 DMPK 3 UTR and positively validates these miRNAs as and miR-148a with the DMPK 3 UTR, we proceeded negative regulators of the DMPK gene. with the same type of experiments to determine the spac- Next, we evaluated the expression of the DMPK protein ing requirements for cooperative miRNA target site inter- and mRNA following the transfection of HEK 293T cells actions. The distance between the miR-206 and 148a sites with plasmids encoding either miR-206 or miR-148a. Real- in the intact DMPK 3 UTR measured between the 5 ends time PCR performed after transfection of the studied miR- of both seeds is 15 nt, which is considered to be the op- NAs did not reveal a considerable decrease in the DMPK timal spacing for efficient target repression ( 5–7). Addi- mRNA (Figure 2C). Although miRNA binding frequently tionally, this miRNA pair was predicted in silico to have leads to the reduction of the cellular levels of targeted tran- the potential to act cooperatively, as measured by the en- scripts (41,42), our observation is consistent with the find- ergy gain achieved through RNA triplex formation in com- ings that no or minimal changes in the respective mRNA parison to two separate miRNA–mRNA duplexes using levels were observed or that these changes were only re- the TriplexRNA database (8). To study the possible co- ported for certain targets (43). By contrast, western blot regulation of DMPK expression by the miR-206/148a pair, analyses performed with either miR-206 or miR-148a re- new sets of reporter constructs were prepared that differed vealed a considerable decrease in the DMPK protein level, in the distance between the b.s. of miR-206 and miR-148a. which was slightly more evident following the overexpres- Three reporters were constructed: (i) a wild-type reporter sion of miR-206 (Figure 2D). In muscle cells, which are the bearing the native sequence of the DMPK 3 UTR encom- primary cell type affected in DM1, miR-206 is highly ex- passing the b.s. of both miRNAs (WT 148/206); (ii) a re- pressed. Therefore, to confirm the DMPK down-regulation porter containing the sequence between miRNA seeds that observed in HEK 293T cells after miR-206 overexpression, were artificially extended to 31 nt (Ext 148/206), a nt length we also addressed the regulation of DMPK expression in that is considered to promote miRNA cooperativity (6)and human muscle cells including DM1 myoblasts (Supplemen- (iii) a reporter in which the sequence between miRNA seeds tary Text 2 and Figure S2). The observed increase of miR- was extended to 85 nt (Long 148/206), a nt length that 206 expression, together with the decrease of DMPK pro- is considered to possess very limited miRNA cooperativ- tein levels, is in agreement with the hypothesis that miR-206 ity (Figure 4A).Inthe case of theExt 148/206 construct, act as a negative regulator of DMPK expression. additional nucleotides were introduced to maintain a na- Finally, we aimed to analyze the binding specificity of tive sequence context, which is required for a proper inter- miR-1 and miR-206, which share the seed sequence and action with both miR-206 and miR-148a. In addition, an are predicted to bind to the same site in the DMPK tran- EcoRI restriction site was introduced, permitting the sub- 9506 Nucleic Acids Research, 2015, Vol. 43, No. 19 Figure 2. Regulation of the DMPK gene by miR-206 and miR-148a. (A) Northern blot detection of miRs 206 and 148a in untreated HEK 293T cells and cells that were transfected with miRNA-coding plasmids (System Biosciences, Cell Biolabs). M denotes the size marker: end-labeled 17, 19, 21, 23 and 25-nt oligoribonucleotides. En and Ex indicate the endogenous and vector-expressed miRNA levels, respectively. Hybridization to U6 RNA served as a loading control. (B) Relative repression of luciferase expression. Reporter constructs carrying a single b.s. for miR-206 and miR-148a were tested. For each luciferase experiment, the miRNA activity on four constructs was measured in parallel: an empty pmirGLO vector (Control), a wild-type potential b.s. for the appropriate miRNA (WT), a mutated b.s. (MUT), and a site with full complementarity (PM). The fireyfl luciferase activity was normalized against that of Renilla luciferase. The standard errors were calculated from nine independent experiments. The asterisks indicate statistical significance at P < 0.001. (C) Relative DMPK mRNA levels. Real-time PCR was performed after the HEK 293T cells were transfected with miR-206 and miR-148a. The bar graphs show the quantification of the DMPK mRNA levels normalized to the actin mRNA level, based on the data from vfi e independent experiments. (D) Western blot analysis of the DMPK protein levels 72 and 96 h after the HEK 293T cells were transfected with miR-206 and miR-148a, as indicated. The GAPDH protein served as a loading control. sequent extension of the spacing between these miRNAs significantly suppressed luciferase expression, demonstrat- b.s. In the case of the longest reporter construct, namely ing their ability to bind miRNAs, but their efficiency varied Long 148/206, an appropriate DNA fragment comprising (Figure 4C). Specifically, after the overexpression of both an irrelevant sequence that further separated the sites for miRNAs, the WT construct carrying the native sequences of miR-206 and miR-148a was cloned into the EcoRI site. the miR-148a and miR-206 sites (WT 148/206) reduced the The secondary structures that can be formed by transcripts luciferase activity to 66% compared with the control, which from every reporter construct, as predicted by the RNA is a better score compared with that observed for the down- metaserver available on the GeneSilico website, showed regulation resulting from the activity of either individual a high degree of secondary structure preservation in the miRNA (compare with Figure 2B and Supplementary Fig- miRNA b.s. (Figure 4B). ure S1A). However, in the case of the Ext 148/206 construct Luciferase assays were performed as described to vali- with the miRNA site separation extended to 31 nt, the re- date the individual miRNA–mRNA interactions. The re- duction of luciferase activity was much stronger (suppres- porters bearing two miRNA sites with increasing separa- sion to 42%). By contrast, in the case of the Long 148/206 tions, i.e., WT, Ext and Long, were tested in parallel follow- construct with miRNA site separation extended to 85 nt, the ing the transfection of HEK 293T cells and the simultane- reduction of luciferase activity was again at a level similar ous overexpression of both miRNAs (an equal amount of to that produced by the WT 148/206 construct (suppres- the plasmids encoding miRs 148a and 206 was transfected sion to 65%). The obtained results indicate that miRNAs into the cells). Importantly, all of the analyzed constructs 148a and 206 may act cooperatively and that their optimal Nucleic Acids Research, 2015, Vol. 43, No. 19 9507 Figure 3. Regulation of the DMPK gene by miR-1. (A) Schematic presentation of the base pairing of miR-1 and miR-206 with their target DMPK sequences. The nucleotides in the seed regions of these miRNAs are marked in green, whereas disparate nucleotides are shown in orange. (B)Relative repression of luciferase expression. Reporter constructs carrying a single b.s. for miR-1 were tested; the miRNA activity on four constructs was measured in parallel (Control, WT, MUT and PM) as described in Figure 2. The standard errors were calculated from eight independent experiments. The asterisk indicates statistical significance ( P < 0.05). (C) Western blot analysis of the DMPK protein levels after the HEK 293T cells were transfected with miR- 1. The GAPDH protein served as a loading control. (D) Relative DMPK mRNA levels. Real-time PCR was performed after the HEK 293T cells were transfected with miR-1. The bar graph shows the quantification of the DMPK mRNA level normalized to the actin mRNA level, based on the data from four experiment repeats. (E) Northern blot detection of miR-1 in the untreated HEK 293T cells and cells that were transfected with the miR-1-coding plasmid (System Biosciences). M denotes the size marker: end-labeled 17, 19, 21, 23 and 25-nt oligoribonucleotides. En and Ex indicate the endogenous and vector-expressed miRNA levels, respectively. Hybridization to U6 RNA served as a loading control. distance between b.s. is in agreement with that previously to structural consequences. Specifically, nucleotides in the studied in other human miRNAs (5–8). region of the miRNA b.s. were predicted to be more tightly The cooperativity between miR-206 and miR-148a in the paired, and a stem rather than a loop was formed at the regulation of DMPK could be further confirmed by the ob- miR-206 b.s., resulting in the lower efficiency of this con- servation that the expression of both the DMPK protein struct compared with Ext 148/206. After the transfection and mRNA was affected following the simultaneous over- of impExt 148/206, the luciferase expression decreased to expression of the cooperating miRNAs in HEK 293T cells 85% (Supplementary Figure S4), which was rather similar (Figure 4D, E). Interestingly, the real-time PCR results re- to the reduction achieved by the overexpression of only one vealed a minor decrease in the DMPK mRNA level after miRNA, suggesting a loss of miRNA cooperativity. In the transfection of the miR-148a/206 pair, which was statisti- same experimental setup, the Ext 148/206 reporter could cally significant and greater than the effect exerted by a sin- suppress luciferase up to 42% (Figure 4C). gle miR-148a or miR-206 (Figure 2C). It is worth noting that the effect of miRNA cooperativity was achieved when Global analysis of the miRNA–mRNA binding sites a half concentration of each miRNA was used (see Mate- Although a few significant advances in the field of miRNA rials and Methods). The DMPK protein was considerably cooperativity have been made (5–8), neither its mecha- repressed when miR-206 and miR-148a were mildly but si- nism nor its biological relevance is yet fully understood. multaneously overexpressed in HEK 293T cells, whereas the Therefore, after discovering the co-regulation of the DMPK protein level was not decreased when a lower concentration transcript via the ubiquitously expressed miR-148a and of either of the single miRNAs was used (Supplementary tissue-specific miR-206, we investigated whether this type Figure S3). of miRNA cooperativity in the regulation of gene expres- The RNA structure itself does not seem to be the main sion could be a common phenomenon that is dependent factor influencing the potential miRNA binding; however, on the miRNA expression profiles, i.e., whether broadly ex- spatial separation and a rather loose structure are prefer- pressed miRNAs and tissue-specific miRNAs could work in able (48–50). To assess the impact of the secondary struc- tandem. ture of the target site on the cooperative binding of the Using the qPCR expression profiles of human miRNAs miRNAs, an impaired reporter construct was also tested, in different tissues (miRNAMap 2.0, (31)), we selected namely the impaired Ext 148/206 (hereafter referred to as miRNA families in which every member of the family was impExt 148/206). The impairment of this reporter con- ubiquitously expressed (94 miRNA families) as well as a cerned its sequence; however, this change in sequence led set of tissue-specific miRNA families (25 miRNA fami- 9508 Nucleic Acids Research, 2015, Vol. 43, No. 19 Figure 4. Potential cooperativity between miR-206 and miR-148a. (A) A schematic presentation of the reporter constructs that were used to study the role of the distance between the b.s. of miR-206 and miR-148a. Three types of reporters are shown with the 5 ends of the miRNA b.s. separated by 15, 31 and 85 nt. The miR-148a and miR-206 seed regions are marked in light red and light green, respectively. (B) Graphical representation of the secondary structures formed by the reporter constructs of different lengths as predicted by RNAmetaserver. WT 148/206, Ext 148/206 and Long 148/206 correspond to constructs with miRNA b.s. spaced by 15, 31 and 85 nt, respectively. (C) Relative repression of luciferase expression by constructs that differ in the distance between the b.s. of miRs 148a and 206. The miRNA activities on the four constructs were assessed in parallel (Control, WT 148/206, Ext 148/206 and Long 148/206) after the simultaneous overexpression of both miRNAs. The standard errors were calculated from six independent experiments. The asterisks indicate statistical significance; a double asterisk denotes P < 0.01, and a triple asterisk denotes P < 0.001. (D)RelativeDMPKmRNAlevels. Real-time PCR was performed after the simultaneous overexpression of miR-148a and miR-206 in HEK 293T cells. The bar graph shows the quantification of the DMPK mRNA level normalized to the actin mRNA level, based on the data from vfi e independent experiments. ( E) Western blot analysis of the DMPK protein levels after the simultaneous overexpression of miR-148a and miR-206 in HEK 293T cells. A representative blot is shown. The GAPDH protein served as a loading control. lies) (see Materials and Methods and Supplementary Ta- three types of miRNA pairs examined (Figure 5). For the ble S1). Next, for all of the human transcripts in the Tar- pairs of miRNAs with one broadly expressed miRNA and getScan database Release 6.2 (see Materials and Methods), one tissue-specific miRNA, we observed an enrichment for we calculated the distances between the putative b.s. for overlapping sites (spacing of less than 7-nt, which causes three types of miRNA pairs: (i) pairs with one broadly ex- the sequence of the ‘seed’ region to be shared by two miR- pressed miRNA and one tissue-specific miRNA; (ii) pairs NAs), but no enrichment of pairs with 7–28-nt spacing, the with two different tissue-specific miRNAs and (iii) pairs distance that enabled cooperativity of miR-148a and miR- with two different broadly expressed miRNAs. Our analysis 206 in the DMPK transcript. This phenomenon suggests revealed different preferences for the distances between the that there may be a tendency for tissue-specific miRNAs Nucleic Acids Research, 2015, Vol. 43, No. 19 9509 Figure 5. Distribution of the frequencies of the pairwise distance between the predicted binding sites of miRNA pairs. The graph shows the distribution of the b.s. distance frequencies between pairs with two broadly expressed miRNAs (blue line), two tissue-specific miRNAs (green line) or one broadly expressed miRNA and one tissue-specific miRNA (red line) in relation to the distribution of the miRNA b.s. distances for all of the miRNA pairs. The dotted lines indicate the 7- and 28-nt pairwise distances. In the upper panel, the schematic demonstrates how the different miRNA pairs were classifie d. to regulate gene expression by competing with broadly ex- found 33 miRNAs with at least six complementary matches pressed miRNAs for binding sites rather than via cooper- to CUG repeats in their seed regions (Figure 1 and Sup- ative regulation by binding to the same transcripts in close plementary Table S4). All of these miRNAs were predicted vicinity. This competition for binding sites could be an ad- to have additional pairing sites to a CUG sequence. It is ditional mechanism for providing and maintaining tissue worth noting that the members of one miRNA family (miR- specificity. By contrast, we observed an enrichment of pairs 15ab/-16/-195/-424/-497/-6838) bear a common seed re- of miRNAs with two tissue-specific miRNAs with 7–28-nt gion that is fully complementary to the CUG repeats. The spacing. This result could be explained by the observation observations that miRNAs from the miR-15/107 group that that tissue-specific miRNAs, which are often expressed in share a 5 AGCAGC sequence (52) play key roles in gene similar types of tissues (brain-, muscle-, liver-, testes- and regulation and have overlapping targets (53,54)are also placenta-specific miRNA families), orchestrate the same of relevance to this issue. These miRNAs are expressed at developmental/physiological processes and miRNA coop- medium-to-high levels across many tissues, and members erativity increases the effectiveness of miRNA regulatory with the 7-nt common sequence AGCAGCA exhibit the mechanisms ((7,51) and references therein). highest expression levels (55). In the last set of experiments, we examined whether the miRNAs that were identified in silico as bearing CAG re- Interaction of miR-15b/16 with CUG repeats in the DMPK peats in their seed regions bind to the CUG expanded re- transcript peats in the DMPK transcript. We used the same exper- imental approach with a luciferase reporter system. The In our prediction of miRNA-mediated regulation of the candidate miRNAs were chosen based on the number of DMPK gene, we also sought to explore the possibility of matches with the CUG repeats in their seeds (Supple- miRNA binding to CUG repeats, which are the underlying mentary Table S4). Specifically, we selected miR-15b and cause of DM1 when they are extensively expanded. Previ- miR-16 (seven matches) and miR-214 and miR-29a (six ous studies have postulated the existence of several miR- matches). First, we co-transfected HEK 293T cells with NAs that can potentially bind to the CUG repeats present reporter constructs carrying either 20 CUG repeats (re- in the DMPK 3 UTR (25). If these interactions occur, the ferred to as WT CUG because there are 20 CUGs in CUG repeats could act as multiple overlapping b.s. for sev- the human DMPK gene sequence deposited in GenBank eral miRNAs and therefore have unique miRNA coopera- (NM 001081560.2)) or 20 CAA repeats (referred to as tive potential. We searched in silico for novel human miR- MUT CUG) and an appropriate miRNA-expressing vec- NAs that can interact with repeated CUG sequences. We 9510 Nucleic Acids Research, 2015, Vol. 43, No. 19 tor. In the luciferase assays, we obtained considerable and DMPK gene, rather than the 20 CTGs anticipated based on statistically significant repression of the luciferase expres- the reference sequence deposited in GenBank. Therefore, sion only after transfection of the reporters and miR- this observation is in agreement with the results of the lu- 15b/16 (luciferase repression equal to 62%). In the case of ciferase assay showing that the CUG-repeat-binding miR- the other tested miRNAs, the luciferase activity was not re- NAs cannot affect the control reporter bearing the 5 CUGs pressed (Figure 6A). (see Figure 6B). To assess whether the presence of abnor- Next, we aimed to determine whether the length of the mally expanded CUG tracts may affect DMPK expression, CUG sequence correlated with the level of transcript down- we performed western blot analyses using fibroblast cells de- regulation. Therefore, we prepared and tested reporter con- rived from DM1 patients (Supplementary Text 3 and Fig- structs that differed with respect to the number of CUG re- ure S6). We did not detect a significant change in the protein peats: the reporters carried 5, 20 (WT), 44, 53 and 72 CUG level in untreated cells (Supplementary Figure S6A), but the triplets, the last two mimicking the expansion of CUGs that DMPK protein level was reduced after overexpression of is observed in DM1 patients. Similarly, in the luciferase as- miR-16 in fibroblasts with a CUG expansion (Supplemen- says, we obtained significant repression of the luciferase ex- tary Figure S6B), which may further confirm the possibility pression after the transfection of miR-15b/16 with most that miRNAs with CAG repeats in their seed regions are of the reporters, with the exception of the reporter with 5 able to bind to the CUG repeats in the DMPK transcript. CUGs (Figure 6B). Moreover, a trend toward stronger re- pression was observed when constructs with a longer repeat Sequestration of miRNAs with a CAG sequence in their seeds length were used; the luciferase repression was equal to 64% by the expanded CUG repeats and 44% of the expression levels for the constructs with 20 and 72 CUGs, respectively. By contrast, the luciferase Having examined the binding of miR-15b/16 and miR-214 activity was not noticeably reduced when other miRNAs to the reporter constructs bearing CUG triplet repeats, we were co-transfected along with the tested reporters, with studied the plausible sequestration of these miRNAs to the the exception of miR-214. The latter considerably repressed expanded CUG repeats exogenously expressed in human luciferase activity, but only in the presence of the longest cells. We hypothesized that the presence of elongated CUG reporter (with 72 CUGs) (repression to 73%). This find- repeats would promote the binding of these miRNAs to ing suggests that 7-mer CAG miRNAs may preferentially the mutant DMPK transcript and that the sequestration of bind to and suppress the expression of transcripts contain- the CUG-repeat-binding miRNAs by abnormally length- ing CUG repeats. Overall, these results indicate that cer- ened mutant mRNAs could impair the miRNA-mediated tain miRNAs can target DMPK mRNA with expanded silencing machinery. To determine whether miRs 16 and triplets, and the degree of miRNA-mediated repression de- 214, which were previously tested in a reporter system, may pends on the number of repeats. The interaction between be arrested by the CUG repeats present in the 3 UTR of the DMPK 3 UTR and miR-15b/16 with a CUG binding the truncated DMPK transcripts, we used fluorescence in motif in their seeds containing seven matching nucleotides situ hybridization (FISH) in HEK 293T cells as a model (AGCAGCA), is a notable example. Interestingly, this in- system in this study. We transfected HEK 293T cells with teraction is even stronger in terms of the DMPK repression plasmid vectors expressing DMPK exons 11–15 contain- achieved by appropriate reporter constructs carrying 20 or ing either 960 CUG or 960 CAG repeats (36), or with the more CUG repeats than the interaction between the ana- control plasmid lacking the repeats. The cells were also co- lyzed conserved miRNAs (non-CUG-repeat-binding miR- transfected with the miR-214 mimic; co-transfection with NAs), namely miRs 148a and 206, when expressed individ- miR-16 was not necessary because it is expressed in HEK ually (compare with Figure 2B and Supplementary Figure 293T cells at a high level (Figure 6C). This experimental ap- S1A). proach permitted the simultaneous detection of both the Some of the CUG-repeat-binding miRNAs are endoge- selected miRNAs and the foci of either CUG- or CAG- nously expressed in cells at high levels (55)(Figure 6C). In triplet repeat mRNAs formed in the transfected cells. Using our experimental luciferase-based system, we overexpressed appropriate complementary probes that were fluorescently miRNAs from plasmids and ensured nearly equal concen- labeled with TYE665 (Cy5) or FAM at their 5 -ends (for trations of all of the tested miRNAs. However, the luciferase details, see Materials and Methods), in the FISH experi- tests used to study miRNA binding to CUG repeats were ment we detected that overexpression of 960 CUG repeat also positive in the absence of miRNA overexpression (Fig- transcripts caused the formation of miR-16-enriched CUG ure 6D). RNA foci, in contrast to the uniform distribution of miR-16 Next, we attempted to evaluate the expression of the in non-transfected cells (Figure 7A and Supplementary Fig- DMPK protein and mRNA following the transfection of ure S5). Interestingly, miR-16-containing CAG RNA foci HEK 293T cells with plasmids encoding miRs 15b/16, 214 were almost completely absent in cells overexpressing 960 and 29a, as well as the positive control transfection with CAG repeat transcripts (Figure 7B). This strongly suggests siRNAs specific to the DMPK sequence ( 56). The west- that miR-16 sequestration was specific for long CUG re- ern blot and real-time PCR analyses did not show a con- peats. The vast majority of the miR-16-enriched RNA foci siderable decrease in the DMPK protein and mRNA lev- were found in the cytoplasm, where 68% were co-localized els after transfection with the miRNAs (Figure 6E, F). Se- with the CUG RNA foci (Figure 7D). The relevant co- quencing of the endogenous DMPK 3 UTR fragment de- localization with CUG RNA inclusions was not observed rived from the HEK 293T cells used here revealed that these for miR-214, the miRNA bearing six matches with CUG cells harbored only 5 CTG repeats in the 3 UTR of the repeats in its seed (Figure 7C), or for the appropriate con- Nucleic Acids Research, 2015, Vol. 43, No. 19 9511 Figure 6. Interaction between miRNAs with CAG sequences in their seed regions and the CUG repeats in the DMPK transcript. (A) Relative repression of luciferase expression after transfection of reporter constructs carrying 20 CUG or 20 CAA repeats (denoted as WT CUG and MUT CUG) and either 9512 Nucleic Acids Research, 2015, Vol. 43, No. 19 trols. Therefore, the miR-16 co-localization with CUG foci cally significant enrichment for the former group, with 22% detected by RNA FISH confirmed our results from the lu- of transcripts with at least one b.s. for CUG-repeat-binding ciferase assays with the use of the same HEK 293T cells. miRNAs versus 13% for all transcripts (P = 0.0018; Yates’ Expanded CUG repeat tracts form both nuclear and cy- continuity corrected chi-squared test) (Figure 8A and Sup- toplasmic aggregates (57,58); however, the pathogenicity as- plementary Table S6). This trend was even stronger for tran- sociated with DM1 has been linked to the fact that mutant scripts that were uniquely deregulated in the CUG trans- DMPK transcripts are retained in the cell nucleus (11,59– genic mice, but not in the Mbnl1 or Clcn1 knockout mice, 60). In our experimental system, in rapidly proliferating which are alternative DM1 mouse models. Additionally, the −/− HEK 293T cells, we observed both nuclear and cytoplasmic transcripts that were uniquely deregulated in the Mbnl1 LR CUG foci, but the co-localization of such foci with miR-16 mice, but not in the HSA mice, did not include any bind- occurred to a greater extent in the cytoplasm (most likely ing sites for the CUG-binding miRNAs. This observation due to a cytoplasmic abundance of miR-16). Nevertheless, supports the hypothesis that the miRNA deregulation ob- because miRNAs function in both the cytoplasm and cell served in animal models of DM1 and DM1 patients may nucleus (reviewed in (61)), we believe that our findings are have a significant impact on changes in the transcriptome relevant and further support the possibility of miRNA se- and may contribute to the pathogenesis of DM1. questration by the abnormally expanded CUG repeats in the mutant DMPK transcripts. More evidence for miRNA DISCUSSION sequestration was indirectly provided by luciferase assays that were performed after the co-expression of appropriate The importance and role of miRNAs in human physiol- reporter constructs and CUG-expanded repeats; an experi- ogy and disease have been increasingly appreciated; how- ment with a reporter carrying a target sequence for miR-16 ever, many miRNA-driven regulatory mechanisms remain revealed that the presence of CUG repeats diminishes the unclear. This study reports that miRNAs regulate the ex- miR-16 effect (Supplementary Text 4 and Figure S7). pression of the DMPK gene and examines how miRNAs Having experimentally demonstrated the binding of the function on a specific target sequence and on a repeated miRNAs to CUG tracts using luciferase assays and having CUG sequence present in the DMPK transcript. observed miR-16-enriched CUG RNA foci by RNA FISH, we sought to examine whether the enhanced interactions miR-206 and miR-148a regulate the DMPK transcript and between the miRNAs and the aberrantly extended CUG re- may functionally cooperate peats in the DMPK 3 UTR could explain the changes in the transcriptome of DM1 mice reported by Thornton and We experimentally assessed the validity of three predicted colleagues (62). We computationally analyzed the miRNA interactions with miR-1, miR-206 and miR-148a. Using regulation of genes with significantly altered expression in luciferase reporter assays and sets of reporter constructs, LR CUG transgenic mice (HSA ) in comparison with the we provide the first evidence of DMPK down-regulation parental line. First, analogous to the human miRNAs, we by miR-206 and miR-148a. No changes were detected in predicted the presence of murine miRNAs with CAG re- the expression levels of the endogenous DMPK mRNA, peats in their seed sequences (Supplementary Table S5). but there was a significant decrease in the DMPK protein Next, we compared the prevalence of the predicted con- levels for both miR-148a and miR-206, which was even served b.s. for those miRNAs in the 3 UTRs of transcripts slightly higher for miR-206. Interestingly, in our experimen- that are altered in CUG transgenic mice, and in the 3 UTRs tal setup, we could not detect any regulation of the DMPK of all of the murine transcripts predicted to have at least one transcript by miR-1, which targets the same DMPK se- b.s. for any miRNAs using TargetScan Mouse Release 6.2. quence and has the same seed as miR-206, but differs some- We observed an overall enrichment for transcripts bearing what at the 3 end. miR-206 has two additional Watson- conserved miRNA b.s. in the CUG transgenic mice (78% of Crick base pairs to the DMPK b.s., which are absent in transcripts with at least one conserved b.s. versus 64% for the case of miR-1 binding. The variety in the 3 -end be- all murine transcripts; P = 0.0003; Yates’ continuity cor- tween miRs 1 and 206 seems to be important for their in- rected chi-squared test). Moreover, we discovered a statisti- hibitory activities against their targets. The impact of the ← −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− − the miR-15b/16-, miR-214- or miR-29a-expressing vectors, as indicated. The standard errors were calculated from twelve, eight and vfi e independent experiments that were performed after overexpression of miRs 15b/16, 214 and 29a, respectively. The asterisks indicate statistical significance ( P < 0.001). (B) Relative repression of luciferase expression by the reporters differing in the number of CUG repeats. Constructs with 5, 20 (WT), 44, 53 and 72 CUGs were tested in parallel. The luminescence was measured after the overexpression of miRs 15b/16 and 214, as indicated. The standard errors were calculated from ten and four independent experiments, respectively. The asterisks indicate statistical significance: a single asterisk denotes P < 0.05, a double asterisk denotes P < 0.01 and a triple asterisk denotes P < 0.001. (C) Northern blot detection of miRs 15b, 16 and 214 and 29a in untreated HEK 293T cells and cells transfected with the miRNA-encoding plasmids (System Biosciences). M denotes the size marker: end-labeled 17, 19, 21, 23 and 25-nt oligoribonucleotides. En and Ex indicate the endogenous and vector-expressed miRNA levels, respectively. The hybridization to U6 RNA serves as a loading control. (D)The relative repression of luciferase expression in HEK 293T cells measured after testing the reporter constructs carrying 20 CUG or 20 CAA repeats (denoted as WT CUG and MUT CUG) and 72 CUGs without miRNA overexpression. The standard errors were calculated from seven independent experiments. The asterisks indicate statistical significance: a single asterisk denotes P < 0.05, a double asterisk denotes P < 0.01 and a triple asterisk denotes P < 0.001. (E) Relative DMPK mRNA levels. Real-time PCR was performed after the HEK 293T cells were transfected with miRs 15b/16, 214, 29 and the appropriate siRNAs as a positive control (ssiDM10 (56)). The bar graphs show the quantification of the DMPK mRNA levels normalized to the actin mRNA level, based on the data from vfi e independent experiments. ( F) Western blot analysis of the DMPK protein levels after transfection of HEK 293T cells with miRs 15b/16, 214 and 29a, as indicated. The GAPDH protein served as a loading control. Nucleic Acids Research, 2015, Vol. 43, No. 19 9513 Figure 7. Visualization of miRNAs and cytoplasmic foci of CUG-triplet repeat mRNAs. (A) Representative RNA-FISH confocal images of cultured HEK 293T cells with endogenous miR-16 expression that were transfected with plasmid vectors expressing DMPK exons 11–15 containing 960 CUG repeats. (B) Representative RNA-FISH confocal images of cultured HEK 293T cells with endogenous miR-16 expression that were transfected with plasmid vectors expressing DMPK exons 11–15 containing 960 CAG repeats. (C) Representative RNA-FISH confocal images of cultured HEK 293T cells that were transfected with plasmid vectors expressing DMPK exons 11–15 containing 960 CUG repeats and the miR-214 mimic. The appropriate miRNAs and CUG repeat-containing RNAs were detected by differentially labeled fluorescent probes as depicted in the figure legend. The nuclei were stained wi th DAPI (blue). Scale bar: 20 m. (D) Graphs showing the statistics for the estimation of the foci found in the cytoplasm (C) and nucleus (N), as well as information about the co-localization of the miRNA and CUG foci. The number of miR-16 foci in cells expressing 960 CUG and 960 CAG repeats were compared using the Mann Whitney test; ***−P-value < 0.001. that two sites of the same or different miRNAs could act 3 -end variability of the miRNAs on target recognition and synergistically; the optimal distance between two miRNA binding strength has been previously reported. For instance, the functionality of the isomiR-214 variants, as opposed to sites was determined to be approximately 7–40 nt between the other tested miRNA isoforms, has been proposed to be neighboring miRNA b.s. (5–7). Potential cooperativity be- due to compensatory base pairing at the 3 end of the miR- tween more distant miRNA sites has also been addressed 214 isomiRs (33). (25,63–64); however, the optimal spacing requirements for The phenomenon of miRNA cooperativity and/or syn- neighboring miRNAs have not been precisely defined. In ergy has not been fully elucidated. It was demonstrated the present study, the distance between the miR-206 and 9514 Nucleic Acids Research, 2015, Vol. 43, No. 19 Figure 8. Potential sequestration of CUG-repeat binding miRNAs. (A) Deregulated genes in the CUG transgenic mice that may be subject to regulation by CUG-repeat-binding miRNAs (62). The left graph shows the percentage of genes bearing putative b.s. for the conserved miRNAs in all murine transcripts LR and the transcripts that are deregulated in the CUG transgenic mice (HSA ). The right graph compares the percentage of genes with the predicted miRNA b.s. for miRNAs with CAG sequences in their seed regions (Supplementary Table S5) in the following groups: all murine transcripts (all mouse genes), LR −/− all transcripts deregulated in the CUG transgenic mice (all HSA ), the transcripts deregulated in CUG transgenic mice, but not in the Mbnl1 or −/− LR −/− −/− Clcn1 mice (unique to HSA ), and the transcripts deregulated only in the Mbnl1 mice, but not in the CUG transgenic mice (unique to Mbnl1 ). All comparisons were made using Yates’ continuity corrected chi-squared test. **−P-value < 0.01, ***−P-value < 0.001. (B) Effects of the extended CUG repeats on miRNA activity. The mutant DMPK transcripts may sequester miRNAs with CAG repeats in their seed sequence. The sequestration of miRNAs could prevent their normal activity, leading to the reduction of accessible miRNA molecules and, consequently, improper repression of their targets. gle miRNAs. Interestingly, Ext 148/206, the reporter con- 148a sites in the intact DMPK 3 UTR measured between the 5 ends of both seeds was 15 nt. Based on the knowl- struct with slightly extended miRNA sites (to 31 nt between edge that these two miRs may act on the DMPK mRNA, the 5 ends of the seed) showed stronger suppression (by we analyzed the binding mechanism of the miR-206/148a approximately 20%). Of the three reporter constructs that pair. We found that this pair may cooperate because the were tested in our study, Ext 148/206 was the most effec- joint binding of miRs 206 and 148a increased the down- tively targeted construct by the simultaneously expressed regulation of DMPK expression. The reporter construct miRs 148a and 206. The observed effect was consistent with carrying the sequence corresponding to the native b.s. of previous findings showing that cooperativity is facilitated miR-148a and miR-206 (WT 148/206) with the 5 seed ends when the miRNA b.s. are directly adjoining (3,65)orare spaced by 15 nt (the lower limit of the cooperative distance) separated by a few additional nucleotides (66). In the case was efficiently suppressed by these miRNAs; and the sup- of the Long 148/206 construct with miRNA seed ends sep- pression was greater than that achieved by either of the sin- arated by 85 nt, the measured repression was restored to the Nucleic Acids Research, 2015, Vol. 43, No. 19 9515 level of the WT reporter. A similar decrease in cooperative that certain miRNAs directly target the expanded CUG re- repression has been reported previously (5,65). It has also peats in the DMPK mRNA and that the degree of miRNA- been assumed that the availability of miRNA sites that were mediated repression increases with the length of the re- separated is reduced due to the lower concentration of b.s. peated sequence. Because human miRNAs are able to de- (25), and that the loss of the directly adjacent b.s. may lead crease mRNA levels (42), mutant mRNAs with expanded to a shift toward independent activities (7). Together, our CUG triplets may serve as targets for therapeutic interven- results indicate that miR-148a and miR-206 are functional tion by reducing the amount of the toxic transcripts. Re- when their b.s. are located at a distance between 15 and 85 cently, several approaches aimed at targeting CUG repeats nt, and this observation may relate to the regulation of other with antisense oligonucleotides and short interfering RNAs genes targeted by these miRNAs. Here, we refer to miRNA (siRNAs) have been proposed (67–70); therefore, close at- cooperativity because the level of joint repression produced tention should also be paid to the potential use of natural by these two miRNA sites was stronger than the repression or artificial miRNA binding to such repeats. induced by only one of them; however, this repression was Having examined the miRNA targeting of the CUG re- still not greater than that expected from the independent peat tracts of different lengths present in the DMPK tran- contribution of two single sites. In particular, the resulting script using the luciferase system, we examined the poten- increase in the level of suppression (by approximately 20%) tial sequestration of these miRNAs by the exogenously ex- resembled the effect obtained using multiple (but the same) pressed mutant DMPK mRNAs in cells. We overexpressed rather than single binding sites in the reporter constructs either CUG or CAG expanded repeats in the context of (32). DMPK exons 11–15 (36) in HEK 293T cells and found Of particular importance is the fact that the studied miR- that miR-16, with the 7-nt seed sequence AGCAGCA, was NAs exhibit different expression profiles; miR-148a is ubiq- located in cytoplasmic foci formed by the mRNAs with uitously expressed in many tissue and cell types, whereas CUG repeats. We propose that the mutant DMPK tran- miR-206 is almost exclusively expressed in muscles. The scripts may serve as molecular sponges for natural miR- synergistic and/or antagonizing activity of miRNAs with NAs with CAG repeats in their seed sequences, sequester- different expression levels could be employed in therapeu- ing them and thereby preventing their physiological activity. tic interventions using miRNA mimics and/or miRNA in- We speculate that in addition to the sequestration of specific hibitors. The natural selective activity of two miRNAs with RNA-binding proteins, such as the MBNL family proteins overlapping binding sites has been previously reported (9). (12,13), the CUG repeats may also bind certain miRNAs Our genome-wide search for pairs of miRNA sites based on and RNA-induced silencing complexes (RISCs). their expression profiles suggests that both competition for The extension of the CUG repeats amplifies the naturally binding sites and miRNA cooperativity by binding to adja- existing overlapping miRNA b.s., and our luciferase assay cent sites are common miRNA regulatory mechanisms. experiments showed that the miRNA binding capacity in- In summary, our findings for miRNA cooperativity pro- creases with the length of the CUG tracts, which may pro- vide mechanistic insights into the miRNA-mediated regu- mote miRNA cooperativity. The mechanism of concerted lation of the DMPK transcript. Interference with miRNA miRNA regulation by two or more different miRNAs with function has known therapeutic potential; therefore, better overlapping or neighboring b.s. has not been fully eluci- knowledge of the miRNA-mediated gene regulation is es- dated, and we can only speculate how these miRNAs are sential because these potent regulatory RNAs can be used associated with the CUG tracts. The footprint of argonaute in various therapeutic strategies. (Ago), a central component of RISCs, was experimentally defined as 45–62 nt with the use of crosslinking immunopre- cipitation of Ago protein-RNA complexes (71), which indi- CUG repeats present in the DMPK 3 UTR may sequester cates that Ago bound to one b.s. covers the other sites po- CUG-repeat-binding miRNAs sitioned in close vicinity and prevents the binding of other The presence of abnormally long CUG repeats in the Ago molecules. Although long tracts of triplet repeats are DMPK 3 UTR can potentially affect the general physio- likely to form tight hairpin structures, the hairpin stems logical miRNA-mediated regulatory mechanisms of gene are easily accessible to RNA binding proteins (reviewed in expression because these expanded CUG repeats can be a (72)), and we think that they are likely to be the associa- target of the miRNA silencing machinery. Using bioinfor- tion sites for RISCs loaded with CUG binding miRNAs. matics, Hon et al. identified a group of 7 miRNAs that were However, we cannot excluded the possibility that miRNAs predicted to bind to CUG repeats via the CAG sequences containing CAG sequences in their seed regions can directly present in their seed regions. It was also calculated that the associate with CUG repeats without being constantly asso- miRNA binding affinity to CUG repeats may increase with ciated with RISCs, because the presence of stable Ago-free the number of CUG repeat sequences (25). miRNA–mRNA duplexes has been recently demonstrated In the present study, we performed a detailed analy- in mammalian cells (73). The increase in the local concen- sis of miRNA binding to CUG repeats in the DMPK tration of CUG binding miRNAs associated with CUG 3 UTR using both computational and experimental ap- tracts could then enhance the preference for Ago bind- proaches. Using luciferase assays, we examined three miR- ing, resulting in sequestration of the miRNA silencing ma- NAs and demonstrated that miR-15b/16 can significantly chinery. In this context, the sequestration of CUG-repeat- down-regulate its target. Importantly, the repression effi- binding miRNAs could cause them to become inactive or ciency of this miRNA increased with the length of the re- simply prevent them from regulating other transcripts that peated CUG sequence. This result corroborates the findings contain these miRNA b.s., leading to a more widespread 9516 Nucleic Acids Research, 2015, Vol. 43, No. 19 impairment of the miRNA-mediated regulation of gene ex- with a Zeiss Axiovert 200M microscope equipped with a pression (Figure 8B). cooled AxioCam HRc camera. It has been demonstrated that CTG expansions are di- rectly linked to alterations in miRNA regulation and RNA FUNDING toxicity in DM1 (24). An analysis of the muscle miRNA transcriptome in a Drosophila model of CUG toxicity re- Polish Ministry of Science and Higher Education vealed 20 alterations in the miRNA profiles triggered by the [N N301 523038 to E.K.]; National Science Center expression of repeats, and the misregulation of several miR- [2011/03/B/NZ1/03259 and 2012/06/A/NZ1/00094 to NAs has been shown to be independent of Mbnl1 seques- W.J.K.]. This publication was also supported by the Polish tration. In addition, the expansion of CUGs in transgenic Ministry of Science and Higher Education, under the Lead- mice has been shown to affect gene expression at the RNA ing National Research Centre (KNOW) program for the level, and the observed effects were mediated mainly, but years 2014-2019. Funding for open access charge: National not entirely, through the sequestration of Mbnl1 (62). Science Center [2012/06/A/NZ1/00094 to W.J.K.]. We consider that the sequestration of the CUG-repeat- Conflict of interest statement. None declared. binding miRNAs may have wider implications for the RNA-mediated regulation of gene expression. Long non- coding RNAs (lncRNAs) were recently shown to play a vital REFERENCES role in connecting many RNA regulatory pathways (74–78). 1. Bartel,D.P. (2009) MicroRNAs: target recognition and regulatory A growing number of direct links between the levels of miR- functions. Cell, 136, 215–233. NAs, mRNAs and lncRNAs are being discovered (recently 2. Eulalio,A., Huntzinger,E., Nishihara,T., Rehwinkel,J., Fauser,M. and Izaurralde,E. (2009) Deadenylation is a widespread effect of reviewed in (79), and lncRNAs are thought to act as com- miRNA regulation. RNA, 15, 21–32. peting endogenous RNAs (ceRNAs) and miRNA sponges 3. Doench,J.G., Petersen,C.P. and Sharp,P.A. (2003) siRNAs can (80)). Based on these reports, we speculate that the seques- function as miRNAs. Genes Dev., 17, 438–442. tration of CUG-repeat-binding miRNAs can affect the ex- 4. Pillai,R.S., Artus,C.G. and Filipowicz,W. (2004) Tethering of human pression levels of lncRNAs with binding sites for these miR- Ago proteins to mRNA mimics the miRNA-mediated repression of protein synthesis. RNA, 10, 1518–1525. NAs. Imbalances between miRNAs and interacting ceR- 5. Grimson,A., Farh,K.K., Johnston,W.K., Garrett-Engele,P., Lim,L.P. NAs due to altered expression levels or the cellular distribu- and Bartel,D.P. (2007) MicroRNA targeting specificity in mammals: tion and mutagenesis of interacting sites have been shown determinants beyond seed pairing. Mol. Cell, 27, 91–105. to contribute to disease pathogenesis (80). It was also pro- 6. Saetrom,P., Heale,B.S., Snove,O. Jr, Aagaard,L., Alluin,J. and Rossi,J.J. (2007) Distance constraints between microRNA target sites posed that competitor RNAs should be among the most dictate efficacy and cooperativity. Nucleic Acids Res., 35, 2333–2342. abundant RNAs in the cell, or that they should contain 7. Rinck,A., Preusse,M., Laggerbauer,B., Lickert,H., Engelhardt,S. dozens of binding sites for a single miRNA species (81,82). and Theis,F.J. (2013) The human transcriptome is enriched for In this scenario, the long CUG repeats extended in DM1 ap- miRNA-binding sites located in cooperativity-permitting distance. pear to be ideal candidates. Therefore, we believe that lncR- RNA Biol., 10, 1125–1135. 8. Schmitz,U., Lai,X., Winter,F., Wolkenhauer,O., Vera,J. and NAs may play a role in the pathogenesis of TREDs, and Gupta,S.K. (2014) Cooperative gene regulation by microRNA pairs future studies are needed to unravel their functions. and their identification using a computational workflow. Nucleic In conclusion, it is of utmost importance to understand Acids Res., 42, 7539–7552. the role of miRNAs in DM1 to identify potential novel ther- 9. Yu,J., Ryan,D.G., Getsios,S., Oliveira-Fernandes,M., Fatima,A. and Lavker,R.M. (2008) MicroRNA-184 antagonizes microRNA-205 to apies. Given the complexity and degree of interactions be- maintain SHIP2 levels in epithelia. Proc. Natl. Acad. Sci. U.S.A., tween miRNAs and their various targets, which may be ei- 105, 19300–19305. ther protein-coding or non-coding transcripts, a better un- 10. Mahadevan,M., Tsilfidis,C., Sabourin,L., Shutler,G., Amemiya,C., derstanding of the rules governing gene regulation by non- Jansen,G., Neville,C., Narang,M., Barcelo,J., O’Hoy,K. et al. (1992) coding RNAs (ncRNAs) and the intricate mechanisms that Myotonic dystrophy mutation: an unstable CTG repeat in the 3 untranslated region of the gene. Science, 255, 1253–1255. are involved in DM1 pathogenesis is necessary. 11. Davis,B.M., McCurrach,M.E., Taneja,K.L., Singer,R.H. and Housman,D.E. (1997) Expansion of a CUG trinucleotide repeat in SUPPLEMENTARY DATA the 3 untranslated region of myotonic dystrophy protein kinase transcripts results in nuclear retention of transcripts. Proc. Natl. Supplementary Data are available at NAR Online. Acad. Sci. U.S.A., 94, 7388–7393. 12. Lin,X., Miller,J.W., Mankodi,A., Kanadia,R.N., Yuan,Y., ACKNOWLEDGEMENT Moxley,R.T., Swanson,M.S. and Thornton,C.A. (2006) Failure of MBNL1-dependent post-natal splicing transitions in myotonic We thank Thomas A. Cooper for kindly providing the dystrophy. Hum. Mol. Genet., 15, 2087–2097. plasmid vectors expressing the highly expanded repeats 13. Mankodi,A., Takahashi,M.P., Jiang,H., Beck,C.L., Bowers,W.J., Moxley,R.T., Cannon,S.C. and Thornton,C.A. (2002) Expanded and Denis Furling for the DM1 myoblasts. We also thank CUG repeats trigger aberrant splicing of ClC-1 chloride channel Krzysztof Sobczak for sharing materials used for myoblast pre-mRNA and hyperexcitability of skeletal muscle in myotonic differentiation and Edyta Jaworska for assistance with the dystrophy. Mol. Cell, 10, 35–44. miRNA profiling. The real-time PCR analyses were per- 14. Sicot,G., Gourdon,G. and Gomes-Pereira,M. (2011) Myotonic dystrophy, when simple repeats reveal complex pathogenic entities: formed using a LightCycler 480 II system (Roche) in the new findings and future challenges. Hum. Mol. Genet., 20, Laboratory of Subcellular Structures Analysis at the Insti- R116–R123. tute of Bioorganic Chemistry PAS in Poznan. The FISH 15. Batra,R., Charizanis,K., Manchanda,M., Mohan,A., Li,M., images were obtained at the Electron and Confocal Mi- Finn,D.J., Goodwin,M., Zhang,C., Sobczak,K., Thornton,C.A. et al. croscopy Core at Adam Mickiewicz University in Poznan (2014) Loss of MBNL leads to disruption of developmentally Nucleic Acids Research, 2015, Vol. 43, No. 19 9517 regulated alternative polyadenylation in RNA-mediated disease. 35. Figura,G., Koscianska,E. and Krzyzosiak,W.J. (2015) In Vitro Mol. Cell, 56, 311–322. Expansion of CAG, CAA, and Mixed CAG/CAA Repeats. Int J Mol 16. Koscianska,E., Kozlowska,E., Jaworska,E. and Krzyzosiak,W. Sci., 16, 18741–18751. (2014) MicroRNA deregulation in trinucleotide repeat expansion 36. Ho,T.H., Savkur,R.S., Poulos,M.G., Mancini,M.A., Swanson,M.S. disorders. In: Arbuthnot,P and Weinberg,MS (eds). Applied RNAi: and Cooper,T.A. (2005) Colocalization of muscleblind with RNA From Fundamental Research to Therapeutic Applications. Caister foci is separable from mis-regulation of alternative splicing in Academic Press, pp. 227–246. myotonic dystrophy. J. Cell Sci., 118, 2923–2933. 17. Latronico,M.V. and Condorelli,G. (2009) MicroRNAs and cardiac 37. Koscianska,E., Starega-Roslan,J., Czubala,K. and Krzyzosiak,W.J. pathology. Nat. Rev. Cardiol., 6, 419–429. (2011) High-resolution northern blot for a reliable analysis of 18. van Rooij,E., Liu,N. and Olson,E.N. (2008) MicroRNAs flex their microRNAs and their precursors. ScientificWorldJournal , 11, muscles. Trends Genet., 24, 159–166. 102–117. 19. Eisenberg,I., Eran,A., Nishino,I., Moggio,M., Lamperti,C., 38. Koscianska,E., Starega-Roslan,J., Sznajder,L.J., Olejniczak,M., Amato,A.A., Lidov,H.G., Kang,P.B., North,K.N., Galka-Marciniak,P. and Krzyzosiak,W.J. (2011) Northern blotting Mitrani-Rosenbaum,S. et al. (2007) Distinctive patterns of analysis of microRNAs, their precursors and RNA interference microRNA expression in primary muscular disorders. Proc. Natl. triggers. BMC Mol. Biol., 12, 14. Acad. Sci. U.S.A., 104, 17016–17021. 39. Wojciechowska,M., Taylor,K., Sobczak,K., Napierala,M. and 20. Greco,S., Perfetti,A., Fasanaro,P., Cardani,R., Capogrossi,M.C., Krzyzosiak,W.J. (2014) Small molecule kinase inhibitors alleviate Meola,G. and Martelli,F. (2012) Deregulated microRNAs in different molecular features of myotonic dystrophy type 1. RNA myotonic dystrophy type 2. PLoS One, 7, e39732. Biol., 11, 742–754. 21. Perbellini,R., Greco,S., Sarra-Ferraris,G., Cardani,R., 40. Witkos,T.M., Koscianska,E. and Krzyzosiak,W.J. (2011) Practical Capogrossi,M.C., Meola,G. and Martelli,F. (2011) Dysregulation Aspects of microRNA Target Prediction. Curr.Mol.Med., 11, and cellular mislocalization of specific miRNAs in myotonic 93–109. dystrophy type 1. Neuromuscul. Disord., 21, 81–88. 41. Baek,D., Villen,J., Shin,C., Camargo,F.D., Gygi,S.P. and Bartel,D.P. 22. Rau,F., Freyermuth,F., Fugier,C., Villemin,J.P., Fischer,M.C., (2008) The impact of microRNAs on protein output. Nature, 455, Jost,B., Dembele,D., Gourdon,G., Nicole,A., Duboc,D. et al. (2011) 64–71. Misregulation of miR-1 processing is associated with heart defects in 42. Guo,H., Ingolia,N.T., Weissman,J.S. and Bartel,D.P. (2010) myotonic dystrophy. Nat. Struct. Mol. Biol., 18, 840–845. Mammalian microRNAs predominantly act to decrease target 23. Gambardella,S., Rinaldi,F., Lepore,S.M., Viola,A., Loro,E., mRNA levels. Nature, 466, 835–840. Angelini,C., Vergani,L., Novelli,G. and Botta,A. (2010) 43. Llorens,F., Banez-Coronel,M., Pantano,L., del Rio,J.A., Ferrer,I., Overexpression of microRNA-206 in the skeletal muscle from Estivill,X. and Marti,E. (2013) A highly expressed miR-101 isomiR myotonic dystrophy type 1 patients. J. Transl. Med., 8, 48. is a functional silencing small RNA. BMC Genom., 14, 104. 24. Fernandez-Costa,J.M., Garcia-Lopez,A., Zuniga,S., 44. Erriquez,D., Perini,G. and Ferlini,A. (2013) Non-coding RNAs in Fernandez-Pedrosa,V., Felipo-Benavent,A., Mata,M., Jaka,O., muscle dystrophies. Int. J. Mol. Sci., 14, 19681–19704. Aiastui,A., Hernandez-Torres,F., Aguado,B. et al. (2012) Expanded 45. Ge,Y. and Chen,J. (2011) MicroRNAs in skeletal myogenesis. Cell CTG repeats trigger miRNA alterations in Drosophila that are Cycle, 10, 441–448. conserved in myotonic dystrophy type 1 patients. Hum. Mol. Genet., 46. Lin,C.Y., Lee,H.C., Fu,C.Y., Ding,Y.Y., Chen,J.S., Lee,M.H., 22, 704–716. Huang,W.J. and Tsai,H.J. (2013) MiR-1 and miR-206 target different 25. Hon,L.S. and Zhang,Z. (2007) The roles of binding site arrangement genes to have opposing roles during angiogenesis in zebrafish and combinatorial targeting in microRNA repression of gene embryos. Nat. Commun., 4, 2829. expression. Genome Biol., 8, R166. 47. Brennecke,J., Stark,A., Russell,R.B. and Cohen,S.M. (2005) 26. Betel,D., Koppal,A., Agius,P., Sander,C. and Leslie,C. (2010) Principles of microRNA-target recognition. PLoS Biol., 3, e85. Comprehensive modeling of microRNA targets predicts functional 48. Kertesz,M., Iovino,N., Unnerstall,U., Gaul,U. and Segal,E. (2007) non-conserved and non-canonical sites. Genome Biol., 11, R90. The role of site accessibility in microRNA target recognition. Nat. 27. Garcia,D.M., Baek,D., Shin,C., Bell,G.W., Grimson,A. and Genet., 39, 1278–1284. Bartel,D.P. (2011) Weak seed-pairing stability and high target-site 49. Long,D., Lee,R., Williams,P., Chan,C.Y., Ambros,V. and Ding,Y. abundance decrease the proficiency of lsy-6 and other microRNAs. (2007) Potent effect of target structure on microRNA function. Nat. Nat. Struct. Mol. Biol., 18, 1139–1146. Struct. Mol. Biol., 14, 287–294. 28. Maragkakis,M., Reczko,M., Simossis,V.A., Alexiou,P., 50. Robins,H., Li,Y. and Padgett,R.W. (2005) Incorporating structure to Papadopoulos,G.L., Dalamagas,T., Giannopoulos,G., Goumas,G., predict microRNA targets. Proc. Natl. Acad. Sci. U.S.A., 102, Koukis,E., Kourtis,K. et al. (2009) DIANA-microT web server: 4006–4009. elucidating microRNA functions through target prediction. Nucleic 51. Guo,Z., Maki,M., Ding,R., Yang,Y., Zhang,B. and Xiong,L. (2014) Acids Res., 37, W273–W276. Genome-wide survey of tissue-specific microRNA and transcription 29. Kozomara,A. and Griffiths-Jones,S. (2014) miRBase: annotating factor regulatory networks in 12 tissues. Sci. Rep., 4, 5150. high confidence microRNAs using deep sequencing data. Nucleic 52. Finnerty,J.R., Wang,W.X., Hebert,S.S., Wilfred,B.R., Mao,G. and Acids Res., 42, D68–D73. Nelson,P.T. (2010) The miR-15/107 group of microRNA genes: 30. Lewis,B.P., Burge,C.B. and Bartel,D.P. (2005) Conserved seed evolutionary biology, cellular functions, and roles in human diseases. pairing, often flanked by adenosines, indicates that thousands of J. Mol. Biol., 402, 491–509. human genes are microRNA targets. Cell, 120, 15–20. 53. Nelson,P.T., Wang,W.X., Mao,G., Wilfred,B.R., Xie,K., 31. Hsu,S.D., Chu,C.H., Tsou,A.P., Chen,S.J., Chen,H.C., Hsu,P.W., Jennings,M.H., Gao,Z. and Wang,X. (2011) Specific sequence Wong,Y.H., Chen,Y.H., Chen,G.H. and Huang,H.D. (2008) determinants of miR-15/107 microRNA gene group targets. Nucleic miRNAMap 2.0: genomic maps of microRNAs in metazoan Acids Res., 39, 8163–8172. genomes. Nucleic Acids Res., 36, D165–D169. 54. Wang,W.X., Kyprianou,N., Wang,X. and Nelson,P.T. (2010) 32. Koscianska,E., Baev,V., Skreka,K., Oikonomaki,K., Rusinov,V., Dysregulation of the mitogen granulin in human cancer through the Tabler,M. and Kalantidis,K. (2007) Prediction and preliminary miR-15/107 microRNA gene group. Cancer Res., 70, 9137–9142. validation of oncogene regulation by miRNAs. BMC Mol. Biol., 8, 55. Wang,W.X., Danaher,R.J., Miller,C.S., Berger,J.R., Nubia,V.G., 79. Wilfred,B.S., Neltner,J.H., Norris,C.M. and Nelson,P.T. (2014) 33. Kozlowska,E., Krzyzosiak,W.J. and Koscianska,E. (2013) Regulation Expression of miR-15/107 family microRNAs in human tissues and of huntingtin gene expression by miRNA-137, -214, -148a, and their cultured rat brain cells. Genomics Proteomics Bioinformatics, 12, respective isomiRs. Int. J. Mol. Sci., 14, 16999–17016. 19–30. 34. Takahashi,N., Sasagawa,N., Suzuki,K. and Ishiura,S. (1999) 56. Langlois,M.A., Boniface,C., Wang,G., Alluin,J., Salvaterra,P.M., Synthesis of long trinucleotide repeats in vitro. Neurosci. Lett., 262, Puymirat,J., Rossi,J.J. and Lee,N.S. (2005) Cytoplasmic and nuclear 45–48. retained DMPK mRNAs are targets for RNA interference in myotonic dystrophy cells. J. Biol. Chem., 280, 16949–16954. 9518 Nucleic Acids Research, 2015, Vol. 43, No. 19 57. Amack,J.D. and Mahadevan,M.S. (2001) The myotonic dystrophy 71. Chi,S.W., Zang,J.B., Mele,A. and Darnell,R.B. (2009) Argonaute expanded CUG repeat tract is necessary but not sufficient to disrupt HITS-CLIP decodes microRNA-mRNA interaction maps. Nature, C2C12 myoblast differentiation. Hum. Mol. Genet., 10, 1879–1887. 460, 479–486. 58. Dansithong,W., Wolf,C.M., Sarkar,P., Paul,S., Chiang,A., Holt,I., 72. Krzyzosiak,W.J., Sobczak,K., Wojciechowska,M., Fiszer,A., Morris,G.E., Branco,D., Sherwood,M.C., Comai,L. et al. (2008) Mykowska,A. and Kozlowski,P. (2012) Triplet repeat RNA structure Cytoplasmic CUG RNA foci are insufficient to elicit key DM1 and its role as pathogenic agent and therapeutic target. Nucleic Acids features. PLoS One, 3, e3968. Res., 40, 11–26. 59. Taneja,K.L., McCurrach,M., Schalling,M., Housman,D. and 73. Janas,M.M., Wang,B., Harris,A.S., Aguiar,M., Shaffer,J.M., Singer,R.H. (1995) Foci of trinucleotide repeat transcripts in nuclei Subrahmanyam,Y.V., Behlke,M.A., Wucherpfennig,K.W., Gygi,S.P., of myotonic dystrophy cells and tissues. J. Cell Biol., 128, 995–1002. Gagnon,E. et al. (2012) Alternative RISC assembly: binding and 60. Timchenko,L.T., Timchenko,N.A., Caskey,C.T. and Roberts,R. repression of microRNA-mRNA duplexes by human Ago proteins. (1996) Novel proteins with binding specificity for DNA CTG repeats RNA, 18, 2041–2055. and RNA CUG repeats: implications for myotonic dystrophy. Hum. 74. Cesana,M., Cacchiarelli,D., Legnini,I., Santini,T., Sthandier,O., Mol. Genet., 5, 115–121. Chinappi,M., Tramontano,A. and Bozzoni,I. (2011) A long 61. Ohrt,T., Muetze,J., Svoboda,P. and Schwille,P. (2012) Intracellular noncoding RNA controls muscle differentiation by functioning as a localization and routing of miRNA and RNAi pathway components. competing endogenous RNA. Cell, 147, 358–369. Curr.Top.Med.Chem., 12, 79–88. 75. Karreth,F.A., Tay,Y., Perna,D., Ala,U., Tan,S.M., Rust,A.G., 62. Osborne,R.J., Lin,X., Welle,S., Sobczak,K., O’Rourke,J.R., DeNicola,G., Webster,K.A., Weiss,D., Perez-Mancera,P.A. et al. Swanson,M.S. and Thornton,C.A. (2009) Transcriptional and (2011) In vivo identification of tumor- suppressive PTEN ceRNAs in post-transcriptional impact of toxic RNA in myotonic dystrophy. an oncogenic BRAF-induced mouse model of melanoma. Cell, 147, Hum. Mol. Genet., 18, 1471–1481. 382–395. 63. Kloosterman,W.P., Wienholds,E., Ketting,R.F. and Plasterk,R.H. 76. Poliseno,L., Salmena,L., Zhang,J., Carver,B., Haveman,W.J. and (2004) Substrate requirements for let-7 function in the developing Pandolfi,P.P. (2010) A coding-independent function of gene and zebrafish embryo. Nucleic Acids Res., 32, 6284–6291. pseudogene mRNAs regulates tumour biology. Nature, 465, 64. Vella,M.C., Choi,E.Y., Lin,S.Y., Reinert,K. and Slack,F.J. (2004) The 1033–1038. C. elegans microRNA let-7 binds to imperfect let-7 complementary 77. Sumazin,P., Yang,X., Chiu,H.S., Chung,W.J., Iyer,A., sites from the lin-41 3 UTR. Genes Dev., 18, 132–137. Llobet-Navas,D., Rajbhandari,P., Bansal,M., Guarnieri,P., Silva,J. 65. Broderick,J.A., Salomon,W.E., Ryder,S.P., Aronin,N. and et al. (2011) An extensive microRNA-mediated network of Zamore,P.D. (2011) Argonaute protein identity and pairing RNA-RNA interactions regulates established oncogenic pathways in geometry determine cooperativity in mammalian RNA silencing. glioblastoma. Cell, 147, 370–381. RNA, 17, 1858–1869. 78. Tay,Y., Kats,L., Salmena,L., Weiss,D., Tan,S.M., Ala,U., Karreth,F., 66. Doench,J.G. and Sharp,P.A. (2004) Specificity of microRNA target Poliseno,L., Provero,P., Di Cunto,F. et al. (2011) selection in translational repression. Genes Dev., 18, 504–511. Coding-independent regulation of the tumor suppressor PTEN by 67. Evers,M.M., Pepers,B.A., van Deutekom,J.C., Mulders,S.A., den competing endogenous mRNAs. Cell, 147, 344–357. Dunnen,J.T., Aartsma-Rus,A., van Ommen,G.J. and van 79. Yoon,J.H., Abdelmohsen,K. and Gorospe,M. (2014) Functional Roon-Mom,W.M. (2011) Targeting several CAG expansion diseases interactions among microRNAs and long noncoding RNAs. Semin. by a single antisense oligonucleotide. PLoS One, 6, e24308. Cell Dev. Biol., 34, 9–14. 68. Sicot,G. and Gomes-Pereira,M. (2013) RNA toxicity in human 80. Tay,Y., Rinn,J. and Pandolfi,P.P. (2014) The multilayered complexity disease and animal models: from the uncovering of a new mechanism of ceRNA crosstalk and competition. Nature, 505, 344–352. to the development of promising therapies. Biochim. Biophys. Acta, 81. Broderick,J.A. and Zamore,P.D. (2014) Competitive endogenous 1832, 1390–1409. RNAs cannot alter microRNA function in vivo. Mol. Cell, 54, 69. Udd,B. and Krahe,R. (2012) The myotonic dystrophies: molecular, 711–713. clinical, and therapeutic challenges. Lancet Neurol., 11, 891–905. 82. Denzler,R., Agarwal,V., Stefano,J., Bartel,D.P. and Stoffel,M. (2014) 70. Sobczak,K., Wheeler,T.M., Wang,W. and Thornton,C.A. (2013) Assessing the ceRNA hypothesis with quantitative measurements of RNA interference targeting CUG repeats in a mouse model of miRNA and target abundance. Mol. Cell, 54, 766–776. myotonic dystrophy. Mol. Ther., 21, 380–387.

Journal

Nucleic Acids Research – Oxford University Press

Published: Oct 30, 2015

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Cooperation meets competition inmicroRNA-mediated DMPK transcript regulation

Cooperation meets competition inmicroRNA-mediated DMPK transcript regulation

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

Loading...

References (86)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies