Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Mar 1.
Published in final edited form as: Nat Struct Mol Biol. 2014 Aug 3;21(9):825–832. doi: 10.1038/nsmb.2862

Regulation of miRNA-mediated gene silencing by miRNA precursors

Biswajoy Roy-Chaudhuri 1,2, Paul N Valdmanis 1,2, Yue Zhang 1,2, Qing Wang 1,2, Qingjun Luo 1,2, Mark A Kay 1,2,3
PMCID: PMC4244528  NIHMSID: NIHMS610614  PMID: 25086740

Abstract

Processing of miRNAs from their precursors to the biologically active mature form is regulated during development and cancer. We show that mouse precursor-miR-151 can bind to and compete with mature miR-151-5p and miR-151-3p for binding sites contained within the complementary regions of the E2f6 mRNA 3′UTR. In agreement, E2f6 mRNA levels were regulated by precursor-miR-151. Conversely, the miR-151-mediated repression of ARHGDIA mRNA was only dependent on the mature miR-151 level as only the mature miRNA was able to bind to the 3′UTR. This suggests that processing of miR-151 can have different effects on separate mRNA targets within a cell. A bioinformatics pipeline revealed additional candidate regions where pre-miRNAs can compete with their mature miRNA counterparts. This was experimentally validated for miR-124 and the SNAI2 3′UTR. Hence, miRNA precursors can serve as post-transcriptional regulators of miRNA activity and are not mere biogenesis intermediates.

Micro RNAs are a class of about 1000 non-coding RNAs that in mammals regulate about half of the protein encoding genes by enhancing their degradation or preventing their translation1,2. Following transcription, the long primary miRNA (pri-miRNAs) transcripts are processed to 60-80 nucleotide precursor miRNAs (pre-miRNAs) and subsequently cleaved to mature ~ 22nt long duplex RNAs3,4. With a few exceptions5,6, the precursor miRNAs have been considered as mere intermediates of the miRNA biogenesis pathway rather than gene regulators by themselves.

The miRNA pathway is controlled by transcriptional and post-transcriptional regulation in a tissue- and developmental stage-specific manner7,8. Various steps of miRNA biogenesis are also specifically regulated during differentiation and tumor development resulting in altered ratios of the mature and intermediate miRNA species in these processes9-12 [also see reviews13-16]. RNA editing17,18 of miRNAs represents an important post-transcriptional mechanism to regulate miRNA expression19,20. The precursor of miR-151, a LINE-2 repetitive element encoded miRNA, isA-to-I edited which interestingly inhibits its further processing by Dicer21. This results in an accumulation of edited precursor and reduced levels of mature miR-151 in the mammalian brain. However, the fate of the edited miR-151 precursor is not known, nor is the importance of miR-151 editing in brain well appreciated.

While various mechanisms have been uncovered which can potentially regulate the expression of mature miRNAs, others have been deciphered that regulate the activity of the mature miRNAs, including RNA editing22, target mimicry, competing endogenous RNAs (ceRNAs), and circular RNAs23-28. Here we show that miRNA precursors act as a new class of post-transcriptional regulators of miRNA activity. Paradoxically, these precursors, including the edited miR-151 precursor can compete with their own mature counterparts to bind to the overlapping miRNA response elements (MRE) in the 3′UTR of target genes to regulate their expression. These specific events may shed new insight to the observed changes in some cancers where components of the miRNA pathway and specific miRNAs are mis-regulated.

Results

miR-151-5p cleaves E2f6 in spite of a seed region mismatch

Repetitive elements in the mammalian genome can act as templates for microRNA biogenesis, particularly if the same repeat integrates in tandem copies29. In the case of miR-151, two head-to-head L2c LINE integration events in intron 21 of the Ptk2 gene lead to expression of a hairpin structure that subsequently becomes a target for Drosha and Dicer enzymes leading to a bona fide miRNA (Fig. 1a). A high degree of complementarity to miR-151 can be found in the 3′UTRs of several genes where L2 LINE integration and duplication events have occurred29,30. Within the 3′UTR of human E2F6, a perfect match exists for mature human miR-151-5p (5p being the guide strand), while one mismatch exists in the seed region of the corresponding mouse E2f6 — miR-151 pair (Fig. 1b). E2f6 is a cell cycle regulatory protein of the E2F family of transcription factors31. Given the presence of the single mismatch in the mouse E2f6 — miR-151 pair in the seed region of miR-151-5p, we asked if miR-151 is able to target E2f6 and if so, establish the mechanism responsible for reducing E2f6 expression. To do so, we performed reporter assays in mouse embryonic fibroblast (MEF) cells by co-transfecting a Pol III based small hairpin (sh) system32 (sh-miR-151-5p) that drove expression of high levels of mature miR-151-5p (and not pre-miR-151) (Supplementary Fig. 1a) and a reporter plasmid cloned with a 902 bp region (out of 1360 nt) of 3′UTR of E2f6 (wt). Overexpression of sh-151-5p resulted in a robust suppression (~80%) relative to a scrambled control (sh-scr) (Fig. 1c). Correction of the one seed-region mismatch present in the E2f6 3′UTR to mimic the perfectly base paired human E2F6 — miR-151-5p pair (per 5p) had no further effect on the repression of E2f6 by miR-151-5p indicating that the sequence difference between human and mouse E2f6 — miR-151-5p pair has no functional consequence. In contrast, placement of four mutations in the E2f6 seed region (mut 5p) reduced the repression to 25% (Fig. 1c). Western blot analysis after co-transfection with an E2f6 expression plasmid containing its entire CDS and 3′UTR, along with sh-miR-151-5p showed that intact E2f6 protein production was substantially inhibited (Fig. 1d) corroborating the results from the reporter assays. Likewise, overexpression of miR-151-5p caused a dramatic reduction in endogenous E2f6 mRNA levels as shown by qPCR (Fig. 1e). Furthermore, a miR-151-5p inhibitor specifically blocked the ability of endogenous miR-151-5p to downregulate the E2f6 3′UTR reporter expression in a dose dependent manner (Fig. 1f). All of these experiments indicate that miR-151-5p is able to target E2f6 3′UTR.

Figure 1.

Figure 1

Open in a new tab

Mmu-miR-151-5p cleaves E2f6 in the absence of a seed match.

(a) Genomic locus of mmu-miR-151 encoded by a LINE2 repeat element. (b) Schematic of the binding site of mmu-miR-151-5p to E2f6 3′UTR. (c) Dual-luciferase reporter assay for the wildtype E2f6 3′UTR (wt) or other mutants (per 5p and mut 5p) in presence of miR-151-5p overexpression (sh-151-5p). (d) Western blot for E2f6in presence of sh-151-5p or a scrambled control (sh-scr). Actin serves as a loading control. Uncropped blot in Supplementary Fig. 6. (e) E2f6 qPCR on miR-151-5p overexpression. Error bars, s.e.m. (n = 3 replicates). (f) Dual-luciferase reporter assay for E2f6 3′UTR or control (Sox4 3′UTR) in presence of an increasing dosage of miR-151-5p inhibitor. (g) Dual-luciferase reporter assay in Ago2−/− cells for E2f6 3′UTR in presence of sh-151-5p and a functional copy of Ago2 or cleavage deficient Ago2 (D597A) or Ago1. (h) 5′-RACE of E2f6. Arrowhead in the E2f6 3′UTR sequence (schematic) indicates the 5′ end of majority of E2f6 cleavage products in mouse lung tissue. Agarose gel showing E2f6 cleaved products (shown by an asterisk) is shown in the top gel (uncropped gel in Supplementary Fig. 6). The bottom gel serves as a RACE reaction control to detect the presence of E2f6 and ARHGDIA cDNAs. (c,f, g) For reporter assays, normalization was done with respect to sh-scr. Error bars, s.e.m. (n = 2 biological replicates, each with 4 technical replicates), ns denotes not significant,***P = 0.001 by two-tailed Student′s t test.

Unlike most mammalian miRNAs that mediate gene knockdown by non-cleavage repression processes, we established using two lines of evidence that miR-151-5p suppressed mouse E2f6 expression by the more robust Ago2-mediated cleavage mechanism. First, in Ago2 null (Ago2−/−) MEFs incapable of cleavage-mediated miRNA knockdown33, we observed only ~ 25% repression (Fig. 1g) by reporter assay. Moreover, we could rescue the robust miR-151-mediated repression of E2f6 when wildtype but not a catalytically inactive form of Ago2 (D597A mutation in the PIWI domain)33 or cleavage incompetent Ago1 was added back to the Ago2−/− cells34 (Fig. 1g). Second, we detected an E2f6 cleavage product in mouse lung tissue using a modified 5′-RACE (rapid amplification of cDNA ends) assay. Sequencing of the 5′-RACE clones confirmed the expected position of the 5′ end of the E2f6 cleavage product (8 of 11 clones from the amplicon with the expected size) which mapped to the 10th nucleotide that pairs with miR-151-5p (Fig. 1h). To confirm that the cleavage product of E2f6 was due to miR-151-5p, MEF cells were co-transfected with a scrambled control (sh-scr) or sh-151-5p along with an E2f6 overexpression plasmid. We detected an E2f6 cleavage product of expected size only in the presence of co-expressed miR-151-5p (Fig. 1h). We did not detect cleavage product when sh-151-3p overexpressing the miR-151 passenger strand (3p) was co-expressed with E2f6 or when sh-151-5p was co-expressed with ARHGDIA (Fig.1h), an established target of miR-151-5p but lacking extensive complementarity with miR-151-5p35. Together, these results demonstrate that miR-151-5p mediated cleavage of E2f6 mRNA is one of the rare instances in mammals where a gene is targeted for cleavage by a miRNA and that too in the absence of a seed match.

miR-151-3p binds to E2f6 3′UTR adjacent to miR-151-5p

During our studies, we noted that in humans, a putative binding site for the mature miR-151-3p arm with a miRNA-like seed region was only 9 nucleotides away from the site where the mature miR-151-5p is expected to bind (Fig. 2a). In mouse, a less canonical central base pairing seed region match, still known to induce non-cleavage repression30 was present (Fig. 2a). Interestingly, we found by reporter assay that the central 3p seed site conferred ~40% repression of E2f6 expression when we overexpressed miR-151-3p in MEF cells (Fig. 2b and Supplementary Fig.1a). Point mutations of this center seed region (mut 3p) reduced the repression potential of the 3p site, while removal of the full 3p binding site abolished the repression entirely (Fig. 2b). Increasing the complementarity of the seed region to include the more canonical 2nt – 8nt region of the mature miRNA (seed 3p) had no further effect on repression over the existing central seed match (Fig. 2b). Absence of Ago2 did not change the miR-151-3p mediated suppression of E2f6 substantially indicating a predominant non-cleavage based mechanism (Fig. 2c). This was also demonstrated by the 5′-RACE assay where we failed to observe RNA fragments diagnostic of miR-151-3p mediated cleavage of E2f6 (Fig. 1h). Thus, miR-151-3p can repress E2f6 gene expression through the canonical miRNA-repressive mechanism.

Figure 2.

Figure 2

Open in a new tab

miR-151-3p suppresses E2f6 expression by binding to E2f6 3′UTR adjacent to where miR-151-5p binds. (a) Schematic of a putative binding site of the miR-151-3p to the E2f6 3′UTR region adjacent to where 5p strand binds, in both mice and humans. The seed regions (nucleotides 2-8) are indicated for both the 5p and 3p arms. (b) Dual-luciferase reporter assay for the wildtype E2f6 3′UTR (wt) or other mutants (mut 3p and seed 3p as shown in the schematic) in presence of a miR-151-3p overexpression (sh-151-3p) or a scrambled control(sh-scr). A reporter construct with deletion of the entire miR-151-3p binding site in E2f6 3′UTR was also included. (c) Dual-luciferase reporter assay in wildtype MEF and Ago2−/− cells for E2f6 3′UTR in presence of sh-151-3p. For comparision, dual-luciferase reporter assay for E2f6 3′UTR in presence of sh-151-5p is also shown. (b,c) For reporter assays, normalization was done with respect to sh-scr. Error bars, s.e.m. (n = 2 biological replicates, each with 4 technical replicates).

Precursor miR-151 antagonizes mature miR-151 activity

The homology of both the pre-miR-151 region and the E2f6 3′UTR binding location is conserved specifically among placental mammals (Supplementary Fig. 2a). We questioned the biological importance served by having both 5p and 3p binding sites of a single miRNA at adjacent positions within a 3′UTR. Interestingly, in mouse, these sites overlap with the most favorable binding location of the miR-151 precursor in the 3′UTR of E2f6 [Minimum free energy of -61.3 Kcal/mol, calculated using RNAhybrid36]. Furthermore, thermodynamic considerations favor the binding of the miR-151 precursor to the E2f6 3′UTR region (Fig. 3a). Therefore, we pursued studies to determine if pre-miR-151 can bind and compete with mature miR-151 for the overlapping sites within the E2f6 3′UTR. Interestingly, pre-miR-151 is a substrate for RNA editing by ADAR1 whereby two adenosines are edited to inosines21 (Fig.3b). These two inosines result in an increase in the steady-state levels of pre-miR-15121 as well as an increase in pre-miR-151 affinity to the E2f6 3′UTR (Supplementary Fig. 2b).

Figure 3.

Figure 3

Open in a new tab

Precursor miR-151 competes with the mature miR-151-5p for binding to E2f6 3′UTR. (a) Thermodynamics of pre-miR-151 binding to E2f6. (b) Schematic of the stem-loop structure of the pre-miR-151 with the 5p arm (blue), 3p arm (purple) and two adenosines (green) substituted to guanosines (orange). (c) Northern analysis of miR-151 processing from pre-miR-151 overexpression plasmid (pEZX-151) or the double mutant form of pre-miR-151 (pEZX-DM). Let-7a serves as a loading control. (d) Dual-luciferase reporter assay for E2f6 3′UTR in presence of only the mature miR-151-5p (sh-miR-151-5p), or both the pre-miR-151 and mature miR-151-5p (pEZX-151 and pEZX-DM). (e) Schematic of the binding site of pre-miR-151 in E2f6 3′UTR and its modifications (E2f6 3p del and E2f6 3p-5p swap).(f) In-vitro gel shift assay with radiolabed (denoted by an asterisk) synthetic pre-miR-151(I) or a control pre-miR-122 and increasing molar concentrations of wildtype E2f6 3′UTR (1, 10 and 100 nM) or its modified forms. (g) Dual-luciferase analysis for E2f6 3′UTR (wt), 3p del or 3p-5p swap reporters with pEZX-151 or pEZX-DM. (h) In-vitro gel shift assay of E2f6 3′UTR bound to miR-151-5p with increasing molar concentrations of a synthetic pre-miR-151 or a control pre-miR-122. Bands below the blue star and orange star represent radiolabeled pre-miR-151 and pre-miR-122 respectively. “*” denotes radiolabeled oligos. (d, g) For reporter assays, normalization was done with respect to a scrambled control (sh-scr). Error bars, s.e.m. (n = 3 biological replicates, each with 3 technical replicates). *P = 0.05, **P = 0.01 by two-tailed Student′s t test.

To perform a competition assay between edited pre-miR-151 and mature miR-151 for the overlapping target, we first evaluated the effect of editing on precursor abundance and E2f6 expression in vivo. We used a Pol II based pEZX construct that contains the endogenous miR-151 stem and loop sequences, pEZX-151, and a modified double mutant version that incorporated two A to G point mutations, pEZX-DM (double mutant), at the same positions where A to I editing occurs in vivo21,22. Unlike the Pol III based sh-system, the Pol II based constructs pEZX-151 and pEZX-DM generated pre-miR-151 (~56 nt) (Fig. 3c) along with a higher molecular weight primary miR-151 or intermediate of the miR-151 biogenesis pathway (Supplementary Fig. 2c). pEZX-151 processed more efficiently to mature miR-151-5p and miR-151-3p compared to pEZX-DM (Fig. 3c). Reporter assays in MEF cells showed that expression from the pEZX-151 construct that resulted in pre-miR-151 accumulation had less E2f6 3′UTR suppressive activity (~40%) (Fig. 3d) compared to the sh-151-5p (~80%) construct (Fig. 3d). Moreover, expression from pEZX-DM that had a higher pre-to mature miR-151 ratio resulted in a drastically reduced repressive activity on the E2f6 3′UTR levels (~10%). This suggested that increasing the pre- to mature miR-151 ratio impeded the suppression ability of miR-151 on E2f6, thereby supporting the possibility of a scenario where pre-miR-151 binding to the E2f6 3′UTR was able to mask the MRE from the mature miR-151-5p. However, the differences in E2f6 3′UTR regulation could be explained in part by different levels of the mature miR-151-5p produced by the different constructs (Fig. 3c). This was especially the case when comparing the Pol II based constructs (pEZX -151 and pEZX-DM) and the Pol III based construct (sh-151-5p) because the latter produced more mature miR151-5p molecules. To rule this out and to further investigate if there was a competition between the precursor and mature miR-151-5p for their overlapping MRE sites, we modified the reporter containing E2f6 3′UTR sequence in two ways. The 3p binding site was removed in the E2f6 3′UTR reporter construct (3p del) so that no thermodynamic advantage was provided to the precursor relative to mature 5p for binding to the target (Fig. 3e). Moreover, the 3p and 5p binding sites were swapped in the reporter construct (3p-5p swap) reducing the potential of the precursor to bind linearly across the E2f6 3′UTR (Fig.3e). Indeed, in-vitro gel shift assays showed that the pre-miR-151 binds to the putative 3′UTR region of E2f6 in a dose-dependent manner (Fig. 3f) but there was no stable interaction between the precursor and the 3p-5p swap or the 3p del target (Fig. 3f). Dual-luciferase results showed that the target site 3p-5p swap or 3p del modifications had little effect on the sh-miR-151 expression systems that did not express the full miR-151 precursor (Supplementary Fig. 2d). However, when pre-miR-151 was present but unable to bind to the modified target sites as with the Pol II based pEZX system, both target site modifications led to significantly more repression of E2f6 compared to the wild type E2f6 3′UTR (wt) (Fig. 3g). Importantly, when relatively more pre-miR-151 accumulated (compare pEZX-DM to pEZX-151; Fig. 3c), E2f6 was suppressed to a greater extent in the modified target sites compared to the unmodified E2f6 3′UTR (Fig.3g). A miR-151 dose response study corroborated these findings (Supplementary Fig.2e). Addition of increasing amounts of sh-miR-151 consistently increased luciferase repression. However, constructs with an active precursor (either the pEZX-151 or pEZX-DM construct) reached steady state levels indicating buffering or competition of the precursor with the mature miRNA. (Supplementary Fig. 2e). In addition, when we designed an LNA/DNA mixmer blocker to weaken the binding of the pre-miR-151 to E2f6 3′UTR region and in turn disrupt the competition between pre-miR-151 and mature miR-151-5p, suppression of E2f6 expression by the pEZX-151 and pEZX-DM constructs was enhanced in presence of the blocker (Supplementary Fig. 2f). These studies do not rule out an alternative explanation that the enhanced pre-151 miRNA induced E2f6 suppression observed in the 3p del reporter construct (Fig. 3g) or by 3p-blocker induced blockade of the 3p binding site (Supplementary Fig. 2f) was the direct result of steric blockage of the 5p target site by its proximal 3p target site. We therefore asked if the presence of miR-151-3p binding site could interfere with E2f6 suppression by miR-151-5p. We used reporter assays with the construct containing the E2f6 3′UTR and provided an increasing ratio of sh-151-5p to sh-151-3p in comparison to the same increasing ratio of sh-151-5p to sh-scr. We observed a greater suppression of E2f6 by miR-151-5p binding in presence of miR-151-3p binding (Supplementary Fig. 3a). Consistent with this finding, we observed greater sh-151-5p directed suppression in the E2f6 3′UTR reporter construct compared to the 3p deletion construct that lacks the miR-151-3p binding site (Supplementary Fig. 3b). These results demonstrated that the miR-151-3p binding site in E2f6 3′UTR is not antagonistic or does not pose steric hindrance to RISC binding at the miR-151-5p binding site.

The above biological repression studies supporting the idea that the miR-151 precursor is able to compete with the mature form and impede suppression by mature miR-151 was further supported by in-vitro gel-shift competition assays (Fig. 3h). Increasing concentrations of a synthetic pre-miR-151 (I) competed with the binding of miR-151-5p to the MRE of E2f6 3′UTR (Fig. 3h). Furthermore, to validate the competition model inside cells, we carried out reporter assays in Dicer knockout cells to look at the effect of mature miR-151-5p alone or with increasing dosage of pre-miR-151 on E2f6 expression. Cells lacking Dicer are unable to process precursor miRNA which we confirmed by Northern blot analysis in Dicer knockout cells (Supplementary Fig. 2g). By increasing the dosage of the double mutant form of pre-miR-151 (pEZX-DM), we found that the suppression of E2f6 by a synthetic miR-151-5p duplex was partially rescued as predicted by the competition model (Supplementary Fig. 2h).

Pri or pre-miR-151 protects E2f6 transcripts in quiescent tissues

Competition between the precursor and mature forms of miR-151 requires the direct binding of the pre-miR-151 to the E2f6 3′UTR. To confirm such an interaction, we pursued two different studies. We first performed a biochemical pull-down assay using a synthetic and A-to-I substituted pre-miR-151 from HEK 293 cells co-transfected with the synthetic edited pre-miR-151 and a E2f6 expression construct and found an enrichment of the E2f6 3′UTR targets in the pulled down fraction. This result established that the pre-miR-151 was able to bind to the E2f6 3′UTR inside cells (Supplementary Fig. 4). The pull-down result was also consistent with our gel-shift assay finding as discussed above (Fig. 3f).

Our second approach was to corroborate the binding of the pre-miR-151 to endogenous E2f6 mRNA in vivo in a primary mammalian tissue. To do this, we performed a modified version of the Chromatin Isolation by RNA Purification (ChIRP) assay37 (outlined in Fig. 4a). Since pre-miR-151 accumulates in the mammalian brain,21 we added biotinylated tiling oligos complementary to the 3′UTR of E2f6, or lacZ mRNA (absent in vertebrates) as a control, to mouse brain homogenates. After streptavidin pull-down, we established that the E2f6 3′UTR tiling oligo-based pull-down samples were highly enriched for the E2f6 3′UTR RNA compared to the neuronal expressed Ctdnep1 mRNA, used as a control (Fig. 4b). We also observed a significant fold enrichment of pre-miR-151 compared to a control pre-miRNA (Fig. 4c) in the E2f6 pulled down samples. This strongly suggested a bonafide in vivo interaction between the pre-miR-151 and E2f6. Interestingly, we did not observe any enrichment of mature miR-151-5p in the E2f6 pulled down fraction (Fig. 4d), which was consistent with the cleavage-based degradation of E2f6 by miR-151-5p. Moreover, these results are also in concordance with the competition model whereby the pre-miR-151 masks the binding site of mature miR-151-5p.

Figure 4.

Figure 4

Open in a new tab

Pre-miR-151 binds to E2f6 in vivo and may protect the E2f6 transcript in quiescent tissues. (a) Schematic of the ChIRP method used to pull-down E2f6 mRNA from mouse brain. Quantitative PCR of (b) E2f6 mRNA and a control Ctdnep1 mRNA, (c) pre-miR-151 and a control pre-miR-124, (d) mature miR-151-5p and a control miR-124, pulled down by the biotinylated tiling oligonucleotides against the E2f6 3′UTR or a control lacZ mRNA. Quantitative PCR of (e) E2f6 mRNA, (f) pri-miR-151, (g) mature miR-151-5p in quiescent and non-quiescent tissues. In each case (e—g), the data are presented as fold induction after normalization to the liver sample (value = 1). (h) Northern analysis of miR-151 processing in various tissues. The blot on the left was probed with a LNA probe against mature miR-151-5p as shown by the schematic above the blot. The primary or intermediate product in the miR-151 biogenesis pathway is indicated by an arrowhead (→) and the mature miR-151-5p is indicated by a circle (○). U6 serves as a loading control. The blot on the right was probed (sequence is provided in Supplementary Table 5) for a region just outside the annotated stem loop structure of mmu-miR-151 (as shown by the schematic above the blot). (i) Quantitative PCR analyses of E2f6, pri-miR-151 and miR-151-5pduring differentiation of muscle cells (C2C12) (b— g, i) For qPCR data, error bars, s.e.m. (n = 2 biological replicates, each with 3 technical replicates).

To begin to unravel the biological importance of this interaction and the resulting unique regulatory circuit controlling expression of a gene important for controlling induction of cell-cycle exit and terminal differentiation38, we surveyed primary mouse tissues for the expression of primary and mature miR-151 and E2f6. We found endogenous expression of E2f6 and pri-miR-151 to be the highest in non-dividing tissues like brain, heart and skeletal muscle (Figs. 4e and 4f). However, the amount of mature miR-151-5p was either lower or similar to other tissues like the lung, spleen and liver, indicative of regulated processing of miR-151 between different tissues (Fig. 4g). Northern analysis of miR-151 processing (Fig. 4h) across these tissues suggests regulated processing of miR-151 and is consistent with the qPCR data as mentioned above. This suggests that the high amount of precursors in non-dividing tissues may help protect the E2f6 3′UTR from cleavage by mature miR-151-5p. To corroborate the correlation between the expression of E2f6 and pri-miR-151, we checked their expression profiles during skeletal muscle differentiation in C2C12 my oblast cells. Both pri-miR-151 and E2f6 expression increased with C2C12 differentiation in a temporal fashion while levels of the mature form of miR-151-5p remained unchanged (Fig. 4i). Based on our model, the high levels of E2f6 mRNA were consistent with the pri or pre-miRNA inhibition of mature miR-151-5p binding.

Pri or pre-miR-151 mediated gene regulation is target specific

While our studies implicated an antagonistic interaction of pre-miR-151 and mature miR-151 for E2f6, we investigated if miR-151 transcripts could have opposing effects on mRNAs that are bonafide targets of miR-151. RhoGDIA (ARHGDIA), a GDP/GTP exchange regulator of the Rho proteins that inhibits cell invasion and migration, is a direct target of miR-151-5p and exhibits an inverse and functional relationship with the miR-151-5p (but not the 3p) expression in hepato cellular carcinoma35. Our predictive thermodynamics analysis using RNA hybrid36 did not favor pre-miR-151 antagonizing miR-151-5p binding in the ARHGDIA mRNA (Fig. 5a) and we confirmed this experimentally using dual-luciferase assays. First, we observed no significant difference in the suppression of ARHGDIA expression with overexpression of either mature miR-151 (from the Pol III system, sh-151-5p) or pre-miR-151 (from the Pol II system, pEZX-151) in the reporter assay (Fig. 5c). Second, target site (ARHGDIA) deletion modifications similar to E2f6 where the pre-miR-151 would have no thermodynamic advantage over miR-151-5p for binding to the target MRE were constructed and tested (Fig. 5b). If there were a competition between the pre-miR-151 and mature miR-151 for an overlapping binding site in ARHGDIA 3′UTR, we would observe a drop in the luciferase activity upon overexpression of pre-miR-151 (pEZX-151) in the target site deletion construct (ARHGDIA 3p del) compared to wildtype construct (ARHGDIA). However, overexpression of pre-miR-151 (pEZX-151) resulted in similar levels of ARHGDIA suppression in both the wildtype and ARHGDIA 3p del target constructs (Fig. 5c). Therefore, unlike in E2f6, pre-miR-151 did not have an antagonistic interaction with miR-151-5p to regulate ARHGDIA expression.

Figure 5.

Figure 5

Open in a new tab

Regulation of miRNA activity by precursor miRNA is target specific.

(a) Thermodynamics of pre-miR-151 binding to ARHGDIA 3′UTR. (b) Schematic of the binding of the mature miR-151-5p and pre-miR-151 to the ARHGDIA 3′UTR and the modified target (ARHGDIA-del). (c) Dual-luciferase reporter assay for wildtype ARHGDIA 3′UTR or ARHGDIA-del construct in presence of sh-151-5p or pEZX-151. Error bars, s.e.m. (n = 3 biological replicates, each with 3 technical replicates).

Predicted widespread gene regulation by pre-miRNAs

Since the ratio of precursor to mature forms varies for many miRNAs during development and cancer globally, we next asked if other miRNA precursors might play a direct regulatory role in target gene expression similar to the pre-miR-151—E2f6 pair. To do so, we developed a bioinformatics pipeline where we combined miRNA target site predictions along with RNA folding analyses (see Supplementary Fig. 5 afora flowchart). Quite surprisingly, we found 138 miRNAs predicted to target 1607 mRNAs that are common to both mice and humans where the pre-miRNAs would be predicted to compete with the corresponding mature miRNAs for the same targets (Fig. 6a and Supplementary Table 1).

Figure 6.

Figure 6

Open in a new tab

Predicted widespread regulation of miRNA-mediated gene silencing by precursor miRNAs. (a) A graphical representation of the predicted miRNA:target pairs common to humans and mice on the basis of the favorable free energy change ΔΔG. (b) Thermodynamics of pre-miR-151 binding to SNAI2 3′UTR. The 5p or passenger strand (orange) and the 3p or guide strand (green) of pre-miR-124 are shown. (c) Dual-luciferase reporter assay for wildtype SNAI2 3′UTR, SNAI2-ran or SNAI2-swap constructs in presence of sh-124 or pEZX-124 constructs. Schematic of predicted binding of pre-miR-124 to either the wildtype SNAI2 3′UTR or its modified forms are depicted on right. Error bars, s.e.m. (n = 3 biological replicates, each with 3 technical replicates). *P = 0.05, ****P = 0.0001 by two-tailed Student′s t test. (d) A model showing competition between the pre-miRNAs and mature miRNAs for the same target. Precursor miRNAs are processed to mature miRNAs and loaded onto RISC resulting in suppression of a set of target genes. In situations where the steady-state levels of the same primary or precursor miRNAs remain high, they can regulate the suppression of other target genes by mature miRNAs in case the pre-miRNAs and their mature counterparts have overlapping and favorable binding sites on target genes.

A second example: the SNAI2 mRNA:miR-124 pair

To obtain functional support for an independent competitive pre-miRNA regulation of a miRNA target, we selected a miRNA:target pair with a high ΔΔG score: the miR-124:SNAI2 pair. This came from a subset of our predicted list of miRNAs and cognate targets, specifically from interactions that have been experimentally validated and curated in the miRTarBase database39(Supplementary Table 2). SNAI2 is a transcription factor overexpressed in various cancers40. Examination of the thermodynamic parameters showed that binding of pre-miR-124 to MRE of mature miR-124 in SNAI2 3′UTR is favorable (Fig. 6b). We realized that unlike the E2f6:miR-151 example, the SNAI2:miR-124 pair would not result in mRNA cleavage but rather the classical miRNA-mediated repression making the relative changes in expression less dramatic. Nevertheless, to validate the predicted competition between miR-124 and pre-miR-124 to target SNAI2 3′UTR, a reporter construct containing SNAI23′UTR was co-transfected with either the mature form of miR-124 (from a PolIII construct, sh-124) or the precursor form of miR-124 (from a PolII construct, pEZX-124). As expected, presence of pre-miR-124 (pEZX-124) resulted in lesser repression of the SNAI2 target than by overexpression of only mature miR-124 (sh-124) suggesting that pre-miR-124 limited the accessibility of mature miR-124 to its MRE in SNAI2 3′UTR (Fig. 6c). To validate the competition, we created reporter constructs similar to those generated for E2f6, with a modified target MRE that would not favor binding of precursor miRNA over the mature miRNA (Fig. 6c). Overexpression of miR-124 from pEZX-124 resulted in greater suppression in all of these modified constructs than the unmodified SNAI23′UTR containing reporter construct, SNAI2 (Fig. 6c) where as mature miR-124 (sh-124) resulted in similar levels of suppression in the unmodified and modified constructs (Supplementary Fig. 5b). These results along with the bioinformatics data suggest that this type of miRNA-mediated gene silencing might be applicable to a substantial number of endogenous miRNAs.

Discussion

We show that a miRNA precursor can physically and functionally block the binding of its mature counter part to a bona-fide target contained within an mRNA. A model is shown in Fig. 6d. In our study, we provide two examples: E2f6, a transcriptional repressor with recognized tissue-specific expression regulated by pre-miR-151, and SNAI2, a known oncogenic factor, by pre-miR-124. Our data also suggest that these miRNA precursors may include both pri- and pre- miRNAs as they share the same sequence required for target gene regulation. In addition to the two examples studied in detail, we developed a bioinformatics pipeline to include about 1600 targets common to both humans and mice that are predicted to be regulated by over 100 miRNA loci.

The miR-151 precursor RNA is derived from the intron of the FAK (or PTK2) gene that encodes a protein tyrosine kinase implicated in cell migration, invasion, adhesion41 and also in various types of cancer42. Our studies suggest that the PTK2 transcription unit may serve as a complex proto-oncogene and both the protein encoding protein tyrosine kinase and miR-151 transcripts will need to be studied in more detail to establish their individual and combined roles in the oncogenic process.

Highly repetitive elements in mammalian genomes can incorporate into non-coding regions of a protein-coding gene (e.g. introns and UTRs) in a relatively robust manner, and provide a rapid process to fine tune gene expression profiles that are necessary during speciation. The repeat-encoded miR-151 can target other single integrants, yet in situations where two copies have integrated, the precursor has the potential to also bind to this region. We found a large number of potential genes targeted by other repeat encoded miRNAs (e.g. miR-28 and miR-340) that could be involved in this type of regulatory process (Supplementary Table 3). However, because many miRNAs only result in small levels of individual mRNA-mediated repression (rather than the Ago2-mediated cleavage of E2f6 by miR-151), it will be more difficult to prove functionality for all cases using the methods employed here. Additionally, it is possible that some of these targets are in evolutionary limbo or indeterminate state showing signs of either positive or negative selection.

The expression of several miRNAs are lost in cancer and other pathological conditions. Impairment of miRNA biogenesis promotes oncogenic transformation and tumorigenesis through downregulation of tumor-suppressor miRNAs and subsequent upregulation of target oncogenes43. Many of the miRNAs lost in tumors and cancer are the result of reduced processing of the transcribed precursor11. Based on our model, the accelerated tumorigenic phenotype may result from both the reduction in expression of the mature tumor-suppressor miRNAs as well as a reduction in the activity of the remaining mature miRNAs because of competition with their respective precursors.

If taken all together, it is likely that precursor miRNAs competing with mature miRNAs and affecting their activity is not an uncommon regulatory event, which can more precisely define important regulatory roles in development and cancer. Although our work has uncovered a new mechanism of regulating miRNA activity and gene expression, further studies are required to underscore the physiological relevance of this model. This model offers additional layers of control over existing mechanisms like miRNA sponges or modifications in miRNA seed regions as the effect of miRNA precursors in gene regulation is target specific. Moreover, since the mature forms are generated from the intermediate precursors, this mechanism can be tightly regulated and can couple or decouple expression from activity of the miRNA resulting in differential regulation of mRNAs containing the same MRE target. Taken together, these types of scenarios may further explain why miRNA levels and target gene repression are not always correlated and suggests that more attention to the primary or precursor miRNAs will be required to better understand pathogenic processes in disease states. Our results challenge the dogma where miRNA precursors are considered as mere non-functional intermediates in the miRNA biogenesis pathway.

Methods

Cell culture and Plasmid constructs

MEF cells and Ago2−/− MEF cells were grown in Dulbecco′s modified Eagle′s media (DMEM; Life Technologies) supplemented with 10% heat-inactivated fetal bovine serum (FBS), 1% L-Glutamine, 1% Non-essential amino acids, 1% sodium pyruvate, 1% Pen-strep (Life Technologies). HEK 293, A549 and Huh7 cells were grown in DMEM with 10% FBS and 1% L-Glutamine. C2C12 mouse skeletal myoblasts were grown and maintained at subconfluent density in DMEM with 20% FBS (GM). To differentiate C2C12 myoblasts into myotubes, GM was replaced with DM (DMEM with 2% heat-inactivated horse serum).

For cloning of the E2f6 reporter construct, the 3′UTR of E2f6 was PCR amplified (see Supplementary Table 4 for the primers) from mouse genomic DNA and cloned downstream of the Renilla luciferase gene in psi-check-2 vector (Promega). This reporter vector contained nucleotides 240 to 1141 (after the stop codon) of the mouse 3′UTR sequence of E2f6 (NM_033270). Similarly, ARHGDIA was PCR amplified from human genomic DNA and cloned in the psi-check-2 vector containing a region of the 3′UTR as previously described35. For cloning of the psi-check constructs containing 3′UTR of SNAI2 and its derivatives (ran and swap), oligos (see Supplementary Table 5) purchased from IDT were chemically synthesized; both strands annealed and inserted downstream of the Renilla luciferase gene.

To generate the shRNA constructs for robust expression of mature miRNAs, oligos (see Supplementary Table 5) were chemically synthesized (IDT); both strands annealed and inserted between BglII and KpnI sites downstream of the U6 Pol III promoter. Pol II promoter (CMV) based constructs (pEZX-MR04) overexpressing precursor mmu-miR-151 and hsa-miR-124-3 were obtained from Gene Copoeia.

Plasmids expressing FLAG tagged human Argonautes (Ago2, Ago1) and EGFP were obtained from Add gene (plasmids 10822, 10820, 10825). The entire length of the E2f6 coding sequence region along with its 3′UTR was cloned in the Argonaute vector backbone between the NotI and EcoRI sites by In-fusion system (Clonetech) [see Supplementary Table 4 for the primers].

Site-directed mutagenesis was performed using QuikChange II XL Site-Directed Mutagenesis Kit (Stratagene) to create point substitutions, deletions or inversions at the putative miR-151 5p or 3p binding regions in the cloned E2f6 3′UTR or ARHGDIA 3′UTR region of the psiCHECK-E2f6 or construct (see Supplementary Table 4 for the primers).

RNA extraction and small RNA northern blots

Total RNA was isolated from MEF or HEK 293 cells using Trizol (Life technologies) 48 hrs after transfection. For C2C12 cells, RNA was Trizol extracted after each day of differentiation for 5 days. Expression of miR-133a, a marker of muscle differentiation44, was monitored to validate the differentiation to myotubes.

For small RNA northerns, ten micrograms of total RNA was electrophoresed on 15% (w/v) acrylamide/7M urea gel and transferred onto a Gene screen Plus (Perkin Elmer) membrane. Hybridization was done overnight in Perfect Hyb Plus (Sigma) buffer at 65°C using LNA probes for miR-151-5p and miR-124 (Exiqon) or at 35°C with an oligonucleotide probe (IDT).

Quantitative PCR

Two micrograms of total RNA were reverse-transcribed using superscript II RT kit (Life technologies) and subjected to gene expression analyses using gene specific Taqman probes (Mm01270320_m1 for E2f6 and Mm03306373_pri for pri-miR-151). For qPCR analysis of miR-151-5p and miR-133a, 100 nanograms of Trizol extracted RNA was reverse-transcribed using miRNA Reverse Transcription kit (Life technologies) followed by qPCR using miRNA Taqman probes (002642 from Life Technologies). To look for the endogenous E2f6 levels on overexpression of miR-151-5p by qPCR, MEF cells were co-transfected with GFP and Sh-151-5p and were subsequently FACS sorted for GFP. For the ChIRP experiment, the pulled down RNAs were reverse-transcribed using Thermoscript (Life technologies) and subjected to SYBR-Green based gene expression analyses (Qiagen) using primers specific for pre-miR-151, pre-miR-124 and U6 (see Supplementary Table 4 for sequences) as described previously45. Quantitative RT-PCR was performed on a CFX384 Real-Time system (BioRad). Fold-change was detected using delta-delta-cT calculations and normalized to mouse beta actin and U6 in gene expression and miRNA expression analyses respectively.

5′-RACE

Animal work was performed in accordance to the guidelines by Administrative Panel on Laboratory Animal Care (APLAC) at Stanford University. Total RNA from lungs of 8 week old B6 mice or from MEF cells co-transfected with Sh-scr, sh-151-5p or sh-151-3p and E2f6 overexpression construct was extracted and subjected to a modified 5′-RACE using the First Choice RLM-RACE kit (Life Technologies). Briefly, RNA (2 ug) was ligated to a synthetic RNA (5′-Adapter) and reverse transcribed using random primers. Next, the cDNA was PCR amplified using a 5′-adapter specific forward primer and a gene specific primer for E2f6 or ARHGDIA followed by a nested PCR with internal primers (see Supplementary Table 4 for the primers). PCR DNA was gel purified, cloned in TOPO vector (Life Technologies) and sequenced.

Reporter assays

Dual-luciferase assays (Promega) were performed 24 hours after transfection according to manufacturer′s protocol and detected by a Modulas Microplate Luminometer (Turner Biosystems). For transfection in a 24-well plate, 250 ng of psiCHECK reporter plasmids were co-transfected with 250 ng of miRNA overexpression plasmids (Sh-constructs or pEZX-constructs) in E10.5 MEF cells using Trans IT-LT1 (Mirus Bio) or in HEK 293 and Huh7 cells using Lipofectamine 2000 (Life Technologies). Cell seeding was performed at a concentration of 2.5 × 104 cells for MEF, 1.25 × 104 cells for Ago2−/− and 5 × 104 cells for HEK 293, A549, Dicer knockout cells and Huh7 per well in a 24-well plate. For reporter assay in Dicer knockout cells, the cells were co-transfected using Lipofectamine 3000 (Life Technologies) with a E2f63′UTR reporter construct (100 ng), 20 nM duplex synthetic miR-151-5p and increasing amounts of pEZX-DM starting from 50 ng. For the miR-151-5p inhibitor study, a miRVANA miRNA inhibitor against miR-151-5p was used (MH11537 from Life Technologies).

A fully phosphorothiolated 18 bp LNA/DNA mixmer (Exiqon) (sequence provided in Supplementary Table 5) containing 7 LNA bases predicted to block the binding of miR-151-3p to the target E2f6 3′UTR site without inducing RNAse H cleavage46 was cotransfected in MEF cells at 50 nM using Trans IT-TKO (Mirus Bio).

Western blotting

48 hrs post-transfection, HEK 293 cells were lysed using M-PER mammalian protein extraction reagent (Pierce) in presence of cOmplete Ultra protease inhibitors (Roche) and samples were run on a NuPAGE Novex 4-12% Bis-Tris polyacrylamide gel (Life Technologies). Western blotting was performed using PVDF membranes and E2f6 was detected with a rabbit polyclonal antibody (ab53061 from Abcam; validation is provided on the manufacturer’s website) titered at a 1:500 dilution. As a loading control, the blot was re-hybridized with an antibody against mouse beta actin.

In-vitro RNA binding assays

A 171 bp region of the E2f6 3′UTR containing the precursor miR-151 binding site was PCR amplified (see Supplementary Table 4 for primers) containing T7 promoter sequence and was in vitro transcribed using the In vitro transcription kit (Life technologies) to generate the target RNA. Other modifications in the E2f6 3′UTR target (3p-5p swap or 3p deletion) were obtained by PCR amplification of the corresponding mutated psiCHECK-E2f6 constructs and subsequent in vitro transcription (see Supplementary Table 4 for primers). The ligands [precursor miR-151(I), precursor miR-122, miR-151-5p] containing a 5′-phosphate group were chemically synthesized (Dharmacon). RNA-RNA interactions were analyzed by electrophoretic mobility shift assays. The synthetic RNA oligos were labeled at the 3′end by PCP end labeling. Binding reactions between the in vitro transcribed RNA targets and radiolabeled synthetic RNA ligands (0.1 nM) were carried out in buffer containing 5 mM HEPES (pH 7.5), 100 mM KCl, 5 mM MgCl2, 0.1 ug/ul tRNA, 3.8 % glycerol at 85°C for 5 min followed by slow cooling and loading on 5% native polyacrylamide gel. For the competition assay, the binding between the in vitro transcribed E2f6 3′UTR and the synthetic miR-151-5p was competed with increasing molar concentration of 1, 10 and 100 nM of a cold synthetic pre-miR-151.

Biotin Pull-down assay

A synthetic pre-miR-151 RNA oligo modified at the 5′ end with a phosphate moiety and biotinylated at the 3′ end was obtained from Dharmacon. An internal uridine (at position 42) was substituted with a photoreactive 4-thio Uridine (4-tU) to confer site-specific pre-miRNA to mRNA crosslinking upon long wave UV irradiation. Pull-down was performed in HEK 293 cells co-transfected with the synthetic RNA oligo and a E2f6 plasmid essentially as described earlier47. GAPDH expression was used as a control for enrichment in the pull-down samples. A second normalization was performed with respect to a non-biotinylated and non 4-tU containing pre-miR-151 RNA oligo.

ChIRP assay

The ChIRP assay was performed as described earlier37 with slight modifications. Briefly, adult mouse brains (C57BL/6, female, 6-8 weeks old) were snap-frozen, ground, cross-linked by 3.7% formaldehyde and sonicated using a Covaris S2 sonicator (Covaris Inc., Woburn, MA, USA) with the following settings: 10% Duty Cycle; Intensity 4; 200 cycles per burst for 4 minutes at 4 °C. The cleared lysate was then hybridized with 12 biotinylated tiling oligos (see Supplementary Table 6 for sequences) complimentary to the 1.2 kb long 3′UTR of E2f6 to specifically pull down E2f6. Biotinylated tiling oligos complimentary to lacZ were used as a negative control for pulldown37. As a control for non-specific mRNA pulldown, the neuronal Dullard (Ctdnep1) gene was used. The brain-enriched miR-124 was used for non-specific miRNA and pre-miRNA controls. U6 RNA levels in each fraction was determined as a normalization control for mRNA/pre-miRNA and miRNA based- qPCR, respectively. Finally, the values were normalized against their respective pre-pulldown input.

Statistical Information

All data are presented as mean ± s.e.m. Data were analyzed using two-tailed Student′s t test using PRISM 6.0. A P value of 0.05 or lower was considered significant.

Uncropped blot and gel

Original images of Western blot (Fig. 1d) and agarose gel (Fig. 1h) can be found in Supplementary Fig. 6.

Supplementary Material

1
2
3

Acknowledgments

We thank members of the Kay laboratory, in particular, L. Lisowski for help with FACS sorting, F. Zhang for mouse dissection and Dicer knockout cell culture, L. Sobkowiak for RNA extraction from mouse tissues, and S. Gu and H.K. Kim for critical comments and helpful discussion. We also thank J. Sage and T. Rando (Stanford University) for providing reagents and A. Lund (University of Copenhagen) for advice on the biotinylated pre-miR-151 pull-down assay. We obtained the catalytically inactive form of Ago2 (Ago2-D597A) as a kind gift from G. Hannon (Cold Spring Harbor Laboratory). This work was supported by US National Institutes of Health(grants DK078424 to M.A.K.). P.N.V. is supported by a Banting Postdoctoral Fellowship from the Canadian Institutes of Health Research.

Footnotes

Author Contributions

B.R-C, P.N.V., M.A.K studied design. B.R-C. performed the experiments with the exception of the ChIRP experiments which were performed by Q.W. and Q.L. Y.Z. performed bioinformatic predictions; B.R-C. P.N.V. and M.A.K. analyzed the results, discussed and wrote the paper.

Competing financial interests

The authors declare no competing financial interests.

References

  • 1.Bushati N, Cohen SM. microRNA functions. Annual review of cell and developmental biology. 2007;23:175–205. doi: 10.1146/annurev.cellbio.23.090506.123406. [DOI] [PubMed] [Google Scholar]
  • 2.Huntzinger E, Izaurralde E. Gene silencing by microRNAs: contributions of translational repression and mRNA decay. Nature reviews Genetics. 2011;12:99–110. doi: 10.1038/nrg2936. [DOI] [PubMed] [Google Scholar]
  • 3.Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell. 2004;116:281–97. doi: 10.1016/s0092-8674(04)00045-5. [DOI] [PubMed] [Google Scholar]
  • 4.Cullen BR. Transcription and processing of human microRNA precursors. Molecular cell. 2004;16:861–5. doi: 10.1016/j.molcel.2004.12.002. [DOI] [PubMed] [Google Scholar]
  • 5.Trujillo RD, Yue SB, Tang Y, O′Gorman WE, Chen CZ. The potential functions of primary microRNAs in target recognition and repression. The EMBO journal. 2010;29:3272–85. doi: 10.1038/emboj.2010.208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Okamura K, Ladewig E, Zhou L, Lai EC. Functional small RNAs are generated from select miRNA hairpin loops in flies and mammals. Genes & development. 2013;27:778–92. doi: 10.1101/gad.211698.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Obernosterer G, Leuschner PJ, Alenius M, Martinez J. Post-transcriptional regulation of microRNA expression. RNA. 2006;12:1161–7. doi: 10.1261/rna.2322506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Roush S, Slack FJ. The let-7 family of microRNAs. Trends in cell biology. 2008;18:505–16. doi: 10.1016/j.tcb.2008.07.007. [DOI] [PubMed] [Google Scholar]
  • 9.Thomson JM, et al. Extensive post-transcriptional regulation of microRNAs and its implications for cancer. Genes & development. 2006;20:2202–7. doi: 10.1101/gad.1444406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Calin GA, et al. Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proceedings of the National Academy of Sciences of the United States of America. 2002;99:15524–9. doi: 10.1073/pnas.242606799. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Lee EJ, et al. Systematic evaluation of microRNA processing patterns in tissues, cell lines, and tumors. RNA. 2008;14:35–42. doi: 10.1261/rna.804508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.De Vito C, et al. A TARBP2-dependent miRNA expression profile underlies cancer stem cell properties and provides candidate therapeutic reagents in Ewing sarcoma. Cancer cell. 2012;21:807–21. doi: 10.1016/j.ccr.2012.04.023. [DOI] [PubMed] [Google Scholar]
  • 13.Davis BN, Hata A. Regulation of MicroRNA Biogenesis: A miRiad of mechanisms. Cell communication and signaling : CCS. 2009;7:18. doi: 10.1186/1478-811X-7-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Treiber T, Treiber N, Meister G. Regulation of microRNA biogenesis and function. Thrombosis and haemostasis. 2012;107:605–10. doi: 10.1160/TH11-12-0836. [DOI] [PubMed] [Google Scholar]
  • 15.Slezak-Prochazka I, Durmus S, Kroesen BJ, van den Berg A. MicroRNAs, macrocontrol: regulation of miRNA processing. RNA. 2010;16:1087–95. doi: 10.1261/rna.1804410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Newman MA, Hammond SM. Emerging paradigms of regulated microRNA processing. Genes & development. 2010;24:1086–92. doi: 10.1101/gad.1919710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Bass BL. RNA editing by adenosine deaminases that act on RNA. Annual review of biochemistry. 2002;71:817–46. doi: 10.1146/annurev.biochem.71.110601.135501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Heale BS, Keegan LP, O′Connell MA. The effect of RNA editing and ADARs on miRNA biogenesis and function. Advances in experimental medicine and biology. 2010;700:76–84. [PubMed] [Google Scholar]
  • 19.Yang W, et al. Modulation of microRNA processing and expression through RNA editing by ADAR deaminases. Nature structural & molecular biology. 2006;13:13–21. doi: 10.1038/nsmb1041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Iizasa H, et al. Editing of Epstein-Barr virus-encoded BART6 microRNAs controls their dicer targeting and consequently affects viral latency. The Journal of biological chemistry. 2010;285:33358–70. doi: 10.1074/jbc.M110.138362. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kawahara Y, Zinshteyn B, Chendrimada TP, Shiekhattar R, Nishikura K. RNA editing of the microRNA-151 precursor blocks cleavage by the Dicer-TRBP complex. EMBO reports. 2007;8:763–9. doi: 10.1038/sj.embor.7401011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Kawahara Y, et al. Redirection of silencing targets by adenosine-to-inosine editing of miRNAs. Science. 2007;315:1137–40. doi: 10.1126/science.1138050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Memczak S, et al. Circular RNAs are a large class of animal RNAs with regulatory potency. Nature. 2013;495:333–8. doi: 10.1038/nature11928. [DOI] [PubMed] [Google Scholar]
  • 24.Poliseno L, et al. A coding-independent function of gene and pseudogene mRNAs regulates tumour biology. Nature. 2010;465:1033–8. doi: 10.1038/nature09144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Tay Y, et al. Coding-independent regulation of the tumor suppressor PTEN by competing endogenous mRNAs. Cell. 2011;147:344–57. doi: 10.1016/j.cell.2011.09.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Cesana M, et al. A long noncoding RNA controls muscle differentiation by functioning as a competing endogenous RNA. Cell. 2011;147:358–69. doi: 10.1016/j.cell.2011.09.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Franco-Zorrilla JM, et al. Target mimicry provides a new mechanism for regulation of microRNA activity. Nature genetics. 2007;39:1033–7. doi: 10.1038/ng2079. [DOI] [PubMed] [Google Scholar]
  • 28.Hansen TB, et al. Natural RNA circles function as efficient microRNA sponges. Nature. 2013;495:384–8. doi: 10.1038/nature11993. [DOI] [PubMed] [Google Scholar]
  • 29.Smalheiser NR, Torvik VI. Mammalian microRNAs derived from genomic repeats. Trends in genetics : TIG. 2005;21:322–6. doi: 10.1016/j.tig.2005.04.008. [DOI] [PubMed] [Google Scholar]
  • 30.Shin C, et al. Expanding the microRNA targeting code: functional sites with centered pairing. Mol Cell. 2010;38:789–802. doi: 10.1016/j.molcel.2010.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Cartwright P, Muller H, Wagener C, Holm K, Helin K. E2F-6: a novel member of the E2F family is an inhibitor of E2F-dependent transcription. Oncogene. 1998;17:611–23. doi: 10.1038/sj.onc.1201975. [DOI] [PubMed] [Google Scholar]
  • 32.Gu S, Jin L, Zhang F, Sarnow P, Kay MA. Biological basis for restriction of microRNA targets to the 3′ untranslated region in mammalian mRNAs. Nat Struct Mol Biol. 2009;16:144–50. doi: 10.1038/nsmb.1552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Liu J, et al. Argonaute2 is the catalytic engine of mammalian RNAi. Science. 2004;305:1437–41. doi: 10.1126/science.1102513. [DOI] [PubMed] [Google Scholar]
  • 34.Diederichs S, Haber DA. Dual role for argonautes in microRNA processing and posttranscriptional regulation of microRNA expression. Cell. 2007;131:1097–108. doi: 10.1016/j.cell.2007.10.032. [DOI] [PubMed] [Google Scholar]
  • 35.Ding J, et al. Gain of miR-151 on chromosome 8q24.3 facilitates tumour cell migration and spreading through downregulating RhoGDIA. Nature cell biology. 2010;12:390–9. doi: 10.1038/ncb2039. [DOI] [PubMed] [Google Scholar]
  • 36.Rehmsmeier M, Steffen P, Hochsmann M, Giegerich R. Fast and effective prediction of microRNA/target duplexes. RNA. 2004;10:1507–17. doi: 10.1261/rna.5248604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Chu C, Qu K, Zhong FL, Artandi SE, Chang HY. Genomic maps of long noncoding RNA occupancy reveal principles of RNA-chromatin interactions. Molecular cell. 2011;44:667–78. doi: 10.1016/j.molcel.2011.08.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Giangrande PH, et al. A role for E2F6 in distinguishing G1/S- and G2/M-specific transcription. Genes & development. 2004;18:2941–51. doi: 10.1101/gad.1239304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Hsu SD, et al. miRTarBase: a database curates experimentally validated microRNA-target interactions. Nucleic acids research. 2011;39:D163–9. doi: 10.1093/nar/gkq1107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Alves CC, Carneiro F, Hoefler H, Becker KF. Role of the epithelial-mesenchymal transition regulator Slug in primary human cancers. Frontiers in bioscience : a journal and virtual library. 2009;14:3035–50. doi: 10.2741/3433. [DOI] [PubMed] [Google Scholar]
  • 41.Peng X, Guan JL. Focal adhesion kinase: from in vitro studies to functional analyses in vivo. Current protein & peptide science. 2011;12:52–67. doi: 10.2174/138920311795659452. [DOI] [PubMed] [Google Scholar]
  • 42.McLean GW, et al. The role of focal-adhesion kinase in cancer - a new therapeutic opportunity. Nature reviews Cancer. 2005;5:505–15. doi: 10.1038/nrc1647. [DOI] [PubMed] [Google Scholar]
  • 43.Kumar MS, Lu J, Mercer KL, Golub TR, Jacks T. Impaired microRNA processing enhances cellular transformation and tumorigenesis. Nature genetics. 2007;39:673–7. doi: 10.1038/ng2003. [DOI] [PubMed] [Google Scholar]
  • 44.Chen JF, et al. The role of microRNA-1 and microRNA-133 in skeletal muscle proliferation and differentiation. Nature genetics. 2006;38:228–33. doi: 10.1038/ng1725. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Schmittgen TD, Jiang J, Liu Q, Yang L. A high-throughput method to monitor the expression of microRNA precursors. Nucleic acids research. 2004;32:e43. doi: 10.1093/nar/gnh040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Kurreck J, Wyszko E, Gillen C, Erdmann VA. Design of antisense oligonucleotides stabilized by locked nucleic acids. Nucleic acids research. 2002;30:1911–8. doi: 10.1093/nar/30.9.1911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Christoffersen NR, et al. p53-independent upregulation of miR-34a during oncogene-induced senescence represses MYC. Cell death and differentiation. 2010;17:236–45. doi: 10.1038/cdd.2009.109. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS610614-supplement-1.doc (35KB, doc)
2
NIHMS610614-supplement-2.doc (3.7MB, doc)
3
NIHMS610614-supplement-3.pdf (113KB, pdf)

RESOURCES