Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Nov 13, 2007; 104(46): 18151–18156.
Published online Nov 7, 2007. doi:  10.1073/pnas.0706673104
PMCID: PMC2084312

Functional screening identifies miR-315 as a potent activator of Wingless signaling


The existence of vast regulatory networks mediated by microRNAs (miRNAs) suggests broad potential for miRNA dysfunction to contribute to disease. However, relatively few miRNA–target interactions are likely to make detectable contributions to phenotype, and effective strategies to identify these few interactions are currently wanting. We hypothesized that signaling cascades represent critical points of susceptibility to miRNA dysfunction, and we developed a strategy to test this theory by using quantitative cell-based screens. Here we report a screen for miRNAs that affect the Wingless (Wg) pathway, a conserved pathway that regulates growth and tissue specification. This process identified ectopic miR-315 as a potent and specific activator of Wg signaling, an activity that we corroborated in transgenic animals. This miR-315 activity was mediated by direct inhibition of Axin and Notum, which encode essential, negatively acting components of the Wg pathway. Genetic interaction tests substantiated both of these genes as key functional targets of miR-315. The ability of ectopic miR-315 to activate Wg signaling was not a trivial consequence of predicted miRNA–target relationships because other miRNAs with conserved sites in the Axin 3′ UTR neither activated Wg outputs nor inhibited an Axin sensor. In summary, activity-based screening can selectively identify miRNAs whose deregulation can lead to interpretable phenotypic consequences.

Keywords: microRNA, signal transduction

MicroRNAs (miRNAs), a large class of ≈22-nt noncoding RNA, mediate extensive gene-regulatory networks whose functional implications are only beginning to be understood (1). Computational studies indicate that a substantial proportion of cellular transcripts are directly regulated by miRNAs (2). Thus, almost any biological process of choice may be plausibly considered to be under miRNA-mediated control. However, few miRNAs are functionally well understood at present. Because individual miRNAs are often predicted to have hundreds of potential targets, it is difficult to decide a priori that any given miRNA–target interaction is particularly meaningful.

The need to identify phenotypically relevant activities and targets of animal miRNAs stems from the finding that miRNA dysfunction is causal to disease and oncogenesis. For example, let-7 family members have been suggested to serve as tumor suppressors by directly inhibiting the Hmg2A and RAS protooncogenes (36). Conversely, the related miR-372/373 miRNAs promote tumorigenesis in combination with oncogenic RAS, at least in part by directly inhibiting the tumor suppressor LATS2 (7). Likewise, the miR-17–92 cluster (8, 9) synergizes with MYC to induce B cell lymphoma.

Because many diseases and cancers are due to the perturbation of dose-sensitive signal-transduction cascades that control cell growth and differentiation, we hypothesized that such pathways provide fertile ground for mining the disease-relevant activities of miRNAs. Indeed, a growing body of work demonstrates biologically important roles for specific miRNAs in regulating the major signaling cascades. For example, aspects of Hedgehog signaling are regulated by miR-214 in zebrafish (10), whereas many components of the Drosophila Notch pathway are regulated by Brd box-, GY box-, and/or K box-family miRNAs (1113). Therefore, we developed a functional screening approach to examine the ability of miRNAs to modulate the transcriptional outputs of signal-transduction cascades. In this report, we used the Wingless (Wg)–Wnt pathway as a testbed for our approach.

The Drosophila morphogen Wg and its vertebrate homologs (Wnts) coordinate a conserved signaling system that directs cell specification, tissue patterning, and cell proliferation. Because precise levels of the Wg–Wnt pathway output are essential for appropriate biological outcomes, net pathway output is carefully balanced by the interplay of positive and negative factors. In fact, Wg–Wnt signaling is modulated at almost every conceivable level, from transcriptional regulation to posttranslational modifications, including lipidation, glycosylation, phosphorylation, and ubiquitination (14). The need to maintain tight control over this pathway is reflected by the fact that inappropriate Wnt pathway activity underlies developmental disorder and disease, including liver, colorectal, breast, and skin cancer (14).

We generated a library of miRNA expression constructs and analyzed their effects on a quantitative Wg reporter assay in Drosophila cells (15). Our approach identified miR-315 as a potent activator of Wg signaling in cultured cells, an activity that we confirmed in transgenic animals. We determined that miR-315 activates Wg signaling by independently repressing two negative regulators of Wg signaling, Axin and Notum. Notably, a specific connection between miR-315 and Wg signaling could not have been anticipated from the outset by inspection of the >400 predicted, conserved targets of miR-315. Our validated strategy for investigating the influence of miRNA gain of function on signaling pathways suggests that this approach efficiently elucidates disease-relevant activities of human miRNAs.


A Plasmid-Based miRNA Expression Library for Use in Cell-Based Assays.

Although bioinformatics has yielded important insights into miRNA biology, miRNA target lists alone are of limited utility in assigning miRNA function. This finding is partly because of the large number of targets predicted for most miRNAs (often 50–200). For example, in the Wg pathway alone, ≈60 miRNA–mRNA target predictions are made by various algorithms (1618) [supporting information (SI) Table 1]. Additional concerns surround the functional significance of many computationally predicted miRNA–target interactions, as well as the extent to which noncanonical targets might be missed by the major computational approaches (2, 19, 20).

To move away from target-motivated hypotheses in favor of activity-based tests of miRNA function, we took advantage of a robust transcriptional reporter for Wg signaling in Drosophila clone 8 cells (15). The Wg signal is transduced by the T cell factor (TCF) family of transcription factors whose activation can be quantified by a firefly luciferase reporter linked to multiple TCF-binding sites (TCF-luc) (21). Although this reporter does not capture all aspects of Wg signaling in vivo (22), TCF-luc is nonetheless a useful monitor of experimental perturbations to Wg signal transduction in cultured cells.

To examine the effect of ectopic miRNAs on TCF-luc, we chose a pri-miRNA expression strategy previously shown to produce active miRNAs in transgenic animals (23). We cloned 400-to 500-nt pri-miRNA fragments centered on the miRNA hairpin into the 3′ UTR of pUAST or UAS-DsRed vectors. This library includes 77/78 miRNA loci currently deposited in miRBase (24), with some miRNAs represented as both single constructs and members of multigene operons (see SI Table 2 for a list of the 75 miRNA expression constructs). Analysis of many such constructs in vivo indicates that the Gal4>UAS-miRNA strategy does not saturate the endogenous miRNA biogenesis or regulatory pathways (13, 23). miRNA expression was induced by cotransfection of a plasmid encoding constitutively expressed Gal4 (ub-Gal4), and Wg pathway activity was measured as the ratio of TCF-luc to a pol III-Renilla luciferase control. Because Wg signaling is actively repressed in unstimulated cells, we performed these assays in naive clone 8 cells, as well as in cells transfected with a Wg expression construct (pAc-Wg) (15).

Expression of miR-315 Strongly Stimulates Basal Wg Pathway Activity.

Most miRNAs failed to alter TCF-luc activity >2-fold in either resting or Wg-stimulated conditions (Fig. 1 and SI Table 2). Of those miRNAs that did, miR-315 caused the most striking change in Wg pathway activity, inducing a 9-fold increase in TCF-luc in the absence of Wg (Fig. 1). Interestingly, although miR-315 caused the third highest increase in TCF-luc activity in the presence of Wg, its effect there was comparatively modest (2-fold induction). Therefore, miR-315 predominantly activated Wg signaling in otherwise unstimulated clone 8 cells.

Fig. 1.
Genome-wide cell-based screen for miRNAs identifies miR-315 as an activator of Wg signaling in unstimulated cells. Naive (−) cells or Wg-expressing (+) clone 8 cells were transfected with individual UAS-DsRed-miRNA expression plasmids, ub-Gal4 ...

To determine where in the Wg pathway miR-315 acts, we performed epistasis tests that asked whether miR-315 could activate TCF-luc after dsRNA-mediated knockdown of various positive components of the Wg pathway. miR-315 induced TCF-luc independently of the Wg ligand, the frizzled receptor, or the LRP/arrow coreceptor, but required the function of the nuclear factors legless/BCL9, pangolin/dTCF, and armadillo/β-catenin (Fig. 2A). These data placed miR-315 activity in cultured cells downstream of the plasma membrane-localized Wg receptors, but upstream of the nuclear components of this signaling pathway.

Fig. 2.
Placement and specificity of miR-315 activity relative to the Wg pathway. (A) Epistasis test. The effect of ectopic miR-315 on TCF-luc was determined in the presence of dsRNAs against various positive components of the canonical Wg signaling pathway. ...

We assessed the specificity of miR-315 activity by using transcriptional reporters for other signaling pathways. We found that miR-315 did not affect a reporter for JAK-STAT signaling and had no effect on a reporter for Hedgehog signaling under either basal or stimulated conditions (Fig. 2B). These data, in combination with the epistasis test with Wg pathway nuclear factors, exclude that ectopic miR-315 operates nonspecifically by some general inhibitor of the basal transcriptional machinery. Finally, we attempted to perform reciprocal experiments to knockdown miR-315 by using 2′-O-methylated antisense oligonucleotides (25), but observed no effect on either Wg (−) or Wg (+) conditions. Subsequent miRNA microarray studies showed that miR-315 is not expressed by any commonly used Drosophila tissue culture cells (clone 8, S2, or Kc cells) (data not shown). Therefore, we could not perform complementary inhibitor studies. Instead, we turned to in vivo animal assays to further probe the functional relationship between miR-315 and the Wg pathway.

miR-315 Activates Wg Signaling in Transgenic Animals.

The induction of a Wg-responsive reporter in cultured cells by miR-315 does not necessarily mean that it can affect Wg signaling in an interpretable manner in the animal. For example, miR-315 might regulate other targets that could together induce confounding phenotypes. Indeed, with hundreds of targets predicted for individual miRNAs (including miR-315), computational studies do not readily predict interpretable consequences for deregulating individual miRNAs. Therefore, we analyzed the consequences of ectopic mir-315 on Drosophila development by using stably integrated UAS-DsRed-miR-315 transgenes.

Patterning of the wing imaginal disk is a well established model system for studying many fundamental cell-signaling pathways, including the Wg pathway. Each of the two wing imaginal discs gives rise to one half of the notum, or back of the fly, and to one of the adult wings. The domain that gives rise to the wing, or wing pouch, can be visualized by expression of Nubbin (Fig. 3A, blue cells). This domain is bisected by a line of Cut-expressing cells termed the “wing margin.” These cells induce the expression of Senseless (Sens) in two flanking rows of cells that include the precursors for sensory organs of the anterior wing margin (Fig. 3A, green cells). During early larval development, Wg signaling antagonizes EGF receptor signaling to specify identity of the wing pouch versus the notum. As a consequence, loss of Wg causes the conversion of wing into notum, whereas a gain of Wg induces the reciprocal conversion of notum into wing (26, 27). Wg also functions later in disk development to specify wing margin.

Fig. 3.
miR-315 is a potent activator of Wg signaling in vivo. (A) Wild-type third instar wing imaginal disk expressing DsRed under the control of ptc-Gal4. W, presumptive wing; n, presumptive notum. Disk was stained for Nubbin (in blue) to mark the wing pouch ...

Misexpression of miR-315 in the wing disk by using the ptc-Gal4 driver, which is active along the anterior–posterior compartment boundary, resulted in a dramatic double-wing pouch disk (Fig. 3B). The ectopic wing pouch expressed Nubbin and displayed features of a normally differentiating wing, including spatially appropriate expression of Cut, Wg, and Sens (Fig. 3B) (data not shown). The respecification of notum as wing induced by ectopic miR-315 was quite similar to the effect of misexpressing Wg (Fig. 3C) (27), although it did not induce ectopic wing margin within the wing pouch proper, as seen with ectopic Wg (Fig. 3C, arrowhead).

Most ptc-Gal4, UAS-DsRed-miR-315 animals died during early pupal development, but culturing at 18°C to minimize Gal4 activity permitted some to survive to pharate stages. Dissection of these animals from their pupal cases revealed transformations of scutellar tissue into ectopic wings (Fig. 3 E and F), which is further evidence for the activation of Wg signaling by ectopic miR-315 in vivo.

miR-315 Directly Targets Axin and Notum, Essential Negative Regulators of the Wg Pathway.

Because miR-315 genetically activates Wg signaling and miRNAs generally function as negative regulators, we hypothesized that miR-315 might target one or more inhibitors of the Wg pathway. In searching for such targets, we paid particular attention to targets containing multiple, canonical, conserved seed matches in their 3′ UTRs (i.e., that display continuous Watson–Crick matches to positions 2–8 from the 5′ end of miR-315). Interestingly, two critical negative regulators of the Wg pathway satisfy these criteria, namely Axin and Notum (Fig. 4A) (16, 17). These candidate targets contain a pair of miR-315 seed matches. One of the Notum sites and both of the Axin sites have been nearly perfectly conserved in all sequenced Drosophilids (Fig. 4A and SI Fig. 5). We also note that both Axin sites exhibit t1A features, which have been suggested to increase target activity (28). The sequenced Drosophilids preserve t1A features in both Axin sites.

Fig. 4.
miR-315 directly targets two negative regulators of the Wg pathway. (A) Axin and Notum each contain a pair of conserved seed matches to miR-315 (AAUCAAA, green shading). Both Axin sites are of the t1A variety, and all four sites exhibit t9W (i.e., A or ...

Axin plays an integral role in the destruction of β-catenin in the absence of Wg signaling (29), whereas Notum (also known as Wingful) is a member of the α/β-hydrolase superfamily that negatively regulates Wg signal reception by modifying the cell-surface proteoglycans, Dally and Dlp (30, 31). To ask whether miR-315 could directly repress Axin or Notum by the candidate-binding sites present in their 3′ UTRs, we constructed heterologous reporters (sensors) consisting of the Renilla luciferase gene fused to the entire Axin or Notum 3′ UTRs. Coexpression with miR-315 resulted in >2-fold repression of both Axin and Notum sensors (Fig. 4 B and C). We also assayed truncated sensors consisting of 3′ UTR segments bounded by the miR-315 sites in each target and found that they recapitulated the quantitative repression by miR-315 mediated by their full-length counterparts. Finally, we introduced two point mutations into each of the miR-315 seed matches of the Axin and Notum sensors. These specific mutations abolished the response of these sensors to miR-315 (Fig. 4 B and C). Therefore, miR-315 directly represses Axin and Notum by the identified seed-match target sites.

As a more biologically relevant test, we assayed the ability of miR-315 to inhibit transgenic Axin and Notum sensors consisting of tubulin-GFP fused to their 3′ UTRs. Expression of the GFP sensor was monitored in wing discs that express UAS-DsRed-miR-315 under the control of ptc-Gal4. As a control, we showed that expression of miR-315 had no effect on the activity of a tub-GFP-miR-7 transgene (Fig. 4 E and F) previously shown to be strongly inhibited by ectopic miR-7 (23). In contrast, both GFP-Axin and GFP-Notum sensors exhibited strong, cell-autonomous inhibition in DsRed/miR-315-positive cells (Fig. 4 G–J). These transgenic tests provide compelling evidence that both target genes can be directly and potently repressed by miR-315 in the animal. We also note that the Axin sensor was repressed more strongly than the Notum sensor, correlating with the t1A features in both Axin target sites.

Notum and Axin Are both Critical Functional Targets of miR-315 with Respect to Wg Signaling.

The sensor data validated the ability of miR-315 to repress Notum and Axin, but did not directly assess how their repression contributes to miR-315-induced phenotypes. Therefore, we asked whether the effects of ectopic miR-315 could be suppressed by coactivation of its targets. Activation of endogenous Notum (by UAS sites integrated into this locus in the EP-Notum allele) by using ptc-Gal4 resulted in complete embryonic lethality (0% hatching of >1,000 embryos). Coexpression with miR-315 suppressed the embryonic lethality of ptc>Notum animals, functional evidence that miR-315 can directly inhibit endogenously expressed Notum transcripts. Wing discs derived from ptc>miR-315+Notum animals also exhibited significant suppression of the notum-to-wing transformation normally induced by miR-315 (Fig. 3D). Further, 100% of ptc>miR-315 discs (n = 11) exhibited a clearly duplicated wing pouch with a Nubbin+ domain at least half the size of the normal wing pouch (and typically approaching the size of the endogenous wing pouch). In contrast, coexpression of miR-315 and Notum resulted in only 50% (n = 12) of discs with evidence of a transformed wing pouch with ectopic Nubbin staining. Therefore, ectopic Notum can reciprocally rescue a mutant phenotype induced by ectopic miR-315.

We also tested for genetic suppression by Axin. When expressed by using ptc-Gal4, UAS-Axin-GFP induced embryo lethality in the absence or presence of coexpressed miR-315, so we could not examine wing discs in this genotype. We developed another assay in which expression of miR-315 by using bx-Gal4 induced male lethality (2.1% survival). In this background, coexpression of Axin-GFP produced 29.2% survival (Fig. 3G). Coexpression of Notum also provided strong rescue, resulting in 41.1% survival. Therefore, lethal activation of Wg signaling induced by ectopic miR-315 can be suppressed by increasing the dosage of either of its targets, Notum or Axin.

Axin Is Repressed by miR-315 but Not Other miRNAs with Conserved Binding Sites.

Because Axin is a limiting factor in degrading the nuclear cofactor, Arm, any miRNA that can inhibit Axin should in principle activate TCF-luc. Many miRNAs are collectively predicted to target the Axin transcript (17, 23), including miR-277 (two sites), miR-315 (two sites), miR-289 (one site), and miR-313/miR-312/miR-311/miR-310 (one site). However, despite that miRNAs such as miR-277 have higher scores according to prediction algorithms, only miR-315 had a robust effect on Wg pathway activity in our screen (Fig. 1). These data suggest that, among these miRNAs, miR-315 is a uniquely potent regulator of Axin.

We next tested directly whether these miRNAs or others could influence Axin sensor activity. Interestingly, only miR-315 effected significant repression of this reporter (Fig. 4D). Although it is possible that these miRNAs might contribute to Axin repression in concert with other factors (such as other cell type-specific miRNAs), these observations indicate that the existence of conserved miRNA seed matches does not reliably translate into functionally or phenotypically significant miRNA-mediated regulation. However, our data demonstrate that, with solid functional data in hand, it is possible to identify valid and biologically relevant interactions from the available target predictions.


miRNA Dysfunction and Disease.

It has been known for decades that duplications and deficiencies across nearly the entire euchromatic portion of the Drosophila genome are viable, fertile, and of relatively normal phenotype (32). This finding indicates that few genes have detectable mutant effects when manipulated across a 50–200% activity range. At the same time, many miRNA-mediated regulatory interactions appear to cause only modest quantitative effects. Thus, although most genes are predicted as miRNA targets, we suspect that few will prove susceptible to phenotypically detectable changes in function as a consequence of altered miRNA activity. We hypothesized that components of signaling pathways, whose activities are well known to be quite dose-sensitive, might harbor a large proportion of such phenotypically relevant miRNA targets.

In considering how deregulated miRNAs might induce disease, we hypothesize that miRNA gain of function will generally prove more debilitating than miRNA loss of function. Endogenous networks are often buffered by multiple layers of regulation, such that the removal of single miRNAs might be reasonably tolerated. In contrast, ectopic miRNAs often induce mutant phenotypes in a dominant fashion. A related consideration is that miRNA loss-of-function effects are limited to locations of endogenous expression, whereas miRNA gain of function is in principle relevant to all tissues.

In this report, we combined high-throughput gain-of-function screening in cell-based assays with directed in vivo biological tests to identify miR-315 as a strong activator of Wg signaling. Our finding that an miRNA can activate Wg signaling in the absence of ligand has implications for disease mechanisms. In humans, hyperactivated Wnt signaling is well known to induce cancer. Correspondingly, negative regulators of Wnt signaling such as Axin and adenomatous polyposis coli are bona fide tumor suppressors (14). Although miR-315 does not appear to be conserved in primary sequence in vertebrates, mammalian Axin and adenomatous polyposis coli contain likely functional sites for other miRNAs (28). That miRNAs could inappropriately activate the Wnt pathway, even in cells that are not engaged in Wnt signaling, is a possibility that deserves study in mammals.

Cell-Based Activity Screens Reveal Biologically Relevant miRNA Activities and Targets.

Few miRNAs had a substantial effect on Wg signaling in our primary screen despite an extensive list of computationally predicted miRNA targets in the Wg pathway. This finding highlights the inadequacy of target-based approaches for analyzing miRNA function and the power of activity-based approaches to investigate miRNAs. We emphasize that a specific functional connection between miR-315 and Wg signaling could not have been anticipated from computational studies because miR-315 is predicted to have an extraordinary number of targets by Drosophila standards (>400) (17). Axin and Notum occupy positions 14 and 210 in the rank-ordered list, yet bulk phenotypes of ectopic miR-315 could be suppressed by elevating the activity of either Notum or Axin. Although the targeting potential of miR-315 likely extends beyond the Wg pathway, our work illustrates how deregulation of an individual miRNA with hundreds of conserved targets can still elicit specific effects on biological processes that are attributable to only a few targets.

In summary, our study highlights that components of signaling pathways represent phenotypically critical targets of miRNAs and describes a biologically validated strategy for identifying miRNAs with the ability to modulate signal-transduction pathways of choice. We believe that similar strategies will permit the rapid elucidation of human miRNAs whose dysfunction will prove relevant during disease.


UAS-miRNA Library.

We initially amplified pri-miRNA fragments consisting of miRNA hairpins and ≈150–200 nt upstream and downstream flanks and cloned them into Gal4-responsive plasmids bearing UAS sites. When exposed to Gal4, such constructs produce hybrid pri-miRNA transcripts as fusions to a noncoding RNA composed primarily of the SV40t 3′ UTR. Because functional pri-miRNAs were subsequently shown to be excised from mRNAs (23), most of our expression constructs contain pri-miRNA fragments cloned into the 3′ UTR of a DsRed-encoding vector. The details of individual constructs are available upon request. Our in vivo analysis of miR-7 and the miR-2 cluster showed that both have identical activity when activated from either UAS-miRNA or UAS-DsRed-miRNA transgenes (13, 23), except that the latter have the technical convenience of producing both active miRNAs and red fluorescent protein as a marker. All 75 UAS-miRNA screening constructs (see SI Table 2) will be publicly available at the Drosophila RNAi Screening Center (DRSC).

miRNA Activity Screening and Target Validation in Cultured Cells.

Primary TCF-luc screening assays were performed in duplicate or triplicate in 100 μl of clone 8 cells plated at 1.5 × 106 cells per ml in 96-well plates (15). Data were averaged and then normalized by using background control wells. All secondary assays, including specificity tests with other signaling pathways and target regulation tests, were performed in triplicate in at least two independent experiments.

We performed transfections by using Effectene (Qiagen, Valencia, CA) with the following modifications to the manufacturer's guidelines: per-well volume of reagents was 50 μl of EC buffer, 0.8 μl of enhancer, and 0.25 μl of Effectene; 25 ng of each construct was used in transfections. Wg assays were performed by using TOP-Flash (15) plus pAct5c-Wg as indicated. For JAK/STAT assays, we used the 10X-STAT92E-luciferase reporter construct (33). For Hh assay, ptcΔ136-luc reporter was used in the presence or absence of pAct5c-Hh as described previously (34). Dual-glo luciferase (Promega, Madison, WI) was used to measure the relative firefly and Renilla luciferase levels. Luciferase activity was measured on an analyst HT plate reader (Molecular Devices, Sunnyvale, CA).

For RNAi epistasis tests, dsRNAs were made and transfected as described (34). dsRNAs were selected from the DRSC collection on the basis of few, if any, predicted off targets and lack of CAR/CAN repeats (35). Their DRSC identification numbers are as follows: arm (DRSC18738), wg (DRSC03636), dTCF (DRSC17176), legless (DRSC17152), arr (DRSC07451), and fz (DRSC11348).

For luciferase sensor tests, wild-type and mutant Axin and Notum 3′ UTRs were amplified from Canton S genomic DNA, cloned into pENTR (Invitrogen, Carlsbad, CA), and transferred into the XhoI and NotI sites of psiCHECK2 (Promega). Full-length sensors consist of genomic segments starting with the stop codon and extending to ≈200 nt downstream of the annotated polyadenylation signal into tub-GFP. Short sensors consist of the 3′ UTR region bounded by the pair of miR-315 sites in each gene. Short mut sensors contain AA→CC point mutations in the miR-315 seed matches. Primers are as follows: notum full length, CACCtcgagTAGgctcaccaaataccctg and cccgcggccgcTTTGCATATAATTATTTTGTCGG; notum short, CACCtcgagCATTTATTCATTCGCATTCTGCAATC and cccGCGGCCGCaagtaTTTGATTttcgttgcacag; notum short mut, cacctcgagCATTTATTCATTCGCATTCTGCAATCatAATCccAGaaatgaaatc and cccGCGGCCGCaagtaTggGATTttcgttgcacaggtacg; axin full length, CACCtcgagTAAtattaagcatttagcgtag and cccgcggccgcAGATGTACAACATGAAGGCCC; axin short, CACCtcgagCAACTGGAGAAGCTATTTCCGAC and cccGCGGCCGcagattatcTTTTGATTttttgtc; and axin short mut, cacctcgaGCAACTGGAGAAGCTATTTCCGACaAATCccAAgctatatatg and CCCGCGGCCGcagattatcTTggGATTttttgtcgagttgaatag.

Drosophila Strains.

For Axin and Notum transgenic sensors, we cloned the full-length 3′ UTR and downstream genomic segments described into tub-GFP (23). UAS-DsRed-miR-315 flies were generated by using the same construct used for tissue culture studies. Transgenic strains were made by BestGene Inc. (Chino Hills, CA) according to standard methods. tub-GFP-miR-7 sensor was a kind gift of Stephen Cohen (Temasek, Singapore) (23). The UAS-Axin-GFP transgene, the EP-Notum insertion, and the alleles AxinS044230, Wingful141, Notum3, and Notum5 have all been described previously (30, 31, 36). Other publicly available strains were obtained from the Bloomington Stock Center (37), including bxMS1096-Gal4, ptc-Gal4, dpp40C.6-Gal4, and UAS-wg.


We performed immunostaining according to standard methods (38) by using the following primary antibodies: mouse α-Nubbin (1:50; gift of Stephen Cohen), mouse α-Cut (1:100; Developmental Studies Hybridoma Bank), mouse α-Wg (1:20; Developmental Studies Hybridoma Bank), guinea pig α-Sens (1:2,000; gift of Hugo Bellen, Baylor University, Houston, TX), and rabbit α-GFP (1:1,250; Molecular Probes). Native DsRed activity was detected. Secondary antibodies were Alexa 488 α-rabbit (1:600), Alexa 488 α-guinea pig (1:600), and Alexa 647 α-mouse (1:600), all from Molecular Probes. Processed discs were counterstained with Hoechst, mounted in Vectashield (Vector Laboratories, Burlingame, CA), and analyzed on an Axioimager Z1 fitted with Apotome (Carl Zeiss, Thornwood, NY).

Supplementary Material

Supporting Information:


We thank Bergin Tam and David Raleigh for assistance with construction of the miRNA expression library; Kent Nybakken (Harvard Medical School, Boston, MA) (Hh), Ram Dasgupta (New York University, New York, NY) (Wg), and Rui Chen (Harvard Medical School, Boston, MA) (JAK/STAT) for cell-based assay plasmids; Hugo Bellen, Konrad Basler (Universität Zürich, Zürich, Switzerland), Nick Tolwinski (Sloan–Kettering Institute, New York, NY), and Stephen Cohen for serum antibodies and fly strains; Nancy Arango for cloning the GFP-Axin and GFP-Notum sensors; Eric Joseph for initial characterization of the UAS-DsRed-miR-315 transgene; the DRSC for dsRNA templates and use of their plate reader; and Gerald Rubin for logistical support during the early phases of this work. This work was supported by the Developmental Studies Hybridoma Bank (Iowa City, IA) under the auspices of the National Institute of Child Health and Human Development, a Damon Runyon Cancer Research Foundation Fellowship (to S.J.S.), a Mildred and Emil Holland Scholarship (to J.W.H.), a Margaret and Herman Sokol Fellowship, and National Institutes of Health Grant GM07739. N.P. is an investigator of the Howard Hughes Medical Institute, and E.C.L. is a fellow of the Leukemia and Lymphoma Society, the Burroughs Wellcome Foundation, the V Foundation for Cancer Research, and the Sidney Kimmel Cancer Foundation.


Drosophila RNAi Screening Center
T cell factor.


The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at www.pnas.org/cgi/content/full/0706673104/DC1.


1. Lai EC. Curr Biol. 2003;13:R925–R936. [PubMed]
2. Rajewsky N. Nat Genet. 2006;38(Suppl 1):S8–S13.
3. Johnson SM, Grosshans H, Shingara J, Byrom M, Jarvis R, Cheng A, Labourier E, Reinert KL, Brown D, Slack FJ. Cell. 2005;120:635–647. [PubMed]
4. Mayr C, Hemann MT, Bartel DP. Science. 2007;315:1576–1579. [PMC free article] [PubMed]
5. Lee YS, Dutta A. Genes Dev. 2007;21:1025–1030. [PMC free article] [PubMed]
6. Shell S, Park SM, Radjabi AR, Schickel R, Kistner EO, Jewell DA, Feig C, Lengyel E, Peter ME. Proc Natl Acad Sci USA. 2007;104:11400–11405. [PMC free article] [PubMed]
7. Voorhoeve PM, le Sage C, Schrier M, Gillis AJ, Stoop H, Nagel R, Liu YP, van Duijse J, Drost J, Griekspoor A, et al. Cell. 2006;124:1169–1181. [PubMed]
8. He L, Thomson JM, Hemann MT, Hernando-Monge E, Mu D, Goodson S, Powers S, Cordon-Cardo C, Lowe SW, Hannon GJ, Hammond SM. Nature. 2005;435:828–833. [PubMed]
9. O'Donnell KA, Wentzel EA, Zeller KI, Dang CV, Mendell JT. Nature. 2005;435:839–843. [PubMed]
10. Flynt AS, Li N, Thatcher EJ, Solnica-Krezel L, Patton JG. Nat Genet. 2007;39:259–263. [PMC free article] [PubMed]
11. Lai EC, Burks C, Posakony JW. Development (Cambridge, UK) 1998;125:4077–4088. [PubMed]
12. Lai EC, Posakony JW. Development (Cambridge, UK) 1997;124:4847–4856. [PubMed]
13. Lai EC, Tam B, Rubin GM. Genes Dev. 2005;19:1067–1080. [PMC free article] [PubMed]
14. Clevers H. Cell. 2006;127:469–480. [PubMed]
15. DasGupta R, Kaykas A, Moon RT, Perrimon N. Science. 2005;308:826–833. [PubMed]
16. Stark A, Brennecke J, Bushati N, Russell RB, Cohen SM. Cell. 2005;123:1133–1146. [PubMed]
17. Grun D, Wang Y-L, Langenberger D, Gunsalus KC, Rajewsky N. PLoS Comp Biol. 2005;1:51–63.
18. Enright AJ, John B, Gaul U, Tuschl T, Sander C, Marks DS. Genome Biol. 2003;5:R1. [PMC free article] [PubMed]
19. Long D, Lee R, Williams P, Chan CY, Ambros V, Ding Y. Nat Struct Mol Biol. 2007;14:287–294. [PubMed]
20. Didiano D, Hobert O. Nat Struct Mol Biol. 2006;13:849–851. [PubMed]
21. Korinek V, Barker N, Willert K, Molenaar M, Roose J, Wagenaar G, Markman M, Lamers W, Destree O, Clevers H. Mol Cell Biol. 1998;18:1248–1256. [PMC free article] [PubMed]
22. Barolo S. Oncogene. 2006;25:7505–7511. [PubMed]
23. Stark A, Brennecke J, Russell RB, Cohen SM. PLoS Biol. 2003;1:E60. [PMC free article] [PubMed]
24. Griffiths-Jones S, Grocock RJ, van Dongen S, Bateman A, Enright AJ. Nucleic Acids Res. 2006;34:D140–D144. [PMC free article] [PubMed]
25. Hutvagner G, Simard MJ, Mello CC, Zamore PD. PLoS Biol. 2004;2:e98. [PMC free article] [PubMed]
26. Wang SH, Simcox A, Campbell G. Genes Dev. 2000;14:2271–2276. [PMC free article] [PubMed]
27. Ng M, Diaz-Benjumea FJ, Vincent JP, Wu J, Cohen SM. Nature. 1996;381:316–318. [PubMed]
28. Lewis BP, Burge CB, Bartel DP. Cell. 2005;120:15–20. [PubMed]
29. Tolwinski NS, Wieschaus E. Development (Cambridge, UK) 2001;128:2107–2117. [PubMed]
30. Giraldez AJ, Copley RR, Cohen SM. Dev Cell. 2002;2:667–676. [PubMed]
31. Gerlitz O, Basler K. Genes Dev. 2002;16:1055–1059. [PMC free article] [PubMed]
32. Lindsley DL, Sandler L, Baker BS, Carpenter ATC, Denell RF, Hall JC, Jacobs PA, Miklos GL, Davis BK, Gethman RC, et al. Genetics. 1972;71:157–184. [PMC free article] [PubMed]
33. Baeg GH, Zhou R, Perrimon N. Genes Dev. 2005;19:1861–1870. [PMC free article] [PubMed]
34. Nybakken K, Vokes SA, Lin TY, McMahon AP, Perrimon N. Nat Genet. 2005;37:1323–1332. [PubMed]
35. Kulkarni MM, Booker M, Silver SJ, Friedman A, Hong P, Perrimon N, Mathey-Prevot B. Nat Methods. 2006;3:833–838. [PubMed]
36. Hamada F, Tomoyasu Y, Takatsu Y, Nakamura M, Nagai S, Suzuki A, Fujita F, Shibuya H, Toyoshima K, Ueno N, Akiyama T. Science. 1999;283:1739–1742. [PubMed]
37. Grumbling G, Strelets V. Nucleic Acids Res. 2006;34:D484–D488. [PMC free article] [PubMed]
38. Lai EC, Rubin GM. Dev Biol. 2001;231:217–233. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...