Functional Dissection of the Human TNRC6 (GW182-Related)...

MOLECULAR AND CELLULAR BIOLOGY, Aug. 2009, p. 4144–4155 Vol. 29, No. 150270-7306/09/$08.00�0 doi:10.1128/MCB.00380-09Copyright © 2009, American Society for Microbiology. All Rights Reserved.

Functional Dissection of the Human TNRC6 (GW182-Related)Family of Proteins�†

David Baillat and Ramin Shiekhattar*Gene Regulation Programme, Centre for Genomic Regulation, UPF, Barcelona, Spain, and Wistar Institute,

3601 Spruce Street, Philadelphia, Pennsylvania 19104

Received 24 March 2009/Returned for modification 28 April 2009/Accepted 15 May 2009

Argonaute (Ago) proteins through their association with small RNAs perform a critical function in theeffector step of RNA interference. The TNRC6 (trinucleotide repeat containing 6) family of proteins have beenshown to stably associate with Agos in mammalian cells. Here, we describe the isolation and functionalcharacterization of TNRC6B- and TNRC6C-containing complexes. We show that TNRC6B and TNRC6Cproteins associate with all four human Agos which are already loaded with microRNAs. Detailed domainanalysis of TNRC6B protein indicated that distinct domains of the protein are required for Ago binding andP-body localization. Functional analysis using reporter constructs responsive to TNRC6B tethered through anMS2-binding domain indicates that neither the Ago-binding nor the P-body localization domains are requiredfor translational silencing. In contrast, the C-terminal domain containing the RNA recognition motif plays acritical role in the silencing mediated by the TNRC6B protein.

MicroRNAs (miRNAs) are �21-nucleotide-long RNAs in-volved in posttranscriptional regulation in metazoans andplants (5). miRNAs are first synthesized by RNA polymerase IIas long primary transcripts in the nucleus (26). The majority ofthese transcripts undergo a first step of maturation whereRNase III Drosha, with the help of the double-stranded RNA-binding domain protein DGCR8, excises the precursormiRNA hairpin of �70 nucleotides (9, 19, 20). These hairpinprecursors are later exported to the cytoplasm by exportin-5(48), where they are further processed by a second RNase IIIenzyme known as Dicer (22, 25). This endonuclease convertsthe pre-miRNA into a �21-bp double-stranded RNA and withthe help of the double-stranded RNA-binding domain proteinTRPB loads one of the strands into the active site of an Ar-gonaute (Ago) family protein (Ago1 to Ago4 in mammals)(7, 18).

Ago family proteins are the central effectors of the posttran-scriptional gene silencing (36). Ago proteins contain two mainstructural domains, the PAZ domain, an oligonucleotide-bind-ing-like fold responsible for the miRNA association through its3� end (29), and a Piwi domain whose fold is related to that ofRNase H (42). In the case of full complementary base pairingbetween the miRNA and its mRNA target, the Piwi domain isable to cleave the target between the residues paired to nucle-otides 10 and 11 of the miRNA (27, 33). This property called“slicing” is routinely used for silencing the expression of aparticular gene in the cell by transfecting short interferingdouble-stranded RNAs (siRNAs). However, the only mamma-lian Ago to possess an endonuclease activity is Ago2, and mostmiRNAs do not have full complementarity with their mRNA

targets, resulting in a primarily slicing-independent mechanismof action in mammals (36, 45).

miRNA-loaded Ago proteins are part of larger complexes re-ferred to as miRNA ribonucleoparticles (miRNPs) or miRNA-induced silencing complexes (miRISCs) (6, 8, 34). These com-plexes repress expression of their target mRNAs through twomechanisms, translation inhibition and mRNA degradation. De-pending on the experimental context, miRNAs can inhibit trans-lation at the level of initiation or elongation (21, 37, 38). Inde-pendently or in conjunction with translational inhibition, it wasalso demonstrated that miRNAs were able to induce poly(A) taildeadenylation by the Ccr4-NotI complex followed by decappingof the mRNA 5� end through Dcp1/Dcp2 and further degradationby the 5�-3� exonuclease Xrn1 (1, 4, 47).

Interestingly, miRNPs and their mRNA targets localize tocytoplasmic foci called P-bodies (28, 41). These structures re-sult from the aggregation of translationally repressed mRNAswith proteins involved in mRNA turnover, such as Dcp1, Dcp2,Xrn1, Ccr4-Not, Lsm1 to Lsm7, or Rck/p54 (2, 15, 23, 46).Moreover, recent experiments have pointed to GW182 (alsoknown as TNRC6A), an Ago-interacting protein, as an impor-tant component of miRNPs not only for localization to theP-bodies but also for translation silencing and degradation ofthe miRNA-targeted mRNAs (4, 13, 28, 39). GW182 belongsto a conserved protein family restricted to metazoans. Caeno-rhabditis elegans homologs AIN-1 and AIN-2 and Drosophilamelanogaster homolog Gawky were also shown to play a majorrole in miRNA silencing and P-body localization of miRISC (4,10, 39, 49). In vertebrates, three paralogs were identified,GW182 (or TNRC6A), KIAA1093 (or TNRC6B), andTNRC6C. They all contain an abnormally high content ofGW/WG repeats, one or more Q/P-rich motifs, and a C-ter-minal RNA recognition motif (RRM) domain. The only ex-ception is the worm protein, which does not display a recog-nizable RRM domain.

In this study, we focused our analysis on the TNRC6B andTNRC6C proteins to better understand their role in transla-

* Corresponding author. Mailing address: Wistar Institute, 3601Spruce Street, Philadelphia, PA 19104. Phone: (215) 898-3896. Fax:(215) 898-3986. E-mail: [email protected].

† Supplemental material for this article may be found at http://mcb.asm.org/.

� Published ahead of print on 26 May 2009.

4144

tional inhibition. We employed biochemical purification ofTNRC6B- and TNRC6C-containing complexes to characterizetheir association with mature miRNAs and their RNA-inducedsilencing activity. Finally, we performed a detailed structure/function analysis to reveal the precise role played by the dif-ferent domains of TNRC6B in Ago protein binding as well asP-body localization and translational inhibition. Our resultspoint to the RRM domain of TNRC6B as an essential domainfor translational inhibition.

MATERIALS AND METHODS

Antibodies. Anti-DCP1a rabbit serum was a kind gift from J. Lykke-Andersen(Molecular, Cellular and Developmental Biology, University of Colorado, Boul-der, CO). Anti-Flag M2 was purchased from Sigma. Anti-Ago2 polyclonal anti-bodies were raised against an internal peptide (KLMRSASFNTDPYVRE,amino acids 381 to 396) of the human Ago2 sequence (Open Biosystems).Anti-mouse immunoglobulin G (IgG) antibodies coupled to Alexa Fluor 488 andanti-rabbit IgG antibodies coupled to Alexa Fluor 568 were purchased fromMolecular Probes. Phosphatase-conjugated anti-IgG secondary antibodies werepurchased from Promega.

Affinity purification of Flag-tagged proteins. Flag-tagged protein-expressingconstructs and a selectable marker for puromycin resistance were cotransfectedin HEK293T (Ago1) or H1299 (TNRC6B and TNRC6C) cells. Transfected cellswere grown in the presence of 2.5 �g/ml puromycin (Sigma) for selection.Individual colonies were isolated and screened for Flag-tagged protein expres-sion. Cytoplasmic extracts (S100) of Flag fusion-expressing cells (50 mg) wereincubated with 250 �l of anti-Flag M2 affinity gel (Sigma) for 2 h at 4°C. Beadswere washed four times with 10 ml of BC500 buffer (20 mM Tris [pH 8], 0.5 MKCl, 10% glycerol, 1 mM EDTA, 1 mM dithiothreitol [DTT], 0.1% NP-40, 0.5mM phenylmethylsulfonyl fluoride, aprotinin, leupeptide, and pepstatin [1 �g/mleach for the last three ingredients]) and one time with 10 ml BC100 buffer (20mM Tris [pH 8], 0.1 M KCl, 10% glycerol, 1 mM EDTA, 1 mM DTT, 0.1%NP-40, aprotinin, leupeptide, and pepstatin [1 �g/ml each for the last threeingredients]). Bound peptides were eluted with 400 �g/ml Flag peptide (Sigma)in BC100 buffer. For size exclusion chromatography, peptides were loaded ontoa Superose 6 HR 10/30 column equilibrated with BC500 buffer. The flow rate wasfixed at 0.35 ml/min, and 0.5-ml fractions were collected.

7mG cap-binding assay. HEK293T cells were transfected with CBP20 or�N1409 cloned into the pFlag-CMV2 (Sigma) vector using MetafectenPro(Biontex Laboratories). Forty-eight hours posttransfection, the cells were lysedin Triton lysis buffer (20 mM Tris [pH 8], 137 mM NaCl, 1 mM EDTA, 1.5 mMMgCl2, 10% glycerol, 1% Triton X-100, 1 mM DTT, aprotinin, leupeptide, andpepstatin [1 �g/ml each for the last three ingredients]). The lysates were incu-bated with 20 �l of anti-Flag M2 beads (Sigma) for 2 h at 4°C using constantrotation. The beads were then washed with three times with 1 ml of Triton lysisbuffer and once with BC100 buffer. Bound proteins were eluted twice by incu-bation with 25 �l of 400 �g/ml Flag peptide (Sigma) in BC100 buffer for 10 minat 4°C. Eluted proteins were incubated with a 10-�l bed volume of 7mGTPSepharose (GE Healthcare) in the presence of 75 �M GpppG or 75 �M7mGpppG (New England Biolabs) in 200 �l of binding buffer (20 mM Tris-HCl,100 mM NaCl, 2.5 mM MgCl2, 1 mM DTT, and 1 mg/ml heparin) for 1 h at 4°Cusing constant rotation. Beads were washed three times with 1 ml binding bufferwithout heparin and eluted with 20 �l binding buffer containing 75 �M7mGpppG. Samples were resolved on a 4 to 12% sodium dodecyl sulfate (SDS)-polyacrylamide gel (Invitrogen), transferred to an Immobilon-P membrane (Mil-lipore), and analyzed by Western blotting using the anti-Flag M2 antibody(Sigma).

Coimmunoprecipitations. HEK293T cells were transfected with equalamounts of the corresponding Myc-tagged Ago2 and Flag-tagged TNRC6Bconstructs (Sigma) vector using MetafectenPro (Biontex Laboratories). Forty-eight hours posttransfection, the cells were lysed in Triton lysis buffer. Thelysates were incubated with 20 �l of anti-Flag M2 beads (Sigma) for 2 h at 4°Cusing constant rotation. The beads were then washed three times with 1 mlBC500 buffer and once with BC100 buffer. Bound proteins were eluted with 20�l of 400 �g/ml Flag peptide (Sigma) in BC100 buffer for 10 min at 4°C. Sampleswere resolved on a 4 to 12% SDS-polyacrylamide gel (Invitrogen), transferred toan Immobilon-P membrane (Millipore), and analyzed by Western blotting usingthe anti-Flag M2 monoclonal antibody (Sigma) and an anti-Myc monoclonalantibody (9E10; Santa Cruz Biotechnology).

Immunohistochemistry. The cells were grown on glass coverslips and treatedadequately. The cells were fixed by 10 min of incubation with 1% formaldehydein phosphate-buffered saline (PBS). After fixation, cells were permeabilized with0.1% Triton X-100 in PBS and blocked with 4% serum and 0.1% Triton X-100in PBS. The primary antibodies were diluted in blocking buffer (anti-Flag,1/2,000; anti-DCP1a, 1/1,000) and incubated with the coverslips for 1 h at roomtemperature. Cells were washed with 0.1% Triton X-100 in PBS and incubatedwith the corresponding secondary antibodies (diluted to 1/1,000) for 1 h at roomtemperature. After the cells were washed, they were mounted with Hard Setmounting medium with 4�,6�-diamidino-2-phenylindole (DAPI) (Vectashield).

Luciferase assays. HEK293 cells growing in 12-well dishes were transfectedwith pGL3-MS2, pRL-CMV, and the suitable MS2 fusion constructs usingMetafectenPro (Biontex Laboratories). Twenty-four hours posttransfection, the cellswere lysed in 200 �l passive lysis buffer (Promega), and firefly and Renillaluciferase activities were measured on 30 �l lysate using the Dual-Luciferasereporter assay kit (Promega). Individual experiments were realized in triplicate,and the ratios presented are the averages of three independent experiments.

RISC cleavage and miRNA processing assays. Both assays were carried outaccording to Gregory et al. (18). The sequences of the GL3 siRNA and hairpinwere taken from reference 31.

Northern blotting for miRNA detection. Flag-tagged TNRC6B and Flag-tagged TNRC6C were immunoprecipitated from total cell extract from approx-imately 107 H1299 cells stably expressing TNRC6 (trinucleotide repeat contain-ing 6) proteins using anti-Flag M2 affinity gel (Sigma). Bound peptides wereeluted with 400 �g/ml Flag peptide (Sigma) in BC100 buffer, and RNAs wereextracted with TRIzol reagent (Invitrogen) and precipitated with 10 �g of tRNAas a carrier. RNAs were resolved on a 15% acrylamide–7 M urea gel andtransferred to a Hybond-N� membrane (GE Healthcare). miRNAs were de-tected by hybridization with short DNA probe complementary to the miRNAsequences labeled with [�-32P]dATP by the method of Behlke et al. (3).

Plasmid constructs. Full-length TNRC6B and related mutants were amplifiedby standard PCR techniques from the KIAA1093 cDNA clone (Kazusa DNAResearch Institute) using vent polymerase (New England Biolabs) and clonedbetween the HindIII and BamHI restriction sites of the pFlag-CMV2 (Sigma)vector. Point mutants were generated using the QuikChange site-directed mu-tagenesis kit (Stratagene). Full-length TNRC6C was amplified by standard PCRtechniques from the KIAA1582 cDNA clone (Kazusa DNA Research Institute)using vent polymerase (New England Biolabs) and cloned between the HindIIIand SalI restriction sites of the pFlag-CMV2 (Sigma) vector. Full-length Ago1was amplified by standard PCR techniques using vent polymerase (New EnglandBiolabs) and cloned between the EcoRI and BamHI restriction sites of thepFlag-CMV2 (Sigma) vector.

MS2 fusion constructs were generated by PCR amplification of the MS2-binding domain and cloning into the HindIII restriction site of the differentpFlag-CMV2 constructs. The pGL3-MS2 construct was generated by cloning anoligonucleotide containing the sequence of two MS2-binding aptamers (5�-CTAGACGCGTACACGATCACGGTACGCTGAATTAGATCTCTGCGGATAGCATGAGGATCACCCATGCTCCCGGGGT-3�) into the XbaI restriction siteof the pGL3-Ctrl vector (Promega).

The EMCV-luciferase-MS2 construct was generated as follows. The encepha-lomyocarditis virus (EMCV) fragment was excised from the EMCV-LEF plas-mid (kind gift of Fatima Gebauer, Centro de Regulación Genòmica, Barcelona,Spain) after digestion by SacI restriction enzyme, filling in by T4 DNA polymer-ase, and further digestion by NcoI restriction enzyme. This fragment was clonedinto the pGL3-MS2 plasmid previously digested by HindIII, filled in by T4 DNApolymerase, and further digested by NcoI.

RESULTS

TNRC6B and TNRC6C form distinct protein complexeswith the four human Argonaute proteins. To identify stablepartners of Ago proteins, we generated HEK293T-derived sta-ble cell lines expressing Flag epitope-tagged human Ago1,Ago2, and Ago3. Following anti-Flag immunoaffinity purifica-tion, the associated proteins were eluted from the affinity ma-trix using Flag peptide. The affinity eluate was then fraction-ated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE)and subjected to colloidal blue staining (Fig. 1a and data notshown). Specific bands were excised, digested with trypsin, andanalyzed using mass spectrometric sequencing. The predomi-

VOL. 29, 2009 FUNCTIONAL DISSECTION OF HUMAN TNRC6 PROTEINS 4145

FIG. 1. The TNRC6 family of proteins interacts with Argonaute proteins. (a) Purification of Ago1-associated protein complexes. (Left)Purification scheme. Flag-Ago1, Flag-tagged Ago1. (Right) Ago1-associated polypeptides. HEK293 cell lines stably expressing a Flag-tagged Ago1protein were established. Polypeptides associated with Ago1 were immunopurified from S100 fractions from these cells using agarose beadsconjugated with anti-Flag M2 monoclonal antibodies. After the polypeptides were washed and eluted, they were resolved by SDS-PAGE and silverstained. The parental HEK293 cell line was used as a control (mock) for the immunopurification. The positions of molecular mass markers areshown to the left of the gels. (b) Schematic alignment of TNRC6 family proteins and their D. melanogaster (dm) and C. elegans (ce) homologs. Grayblocks represent domains of higher amino acid sequence identity. The positions of domains I to IV, Q/P-rich domain (Q/P), RNA recognition motif(RRM), and ubiquitin-associated motif (UBA) are shown. The percent identity (percent similarity in parentheses) between the conserved domainsof TNRC6C and Gawky is indicated. aa, amino acids. (c) Purification of TNRC6B- and TNRC6C-associated protein complexes. (Right)Purification of TNRC6B- and TNRC6C-associated polypeptides. H1299 cell lines stably expressing a Flag-tagged TNRC6C or TNRC6B proteinwere established. Polypeptides associated with TNRC6C or TNRC6B were immunopurified from S100 fractions from these cells using agarosebeads conjugated to anti-Flag M2 monoclonal antibodies. After the polypeptides were washed and eluted, they were resolved by SDS-PAGE andsilver stained. The positions of major degradation fragments of Flag-tagged TNRC6B (Flag-TNRC6B) and TNRC6C are indicated by asterisksbesides the gels.

4146 BAILLAT AND SHIEKHATTAR MOL. CELL. BIOL.

nant sequences obtained with all three Agos corresponded tothe three human GW182-related proteins (also known asTNRC6 [trinucleotide repeat containing 6]; see Fig. 1b for aschematic alignment of the proteins). We focused on TNRC6Band TNRC6C, since a number of studies had characterizedGW182 (TNRC6A) previously (4, 10, 28, 39). To confirm theassociation of TNRC6B and TNRC6C with the Agos, we gen-erated stable cell lines expressing Flag epitope-taggedTNRC6B and TNRC6C. Mass spectrometric sequencing ofaffinity eluates confirmed the presence of all four human Agoproteins in TNRC6B and TNRC6C affinity eluates (Fig. 1c).We next subjected the affinity eluates to size exclusion chro-matography using a Superose 6 column. As shown in Fig. 2aand b, TNRC6B and the four Agos coelute as a complex ofabout 700 kDa. Our results are consistent with a previousreport of association of TNRC6B with Agos (34). Importantly,we did not find Agos in association with other Agos. Moreover,TNRC6 proteins did not associate with other TNRC6 proteins.Therefore, it is likely that each TNRC6 protein forms a com-plex with an individual Ago, resulting in the existence of 12distinct TNRC6-Ago complexes.

The TNRC6 proteins are associated with Agos loaded withmature miRNAs. As TNRC6 complexes contain Ago2, theRNase H fold endonuclease responsible for siRNA-directedcleavage of mRNAs (40, 42), we assessed the enzymatic activityof the TNRC6 complexes. We measured their activity in anAgo2-mediated targeted RNA cleavage assay. Ago2 activitywas triggered by a siRNA complementary to let-7 and directedagainst a small 22-nucleotide radiolabeled probe fully comple-mentary to its sequence (18). As shown in Fig. 2c, the cleavageactivity nicely coeluted with the TNRC6B/Ago2-containingfractions.

Interestingly, while Dicer-containing complexes displayedRISC activity following the addition of exogenously addedsiRNAs in vitro (10), the cleavage activity obtained fromTNRC6-containing complexes did not require additionalsiRNA trigger (Fig. 3a). This result suggested that Agos inassociation with TNRC6 proteins are already loaded withmiRNAs (30). Indeed, Northern blot analysis confirmed the pres-ence of different miRNAs in association with the TNRC6B andTNRC6C complexes (Fig. 3b). Similar to the TNRC6B com-plex and in accordance with its association with mature

FIG. 2. Analysis of the TNRC6B-associated complexes. (a) SDS-PAGE analysis of the TNRC6B-associated complex fractionated on aSuperose 6 gel filtration column after silver staining. The Superose 6 fractions (fractions 14 to 38) are shown above the gel. The positions ofmolecular mass markers are indicated to the left of the gel. The positions of major contaminant polypeptides are indicated by asterisks. (b) Thesame fractions were analyzed by Western blotting with anti-Flag M2 and anti-Ago2 antibodies. (c) RISC activity after preincubation of theSuperose 6 fractions with an siRNA trigger corresponding to the let-7 sequence.


miRNAs, the TNRC6C complex displayed targeted cleavage inthe absence of any exogenously added trigger (Fig. 3c).

The presence of the target mRNA does not influence theloading of TNRC6-containing complexes. We next assessed the

mechanism by which TNRC6B protein is loaded with miRNAsin vivo. Cells stably expressing Flag-tagged TNRC6B weretransfected with either a synthetic construct mimicking maturemiRNA or a precursor RNA hairpin. We first confirmed the

FIG. 3. Ago proteins interacting with TNRC6B and TNRC6C are already loaded with mature miRNAs and active. (a) Comparison of RISCactivity between the TNRC6B- and Dicer-associated complexes with (�) or without (�) preincubation with a siRNA trigger (let-7). Flag-TNRC6B,Flag-tagged TNRC6B. (b) Northern blot analysis of TNRC6B- and TNRC6C-associated complexes. RNAs were extracted from the TNRC6B- andTNRC6C-associated complexes, precipitated, and resolved on a 15% acrylamide–7 M urea gel. After the RNAs were transferred from the gel toa membrane, the membranes were hybridized with 32P-labeled probes corresponding to the indicated miRNAs. The parental HEK293 cell line wasused as a control (mock) for the immunopurification. Flag-IP, Flag immunoprecipitation. (c) Comparison of RISC activity corresponding to thelet-7 miRNA between TNRC6B- and TNRC6C-associated complexes in the absence of any exogenous siRNA trigger. (d) Pre-miRNA processingassay. Dicer- and TNRC6B-associated complexes were immunopurified from stably expressing cells. Increasing amounts (in microliters) weretested for processing of pre-miR30a into mature miR30a. (e) Cells stably expressing TNRC6B were transfected with either 70-bp hairpins orsiRNA corresponding to the luciferase sequence. After 24 h, the TNRC6B-associated complexes were purified by immunoaffinity, and their RISCactivity was assayed against a luciferase probe (Luc probe) and a let7 probe. For controls, untransfected cells were used and the specificity wasassessed by adding 2�-O-methyl (2�OMe) oligonucleotides complementary to the luciferase siRNA sequence. GL3 hp, luciferase hairpin. (f) Sameas panel e except the luciferase hairpin was cotransfected with increasing amounts of a plasmid expressing the target sequence (pGL3-ctrl;Promega).


absence of Dicer and its pre-miRNA processing activity in theTNRC6B-immunoprecipitated complexes (Fig. 3d). Next, wemeasured the RISC activity of the eluted fractions, which in-dicated that TNRC6B-containing complexes could be effi-ciently triggered by either an siRNA or a RNA hairpin corre-sponding to precursor miRNA (Fig. 3e). Moreover, eithermode of triggering the RISC activity could be efficiently inhib-ited using a 2�-O-methyl construct (Fig. 3e). Consistent withour previous analysis of Dicer-TRBP-Ago complexes (7, 18),the present results indicated that a double-stranded siRNAcould be efficiently loaded into a TNRC6B-containing com-plex. To further assess whether the presence of increasingamounts of target influences the loading of the siRNA into aTNRC6B complex, we titrated in various amounts of synthetictarget with perfect complementarity to the siRNA. These ex-periments revealed that the siRNA loading of the TNRC6Bcomplex is not influenced by the presence of the target mRNA(Fig. 3f).

The conserved glutamine/proline-rich domain (domain III)of TNRC6B is responsible for P-body localization. SinceTNRC6A was shown to be an integral component of P-bodies(14), we assessed the cytoplasmic localization of TNRC6B andTNRC6C. TNRC6B and TNRC6C proteins were expressed inHeLa cells and examined for their cytoplasmic localizationcompared to that of the decapping enzyme subunit DCP1a.DCP1a has previously been shown to be enriched in P-bodiesand is used extensively as a P-body marker (16). Similar topreviously reported localization of GW182 to P-bodies, wefound that both TNRC6B and TNRC6C proteins were en-riched in P-bodies (see Fig. S1a in the supplemental material).Interestingly, increased expression of TNRC6B and TNRC6Cresulted in enhanced P-body size and number (data notshown), consistent with previous reports examining the ectopicexpression of other P-body protein components (12). We nextexamined the domain(s) in TNRC6B responsible for localiza-tion to P-bodies. Previous studies had pointed to the C-termi-nal part of GW182/Gawky as being critical for P-body local-ization (4). Using N- and C-terminal truncations of TNRC6B(Fig. 4a), we narrowed down the region necessary for P-bodylocalization to the conserved C-terminal glutamine/proline-rich (Q/P-rich) domain (Fig. 4b). This glutamine/proline-richdomain (domain III) is reminiscent of prion-related or poly-glutamine domains involved in the formation of physiologicalor pathological protein aggregates (35). Our results suggestthat such domains may contribute to accumulation of someproteins in P-body-like cytoplasmic structures.

Distinct domains in TNRC6B are responsible for Ago2 bind-ing and P-body localization. To analyze whether there is adistinct domain in the TNRC6B protein that mediates theassociation with Agos, we used our multiple C- and N-terminaltruncations of TNRC6B to map for Ago binding (Fig. 4a and5a). We found that in contrast to P-body localization in whichdomain III played the critical role, domain I is necessary andsufficient for the association of TNRC6B with Ago2 (Fig. 5a).Removal of this domain from the full-length protein abrogatedAgo2 binding but had no effect on TNRC6B localization to theP-bodies (Fig. 5b). Moreover, ectopic expression of domain Icould be coimmunoprecipitated with endogenous Ago2 butfailed to localize to the P-body (Fig. 5a and b). We also con-firmed that similar to TNRC6A (4), TNRC6B is interacting

with the Piwi domain of Ago2 (see Fig. S1b in the supplemen-tal material). Taken together, these results indicates that Ago2binding to TNRC6B is functionally uncoupled from P-bodylocalization.

TNRC6B domain I corresponds to the WG/GW-rich regionof GW182 (TNRC6A) that was recently described as an evo-lutionarily conserved platform for Ago protein recruitment(11, 44). However, after alignment of domain I sequencescorresponding to TNRC6 proteins from different organisms(Fig. 5c), it appeared that only a few WG/GW pairs are con-served, suggesting that these conserved amino acids might havea specific role in the interaction between TNRC6 and Agoproteins. To test this contention, we mutated the highly con-served tryptophan 623 to alanine (W623A) in full-lengthTNRC6B as well as in the isolated domain I from TNRC6Band repeated the Ago-binding experiments. Interestingly, theW623A mutation completely abrogated the interaction be-tween TNRC6B and Ago2 (Fig. 5c, gels). These results suggesta more specific mechanism of molecular recognition betweenTNRC6B and Ago2 rather than the existence of a large non-specific interaction platform.

Inhibition of translation by TNRC6B does not require eitherassociation with Agos or localization to P-bodies. D. melano-gaster Gawky (GW182-like) protein has been reported tostrongly repress translation when tethered to a luciferase re-porter construct, leading to a decrease in the mRNA transcript(4). We used a similar tethering methodology to investigate thetranslation-inhibitory properties of TNRC6B and TNRC6C.Similar to GW182, MS2 fusions of TNRC6B and TNRC6Cshowed a strong translational inhibitory activity (Fig. 6a) whenbrought to a reporter construct containing MS2-binding sites.However, other proteins shown to play a role in miRNA-mediated silencing displayed a weaker inhibitory potential insuch an assay (Fig. 6a). Interestingly, while luciferase expres-sion was inhibited by nearly 80%, we did not detect any changein the mRNA levels (see Fig. S2a in the supplementalmaterial).

We next investigated whether the domains shown to beinvolved in Ago binding (domain I) and P-body localization(domain III) contributed to translation inhibition in suchtethering experiments. Multiple MS2 fusion constructs ofTNRC6B were developed and tested for their effects on trans-lation (Fig. 6b). Surprisingly, the deletion of neither the Agobinding nor the P-body localization had an impact on transla-tion inhibition (Fig. 6c), suggesting that their associated func-tions are uncoupled from translation inhibition. Indeed, directrecruitment of Agos through tethering of domain I failed toexert translational repression (Fig. 6c). Consistent with thesedata, a single point mutation (W623A) in either full-lengthTNRC6B or domain I, which abrogates its association withAgos, maintains the full extent of translational inhibition (Fig.6d) (see Fig. S2b in the supplemental material). These resultsindicate that once tethered to a reporter construct, TNRC6Bdoes not require the association with Agos or localization toP-bodies to bring about translational inhibition.

The C-terminal domain of TNRC6B, comprising an RRM,exerts a strong translation inhibition potential. The abovefindings led us to look for the domain(s) in the TNRC6Bprotein responsible for mediating its translational inhibition.MS2 fusion constructs of C-terminal truncations were engi-


neered, and their translational inhibitory activities were as-sessed (Fig. 6b and 7a). While full-length TNRC6B resulted ina sixfold inhibition, truncation of the RNA recognition motif(�C1415) relieved the inhibition to threefold. Further C-ter-

minal deletions resulted in a gradual decrease in translationalinhibition (Fig. 7a). Consistent with these findings, the tether-ing of the RRM-containing C-terminal fragment from posi-tions 1409 to 1722 to the luciferase messenger resulted in an

FIG. 4. The glutamine/proline-rich domain localizes TNRC6B to P-bodies. (a) Schematic representation of the Flag-tagged TNRC6B deletionmutants used. The positions of domains I to IV, the Q/P-rich motif, and RRM are shown. aa, amino acids. (b) Cellular localization of TNRC6B.HeLa cells were transfected with the indicated constructs. After fixation and permeabilization, the cells were probed with anti-Flag M2 monoclonalantibodies (revealed with anti-mouse IgG antibodies coupled to Alexa Fluor 488) and with anti-DCP1a polyclonal antibodies (revealed withanti-rabbit IgG antibodies coupled to Alexa Fluor 568).


FIG. 5. A single point mutation in domain I of TNRC6B abrogates Ago binding. (a) HEK293 cells were transfected with the different TNRC6Bconstructs described in the legend to Fig. 4a. Total cell extracts were immunoprecipitated with anti-Flag M2 antibodies. Immunoprecipitates wereresolved by SDS-PAGE and analyzed by Western blotting with anti-Flag M2 antibodies and anti-Ago2 antibodies. Flag-IP, Flag immunoprecipi-tation. (b) HeLa cells were transfected with the indicated constructs. After fixation and permeabilization, the cells were probed with anti-Flag M2monoclonal antibodies (revealed with anti-mouse IgG antibodies coupled to Alexa Fluor 488) and with anti-DCP1a polyclonal antibodies (revealedwith anti-rabbit IgG antibodies coupled to Alexa Fluor 568). (c) (Top) Protein alignment for the highly conserved region in the Ago-bindingdomain of TNRC6B proteins from human (Homo sapiens [Hs]), zebra fish (Danio rerio [Dr]), and fruit fly (Drosophila melanogaster [Dm]) sources.(Bottom) Same as in panel b using the wild-type (WT) and W623A point mutant (mut) of full-length TNRC6B and D-I. ctrl, control.

4151

FIG. 6. TNRC6B exerts a strong Ago-independent translational inhibition. (a) HEK293 cells were transfected with pGL3-MS2, pRL-CMV, andthe various MS2 fusion constructs indicated. Twenty-four hours posttransfection, the cells were lysed, and firefly and Renilla luciferase activitieswere measured and plotted as the firefly luciferase/Renilla luciferase ratio. (b) Schematic representations of the TNRC6B-MS2 fusion constructsand the luciferase constructs used in this figure. The positions of MS2 domain, domains I to IV, Q/P-rich motif, and RRM are shown. aa, aminoacids; SV40, simian virus 40; MBS x2, two copies of an MS2 binding site. (c and d) Same as in panel a. CTRL, control.

4152

inhibition comparable to that observed with the full-lengthTNRC6B (Fig. 7a), suggesting a critical role for this region inTNRC6B-associated translational inhibition.

We envisioned a possible mechanism for translation inhibi-tion through a competition between TNRC6B and eIF4E, thecap-binding protein, for binding to the cap and thus preventingefficient translation initiation. To assess this possibility, wemeasured the ability of the RRM domain of TNRC6B forbinding to the cap in vitro. While we observed the associationof CBP20, an RRM-containing subunit of the nuclear cap-binding complex (32), with the cap, we failed to detect theassociation of TNRC6B RRM domain (�N1409) with the 7mGcap (Fig. 7b). As an alternative approach for investigating thecap-dependent mechanism of action, we assessed the extent oftranslational repression when TNRC6B is tethered to a re-porter construct directed for translation by the internal ribo-some entry site of the EMCV. Since the translation of thisconstruct does not require eIF4E, TNRC6B effects on trans-

lation cannot be mediated through an eIF4E-mediated mech-anism. Indeed, the addition of MS2-TNRC6B to this constructresulted in the same extent of translational inhibition as seenfrom an eIF4E-dependent reporter, indicating that the inhibi-tion does not seem to be directed through TNRC6B inhibitionof eIF4E binding (Fig. 7c).

In addition to CBP20, the RRM domain is present in mul-tiple RNA-binding proteins. Indeed, a number of differentRRM domains were shown to bind RNA directly (32). Toassess whether the repression of translation by TNRC6B RRMdomain requires the integrity of the RRM domain, we intro-duced mutations (H1642P and Y1678P) in aromatic residuesthat could potentially play a role in protein-RNA or protein-protein interactions according to the known structures of var-ious RRM domains (32). Analysis of these mutants supporteda role for the TNRC6B RRM domain in mediating the trans-lational repression (Fig. 7d). Taken together, our results indi-cate that while the association of the RRM domain with the

FIG. 7. Translation inhibition exerted by the RRM-containing C-terminal domain of TNRC6B. (a) HEK293 cells were transfected withpGL3-MS2, pRL-CMV, and the various MS2 fusion constructs indicated. Twenty-four hours posttransfection, the cells were lysed, and firefly andRenilla luciferase activities were measured and plotted as the change in repression from the relative luciferase activity measured with the MS2domain alone. (b) Cap-binding assay. HEK293 cells were transfected with Flag-tagged CBP20 or Flag-tagged �N1409. After 48 h, the cells werelysed, and Flag proteins were immunopurified. Eluted proteins were tested for cap binding using 7mGTP beads in the presence of specific(7mGpppG) or nonspecific (GpppG) competitor. Bound proteins were eluted with 7mGpppG and analyzed by Western blotting using anti-Flagantibodies. (c) (Top) Schematic representation of the EMCV internal ribosome entry site (IRES) luciferase construct. luc, luciferase; SV40, simianvirus 40; MBS x2, two copies of an MS2 binding site. (Bottom) HEK293 cells were transfected with EMCV-luc-MS2 construct, pRL-CMV, andthe MS2 fusion constructs indicated. Twenty-four hours posttransfection, the cells were lysed, and firefly and Renilla luciferase activities weremeasured and plotted as the firefly luciferase/Renilla luciferase ratio. (d) Same as in panel a but relative luciferase activity (firefly luciferase/Renillaluciferase) is plotted.


cap may not be a mechanism for TNRC6B repression of trans-lation, the RRM domain is playing a structural role in suchinhibition, perhaps through associating with RNA or a proteinpartner.

DISCUSSION

Our results indicate that each of the TNRC6 family mem-bers can form a complex with one of the four Ago proteins.Therefore, our data suggest the presence of 12 different dis-tinct TNRC6-Ago complexes in human cells. Since we do notknow the exact temporal localization of TNRC6 and Agos indifferent tissues, it is tempting to speculate that they may beexpressed in a tissue-specific and developmentally regulatedfashion. More-detailed analyses of TNRC6 and Ago proteinexpression patterns are needed to fully address individual com-plex expression profiles and functions. Recent results in em-bryonic stem cells (43) suggest that in cells lacking all four Agoproteins, the expression of any of the Agos was sufficient torescue miRNA-directed silencing. However, a recent study ofDrosophila suggests that different Ago proteins can exert theiraction through different mechanisms (24). Taken together,these results suggest that while the system may allow redun-dant mechanisms of action, different Agos may use differentrepression pathways. Indeed, there is ample evidence pointingto the abilities of TNRC6 and Agos to induce both transla-tional repression and RNA destabilization through deadeny-lation.

We do not find a stable association of TNRC6 proteins withDicer. However, we were able to load TNRC6B using synthetichairpins, indicating that the loading of small siRNAs requiresprior Dicer participation, the only enzyme known to convertpre-miRNA hairpins into mature miRNAs. Moreover, sincethe Agos associated with TNRC6 proteins are already loadedwith miRNAs, it is reasonable to surmise that the associationof miRNA with Agos may enhance their interaction with theTNRC6 family of proteins. Indeed, we were unable to findTNRC6 complexes that could be loaded in vitro using single-stranded siRNAs (data not shown), which is consistent withtheir full occupancy with miRNAs. Once loaded, these com-plexes could associate with the targets of each miRNA. Inter-estingly, our studies indicate that the system does not incor-porate a feedback pathway, as the increased presence of amiRNA target does not affect the loading of a specific miRNAinto an Ago complex. It would be interesting to investigatewhether and how the miRISC complex associated with itsmRNA target is recycled following its localization to P-bodies.

Our detailed analysis of TNRC6B indicated that this proteinuses a specific domain, known as domain I, to interact with theAgo proteins. Importantly, a single point mutation in this do-main is sufficient to completely abrogate Ago binding. A sim-ilar domain in the plant large subunit of RNA polymerase IVand the TAS3 protein in Schizosaccharomyces pombe had beenpreviously proposed to contribute to Ago binding (11). Theseresults point to an evolutionarily conserved protein-proteininteraction interface in different components of RNA interfer-ence machinery across species. Interestingly, this domainwhich was initially present in nuclear proteins in plants andyeasts evolved to play a role in the cytoplasm in animals.

While we find domain I to be a major determinant in Ago

interaction, a different domain in TNRC6B, domain III, wasrequired for the localization of this protein to P-bodies. This isconsistent with previous studies mapping GW182 localizationto P-bodies (4). The intriguing finding is the resemblance ofproline/glutamine-rich domain III to glutamine-rich domain ofthe prion-like domain-containing proteins (17). While thisfinding suggests that the proline/glutamine-rich domain inother P-body components may serve as a P-body localizationsignal, our deletion analysis of number of other proteins local-ized to P-bodies indicated that such a domain does not serve auniversal role in P-body localization (data not shown).

Importantly, beside Agos, we were unable to find otherstable partners of TNRC6B and TNRC6C proteins. In addi-tion, we did not find any evidence for the stable association ofeither the translation or RNA deadenylation machinery withTNRC6B or TNRC6C proteins. However, these results do notrule out the possibility of a transient functional interactionbetween TNRC6 proteins and the machineries involved intranslational regulation and RNA metabolism. Indeed, ourfunctional analysis of TNRC6B which was tethered to thereporter constructs suggests a crucial role for this protein intranslational inhibition. Our domain mapping studies indicatethat the tethering of the C-terminal domain of TNRC6B con-sisting of an RRM domain is able to repress translation asefficiently as the full-length protein. Interestingly, the repres-sion of translation by the RRM domain does not require theinteraction with Agos or the localization to the P-bodies as wasdemonstrated by our mapping studies. Since this domain isconserved in all three mammalian TNRC6 proteins as well asthe Drosophila GW182, it may contribute to an evolutionarilyconserved mechanism of repression. The only caveat is theabsence of a detectable RRM domain in the worm proteinsAIN-1 and AIN-2 (10, 49). However, future structural analysisof the worm proteins may uncover a RRM-like domain inthese proteins.

While we do not know the precise mode by which the RRMdomain may contribute to silencing, it is likely that it mayoperate via multiple mechanisms. Other RRM domains havebeen shown to be versatile modules, capable of RNA binding,7mG cap binding, and providing a platform for protein-proteininteractions (32). We show that TNRC6B was unable to bindthe mRNA 5� cap. Moreover, the presence of a single RRMdomain is shown to be more in line with a mechanism evokingprotein-protein interaction, as several RRM domains are usu-ally required for RNA binding (32). The characterization ofthe detailed mechanism by which the RRM domains ofTNRC6 proteins are required for translational repressionshould provide further insight into the action of miRNAs.

ACKNOWLEDGMENTS

This work was supported by a grant from the National Institute ofGeneral Medicine, GM-079091, and a grant from the Mathers Foun-dation to R.S. R.S. is thankful for support from Institució Catalana deRecerca i Estudis Avançats.

REFERENCES

1. Bagga, S., J. Bracht, S. Hunter, K. Massirer, J. Holtz, R. Eachus, and A. E.Pasquinelli. 2005. Regulation by let-7 and lin-4 miRNAs results in targetmRNA degradation. Cell 122:553–563.

2. Bashkirov, V. I., H. Scherthan, J. A. Solinger, J. M. Buerstedde, and W. D.Heyer. 1997. A mouse cytoplasmic exoribonuclease (mXRN1p) with prefer-ence for G4 tetraplex substrates. J. Cell Biol. 136:761–773.


3. Behlke, M. A., S. A. Dames, W. H. McDonald, K. L. Gould, E. J. Devor, andJ. A. Walder. 2000. Use of high specific activity StarFire oligonucleotideprobes to visualize low-abundance pre-mRNA splicing intermediates in S.pombe. BioTechniques 29:892–897.

4. Behm-Ansmant, I., J. Rehwinkel, T. Doerks, A. Stark, P. Bork, and E.Izaurralde. 2006. mRNA degradation by miRNAs and GW182 requires bothCCR4:NOT deadenylase and DCP1:DCP2 decapping complexes. GenesDev. 20:1885–1898.

5. Carthew, R. W., and E. J. Sontheimer. 2009. Origins and mechanisms ofmiRNAs and siRNAs. Cell 136:642–655.

6. Caudy, A. A., M. Myers, G. J. Hannon, and S. M. Hammond. 2002. FragileX-related protein and VIG associate with the RNA interference machinery.Genes Dev. 16:2491–2496.

7. Chendrimada, T. P., R. I. Gregory, E. Kumaraswamy, J. Norman, N. Cooch,K. Nishikura, and R. Shiekhattar. 2005. TRBP recruits the Dicer complex toAgo2 for microRNA processing and gene silencing. Nature 436:740–744.

8. Chu, C. Y., and T. M. Rana. 2006. Translation repression in human cells bymicroRNA-induced gene silencing requires RCK/p54. PLoS Biol. 4:e210.

9. Denli, A. M., B. B. Tops, R. H. Plasterk, R. F. Ketting, and G. J. Hannon.2004. Processing of primary microRNAs by the microprocessor complex.Nature 432:231–235.

10. Ding, L., A. Spencer, K. Morita, and M. Han. 2005. The developmentaltiming regulator AIN-1 interacts with miRISCs and may target the argonauteprotein ALG-1 to cytoplasmic P bodies in C. elegans. Mol. Cell 19:437–447.

11. El-Shami, M., D. Pontier, S. Lahmy, L. Braun, C. Picart, D. Vega, M. A. Hakimi,S. E. Jacobsen, R. Cooke, and T. Lagrange. 2007. Reiterated WG/GW motifsform functionally and evolutionarily conserved ARGONAUTE-binding plat-forms in RNAi-related components. Genes Dev. 21:2539–2544.

12. Eulalio, A., I. Behm-Ansmant, and E. Izaurralde. 2007. P bodies: at thecrossroads of post-transcriptional pathways. Nat. Rev. Mol. Cell Biol. 8:9–22.

13. Eulalio, A., E. Huntzinger, and E. Izaurralde. 2008. GW182 interaction withArgonaute is essential for miRNA-mediated translational repression andmRNA decay. Nat. Struct. Mol. Biol. 15:346–353.

14. Eystathioy, T., E. K. Chan, S. A. Tenenbaum, J. D. Keene, K. Griffith, andM. J. Fritzler. 2002. A phosphorylated cytoplasmic autoantigen, GW182,associates with a unique population of human mRNAs within novel cyto-plasmic speckles. Mol. Biol. Cell 13:1338–1351.

15. Eystathioy, T., A. Jakymiw, E. K. Chan, B. Seraphin, N. Cougot, and M. J.Fritzler. 2003. The GW182 protein colocalizes with mRNA degradationassociated proteins hDcp1 and hLSm4 in cytoplasmic GW bodies. RNA9:1171–1173.

16. Fenger-Gron, M., C. Fillman, B. Norrild, and J. Lykke-Andersen. 2005.Multiple processing body factors and the ARE binding protein TTP activatemRNA decapping. Mol. Cell 20:905–915.

17. Gilks, N., N. Kedersha, M. Ayodele, L. Shen, G. Stoecklin, L. M. Dember,and P. Anderson. 2004. Stress granule assembly is mediated by prion-likeaggregation of TIA-1. Mol. Biol. Cell 15:5383–5398.

18. Gregory, R. I., T. P. Chendrimada, N. Cooch, and R. Shiekhattar. 2005.Human RISC couples microRNA biogenesis and posttranscriptional genesilencing. Cell 123:631–640.

19. Gregory, R. I., T. P. Chendrimada, and R. Shiekhattar. 2006. MicroRNAbiogenesis: isolation and characterization of the microprocessor complex.Methods Mol. Biol. 342:33–47.

20. Han, J., Y. Lee, K. H. Yeom, J. W. Nam, I. Heo, J. K. Rhee, S. Y. Sohn, Y.Cho, B. T. Zhang, and V. N. Kim. 2006. Molecular basis for the recognitionof primary microRNAs by the Drosha-DGCR8 complex. Cell 125:887–901.

21. Humphreys, D. T., B. J. Westman, D. I. Martin, and T. Preiss. 2005.MicroRNAs control translation initiation by inhibiting eukaryotic initiationfactor 4E/cap and poly(A) tail function. Proc. Natl. Acad. Sci. USA 102:16961–16966.

22. Hutvagner, G., J. McLachlan, A. E. Pasquinelli, E. Balint, T. Tuschl, andP. D. Zamore. 2001. A cellular function for the RNA-interference enzymeDicer in the maturation of the let-7 small temporal RNA. Science 293:834–838.

23. Ingelfinger, D., D. J. Arndt-Jovin, R. Luhrmann, and T. Achsel. 2002. Thehuman LSm1-7 proteins colocalize with the mRNA-degrading enzymesDcp1/2 and Xrnl in distinct cytoplasmic foci. RNA 8:1489–1501.

24. Iwasaki, S., T. Kawamata, and Y. Tomari. 2009. Drosophila Argonaute1 andArgonaute2 employ distinct mechanisms for translational repression. Mol.Cell 34:58–67.

25. Ketting, R. F., S. E. Fischer, E. Bernstein, T. Sijen, G. J. Hannon, and R. H.Plasterk. 2001. Dicer functions in RNA interference and in synthesis of small

RNA involved in developmental timing in C. elegans. Genes Dev. 15:2654–2659.

26. Lee, Y., M. Kim, J. Han, K. H. Yeom, S. Lee, S. H. Baek, and V. N. Kim. 2004.MicroRNA genes are transcribed by RNA polymerase II. EMBO J. 23:4051–4060.

27. Liu, J., M. A. Carmell, F. V. Rivas, C. G. Marsden, J. M. Thomson, J. J.Song, S. M. Hammond, L. Joshua-Tor, and G. J. Hannon. 2004. Argonaute2is the catalytic engine of mammalian RNAi. Science 305:1437–1441.

28. Liu, J., F. V. Rivas, J. Wohlschlegel, J. R. Yates III, R. Parker, and G. J.Hannon. 2005. A role for the P-body component GW182 in microRNAfunction. Nat. Cell Biol. 7:1261–1266.

29. Ma, J. B., K. Ye, and D. J. Patel. 2004. Structural basis for overhang-specificsmall interfering RNA recognition by the PAZ domain. Nature 429:318–322.

30. Ma, J. B., Y. R. Yuan, G. Meister, Y. Pei, T. Tuschl, and D. J. Patel. 2005.Structural basis for 5�-end-specific recognition of guide RNA by the A.fulgidus Piwi protein. Nature 434:666–670.

31. Maniataki, E., and Z. Mourelatos. 2005. A human, ATP-independent, RISCassembly machine fueled by pre-miRNA. Genes Dev. 19:2979–2990.

32. Maris, C., C. Dominguez, and F. H. Allain. 2005. The RNA recognitionmotif, a plastic RNA-binding platform to regulate post-transcriptional geneexpression. FEBS J. 272:2118–2131.

33. Meister, G., M. Landthaler, A. Patkaniowska, Y. Dorsett, G. Teng, and T.Tuschl. 2004. Human Argonaute2 mediates RNA cleavage targeted bymiRNAs and siRNAs. Mol. Cell 15:185–197.

34. Meister, G., M. Landthaler, L. Peters, P. Y. Chen, H. Urlaub, R. Luhrmann,and T. Tuschl. 2005. Identification of novel argonaute-associated proteins.Curr. Biol. 15:2149–2155.

35. Nagai, Y., T. Inui, H. A. Popiel, N. Fujikake, K. Hasegawa, Y. Urade, Y. Goto,H. Naiki, and T. Toda. 2007. A toxic monomeric conformer of the polyglu-tamine protein. Nat. Struct. Mol. Biol. 14:332–340.

36. Peters, L., and G. Meister. 2007. Argonaute proteins: mediators of RNAsilencing. Mol. Cell 26:611–623.

37. Petersen, C. P., M. E. Bordeleau, J. Pelletier, and P. A. Sharp. 2006. ShortRNAs repress translation after initiation in mammalian cells. Mol. Cell21:533–542.

38. Pillai, R. S., S. N. Bhattacharyya, C. G. Artus, T. Zoller, N. Cougot, E.Basyuk, E. Bertrand, and W. Filipowicz. 2005. Inhibition of translationalinitiation by Let-7 microRNA in human cells. Science 309:1573–1576.

39. Rehwinkel, J., I. Behm-Ansmant, D. Gatfield, and E. Izaurralde. 2005. Acrucial role for GW182 and the DCP1:DCP2 decapping complex in miRNA-mediated gene silencing. RNA 11:1640–1647.

40. Rivas, F. V., N. H. Tolia, J. J. Song, J. P. Aragon, J. Liu, G. J. Hannon, andL. Joshua-Tor. 2005. Purified Argonaute2 and an siRNA form recombinanthuman RISC. Nat. Struct. Mol. Biol. 12:340–349.

41. Sen, G. L., and H. M. Blau. 2005. Argonaute 2/RISC resides in sites ofmammalian mRNA decay known as cytoplasmic bodies. Nat. Cell Biol.7:633–636.

42. Song, J. J., S. K. Smith, G. J. Hannon, and L. Joshua-Tor. 2004. Crystalstructure of Argonaute and its implications for RISC slicer activity. Science305:1434–1437.

43. Su, H., M. I. Trombly, J. Chen, and X. Wang. 2009. Essential and overlap-ping functions for mammalian Argonautes in microRNA silencing. GenesDev. 23:304–317.

44. Till, S., E. Lejeune, R. Thermann, M. Bortfeld, M. Hothorn, D. Enderle, C.Heinrich, M. W. Hentze, and A. G. Ladurner. 2007. A conserved motif inArgonaute-interacting proteins mediates functional interactions through theArgonaute PIWI domain. Nat. Struct. Mol. Biol. 14:897–903.

45. Tolia, N. H., and L. Joshua-Tor. 2007. Slicer and the argonautes. Nat. Chem.Biol. 3:36–43.

46. van Dijk, E., N. Cougot, S. Meyer, S. Babajko, E. Wahle, and B. Seraphin.2002. Human Dcp2: a catalytically active mRNA decapping enzyme locatedin specific cytoplasmic structures. EMBO J. 21:6915–6924.

47. Wu, L., J. Fan, and J. G. Belasco. 2006. MicroRNAs direct rapid deadeny-lation of mRNA. Proc. Natl. Acad. Sci. USA 103:4034–4039.

48. Yi, R., Y. Qin, I. G. Macara, and B. R. Cullen. 2003. Exportin-5 mediates thenuclear export of pre-microRNAs and short hairpin RNAs. Genes Dev.17:3011–3016.

49. Zhang, L., L. Ding, T. H. Cheung, M. Q. Dong, J. Chen, A. K. Sewell, X. Liu,J. R. Yates III, and M. Han. 2007. Systematic identification of C. elegansmiRISC proteins, miRNAs, and mRNA targets by their interactions withGW182 proteins AIN-1 and AIN-2. Mol. Cell 28:598–613.


Functional Dissection of the Human TNRC6 (GW182-Related)...

Documents

Transcript of Functional Dissection of the Human TNRC6 (GW182-Related)...