Erratum

The role of GW/P-bodies in RNA processing and silencingAndrew Jakymiw, Kaleb M. Pauley, Songqing Li, Keigo Ikeda, Shangli Lian, Theophany Eystathioy, MinoruSatoh, Marvin J. Fritzler and Edward K. L. Chan

Journal of Cell Science 120, 1702 (2007) doi:10.1242/jcs.03452

There was an error published in J. Cell Sci. 120, 1317-1323.

We apologise for two errors that occurred in the online and pdf versions of this article. The printed version is correct.

On p. 1317, in the Summary, the sentence ‘Formation of GW bodies appears to depend on both specific protein factors and RNA,in particular, microRNA.’ appeared twice. The correct version of the summary is shown below.

SummaryGW bodies, also known as mammalian P-bodies, are cytoplasmic foci involved in the post-transcriptional regulation ofeukaryotic gene expression. Recently, GW bodies have been linked to RNA interference and demonstrated to be importantfor short-interfering-RNA- and microRNA-mediated mRNA decay and translational repression. Evidence indicates thatboth passenger and guide strands of short-interfering RNA duplexes can localize to GW bodies, thereby indicating thatRNA-induced silencing complexes may be activated within these cytoplasmic centers. Work over the past few years hassignificantly increased our understanding of the biology of GW bodies, revealing that they are specialized cell componentsthat spatially regulate mRNA turnover in various biological processes. Formation of GW bodies appears to depend onboth specific protein factors and RNA, in particular, microRNA. Here, we propose a working model for GW body assemblyin terms of its relationship to RNA interference. In this process, one or more heteromeric protein complexes accumulatein successive steps into larger ribonucleoprotein structures.

On p. 1319, right column, first paragraph, the word order of the penultimate sentence was incorrect and should read:

In particular, studies in Drosophila indicate that GW182 interacts with Ago1 and promotes miRNA-mediated degradation of asubset of mRNA targets (Behm-Ansmant et al., 2006).

1317Commentary

IntroductionThe control of mRNA stability plays key roles in both the post-transcriptional regulation of eukaryotic gene expression(Keene and Lager, 2005; Wilusz and Wilusz, 2004) and mRNAquality control (Fasken and Corbett, 2005). The latter involvesthe recognition and rapid degradation of aberrant mRNAs andtakes place when translation termination occurs too early(nonsense-mediated decay) or fails to occur (non-stop decay)(Fasken and Corbett, 2005) or when translation elongationstalls (no-go decay) (Doma and Parker, 2006). In eukaryotes,mRNA turnover is regulated by two major mechanisms. Oneinvolves the multisubunit exosome, where transcripts aredegraded by 3�-to-5� exonucleases (for a review, see van Hoofand Parker, 1999). The second mechanism involvescytoplasmic compartments termed GW bodies (GWBs), whichspatially control mRNA turnover by the 5�-to-3� mRNA decaymachinery. These discrete cytoplasmic foci, also called Dcp-containing bodies or processing (P)-bodies, constitute sites ofmRNA degradation, storage and translational repression(Brengues et al., 2005; Coller and Parker, 2005; Cougot et al.,2004; Eystathioy et al., 2002; Eystathioy et al., 2003; Shethand Parker, 2003; Van Dijk et al., 2002). Recently, they havealso been shown to function in RNA interference (RNAi)(Jakymiw et al., 2005; Liu et al., 2005a; Liu et al., 2005b;Meister et al., 2005; Pillai et al., 2005; Sen and Blau, 2005).

RNAi is a post-transcriptional silencing mechanism in whichsmall double-stranded RNA molecules induce sequence-specificdegradation and/or translational repression of homologous

mRNAs (for reviews, see Filipowicz et al., 2005; Meister andTuschl, 2004; Rana, 2007; Sen and Blau, 2006; Valencia-Sanchez et al., 2006). The discovery that RNAi effector proteins(Jakymiw et al., 2005; Liu et al., 2005a; Liu et al., 2005b; Senand Blau, 2005) and small RNAs localize to GWBs (Jakymiw etal., 2005; Pauley et al., 2006; Pillai et al., 2005), and that GWBassembly appears to be required for the proper functioning of theRNAi pathway (Jakymiw et al., 2005; Liu et al., 2005a; Meisteret al., 2005), suggests that these foci are specifically involved inshort-interfering RNA (siRNA)- and microRNA (miRNA)-mediated mRNA degradation and/or translational repression†.

GW bodies, also known as mammalian P-bodies, arecytoplasmic foci involved in the post-transcriptionalregulation of eukaryotic gene expression. Recently, GWbodies have been linked to RNA interference anddemonstrated to be important for short-interfering-RNA-and microRNA-mediated mRNA decay and translationalrepression. Evidence indicates that both passenger andguide strands of short-interfering RNA duplexes canlocalize to GW bodies, thereby indicating that RNA-induced silencing complexes may be activated within thesecytoplasmic centers. Formation of GW bodies appears todepend on both specific protein factors and RNA, inparticular, microRNA. Work over the past few years has

significantly increased our understanding of the biology ofGW bodies, revealing that they are specialized cellcomponents that spatially regulate mRNA turnover invarious biological processes. The formation of GW bodiesappears to depend on both specific protein factors andRNA, in particular, microRNA. Here, we propose aworking model for GW body assembly in terms of itsrelationship to RNA interference. In this process, one ormore heteromeric protein complexes accumulate insuccessive steps into larger ribonucleoprotein structures.

Key words: GW bodies, P-bodies, RNA interference, mRNAdegradation, microRNA, short-interfering RNA

Summary

The role of GW/P-bodies in RNA processing andsilencingAndrew Jakymiw1, Kaleb M. Pauley1, Songqing Li1, Keigo Ikeda1, Shangli Lian1, Theophany Eystathioy2,Minoru Satoh3, Marvin J. Fritzler2 and Edward K. L. Chan1,*1Department of Oral Biology, University of Florida, Gainesville, FL 32610, USA2Department of Biochemistry and Molecular Biology, University of Calgary, Calgary, Alberta, T2N 4N1, Canada3Division of Rheumatology and Clinical Immunology, Department of Medicine, and Department of Pathology, Immunology, and LaboratoryMedicine, University of Florida, Gainesville, FL 32610, USA*Author for correspondence (e-mail: echan@ufl.edu)

Accepted 21 February 2007Journal of Cell Science 120, 1317-1323 Published by The Company of Biologists 2007doi:10.1242/jcs.03429

†siRNAs are small RNAs of ~21 nucleotides in length and are derived from the progressivecleavage of long, perfectly complementary double-stranded RNAs (dsRNAs) by anRNase-III-type endonuclease, Dicer. They can originate from long dsRNAs transientlyintroduced into cells by transfection or stably expressed hairpin-containing dsRNAprecursors derived from DNA constructs. They assemble into an RNA-protein complexknown as the RNA-induced silencing complex (RISC or siRISC), which includesArgonaute 2 (Ago2), a key component of RNAi that possesses endonuclease activity.RISC then targets and cleaves perfectly complementary mRNAs, generating 5� and 3�fragments, which are subsequently degraded. MiRNAs are similar in size to siRNAs butoriginate from hairpin-containing precursors encoded by the genome. These non-codingprecursors of miRNAs have double-stranded regions with imperfect complementarity andare sequentially processed by RNase-III-type enzymes Drosha and Dicer into maturemiRNAs. The mature miRNAs then assemble into an RNA-protein complex referred toas the miRNA ribonucleoprotein complex (miRNP or miRISC), which is structurallysimilar to RISC and contains at least one Argonaute protein. Multiple copies of miRNPsare directed to the 3�-UTR of certain mRNAs that have imperfect complementarity to thebound miRNAs. In plants, miRNAs cleave their regulated target mRNA, whereas inmammals miRNAs promote translational repression of the targeted mRNA. Until recently,this was considered the major distinction between siRISC and miRISC in mammaliancells. However, new evidence suggests that miRNAs can also regulate mRNA degradationsimilarly to siRNAs (Bagga et al., 2005; Jing et al., 2005; Yekta et al., 2004), thus blurringthe distinction between the two small-RNA-mediated silencing complexes.

Jour

nal o

f Cel

l Sci

ence

1318

GWBs are enriched in mRNA decay factors and pools ofstored messenger ribonucleoproteins (mRNPs) (Bruno andWilkinson, 2006; Sheth and Parker, 2006). Moreover, they aredynamic structures, whose size and number appear to dependon specific intracellular processes. For instance, GWBs vary insize and number throughout the cell cycle, the largest of whichare observed in late S and G2 phase (Yang et al., 2004).Furthermore, stress (Kedersha et al., 2005; Teixeira et al.,2005), cell proliferation (Yang et al., 2004), blocking mRNAdecay (Andrei et al., 2005; Cougot et al., 2004; Sheth andParker, 2003) and inhibition of translational initiation(Brengues et al., 2005; Sheth and Parker, 2003; Teixeira et al.,2005) all increase the size and number of GWBs. Conversely,blocking transcription, deadenylation of mRNAs ortranslational elongation decreases the size and number ofGWBs (Cougot et al., 2004; Sheth and Parker, 2003). Recentstudies have highlighted the importance of these structures inthe regulation of mRNA turnover and provided insights intotheir involvement in RNA silencing. Here, we review thecurrent understanding of GWB function, focusing on its recentlinks to RNAi, and discuss the role of these cytoplasmic fociin RNA-induced silencing complex (RISC) activation. We alsoexamine the requirements for GWB formation and disassemblyand propose a working model for their genesis.

GWBs uncoveredGWBs were first identified and characterized in studies usingan autoimmune serum from a human patient with motor andsensory neuropathy (Eystathioy et al., 2002). They were namedas such because they harbor the mRNA-binding proteinGW182. GW182 is characterized by glycine (G) andtryptophan (W) repeat-rich domains and a canonical RNA-recognition motif and has been demonstrated to associate witha specific subset of transcripts in HeLa cells (Eystathioy et al.,2002). The finding that a set of mammalian proteins involvedin mRNA degradation – Dcp1/2, Xrn1 and LSm1-7 – localizeto similar prominent cytoplasmic foci (Bashkirov et al., 1997;Eystathioy et al., 2003; Ingelfinger et al., 2002; Van Dijk et al.,2002), and that GWBs are active sites of mRNA decay inhuman cells (Cougot et al., 2004), indicated that GWBs areinvolved in the spatial regulation of mRNA turnover, and morespecifically, a selective 5�-to-3� mRNA degradation pathway.Concurrent studies identified GWB-related structures referredto as P-bodies in budding yeast, and these were found to playa role in mRNA decapping and 5�-to-3� decay (Sheth andParker, 2003). In particular, Sheth and Parker demonstratedthat insertion of a poly(G) tract into mRNAs, which blocks 5�-to-3� exonuclease activity, causes accumulation of decayintermediates within P-bodies, suggesting that the mRNAdecay process is associated with them (Sheth and Parker,2003).

These bodies are thought to be conserved in budding yeastand mammals because both the yeast P-bodies and mammalianGWBs contain activators of decapping and decapping enzymes(Cougot et al., 2004; Eystathioy et al., 2003; Ingelfinger et al.,2002; Lykke-Andersen, 2002; Segal et al., 2006; Sheth andParker, 2003; Van Dijk et al., 2002). Note, however, that GWBsdiffer from yeast P-bodies in that they contain translationinitiation factors and other proteins that have no yeastcounterparts, such as factors involved in RNAi [e.g. GW182and Argonaute (Ago) proteins] (Anderson and Kedersha, 2006;

Segal et al., 2006). Furthermore, there are functionaldifferences between GWBs and yeast P-bodies in terms of theirresponses to stress and cell growth (Schneider et al., 2006). Forexample, P-bodies increase in size and number during growthlimitation, increased cell density, and stress (Teixeira et al.,2005), whereas GWBs increase in size and number inproliferating cells (Yang et al., 2004) and dynamically interactwith stress granules‡ in stressed mammalian cells (Kedersha etal., 2005). Therefore, one needs to be cautious whengeneralizing about these structures because differences doexist, not only between species but even within a population ofGWBs in a single mammalian cell (Fig. 1). Interestingly, arecent report similarly found that some GW/P-bodies thatcontain Mex-3B, a newly described class of human RNA-binding protein that localizes to GW/P-bodies, are also devoidof Dcp1 (Buchet-Poyau et al., 2007). The function of aparticular GW/P-body may therefore depend on itsorganization or composition.

GW body functionmRNA degradationThe decay of mRNA in GW/P-bodies is thought to occur bya 5�-to-3� exonucleolytic process, which first requires theremoval of the 3�-poly(adenosine) [poly(A)] tail –deadenylation. Subsequently, the 5� mRNA cap is irreversiblyremoved by a decapping complex and the body of the mRNAis degraded by the 5�-to-3� exonuclease Xrn1 (for a review,see Coller and Parker, 2004). Evidence also suggests thatGW/P-bodies can regulate translation (Bhattacharyya et al.,2006; Coller and Parker, 2005; Ferraiuolo et al., 2005; Pillaiet al., 2005) and store mRNAs (Bhattacharyya et al., 2006;Brengues et al., 2005; Pillai et al., 2005). Moreover, GW/P-bodies also appear to be involved in other mRNA decaypathways, such as nonsense mediated decay (NMD), AU-richelement (ARE)-mediated decay (AMD) and stress-induceddecay (SID).

NMD is a surveillance mechanism that removes aberrantmRNA transcripts containing premature termination codons.Several factors involved in NMD, including UPF1-3, SMG5and SMG7, as well as reporter mRNAs harboring nonsensemutations have been found to localize to GW/P-bodies(Fukuhara et al., 2005; Sheth and Parker, 2006; Unterholznerand Izaurralde, 2004), which links these foci and the NMDprocess. GWBs have also been linked to AMD, a decay processinvolving messages containing AREs in the 3�-untranslatedregion (UTR). In particular, depletion of three GWB proteins,Xrn1, LSm1 or 4E-T, has been demonstrated to inhibit AMD(Ferraiuolo et al., 2005; Stoecklin et al., 2006). Moreover,evidence also indicates that the ARE-binding protein TTP,which is known to destabilize ARE-containing mRNAs,localizes to GWBs (Kedersha et al., 2005) and interacts withand activates the decapping complex (Fenger-Gron et al., 2005;Lykke-Andersen and Wagner, 2005). Finally, the observationthat the size and number of P-bodies increases during stress inbudding yeast cells (Teixeira et al., 2005) and that GWBs canbe induced to form and interact with stress granules by specificstresses in mammalian cells demonstrates their importance inSID (Kedersha et al., 2005). However, because budding yeast

Journal of Cell Science 120 (8)

‡Stress granules are large RNP particles that store non-translating mRNAs when cells areexposed to environmental stresses, but are absent in budding yeast cells.

Jour

nal o

f Cel

l Sci

ence

1319GW/P-bodies in RNA processing and silencing

cells lack stress granules, the mechanism of SID probablydiffers from that in higher eukaryotes.

In mammalian cells, GWBs and stress granules have beenreported to be compositionally and morphologically distinctstructures (Cougot et al., 2004; Kedersha et al., 2005);however, evidence indicates that they are functionally andspatially linked (Kedersha et al., 2005). Interestingly, bothentities share an assortment of proteins (e.g. CPEB1, Rck/p54,FAST, Xrn1, eIF4 and TTP) and a single class of reportermRNA has been observed within both GWBs and stressgranules (Kedersha et al., 2005; Wilczynska et al., 2005).Moreover, overexpression of TTP and CPEB1, a translationalregulator, induces the fusion of stress granules and GWBs,which suggests that these proteins regulate the dynamicinteraction between these two structures (Kedersha et al., 2005;Wilczynska et al., 2005). One model is that stress granulesserve as mRNA triage sites where transcripts are sorted forstorage, re-initiation of translation, or degradation, and thatmRNAs targeted for decay are exported from stress granulesto GWBs (Kedersha et al., 2005). Regardless of themechanistic differences between yeast and higher eukaryotes,it is evident that GW/P-bodies have multiple functions, theultimate goal being mRNA decay and/or storage. It is thereforenot surprising that these foci comprise multiple factors and thattheir specific structural composition and organization probablydetermines their mode of action.

RNAiSeveral groups recently demonstrated a link between GWBsand RNAi (Jakymiw et al., 2005; Liu et al., 2005a; Liu et al.,2005b; Pillai et al., 2005; Sen and Blau, 2005). In particular,the Ago family of proteins, Ago1-4, which are components ofRISC, the key effector complex of RNAi, were found to beconcentrated in GWBs (Jakymiw et al., 2005; Liu et al., 2005a;Liu et al., 2005b; Sen and Blau, 2005). Moreover, GWBs alsoappeared to be sites involved in miRNA-mediated repressionof targeted mRNAs (Liu et al., 2005b).

In addition, we and others demonstrated that hAgo2, thecatalytic engine of RNA silencing, associates with severalcomponents of GWBs, including GW182 (Jakymiw et al.,

2005; Liu et al., 2005a; Liu et al., 2005b). Moreover,depletion of GW182, which disrupts GWBs (Yang et al.,2004), perturbs both siRNA- and miRNA-mediatedrepression (Jakymiw et al., 2005; Liu et al., 2005a). Dominantinterfering GW182 and hAgo2 mutants also disrupt GWBformation and similarly inhibit RNA silencing (Jakymiw etal., 2005). The impairment in RNAi appears to depend onblocking the localization of hAgo2 to GWBs (Jakymiw et al.,2005). Indeed, hAgo2 constructs containing point mutationsthat prevent siRNA binding and localization to GWBs do notrepress target reporter mRNA, despite being tethered to thetarget (Liu et al., 2005a). Additional evidence for theinvolvement of GW182 and GWB in RNAi comes fromstudies of the Drosophila melanogaster ortholog andCaenorhabditis elegans functional analog (Ding et al., 2005;Rehwinkel et al., 2005; Schneider et al., 2006). In particular,studies in Drosophila indicate that GW182 interacts withAgo1 and promotes degradation of a subset of miRNA-mediated-decay mRNA targets (Behm-Ansmant et al., 2006).GW182 may thus function as a molecular scaffold thatbridges components of the miRNA pathway with mRNAdecay enzymes (Behm-Ansmant et al., 2006).

A point of contention is whether GW182 and GWBs areimportant for both slicer-dependent mechanisms (i.e. siRNA-mediated RNAi) and miRNA-mediated repression or only thelatter. Work from our laboratory has demonstrated thattransfected fluorophore-labeled siRNAs associate with GW182protein complexes and localize to GWBs and that GW182 andGWBs play an important role in slicer-mediated functions(Jakymiw et al., 2005). Moreover, Liu et al. found thatsuppression of GW182 similarly impairs, albeit not aseffectively, the ability of an siRNA to silence its target bymRNA cleavage (Liu et al., 2005a). In addition, silencing ofthe GWB protein TNRC6B, a GW182 paralog, inhibitscleavage of an mRNA reporter gene containing a target siteperfectly complementary to an endogenous miRNA (Meisteret al., 2005). TNRC6B thus appears to be important for slicer-dependent mechanisms and subsequent mRNA degradation.

Other work has indicated a greater role for GW182 andGWBs in miRNA-mediated translational repression by

Fig. 1. GW bodies (GWBs) are heterogeneous structureswith obvious differences in protein composition. (Top) HEp-2 cells stained with human anti-GWB serum(a prototype serum often used as a GWB marker andknown to contain antibodies to GW182, hAgo2 and Ge-1,but not Dcp1; green) and rabbit anti-Dcp1 antibody (oftenused as a marker for P-bodies; red) demonstrate that notall GWBs contain Dcp1. (Bottom) HEp-2 cells stainedwith mouse anti-hAgo2 monoclonal (green) and rabbitanti-RCK/p54 (red) polyclonal antibodies demonstratethat not all foci containing RCK/p54 contain hAgo2.Conversely, not all foci that stain for hAgo2 containRCK/p54. Arrows, GWBs that do not contain bothprotein factors. Arrowheads, GWBs that do contain bothprotein factors. Nuclei (blue) were counterstained withDAPI. Bar, 10 �m.

Jour

nal o

f Cel

l Sci

ence

1320

demonstrating, using various reporter systems, that depletionof GW182 impairs miRNA function (Chu and Rana, 2006; Liuet al., 2005a; Rehwinkel et al., 2005). Furthermore, depletionof GW182 has also been demonstrated to result in alterationsof mRNA expression profiles very similar to those seen in cellsdepleted of the Drosophila miRNA effector Ago1 and not thesiRNA effector Ago2, which suggests that GW182 functionsin the miRNA pathway (Behm-Ansmant et al., 2006). Whetherthe above disparities are because of the use of different reportersystems (i.e. exogenously introduced versus endogenousreporter systems), species variations or GW182 redundancy(three paralogs have been identified in humans) will requirefurther study. Interestingly, in the study implicating GW182 inslicer-mediated function, we used siRNAs with a fluorophoreconjugated to the 5�-end of the guide strand (Jakymiw et al.,2005). The fluorophore might therefore have interfered withthe siRNA activity, making it behave more like an miRNA byproducing imperfect base-pairing between siRNA and target.Regardless, the studies collectively demonstrate a role forGW182 and GWBs in RNAi. More work will be needed todetermine whether GWBs represent converging sites forsiRISC and/or miRISC.

MiRNAs have similarly been identified within GWBs anddemonstrated to associate with GW182 protein complexes(Pauley et al., 2006; Pillai et al., 2005). The identification ofsiRNA and/or miRNA within GWBs suggests that RISCactivation, activity and/or recycling may occur within GWBs.Are GWBs sites of RISC activation and activity? Evidencesupporting this possibility includes the observation that Ago2,a known GWB component, is directly involved in passenger-strand cleavage of double-stranded siRNAs during RISCassembly (Matranga et al., 2005; Rand et al., 2005). Also,following transient transfection of HeLa cells with afluorophore-labeled passenger-strand siRNA duplex that hasno endogenous mRNA target, the siRNA localizes to GWBs –similarly to a fluorophore-labeled-guide-strand siRNA duplexthat has an endogenous mRNA target (Jakymiw et al., 2005).This suggests that the passenger strand and the guide strandlocalize to GWBs independently of the mRNA target and thatpassenger-strand cleavage and incorporation of the antisense-

strand into RISC may occur within GWBs. Docking ofsiRNA/miRNA duplexes into RISC and its subsequentactivation could therefore be early events in GWB formation.Interestingly, in fission yeast cells, a Dicer ortholog localizesto structures resembling GWBs (Carmichael et al., 2006),which suggests that GWB formation may take place evenbefore RISC activation.

GW body formation and structureImmunogold electron microscopy of GWBs identifiescytoplasmic electron-dense structures that are 100-300 nm indiameter and which lack a membrane (Eystathioy et al., 2002;Yang et al., 2004). Closer inspection reveals that they comprise8-10 nm strands or fibrils (Yang et al., 2004). Currently, themechanism of GWB formation is not well understood. Itremains unclear whether GWBs form de novo or whethermRNAs and their associated proteins are targeted to pre-existing structures. Furthermore, we do not know whetherGWB components are targeted to GWBs independently or aspart of larger complexes that form higher-order structures thatcan be visualized by conventional light microscopy. RNAi-mediated depletion of specific GWB factors in human cells hasdemonstrated an interdependence of each of the proteins fortheir accumulation in GWBs (see Table 1). GWBs may thusform by the assembly of one or more heteromeric proteincomplexes on mRNAs that can amass into larger mRNPstructures. The finding that specific enzymes involved indecapping and subsequent 5�-to-3� degradation are dispensablefor GWB formation (Andrei et al., 2005) – unlike factorsinvolved at earlier stages of mRNA decay (e.g. mRNA 3�-endtrimming) – indicates that earlier stages are more crucial forGWB assembly.

Fig. 2 shows our working model, in which several factors,including RISC, initially interact with the target mRNA to forma specific RNP structure dependent on the type of decay orstorage process that will occur (e.g. siRNA-mediated decayversus miRNA-mediated translational repression). This resultsin the recruitment of other protein complexes, which dependon the composition of the initial RNP; so their finalcomposition or structural organization promotes the proper

Journal of Cell Science 120 (8)

Table 1. Suppression of specific GWB factors demonstrates their interdependence on accumulation within and forassembly of GWBs

Target* Cell function GWB foci† GWB markers examined References

GW182 RNAi – LSm4, hAgo2, hDcp1a, hDcp2 (Jakymiw et al., 2005; Liu et al., 2005a;Yang et al., 2004)

Xrn1 5�-to-3� exonuclease + hDcp2 (Andrei et al., 2005; Cougot et al., 2004)Ccr4 Deadenylation – LSm1, RCK/p54, eIF4E, eIF4E-T (Andrei et al., 2005)LSm1 Decapping – Ccr4, RCK/p54, eIF4E, eIF4E-T (Andrei et al., 2005; Chu and Rana, 2006)LSm4 Decapping – GW182, hDcp1a (Kedersha et al., 2005)hDcp2 Decapping + LSm1, RCK/p54, eIF4E, eIF4E-T, Ge-1 (Andrei et al., 2005; Yu et al., 2005)RCK/p54 Decapping; translation – hAgo2, LSm1, Ccr4, eIF4E, eIF4E-T (Andrei et al., 2005; Chu and Rana, 2006)

control; RNAieIF4E-T Translation control – Ccr4, RCK/p54, eIF4E, LSm1, hDcp1 (Andrei et al., 2005; Ferraiuolo et al., 2005)RAP55 Translation control – hDcp1a (Tanaka et al., 2006; Yang et al., 2006)Ge-1/Hedls Decapping – hDcp1a, hDcp2 (Yu et al., 2005)miRNA‡ RNAi – GW182, hAgo2, hDcp1a (Pauley et al., 2006)mRNA§ Translation – hDcp1a, LSm1, RCK/p54, (Andrei et al., 2005; Cougot et al., 2004)

eIF4E, eIF4E-T

GWB, GW body; miRNA, microRNA; RNAi, RNA interference; siRNA, short-interfering RNA; *proteins silenced through siRNA-mediated gene-specificknockdown; †detected using fluorescence microscopy; ‡miRNA suppression was induced indirectly through siRNA-mediated knockdown of Drosha or DGCR8;§accumulation within GWBs inhibited using cycloheximide or actinomycin D.

Jour

nal o

f Cel

l Sci

ence

1321GW/P-bodies in RNA processing and silencing

decay mechanism and/or storage of the mRNA. Such a modelwould explain why disruption of GWBs by depletion of otherGWB components, such as LSm1 or RCK/p54, does notnecessarily translate into impaired RISC activity (Chu andRana, 2006), whereas silencing of GW182 does (Jakymiw etal., 2005). One possibility is that GW182 exists in a smallerRNP complex with RISC components (e.g. hAgo2) that isundetectable by fluorescence microscopy and supersedes therequirement of LSm1 and RCK/p54 for initial targetrecognition and endonucleolytic cleavage or translationrepression. Moreover, not all silencing of GWB factors impairsformation of GWBs equivalently. Interestingly, Andrei et al.have demonstrated that silencing of RCK/p54 has a lesspronounced effect on the disassembly of GWBs compared withother factors (Andrei et al., 2005).

The integrity of GWBs also appears to depend on RNA.Studies in yeast and mammals indicate that mRNA is an

integral component of GW/P-bodies (Andrei et al., 2005;Cougot et al., 2004; Eystathioy et al., 2002; Teixeira et al.,2005). Furthermore, silencing of Drosha or its required partnerprotein, DGCR8, which together comprise the microprocessorcomplex and are responsible for processing long nuclearprimary miRNA (pri-miRNA) transcripts to ~70-nucleotidehairpin precursor miRNA (pre-miRNA), depletes cells ofmature miRNA and causes the disappearance of GWBs (Pauleyet al., 2006). This suggests that miRNAs are required for theformation of GWBs. Although the possibility that the effect isindirect cannot be ruled out, this seems unlikely becausesiRNA transfected into Drosha-deficient cells can serve as asurrogate for miRNA and drive the reappearance of GWBs.

The dependence of GWBs on miRNA means that many ofthe functions attributed to GWBs similarly may depend onmiRNAs in mammalian cells. Processes such as translationalrepression, mRNA decay and storage depend on miRNAs.

Fig. 2. A model linking RNAi and GW body (GWB) assembly and function. RNAi activity is triggered by siRNA/miRNA duplexes, which arefirst processed from either double-stranded RNA (dsRNA) or precursor microRNA (pre-miRNA), respectively, by the RNase-III-typeendonuclease Dicer. These duplexes are then incorporated into RISC, where the passenger strand (black) is either cleaved and degraded orremoved by the bypass mechanism (Matranga et al., 2005). This activation process and assembly of the guide strand (red) into RISC is thoughtto initiate early stages of GWB formation. Subsequent targeting by RISC results in further recruitment of one or more heteromeric proteincomplexes (which could include GW182 and RCK/p54) on the mRNA, which forms a specific RNP structure that causes post-transcriptionalinhibition of gene expression (through siRNA-mediated cleavage or miRNA-mediated translational repression, depending on the degree ofcomplementarity between the guide-strand RNA and its target mRNA). The targeted mRNA is eventually degraded by further recruitment of 5�-to-3� mRNA decay factors, which include the deadenylase Ccr4, decapping factors (LSm1-7 ring, Dcp complex) and the 5�-to-3� exonucleaseXrn1. The accumulation of RNA decay factors during degradation could be responsible for the size increases in GWBs, which makes themvisible by conventional light microscopy. ORF, open reading frame.

Jour

nal o

f Cel

l Sci

ence

1322

Furthermore, miRNAs are involved in ARE-mediateddegradation events (Jing et al., 2005). All of these processesare associated with GWBs. The primary function of GWBsmight therefore be to provide a microenvironment for miRNA-mRNA interactions that lead to translational inhibition and/ormRNA degradation. Whether NMD processes share thisrequirement for miRNAs requires further study. Nevertheless,if this hypothesis holds, a more appropriate name for these fociin mammalian cells may be miRNA-induced bodies (miRBs).

Conclusions and perspectivesRecent cell biology and biochemical findings havesignificantly enhanced our understanding of the spatialregulation of mRNA decay and/or storage within eukaryoticcells, and it is now evident that dynamic cytoplasmic foci,GW/P-bodies, are crucial for these processes. The discovery ofa functional link between GWBs and RNAi has beenparticularly instrumental in allowing us to decipher thecomplexities of how small RNAs and their cognate proteins areinvolved in post-transcriptional gene regulation (for reviews,see Engels and Hutvagner, 2006; Eulalio et al., 2007; Jacksonand Standart, 2007; Pillai et al., 2007; Rana, 2007).

Our understanding of the cell biology of RNAi and itsrelationship to GWBs is still limited, however. During thecourse of writing this review, several articles were publisheddemonstrating that a large fraction of miRNAs and Agoproteins reside in the cytoplasm (Leung et al., 2006; Maroneyet al., 2006; Nottrott et al., 2006), in particular on mRNAsbeing actively translated by polyribosomes (Maroney et al.,2006; Nottrott et al., 2006). Furthermore, inhibiting translationinitiation or inducing stress has been shown to result inlocalization of miRNA and Ago to stress granules (Leung etal., 2006). Data from these studies suggest that miRNPs firstassociate with and suppress actively translating mRNAs in thecytoplasm prior to completely dropping off ribosomes. UponmiRNA-mediated repression, these mRNA are then believed tobe targeted to either stress granules for storage or sorted andshuttled to GWBs for decay. Given the close relationshipbetween GWBs and stress granules, this would not besurprising; however, one needs to be cautious in interpretingthe stress granule data because stress granules are generallyobserved only during a stress response, whereas GWBs arepresent continuously (Teixeira et al., 2005). Alternatively,inhibition of actively translating polysomes on mRNAs bymiRISCs might directly trigger the recruitment of RNPcomplexes involved in either early stages of GWB formationand/or GWB targeting. Much more work will be needed if weare to completely understand the interactions between GWBsand other mRNP structures and the significance of theirheterogeneous composition.

Finally, although numerous studies have demonstrated thatRNAi occurs in the cytoplasm of a cell, a growing number ofreports suggest that it occurs within the nucleus as well. Inparticular, recent studies in human cells have implicated theRNAi effector proteins hAgo1 and hAgo2 in the post-transcriptional modulation of gene expression in the nucleus(Robb et al., 2005) and transcriptional silencing (Janowski etal., 2006; Kim et al., 2006). Furthermore, transfection ofsiRNAs targeting small nuclear RNAs (e.g. 7SK and U6 RNA)leads to their translocation into the nucleus and silencing of thetarget genes (Berezhna et al., 2006). How the cytoplasmic and

nuclear RNAi pathways are interrelated and the relationshipbetween GWBs and nuclear RNAi are currently unclear.Biochemical evidence suggests that these processes are linkedby a common requirement for Ago proteins; however,immunofluorescence studies using a newly developed mousemonoclonal antibody specific for hAgo2 show no evidence ofnuclear localization of this endogenous protein (Ikeda et al.,2006), which is consistent with our earlier work usingpolyclonal autoimmune sera (Jakymiw et al., 2006).Regardless, the recent findings that GWBs appear to functionin siRNA/miRNA-mediated forms of post-transcriptionalregulation, the apparent requirement for GWBs in Drosophila(Schneider et al., 2006) and C. elegans (Ding et al., 2005)development, and the knowledge that small RNAs regulatemany cellular activities (Ambros, 2004), includingdifferentiation, stem cell division, and apoptosis, underline theimportance of GWBs for many biological processes.

We thank Jens Lykke-Andersen (University of Colorado) and TomHobman (University of Alberta) for their generosity in providingvaluable antibody reagents. We apologize to colleagues whoseinteresting work could not be cited owing to space limitations. Thiswork was supported in part by the National Institutes of Health GrantAI47859 and the Canadian Institutes for Health Research Grant MOP-38034 and the Canadian Breast Cancer Research Foundation Grant16992.

ReferencesAmbros, V. (2004). The functions of animal microRNAs. Nature 431, 350-355.Anderson, P. and Kedersha, N. (2006). RNA granules. J. Cell Biol. 172, 803-808.Andrei, M. A., Ingelfinger, D., Heintzmann, R., Achsel, T., Rivera-Pomar, R. and

Luhrmann, R. (2005). A role for eIF4E and eIF4E-transporter in targeting mRNPs tomammalian processing bodies. RNA 11, 717-727.

Bagga, S., Bracht, J., Hunter, S., Massirer, K., Holtz, J., Eachus, R. and Pasquinelli,A. E. (2005). Regulation by let-7 and lin-4 miRNAs results in target mRNAdegradation. Cell 122, 553-563.

Bashkirov, V. I., Scherthan, H., Solinger, J. A., Buerstedde, J. M. and Heyer, W. D.(1997). A mouse cytoplasmic exoribonuclease (mXRN1p) with preference for G4tetraplex substrates. J. Cell Biol. 136, 761-773.

Behm-Ansmant, I., Rehwinkel, J., Doerks, T., Stark, A., Bork, P. and Izaurralde, E.(2006). mRNA degradation by miRNAs and GW182 requires both CCR4:NOTdeadenylase and DCP1:DCP2 decapping complexes. Genes Dev. 20, 1885-1898.

Berezhna, S. Y., Supekova, L., Supek, F., Schultz, P. G. and Deniz, A. A. (2006).siRNA in human cells selectively localizes to target RNA sites. Proc. Natl. Acad. Sci.USA 103, 7682-7687.

Bhattacharyya, S. N., Habermacher, R., Martine, U., Closs, E. I. and Filipowicz, W.(2006). Relief of microRNA-mediated translational repression in human cells subjectedto stress. Cell 125, 1111-1124.

Brengues, M., Teixeira, D. and Parker, R. (2005). Movement of eukaryotic mRNAsbetween polysomes and cytoplasmic processing bodies. Science 310, 486-489.

Bruno, I. and Wilkinson, M. F. (2006). P-bodies react to stress and nonsense. Cell 125,1036-1038.

Buchet-Poyau, K., Courchet, J., Hir, H. L., Seraphin, B., Scoazec, J. Y., Duret, L.,Domon-Dell, C., Freund, J. N. and Billaud, M. (2007). Identification andcharacterization of human Mex-3 proteins, a novel family of evolutionarily conservedRNA-binding proteins differentially localized to processing bodies. Nucleic Acids Res.In press.

Carmichael, J. B., Stoica, C., Parker, H., McCaffery, J. M., Simmonds, A. J. andHobman, T. C. (2006). RNA interference effector proteins localize to mobilecytoplasmic puncta in Schizosaccharomyces pombe. Traffic 7, 1032-1044.

Chu, C. Y. and Rana, T. M. (2006). Translation repression in human cells by microRNA-induced gene silencing requires RCK/p54. PLoS Biol. 4, e210.

Coller, J. and Parker, R. (2004). Eukaryotic mRNA decapping. Annu. Rev. Biochem. 73,861-890.

Coller, J. and Parker, R. (2005). General translational repression by activators of mRNAdecapping. Cell 122, 875-886.

Cougot, N., Babajko, S. and Seraphin, B. (2004). Cytoplasmic foci are sites of mRNAdecay in human cells. J. Cell Biol. 165, 31-40.

Ding, L., Spencer, A., Morita, K. and Han, M. (2005). The developmental timingregulator AIN-1 interacts with miRISCs and may target the argonaute protein ALG-1to cytoplasmic P bodies in C. elegans. Mol. Cell 19, 437-447.

Doma, M. K. and Parker, R. (2006). Endonucleolytic cleavage of eukaryotic mRNAswith stalls in translation elongation. Nature 440, 561-564.

Journal of Cell Science 120 (8)

Jour

nal o

f Cel

l Sci

ence

1323GW/P-bodies in RNA processing and silencing

Engels, B. M. and Hutvagner, G. (2006). Principles and effects of microRNA-mediatedpost-transcriptional gene regulation. Oncogene 25, 6163-6169.

Eulalio, A., Behm-Ansmant, I. and Izaurralde, E. (2007). P bodies: at the crossroadsof post-transcriptional pathways. Nat. Rev. Mol. Cell Biol. 8, 9-22.

Eystathioy, T., Chan, E. K., Tenenbaum, S. A., Keene, J. D., Griffith, K. and Fritzler,M. J. (2002). A phosphorylated cytoplasmic autoantigen, GW182, associates with aunique population of human mRNAs within novel cytoplasmic speckles. Mol. Biol.Cell 13, 1338-1351.

Eystathioy, T., Jakymiw, A., Chan, E. K., Seraphin, B., Cougot, N. and Fritzler, M.J. (2003). The GW182 protein colocalizes with mRNA degradation associated proteinshDcp1 and hLSm4 in cytoplasmic GW bodies. RNA 9, 1171-1173.

Fasken, M. B. and Corbett, A. H. (2005). Process or perish: quality control in mRNAbiogenesis. Nat. Struct. Mol. Biol. 12, 482-488.

Fenger-Gron, M., Fillman, C., Norrild, B. and Lykke-Andersen, J. (2005). Multipleprocessing body factors and the ARE binding protein TTP activate mRNA decapping.Mol. Cell 20, 905-915.

Ferraiuolo, M. A., Basak, S., Dostie, J., Murray, E. L., Schoenberg, D. R. andSonenberg, N. (2005). A role for the eIF4E-binding protein 4E-T in P-body formationand mRNA decay. J. Cell Biol. 170, 913-924.

Filipowicz, W., Jaskiewicz, L., Kolb, F. A. and Pillai, R. S. (2005). Post-transcriptionalgene silencing by siRNAs and miRNAs. Curr. Opin. Struct. Biol. 15, 331-341.

Fukuhara, N., Ebert, J., Unterholzner, L., Lindner, D., Izaurralde, E. and Conti, E.(2005). SMG7 is a 14-3-3-like adaptor in the nonsense-mediated mRNA decaypathway. Mol. Cell 17, 537-547.

Ikeda, K., Satoh, M., Pauley, K. M., Fritzler, M. J., Reeves, W. H. and Chan, E. K.(2006). Detection of the argonaute protein Ago2 and microRNAs in the RNA inducedsilencing complex (RISC) using a monoclonal antibody. J. Immunol. Methods 317, 38-44.

Ingelfinger, D., Arndt-Jovin, D. J., Luhrmann, R. and Achsel, T. (2002). The humanLSm1-7 proteins colocalize with the mRNA-degrading enzymes Dcp1/2 and Xrnl indistinct cytoplasmic foci. RNA 8, 1489-1501.

Jackson, R. J. and Standart, N. (2007). How do microRNAs regulate gene expression?Sci. STKE 2007, re1.

Jakymiw, A., Lian, S., Eystathioy, T., Li, S., Satoh, M., Hamel, J. C., Fritzler, M. J.and Chan, E. K. (2005). Disruption of GW bodies impairs mammalian RNAinterference. Nat. Cell Biol. 7, 1267-1274.

Jakymiw, A., Ikeda, K., Fritzler, M. J., Reeves, W. H., Satoh, M. and Chan, E. K.(2006). Autoimmune targeting of key components of RNA interference. Arthritis Res.Ther. 8, R87.

Janowski, B. A., Huffman, K. E., Schwartz, J. C., Ram, R., Nordsell, R., Shames, D.S., Minna, J. D. and Corey, D. R. (2006). Involvement of AGO1 and AGO2 inmammalian transcriptional silencing. Nat. Struct. Mol. Biol. 13, 787-792.

Jing, Q., Huang, S., Guth, S., Zarubin, T., Motoyama, A., Chen, J., Di Padova, F.,Lin, S. C., Gram, H. and Han, J. (2005). Involvement of microRNA in AU-richelement-mediated mRNA instability. Cell 120, 623-634.

Kedersha, N., Stoecklin, G., Ayodele, M., Yacono, P., Lykke-Andersen, J., Fitzler, M.J., Scheuner, D., Kaufman, R. J., Golan, D. E. and Anderson, P. (2005). Stressgranules and processing bodies are dynamically linked sites of mRNP remodeling. J.Cell Biol. 169, 871-884.

Keene, J. D. and Lager, P. J. (2005). Post-transcriptional operons and regulons co-ordinating gene expression. Chromosome Res. 13, 327-337.

Kim, D. H., Villeneuve, L. M., Morris, K. V. and Rossi, J. J. (2006). Argonaute-1 directssiRNA-mediated transcriptional gene silencing in human cells. Nat. Struct. Mol. Biol.13, 793-797.

Leung, A. K., Calabrese, J. M. and Sharp, P. A. (2006). Quantitative analysis ofArgonaute protein reveals microRNA-dependent localization to stress granules. Proc.Natl. Acad. Sci. USA 103, 18125-18130.

Liu, J., Rivas, F. V., Wohlschlegel, J., Yates, J. R., III, Parker, R. and Hannon, G. J.(2005a). A role for the P-body component GW182 in microRNA function. Nat. CellBiol. 7, 1261-1266.

Liu, J., Valencia-Sanchez, M. A., Hannon, G. J. and Parker, R. (2005b). MicroRNA-dependent localization of targeted mRNAs to mammalian P-bodies. Nat. Cell Biol. 7,719-723.

Lykke-Andersen, J. (2002). Identification of a human decapping complex associatedwith hUpf proteins in nonsense-mediated decay. Mol. Cell. Biol. 22, 8114-8121.

Lykke-Andersen, J. and Wagner, E. (2005). Recruitment and activation of mRNA decayenzymes by two ARE-mediated decay activation domains in the proteins TTP andBRF-1. Genes Dev. 19, 351-361.

Maroney, P. A., Yu, Y., Fisher, J. and Nilsen, T. W. (2006). Evidence that microRNAsare associated with translating messenger RNAs in human cells. Nat. Struct. Mol. Biol.13, 1102-1107.

Matranga, C., Tomari, Y., Shin, C., Bartel, D. P. and Zamore, P. D. (2005). Passenger-strand cleavage facilitates assembly of siRNA into Ago2-containing RNAi enzymecomplexes. Cell 123, 607-620.

Meister, G. and Tuschl, T. (2004). Mechanisms of gene silencing by double-strandedRNA. Nature 431, 343-349.

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

Nottrott, S., Simard, M. J. and Richter, J. D. (2006). Human let-7a miRNA blocksprotein production on actively translating polyribosomes. Nat. Struct. Mol. Biol. 13,1108-1114.

Pauley, K. M., Eystathioy, T., Jakymiw, A., Hamel, J. C., Fritzler, M. J. and Chan,E. K. (2006). Formation of GW bodies is a consequence of microRNA genesis.EMBO Rep. 7, 904-910.

Pillai, R. S., Bhattacharyya, S. N., Artus, C. G., Zoller, T., Cougot, N., Basyuk, E.,Bertrand, E. and Filipowicz, W. (2005). Inhibition of translational initiation byLet-7 MicroRNA in human cells. Science 309, 1573-1576.

Pillai, R. S., Bhattacharyya, S. N. and Filipowicz, W. (2007). Repression of proteinsynthesis by miRNAs: how many mechanisms? Trends Cell Biol. 17, 118-126.

Rana, T. M. (2007). Illuminating the silence: understanding the structure and functionof small RNAs. Nat. Rev. Mol. Cell Biol. 8, 23-36.

Rand, T. A., Petersen, S., Du, F. and Wang, X. (2005). Argonaute2 cleaves the anti-guide strand of siRNA during RISC activation. Cell 123, 621-629.

Rehwinkel, J., Behm-Ansmant, I., Gatfield, D. and Izaurralde, E. (2005). A crucialrole for GW182 and the DCP1:DCP2 decapping complex in miRNA-mediated genesilencing. RNA 11, 1640-1647.

Robb, G. B., Brown, K. M., Khurana, J. and Rana, T. M. (2005). Specific and potentRNAi in the nucleus of human cells. Nat. Struct. Mol. Biol. 12, 133-137.

Schneider, M. D., Najand, N., Chaker, S., Pare, J. M., Haskins, J., Hughes, S. C.,Hobman, T. C., Locke, J. and Simmonds, A. J. (2006). Gawky is a component ofcytoplasmic mRNA processing bodies required for early Drosophila development.J. Cell Biol. 174, 349-358.

Segal, S. P., Dunckley, T. and Parker, R. (2006). Sbp1p affects translationalrepression and decapping in Saccharomyces cerevisiae. Mol. Cell. Biol. 26, 5120-5130.

Sen, G. L. and Blau, H. M. (2005). Argonaute 2/RISC resides in sites of mammalianmRNA decay known as cytoplasmic bodies. Nat. Cell Biol. 7, 633-636.

Sen, G. L. and Blau, H. M. (2006). A brief history of RNAi: the silence of the genes.FASEB J. 20, 1293-1299.

Sheth, U. and Parker, R. (2003). Decapping and decay of messenger RNA occur incytoplasmic processing bodies. Science 300, 805-808.

Sheth, U. and Parker, R. (2006). Targeting of aberrant mRNAs to cytoplasmicprocessing bodies. Cell 125, 1095-1109.

Stoecklin, G., Mayo, T. and Anderson, P. (2006). ARE-mRNA degradation requiresthe 5�-3� decay pathway. EMBO Rep. 7, 72-77.

Tanaka, K. J., Ogawa, K., Takagi, M., Imamoto, N., Matsumoto, K. andTsujimoto, M. (2006). RAP55, a cytoplasmic mRNP component, repressestranslation in Xenopus oocytes. J. Biol. Chem. 281, 40096-40106.

Teixeira, D., Sheth, U., Valencia-Sanchez, M. A., Brengues, M. and Parker, R.(2005). Processing bodies require RNA for assembly and contain nontranslatingmRNAs. RNA 11, 371-382.

Unterholzner, L. and Izaurralde, E. (2004). SMG7 acts as a molecular link betweenmRNA surveillance and mRNA decay. Mol. Cell 16, 587-596.

Valencia-Sanchez, M. A., Liu, J., Hannon, G. J. and Parker, R. (2006). Control oftranslation and mRNA degradation by miRNAs and siRNAs. Genes Dev. 20, 515-524.

Van Dijk, E., Cougot, N., Meyer, S., Babajko, S., Wahle, E. and Seraphin, B.(2002). Human Dcp2: a catalytically active mRNA decapping enzyme located inspecific cytoplasmic structures. EMBO J. 21, 6915-6924.

van Hoof, A. and Parker, R. (1999). The exosome: a proteasome for RNA? Cell 99,347-350.

Wilczynska, A., Aigueperse, C., Kress, M., Dautry, F. and Weil, D. (2005). Thetranslational regulator CPEB1 provides a link between dcp1 bodies and stressgranules. J. Cell Sci. 118, 981-992.

Wilusz, C. J. and Wilusz, J. (2004). Bringing the role of mRNA decay in the controlof gene expression into focus. Trends Genet. 20, 491-497.

Yang, W. H., Yu, J. H., Gulick, T., Bloch, K. D. and Bloch, D. B. (2006). RNA-associated protein 55 (RAP55) localizes to mRNA processing bodies and stressgranules. RNA 12, 547-554.

Yang, Z., Jakymiw, A., Wood, M. R., Eystathioy, T., Rubin, R. L., Fritzler, M. J.and Chan, E. K. (2004). GW182 is critical for the stability of GW bodies expressedduring the cell cycle and cell proliferation. J. Cell Sci. 117, 5567-5578.

Yekta, S., Shih, I. H. and Bartel, D. P. (2004). MicroRNA-directed cleavage ofHOXB8 mRNA. Science 304, 594-596.

Yu, J. H., Yang, W. H., Gulick, T., Bloch, K. D. and Bloch, D. B. (2005). Ge-1 is acentral component of the mammalian cytoplasmic mRNA processing body. RNA 11,1795-1802.

Jour

nal o

f Cel

l Sci

ence