RNAi Collection - Science · Science RNAi Collection Supported by Sigma-Aldrich RNA interference...

A booklet from Science, produced by the AAAS/Science Business Office

Sponsored by

RNAi Collection

Member of the RNAi Consortium

MISSION is a trademark belonging toSigma-Aldrich Co. and its affiliateSigma-Aldrich Biotechnology LP.The RNAi Consortium shRNA library is produced and distributed underlicense from the Massachusetts Instituteof Technology.

Accelerating Customers' Success through Leadership in Life Science, High Technology and ServiceS I G M A - A L D R I C H C O R P O R AT I O N • B O X 1 4 5 0 8 • S T. L O U I S • M I S S O U R I 6 3 1 7 8 • U S A

Accelerate!

INNOVATION @ WORK

RNAi is revolutionizing life science research and is one of the mostsignificant scientific advances in recent years. Accelerating thediscovery of potential new targets for diagnosis and therapeutics, and providing new insights into gene function and pathway analysis.

Sigma can accelerate your results with an impressive array of RNAitechnologies, combined to provide products that will turbo chargeyour research. Through in-house development, licensing andcollaboration, Sigma has built the most comprehensive range availableof RNAi research tools and is the only provider of both custom siRNAservices and shRNA genome-wide libraries.

Whether you are determining gene function, analyzing signaltransduction or screening for potential drug targets, why notdiscover how Sigma can accelerate the process.

with Sigma, the new leader in RNAiaccelerate your rate of life science discovery

sigma.com/rnai

2 Introduction: SCIENCE RNAi

Keith Jolliff

3 Introduction:The RNAi WorldGuy Riddihough

4 A Species of Small Antisense RNA in

Posttranscriptional Gene Silencing in PlantsAndrew J. Hamilton and David C. BaulcombeScience 29 October 1999 286: 950–952

8 Argonaute2 Is the Catalytic Engine of Mammalian RNAiJidong Liu, Michelle A. Carmell, Fabiola V. Rivas, Carolyn G. Marsden,J. Michael Thomson, Ji-Joon Song, Scott M. Hammond, Leemor Joshua-Tor, Gregory J. HannonScience 3 September 2004 305: 1437–1441; published online 29 July

2004

16 Structural Basis for Double-Stranded RNA

Processing by DicerIan J. MacRae, Kaihong Zhou, Fei Li, Adrian Repic, Angela N. Brooks,W. Zacheus Cande, Paul D. Adams, Jennifer A. DoudnaScience 13 January 2006 311: 195–198

22 RNAi-Mediated Targeting of Heterochromatin by the

RITS ComplexAndré Verdel, Songtao Jia, Scott Gerber, Tomoyasu Sugiyama, StevenGygi, Shiv I. S. Grewal, Danesh Moazed30 January 2004 303: 672–676; published online 2 January 2004

30 Advertising Supplement: Sigma-Aldrich Corporation

COVER: Leaves from Nicotiana benthamiana showing RNA silencing of gene expression.

Lighter areas are those in which the amount of green fluorescent protein is enhanced as a

result of RNA silencing of a suppressor. [Credit: Shou-Wei Ding] [Image from the cover of

Science 296, 1319 (2002)]

T A B L E O F C O N T E N T S

Copyright © 2006 by The American Association for the Advancement of Science. All

rights reserved.

Science RNAi Collection

Supported by Sigma-Aldrich

RNA interference (RNAi) has emerged as one of the most promising tools forbiological research. It is thought that RNAi may have evolved originally as a de-fense mechanism against foreign parasitic nucleic acid sequences and functions byspecifically knocking down the activity of target genes. RNAi has become a keyresearch tool, and is now routinely used by scientists as it offers many advantagesover more traditional knockdown technologies.

The ability of RNAi to target specific genes for silencing has extended its useinto the development of new experimental therapeutic approaches for various dis-eases, such as cancer, neurodegeneration, and AIDS. Although encouraging resultshave been achieved in animal models, the development of new and innovative tech-niques is crucial to the acceleration of RNAi-based approaches.

Sigma-Aldrich is dedicated to providing the scientific community with the mostadvanced and innovative tools in RNAi and functional genomics. This commitmentis exemplified by our collaboration with leading RNAi scientists, including TheRNAi Consortium (TRC), a collaborative group of 11 world-renowned academicand corporate life science research groups, including the Massachusetts Institute ofTechnology, Harvard Medical School, and the Broad Institute.

Our aim is to create a portfolio of comprehensive tools and make them broadlyavailable to scientists worldwide. Underlying this strategy, our product range inRNAi and functional genomics has been substantially enhanced by the introductionof novel technologies, including the first lentiviral MISSION™ shRNA gene familysets (in collaboration with TRC) and the revolutionary TargeTron™ gene disruptiontechnology (in partnership with InGex).

The utility of RNAi knockdown approaches is attracting scientists to use anddevelop the technology further. The rate of progress is astonishing and the ground-breaking research assembled in this Science RNAi Collection represents the breadthand depth of progress in RNAi over recent years. We are pleased to have this oppor-tunity to work together with Science to sponsor the Science RNAi Collection.

Keith JolliffGlobal Director of Research Biotech Marketing

Page 2

I N T R O D U C T I O N

I N T R O D U C T I O N

Page 3

The RNAi WorldRNA interference (RNAi) is an evolutionary ancient mechanism for silencing

gene expression. It is found in plants and animals and thus was presumably present in the common ancestor of both. RNAi may have arisen as a host defense mecha-nism against viruses and other foreign nucleic acids. This role persists: RNAi acts as an innate and adaptive “immune” response against viruses in plants and animals,and in turn viruses have evolved counter measures to abrogate the RNAi response.

RNAi forms the mechanistic core of a number of closely related endogenous RNA silencing pathways: for example, posttranscriptional gene silencing, quelling,and transcriptional gene silencing. Common to all of them is the presence of short ~21 to 25 nt RNAs generated by the action of members of the Dicer enzyme family.Those of the RNAi pathway, derived from viral and cellular dsRNA, are known as small interfering RNAs (siRNAs), whereas a large family of small noncoding RNAgenes found in the genomes of animals (a thousand and counting in humans) andplants and their viruses are processed into micro-(mi)RNAs.

These various small RNAs are bound by Argonaut (Ago) protein family mem-bers, part of the RNA-induced silencing complex (RISC). RISC binds target RNAs through sequence complementarity with the bound small RNA, and silencing is effected through degradation (siRNAs) and/or repression of translation of the target RNA (miRNAs can use both mechanisms). siRNAs can also direct transcription-al gene silencing through the formation of repressive heterochromatin and/or themethylation of DNA. This mode of silencing provides a form of genome protec-tion, in suppressing the action of transposons, retro-elements and other repeatedsequences.

RNAi-driven resistance to an initial viral infection in a single leaf can spreadthroughout the whole plant, and although the nature of the mobile signal is not yet known, it presumably involves a species of nucleic acid homologous to theviral target. Similarly, spreading of the RNAi signal beyond the initial site of in-oculation is seen in the nematode C. elegans, which can also take up RNA directly from the environment—these process involving the membrane proteins Sid-1 andSid-2. Furthermore, RNAi-induced silencing in C. elegans can be inherited for many generations, although the inheritance does not itself require genes in the RNAipathway.

Tapping into the RNAi pathway has provided the biological research community with a powerful tool for manipulating gene expression. Genomewide RNAi-basedscreens where RNAi-targeted genes are decreased in abundance in tissue culturecells are now common, with results providing insights into fundamental cellular processes as well as fueling research into the treatment of human disease (withspecific miRNAs themselves implicated in tumorigenesis). Indeed, siRNAs as drugs are now moving into clinical trials. Issues of delivery remain a problem, althoughvarious chemical modifications and conjugations can prolong half-lives and enhancecell up-take. Surprisingly, mucosal tissues are particularly permissive in their ability to absorb even unmodified RNAs, which retain potent biological activity.

In a few short years RNAi has become a standard technique in the molecular genetic toolkit and a highly active area of basic research in its own right. It may also become part of our pharmaceutical armory in the future and will very likely continue to surprise us with its functions in the cell.

Guy RiddihoughSenior Editor, Science

A Species of Small Antisense

RNA in Posttranscriptional

Gene Silencing in Plants

Andrew J. Hamilton and David C. Baulcombe*

Posttranscriptional gene silencing (PTGS) is a nucleotide sequence–

specific defense mechanism that can target both cellular and viral mRNAs.

Here, three types of transgene-induced PTGS and one example of virus-in-

duced PTGS were analyzed in plants. In each case, antisense RNA comple-

mentary to the targeted mRNA was detected. These RNA molecules were of a

uniform length, estimated at 25 nucleotides, and their accumulation required

either transgene sense transcription or RNA virus replication. Thus, the 25-

nucleotide antisense RNA is likely synthesized from an RNA template and may

represent the specificity determinant of PTGS.

Posttranscriptional gene silencing occurs in plants and fungi transformed with foreignor endogenous DNA and results in the reduced accumulation of RNA molecules withsequence similarity to the introduced nucleic acid (1, 2). Double-stranded RNA inducesa similar effect in nematodes (3), insects (4), and protozoa (5). PTGS can be suppressedby several virus-encoded proteins (6) and is closely related to RNA-mediated virusresistance and cross-protection in plants (7, 8). Therefore, PTGS may represent a natu-ral antiviral defense mechanism and transgenes might be targeted because they, or theirRNA, are perceived as viruses. PTGS could also represent a defense system againsttransposable elements and may function in plant development (9–11).

To account for the sequence specificity and posttranscriptional nature of PTGS, ithas been proposed that antisense RNA forms a duplex with the target RNA, therebypromoting its degradation or interfering with its translation (12). If these hypotheti-cal antisense RNA molecules are of a similar size to typical mRNAs, they would havebeen readily detected by routine RNA analyses. However, there have been no reports ofsuch antisense RNA that is detected exclusively in plants or animals exhibiting PTGS.Nevertheless, PTGS-specific antisense RNA may exist, but may be too short for easydetection. We carried out analyses specifically to detect low molecular weight antisenseRNA in four classes of PTGS in plants (13). The first class tested was transgene-inducedPTGS of an endogenous gene (“cosuppression”). We used five tomato lines (T1.1, T1.2,T5.1, T5.2, and T5.3), each transformed with a tomato 1-aminocyclopropane- 1-car-boxylate oxidase (ACO) cDNA sequence placed downstream of the cauliflower mo-saic virus 35S promoter (35S). Two lines (T5.2 and T5.3) exhibited PTGS of theendogenous ACO mRNA (Fig. 1A). Low molecular weight nucleic acids purified fromthe five lines were separated by denaturing polyacrylamide gel electrophoresis, blotted,and hybridized to an ACO sense (antisense-specific) RNA probe (Fig. 1B). A discrete,ACO antisense RNA (14) of 25 nucleotides (nt) was present in both PTGS lines butabsent from the nonsilencing lines. Twenty-five–nucleotide ACO RNA of sense polarityand at the same abundance as the 25-nt ACO antisense RNA was also present only inthe PTGS lines (Fig. 1C).

PTGS induced by transgenes can also occur when a transgene does not havehomology to an endogenous gene (1). Therefore, we tested whether this type of PTGSwas also associated with small antisense RNA. We analyzed three tobacco lines carry-

Sainsbury Laboratory, John Innes Centre, Colney Lane, Norwich NR4 7UH, UK.* To whom correspondence should be addressed. E-mail: [email protected]

ing 35S-ßglucuronidase (GUS) transgenes. Two of these lines, T4 (15) and 6b5 (16),exhibited PTGS of GUS. The third line (6b5×271) tested was produced by crossing6b5 with line 271 (17), in which there is a transgene suppressor of the 35S promoter in6b5. There was no PTGS of GUS in 6b5×271 because of the transcriptional suppres-sion of the 35S GUS transgene (18). Hybridization with a GUS-specific probe revealedthat low molecular weight GUS antisense RNA was present in T4 and 6b5 (Fig. 2, lanes1 and 2) but absent from line 6b5×271 (Fig. 2, lane 3). The amount of antisense RNAcorrelated with the extent of PTGS: Line 6b5 has stronger PTGS of GUS than line T4(18) and had more GUS antisense RNA (Fig. 2). As for PTGS of ACO in tomato, theGUS antisense RNA was a discrete species of ~25 nt.

In some examples of PTGS, silencing is initiated in a localized region of theplant. A signal molecule is produced at the site of initiation and mediates systemicspread of silencing to other tissues of the plant (19, 20). Weinvestigated whether system-ic PTGS of a transgene en-coding the green fluorescentprotein (GFP) is associatedwith 25-nt GFP antisenseRNA. PTGS was initiated inNicotiana benthamiana ex-pressing a GFP transgene byinfiltration of a single leafwith Agrobacterium tumefa-ciens containin GFP sequenc-es in a binary plant transfor-mation vector (19). Two to 3 weeks after this infiltration, the GFPfluorescence disappeared owing to systemic spread of PTGS as de-scribed (11, 20). We detected 25-nt GFP antisense RNA in systemictissues exhibiting PTGS of GFP. It was not detected in equivalent leaves of plants thathad not been infiltrated or in nontransformed plants that had been infiltrated with A.tumefaciens (Fig. 3).

A natural manifestation of PTGS is the RNA-mediated defense induced in virus-infected cells (8). Therefore we investigated whether virus-specific, 25-nt RNA couldbe detected in a virus-infected plant. Twenty-five–nucleotide RNA complementaryto the positive strand (genomic) of potato virus X (PVX) was detected 4 days afterinoculation of N. benthamiana and continued to accumulate for at least another 6 days

Fig. 1. Twenty-fi ve–nucleotide ACO antisense andsense RNA in PTGS lines. (A) Endogenous ACO mRNAabundance in fi ve tomato lines containing 35S-ACOtransgenes. ACO mRNA was amplified by reverse tran-scriptase–polymerase chain reaction and detectedby hybridization with labeled ACO cDNA. (B and C)Low molecular weight RNA from the same fi ve lines and a30-nt ACO antisense RNA were fractionated, blotted, andhybridized with either ACO sense RNA (B) or antisenseRNA (C) transcribed from full-length ACO cDNA. The low hybridization temperature permitted some nonspecific hy-bridization to tRNA and small ribosomal RNA species,which constitute most of the RNA mass in these fractions.The oligonucleotide hybridized only to the antisense-spe-cific probe (B). Twenty-fi ve–nucleotide, PTGS-specificRNA is indicated.

Fig. 2. Twenty-fi ve–nucleotide an-tisense GUS RNA is dependent on transcription from the 35Spromoter. Twenty-fi ve–nucleotideGUS antisense RNA was detectedby hybridization with hydrolyzedGUS sense RNA transcribed fromthe 3´ 700 base pairs of theGUS cDNA.

in the inoculated leaf (Fig. 4). Twenty-five–nucleotide PVX RNA accumulated to asimilar extent in systemically infected leaves but was not detected in mock-inoculatedleaves.

Thus, 25-nt antisense RNA, complementary to targeted mRNAs, accumulates infour types of PTGS. We have detected 25-nt RNA in other examples of PTGS (22),and never detected 25-nt RNA in the absence of PTGS. This correlation and the prop-erties of 25-nt RNA are consistent with a direct role for 25-nt RNA in PTGS inducedby transgenes or viruses (12). Twenty-five–nucleotide RNA species also serve as mo-lecular markers for PTGS. Their presence could be used to confirm other examples oftransgene- or virus-induced PTGS and perhaps also to identify endogenous genes thatare targeted by PTGS in nontransgenic plants.

The 25-nt antisense RNA species are not degradation products of the target RNAbecause they have antisense polarity. A more likely source of these RNAs is thetranscription of an RNA template. This is consistent with the presence of the 25-nt PVXRNA in PVX-infected cells that do not contain a DNA template (Fig 4, “syst. leaf ”).The dependency of 25-nt GUS antisense RNA accumulation on sense transcription of aGUS transgene also supports the RNA template model (Fig. 2).An RNAdependent RNA

Fig. 3. Twenty-fi ve–nucleotide antisense GFP RNA in systemi-cally silenced tissue. Lower leaves of untransformed N. ben-thamiana (WT) and N. benthamiana carrying an active 35S-GFP transgene (35S-GFP) were infiltrated with A. tumefacienscontaining the same 35S-GFP transgene in a binary vector.RNA from upper, noninfiltrated leaves of these plants (inf.)and from equivalent leaves of noninfiltrated plants (–) was hybridized with GFP sense RNA transcribed from a full-lengthGFP cDNA. Only the transgenic N. benthamiana infiltrated withthe A. tumefaciens accumulated 25-nt GFP antisense RNA.

Fig. 4. Twenty-fi ve–nucleotide antisense PVX RNAaccumulates during virus replication. RNA was extracted from inoculated leaves after 2, 4, 6, and10 days and from systemic (syst.) leaves after 6and 10 days (d.p.i.: days post inoculation). RNAwas extracted from mock-inoculated leaves after2 days. Twenty-fi ve–nucleotide PVX antisense RNAwas detected by hybridization with PVX sense RNAtranscribed from a full-length PVX cDNA.

polymerase, as required by this model, is required for PTGS in Neurospora crassa (23).With the present data, we cannot distinguish whether the antisense RNA is made di-rectly as a 25-nt species or as longer molecules that are subsequently processed. Theprecise role of 25-nt RNA in PTGS remains to be determined. However, because theyare long enough to convey sequence specificity yet small enough to move throughplasmodesmata, it is possible that they are components of the systemic signal and speci-ficity determinants of PTGS.

References and Notes1. H. Vaucheret et al., Plant J. 16, 651 (1998).2. C. Cogoni and G. Macino, Trends Plant Sci. 2, 438 (1997).3. A. Fire et al., Nature 391, 806 (1998).4. J. R. Kennerdell and R. W. Carthew, Cell 95, 1017 (1998).5. H. Ngo, C. Tschudi, K. Gull, E. Ullu, Proc. Natl. Acad. Sci. U.S.A. 95, 14687 (1998).6. G. Pruss, X. Ge, X. M. Shi, J. C. Carrington, V. B. Vance, Plant Cell 9, 859 (1997); R.

Anandalakshmi et al., Proc. Natl. Acad. Sci. U.S.A. 95, 13079 (1998); K. D. Kasschau and J. C. Carrington, Cell 95, 461 (1998); G. Brigneti et al., EMBO J. 17, 6739 (1998);C. Beclin, R. Berthome, J.-C. Palauqui, M. Tepfer, H. Vaucheret, Virology 252, 313(1998).

7. J. A. Lindbo, L. Silva-Rosales, W. M. Proebsting, W. G. Dougherty, Plant Cell 5,1749 (1993).

8. F. Ratcliff, B. D. Harrison, D. C. Baulcombe, Science 276, 1558 (1997); S. N. Covey, N.S. Al-Kaff, A. Langara, D. S. Turner, Nature 385, 781 (1997); F. Ratcliff, S. MacFarlane,D. C. Baulcombe, Plant Cell 11, 1207 (1999).

9. R. B. Flavell, Proc. Natl. Acad. Sci. U.S.A. 91, 3490 (1994).10. R. A. Jorgensen, R. G. Atkinson, R. L. S. Forster, W. J. Lucas, Science 279, 1486

(1998).11. O. Voinnet, P. Vain, S. Angell, D. C. Baulcombe, Cell 95, 177 (1998).12. D. Grierson, R. G. Fray, A. J. Hamilton, C. J. S. Smith, C. F. Watson, Trends Biotechnol.

9, 122 (1991); W. G. Dougherty and T. D. Parks, Curr. Opin. Cell Biol. 7, 399 (1995);D. C. Baulcombe and J. J. English, Curr. Opin. Biotechnol. 7, 173 (1996).

13. Total RNA was extracted from leaves of tomato, tobacco, and N. benthamiana asdescribed [E. Mueller, J. E. Gilbert, G. Davenport, G. Brigneti, D. C. Baulcombe,Plant J. 7, 1001 (1995)]. From these preparations, low molecular weight RNA was enriched by ion-exchange chromatography on Qiagen columns after removal ofhigh molecular weight species by precipitation with 5% polyethylene glycol 8000–0.5 M NaCl (for tobacco and N. benthamiana) or by filtration through Centricon100 concentrators (Amicon) (for tomato). Low molecular weight RNA was sepa-rated by electrophoresis through 15% polyacrylamide–7 M urea–0.5× tris-borateEDTA gels, transferred onto Hybond Nx filters (Amersham), and fixed by ultraviolet cross-linking. Prehybridization was in 45% formamide, 7% SDS, 0.3 M NaCl, 0.05M Na2HPO4–NaH2PO4 (pH 7), 1× Denhardt’s solution, and sheared, denatured,salmon sperm DNA (100 mg/ml) at between 30° and 40°C. Hybridization was inthe same solution with single-stranded RNA probes transcribed with α-32P-labeleduridine triphosphate. Before addition to the filters in the prehybridization solution,probes were hydrolyzed to lengths averaging 50 nt. Hybridization was for 16 hours at 30°C (ACO probes), 35°C (GUS probe), or 40°C (GFP and PVX probes). Sizes ofRNA molecules were estimated by comparison with 33P-phosphorylated DNA oli-gonucleotides run on the same gels but imaged separately. Additionally, samples from different types of PTGS, including those shown, were frequently run on thesame gel. Alignment of the filters after hybridization with different specific probes confirmed that the PTGS-specific signals were identical in size. The probes used arein each case sequence specific. We have observed no cross-hybridization between25-nt signals in different PTGS systems using either filter hybridization or RNAaseprotection (www.sciencemag.org/feature/data/1042575.shl). We do not have an ex-act measurement of the amount of 25-nt RNA per cell, but given the short exposuretimes routinely used to detect these molecules and taking into account their size,they are likely to be abundant in cells exhibiting PTGS.

14. The 25-nt ACO antisense signal was completely abolished by pretreatment with ei-ther RNAaseONE (Promega) or NaOH.

15. S. L. A. Hobbs, T. D. Warkentin, C. M. O. DeLong, Plant Mol. Biol. 21, 17 (1993).

16. T. Elmayan and H. Vaucheret, Plant J. 9, 787 (1996).17. H. Vaucheret, C. R. Acad. Sci. Paris 316, 1471 (1993).18. J. J. English, G. F. Davenport, T. Elmayan, H. Vaucheret, D. C. Baulcombe, Plant J.

12, 597 (1997).19. O. Voinnet and D. C. Baulcombe, Nature 389, 553 (1997).20. J.-C. Palauqui and S. Balzergue, Curr. Biol. 9, 59 (1999).21. A high-titer, synchronized PVX infection on leaves of untransformed N. benthami-

ana was initiated by infiltration of single leaves with A. tumefaciens containing abinary plasmid incorporating a 35S-PVX-GFP sequence. Once transcribed, the PVXRNA replicon is independent of the 35S-PVX-GFP DNA, replicates to high levels,and moves systemically through the plant. The A. tumefaciens does not spreadbeyond the infiltrated patch and is not present in systemic leaves (20). The GFP re-porter in the virus was used to allow visual monitoring of infection progress. We haveobtained similar signals with wild-type PVX inoculated as virions in sap taken froman infected plant.

22. The other examples of PTGS tested were in N. benthamiana (spontaneous silencing ofa 35S-GFP transgene), tomato (35S-ACO containing an internal direct and invertedrepeat), petunia (cosuppression of chalcone synthase transgenes and endogenes),and Arabidopsis thaliana (PTGS of 35S-GFP by a 35S-PVX-GFP transgene).

23. C. Cogoni and G. Macino, Nature 399, 166 (1999).24. We thank D. Grierson, C. DeLong, H. Vaucheret, and R. Hellens for transgenic plants.

We are also grateful to O. Voinnet, D. Bradley, A. Bendahmane, and Ratcliff for helpfulcomments and suggestions. This work was carried out under M.A.F.F. licence PHL24A/2921. Funded by the Biotechnology and Biological Sciences Research Council

11 June 1999; accepted 14 September 1999.

Argonaute2 Is the Catalytic Engine of

Mammalian RNAi

Jidong Liu,1* Michelle A. Carmell,1,2* Fabiola V. Rivas,1 Carolyn

G. Marsden,1 J. Michael Thomson,3 Ji-Joon Song1, Scott M.

Hammond,3 Leemor Joshua-Tor,1 Gregory J. Hannon1†

Gene silencing through RNA interference (RNAi) is carried out by RISC,

the RNA-induced silencing complex. RISC contains two signature com-

ponents, small interfering RNAs (siRNAs) and Argonaute family proteins.

Here, we show that the multiple Argonaute proteins present in mammals

are both biologically and biochemically distinct, with a single mammalian

family member, Argonaute2, being responsible for messenger RNA cleav-

age activity. This protein is essential for mouse development, and cells

lacking Argonaute2 are unable to mount an experimental response to

siRNAs. Mutations within a cryptic ribonuclease H domain within

Argonaute2, as identified by comparison with the structure of an archeal

Argonaute protein, inactivate RISC. Thus, our evidence supports a model

in which Argonaute contributes “Slicer” activity to RISC, providing the cat-

alytic engine for RNAi.

1Cold Spring Harbor Laboratory, Watson School of Biological Sciences, 1 BungtownRoad, Cold Spring Harbor, NY 11724, USA. 2Program in Genetics, Stony BrookUniversity, Stony Brook, NY 11794, USA. 3Department of Cell and DevelopmentalBiology, University of North Carolina, Chapel Hill, NC 27599, USA.*These authors contributed equally to this work.†To whom correspondence should be addressed. E-mail: [email protected]

Science 286, 950-952 (1999).

and the Gatsby Charitable Foundation.

The presence of double-stranded RNA (dsRNA) in most eukaryotic cells provokesa sequence-specific silencing response known as RNA interference (RNAi) (1, 2).The dsRNA trigger of this process can be derived from exogenous sources or transcribedfrom endogenous noncoding RNA genes that produce microRNAs (miRNAs) (1, 3).

RNAi begins with the conversion of dsRNA silencing triggers into small RNAs of~21 to 26 nucleotides (nts) in length (4). This is accomplished by the processing of trig-gers by specialized ribonuclease III (RNase III)–family nucleases, Dicer and Drosha(5, 6). Resulting small RNAs join an effector complex, known as RISC (RNA-inducedsilencing complex) (7). Silencing by RISC can occur through several mechanisms. Inflies, plants, and fungi, dsRNAs can trigger chromatin remodeling and transcriptionalgene silencing (8–11). RISC can also interfere with protein synthesis, and this is thepredominant mechanism used by miRNAs in mammals (12, 13). However, the best stud-ied mode of RISC action is mRNA cleavage (14, 15). When programmed with a smallRNA that is fully complementary to the substrate RNA, RISC cleaves that RNA at adiscrete position, an activity that has been attributed to an unknown RISC component,“Slicer” (16, 17). Whether or not RISC cleaves a substrate can be determined by thedegree of complementarity between the siRNA and mRNA, as mismatched duplexesare often not processed (16). However, even for mammalian miRNAs, which normallyrepress at the level of protein synthesis, cleavage activity can be detected with a substratethat perfectly matches the miRNA sequence (18). This result prompted the hypothesisthat all RISCs are equal, with the outcome of the RISC-substrate interaction beingdetermined largely by the character of the interaction between the small RNA and itssubstrate.

RISC contains two signature components. The first is the small RNA, which co-fractionated with RISC activity in Drosophila S2 cell extracts (7), and whose presencecorrelated with dsRNA-programmed mRNA cleavage in Drosophila embryo lysates(14, 15). The second is an Argonaute (Ago) protein, which was identified as a com-ponent of purified RISC in Drosophila (19). Subsequent studies have suggested thatArgonautes are also key compnents of RISC in mammals, fungi, worms, protozoans,and plants (17, 20).

Argonautes are often present as multiprotein families and are identified by two char-acteristic domains, PAZ and PIWI (21). These proteins mainly segregate into two sub-families, comprising those that are more similar to either Arabidopsis Argonaute1 orDrosophila Piwi. The Argonaute family was first linked to RNAi through genetic studiesin Caenorhabditis elegans, which identified Rde-1 as a gene essential for silencing (22).Our subsequent placement of a Drosophila Argonaute protein in RISC (19) promptedus to explore the roles of this protein family. Toward this end, we have undertaken bothbiochemical and genetic studies of the Ago1 subfamily proteins in mammals.

Mammals contain four Argonaute1 subfamily members, Ago1 to Ago4 [nomencla-ture as in (23); see fig. S1]. We have previously shown that different Argonaute familymembers in Drosophila preferentially associate with different small RNAs, with Ago1preferring miRNAs and Ago2 siRNAs (24). Recent studies of Drosophila melanogaster(dm) Ago1 and dmAgo2 mutants have strengthened these conclusions (25). To assesswhether mammalian Ago proteins specialized in their interactions with small RNAs, weexamined Ago-associated miRNA populations by microarray analysis. Ago1-, Ago2-and Ago3- associated RNAs were hybridized to microarrays that report the expressionstatus of 152 human microRNAs. Patterns of associated RNAs were identical within ex-perimental error in each case (Fig. 1A). Additionally, each of the tagged Ago proteins as-sociated similarly with a cotransfected siRNA (Fig. 1C).

Previous studies have used tagged siRNAs to affinity purify Argonaute-contain-ing RISC (17). These preparations, containing mixtures of at least two mammalianArgonautes, were capable of cleaving synthetic mRNAs that were complementary tothe tagged siRNA. We examined the ability of purified complexes containing individ-ual Argonaute proteins to catalyze similar cleavages. Unexpectedly, irrespective of thesiRNA sequence, only Ago2-containing RISC was able to catalyze cleavage (Fig. 1B

and fig. S2). All three Ago proteins were similarly expressed and bound similar amountsof transfected siRNA (Fig. 1, C and D).

These results demonstrated that mammalian Argonaute complexes are biochemical-ly distinct, with only a single family member being competent for mRNA cleavage. Toexamine the possibility that Ago proteins might also be biologically specialized, we dis-rupted the mouse Ago2 gene by targeted insertional mutagenesis (fig. S3 and Fig. 2A)(26). Intercrosses of Ago2 heterozygotes produced only wild-type and heterozygousoffspring, strongly suggesting that disruption of Ago2 produced an embryonic-lethalphenotype.

Ago2-deficient mice display several developmental abnormalities beginning approx-imately halfway through gestation. Both gene-trap and in situ hybridization data of day9.5 embryos show broad expression of Ago2 in the embryo, with some hot spots of ex-pression in the forebrain, heart, limb buds, and branchial arches (Fig. 2, F and G). Themost prominent phenotype is a defect in neural tube closure (Fig. 2, D and E), oftenaccompanied by apparent mispatterning of anterior structures, including the forebrain(Fig. 2, C and D). Roughly half of the embryos display complete failure of neural tubeclosure in the head region (Fig. 2E), while all embryos display a wavy neural tube inmore caudal regions. Mutant embryos also suffer from apparent cardiac failure. Thehearts are enlarged and often accompanied by pronounced swelling of the pericardialcavity (Fig. 2C). By day 10.5, mutant embryos are severely developmentally delayedcompared with wild-type and heterozygous littermates (Fig. 2B). This large differencein size, like the apparent cardiac failure, may be accounted for by a general nutritionaldeficiency caused by yolk sac and placental defects (27), as histological analysis revealsabnormalities in these tissues.

Not all Argonaute proteins are required for successful mammalian development (28,29). Thus, it is unclear why Ago2 should be required for development, while other Agoproteins are dispensable. Ago subfamily members are expressed in overlapping patternsin humans (30). In situ hybridization demonstrates overlapping expression patterns for

Fig. 1. Only mammalian Ago2 canform cleavage-competent RISC. (A)The miRNA populations associated withAgo1, Ago2, and Ago3 were measured by microarray analysis as described in (44).The heat map shows normalized log-ratiovalues for each data set, with yellow rep-resenting increased relative amounts and blue indicating decreased amounts relative to the median. The top 25 logratios are shown in the expanded region.In each panel, “control” indicates parallelanalysis of cells transfected with a vectorcontrol. (B) The 293T cells were transfect-ed with a control vector or with vectors encoding myc-tagged Ago1, Ago2, orAgo3, along with an siRNA that targets firefl y luciferase. Immunoprecipitates weretested for siRNA-directed mRNA cleav-age as described in (44). Positions of 5´and 3´ cleavage products are shown. (C)Immunoprecipitates as in (B) were testedfor in vivo siRNA binding by Northern blot-ting of Ago immunoprecipitates (44). (D)Western blots of transfected cell lysates show similar levels of expression for eachrecombinant Argonaute protein.

Ago2 and Ago3 in mouse embryos (Fig. 2F and fig. S4). Considered together withthe essentially identical patterns of miRNA binding, our results suggest the possibil-ity that the ability of Ago2 to assemble into catalytically active complexes might becritical for mouse development. Although most miRNAs regulate gene expression at the level of protein synthesis, recently miR196 has been demonstrated to cleave themRNA encoding HoxB8, a developmental regulator (31). Evolutionary conservation ofan essential cleavage-competent RISC in organisms in which miRNAs predominantly act by translational regulation raises the possibility that target cleavage by mammalianmiRNAs might be more important and widespread than previously appreciated.

Numerous studies have indicated that experimentally triggered RNAi in mammali-an cells proceeds through siRNA-directed mRNA cleavage because in many, but not all,cases, reiterated binding sites are necessary for repression at the level of protein synthe-sis [see, for example (13, 32, 33)]. If Ago2 were uniquely capable of assembling intocleavage-competent complexes in mice, then embryos or cells lacking Ago2 might beresistant to experimental RNAi. To address this question, we prepared mouse embryofibroblasts (MEFs) from E10.5 embryos from Ago2 heterozygous intercrosses. Reversetranscription polymerase chain reaction (RT-PCR) analysis and genotyping revealedthat we were able to obtain wild-type, mutant, and heterozygous MEF populations.Importantly, MEFs also express other Ago proteins, including Ago1 and Ago3 (Fig. 3A).Ago2-null MEFs were unable to repress gene expression in response to an siRNA (Fig.3B and fig. S5). This defect could be rescued by the addition of a third plasmid that en-coded human Ago2 but not by a plasmid encoding human Ago1 (Fig. 3B). In contrast,responses were intact for a reporter of repression at the level of protein synthesis, medi-ated by an siRNA binding to multiple mismatched sites (32) (Fig. 3C).

Fig. 2. Argonaute2 is essentialfor mouse development. (A)Total RNA from wild-type or mu-tant embryos was tested for ex-pression of Ago1, Ago2, or Ago3by RT-PCR. Actin was also ex-amined as a control. (B) At dayE10.5, Ago2-null embryos showsevere developmental delay ascompared with heterozygous and wild-type littermates. Theseembryos also show a variety ofdevelopmental defects, includ-ing swelling inside the pericar-dial membrane (indicated by arrow) (h, heart) (C) and failure to close the neural tube(D and E). Arrows in (D) indicate the edges of the neural tube that has failed to close.In caudal regions, where the neural tube does close, it has an abnormal appearance,being wavy as compared with wild-type embryos (E) (compare wild-type and Ago2 –/–). Ago2 is expressed in most tissues of the developing embryo as measured by insitu hybridization (F) or by analysis of an Ago2 gene-trap animal (G). In (F), f is fore-brain, b is branchial arches, h is heart, and lb is limb bud, all of which are relative hot spots for Ago2 mRNA. In (G), the left embryo shows similar patterns when staining for the gene-trap marker, β-galactosidase, proceeds for only a short period. Longer incu-bation (G, right) gives uniform staining throughout the embryo.

Because Ago2 is exceptional in its ability to form cleavage-competent complexes, weset out to map the determinants of this capacity. Deletion analysis indicated that an in-tact Ago2 was required for RISC activity (fig. S6). We therefore used the sequence ofhighly conserved but cleavage-incompetent Ago proteins as a guide to the constructionof Ago2 mutants. A series of point mutations included H634P, H634A, Q633R, Q633A,H682Y, L140W, F704Y, and T744Y. Whereas all of these mutations retain siRNA-bind-ing activity and most retain cleavage activity, changes at Q633 and H634 have a pro-found effect on target cleavage (Fig. 4). Both the Q633R and H634P mutations, in whichresidues were changed to corresponding residues in Ago1 and Ago3, abolished catalysis.Changing H634 to A also inactivated Ago2, whereas a similar change, Q633A, was per-missive for cleavage. Thus, even relatively conservative changes can negate the abilityof Ago2 to form cleavage-competent RISC.

Several possibilities could explain a lack of cleavage activity for Ago2 mutants. Suchmutations could interfere with the proper folding of Ago2. However, this seems un-likely because those same residues presumably permit proper folding in closely relat-ed Argonaute proteins, and mutant Ago2 proteins retained the ability to interact withsiRNAs. Alternatively, cleavage-incompetent Ago2 mutants could lose the ability to in-teract with the putative Slicer. Finally, Ago2 itself might be Slicer, with our conservativesubstitutions altering the active center of the enzyme in a way that prevents cleavage.

The last possibility predicted that we might reconstitute an active enzyme with rel-atively pure Ago2 protein. We immunoaffinity purified Ago2 from 293T cells and at-tempted to reconstitute RISC in vitro. Incubation with the double-stranded siRNA pro-duced no appreciable activity, whereas Ago2 could be successfully programmed withsingle-stranded siRNAs to cleave a complementary substrate (Fig. 5A). Formation ofthe active enzyme was unaffected by first washing the immunoprecipitates with up to2.5 M NaCl or 1 M urea. A 21-nt single-stranded DNA was unable to direct cleavage(Fig. 5A). Programming could be accomplished with different siRNAs that direct activ-ity against different substrates (fig. S7). RISC is formed though a concerted assemblyprocess in which the RISC-loading complex (RLC) acts in an adenosine triphosphate(ATP)–dependent manner to place one strand of the small RNA into RISC (34–36). Invitro reconstitution occurs in the absence of ATP, which suggests that Ago2 could be

Fig. 3. Argonaute2 is essentialfor RNAi in MEFs. (A) RT-PCRof mRNA prepared from wild-type or Ago2�/�MEFs revealsconsistent expression of Ago1and Ago3 but a specific lack ofAgo2 expression in the null

Fig. 3. Argonaute2 is essential for RNAi in MEFs. (A) RT-PCR of mRNA prepared fromwild-type or Ago2–/– MEFs reveals consistent expression of Ago1 and Ago3 but a spe-cific lack of Ago2 expression in the null MEF. Actin mRNA serves as a control. (B) Wild-type and mutant MEFs were cotransfected with plasmids encoding Renilla and firefl y luciferases, either with or without firefl y siRNA. Ratios of firefl y to Renilla activity, normal-ized to 1 for the no-siRNA control, were plotted. For each genotype, the ability of Ago1and Ago2 to rescue suppression was tested by cotransfection with expression vectors encoding each protein as indicated. (C) NIH-3T3 cells, wild-type MEFs, or Ago2 mu-tant MEFs were tested as described in (B) (except that Renilla/firefl y ratios are plotted)for their ability to suppress a reporter of repression at the level of protein synthesis. Inthis case, the Renilla luciferase mRNA contains multiple imperfect binding sites fora CXCR4 siRNA. Cells were transfected with a mixture of firefl y and Renilla luciferaseplasmids with or without the siRNA.

programmed with siRNAs without a need for the normal assembly process (Fig. 5A).However, in vitro reconstitution of RISC still requires the essential characteristics ofan siRNA. For example, single-stranded siRNAs that lack a 5´ phosphate group cannotreconstitute an active enzyme.

Although consistent with the possibility that the catalytic activity of RISC is car-ried within Ago2, these results do not rule out the possibility that a putative Slicer copu-rifies with Ago2. To demonstrate more conclusively that Ago2 is Slicer, we turned to thecrystal structure of an Argonaute protein from an archebacterium, Pyrococcus furiosus(37). This structure revealed that the PIWI domain folds into a structure analogous tothe catalytic domain of RNase H and avian sarcoma virus (ASV) integrase. The notionthat such a domain would lie at the center of RISC cleavage is consistent with previousobservations. RNase H and integrases cleave their substrates, leaving 5´ phosphate and3´ hydroxyl groups through a metal-catalyzed cleavage reaction (38, 39). Notably, previ-ous studies have strongly indicated that the scissile phosphate in the targeted mRNA is

cleaved via a metal ion in RISC to give the same phosphate polarity (40). Our in vitrodata are consistent with the reconstituted RISC also requiring a divalent metal (fig. S8).

The active center of RNase H and its relatives consists of a catalytic triad of threecarboxylate groups contributed by aspartic or glutamic acid (38, 39). These amino acidresidues coordinate the essential metal and activate water molecules for nucleolytic at-tack. Reference to the known structure of RNase H reveals two aspartate residues inthe archeal Ago protein present at the precise spatial locations predicted for formationof an RNase H–like active site (37). These align with identical residues in the humanAgo2 protein (fig. S9). Therefore, to test whether the PIWI domain of Ago2 providescatalytic activity to RISC, we changed the two conserved aspartates, D597 and D669,to alanine, with the prediction that either mutation would inactivate RISC cleavage.Consistent with our hypothesis, the mutant Ago2 proteins were incapable of assemblinginto a cleavage-competent RISC in vitro or in vivo, despite retaining the ability to bindsiRNAs (Fig. 5, B to D).

Considered together, our data provide strong support for the notion thatArgonaute proteins are the catalytic components of RISC. First, the ability to form anactive enzyme is restricted to a single mammalian family member, Ago2. This conclu-sion is supported both by biochemical analysis and by genetic studies in mutant MEFs.

Fig. 4. Mapping the require-ments for assembly of cleavage-competent RISC.Ago1, Ago2, ormutants of Ago2 were expressedas myc-tagged fusion proteins in293T cells. In all cases, expres-sion constructs were cotrans-fected with a luciferase siRNA.Western blotting indicated sim-ilar expression for each mutant.Immunoprecipitates containingindividual proteins were testedfor cleavage activity against a lu-ciferase mRNA (44). Positions of5´ and 3´ cleavage products areindicated. SiRNA binding was examined for each mutant by Northern blotting of immunopre-cipitates or by staining of immu-noprecipitates with Sybr Gold(Molecular Probes, Eugene, Oregon). Representatives for these assays are shown. Inno case did we detect a defect in interaction of mutants with siRNAs.

Second, single amino acid substitutions within Ago2 that convert residues to thosepresent in closely related proteins negate RISC cleavage. Third, the structure of the P.furiosis Argonaute protein reveals provocative structural similarities between the PIWIdomain and the RNase H domains, providing a hypothesis for the method by whichArgonaute cleaves its substrates. We tested this hypothesis by introducing mutations inthe predicted Ago2 active site. It is extremely unlikely that such mutations could affectinteractions with other proteins, because they are buried within a cleft of Ago.

Our studies indicate that the Argonaute proteins that are unable to form cleavage-competent RISC differ from Ago2 at key positions that do not include the putative metal-coordinating residues themselves. However, we cannot yet, based either on biochemicalor structural studies, provide a precise explanation for the catalytic defects in these pro-teins. It is conceivable that Ago1 and Ago3 fail to coordinate the catalytic metal or thatthe structure of the active site is distorted sufficiently that a bound metal is unable to ac-cess the scissile phosphate. Alternatively, catalytic mechanisms with two metal ions havebeen proposed for RNase H (38, 39), which leaves open the possibility that catalyticallyinert Ago family members might lack structures essential to bind the second metal ion.

The relationship between the nuclease domain in PIWI and conserved nuclease do-mains in viral reverse transcriptases, transposases, and viral integrases has potential evo-lutionary implications. In Drosophila, plants, and C. elegans, the RNAi pathway has amajor role in controlling parasitic nucleic acids such as viruses and transposons (41–43).The fact that the RNAi machinery shares a core structural domain with viruses and trans-posons suggests that this nucleic acid immune system may have arisen in part by pirat-ing components from the replication and movement machineries of the very elements

Fig. 5. Argonaute2 is a candidate for Slicer. (A) Ago2 protein was immunoaffinity puri-fied from transiently transfected 293T cells. The preparation contained two major pro-teins (protein gel), in addition to heavy and light chains. These were identified by mass spectrometry as Ago2 and HSP90. Immunoprecipitates were mixed (44) in vitro withsingle- or double-stranded siRNAs or with a 21-nt DNA having the same sequenceas the siRNA. Reconstituted RISC was tested for cleavage activity with a uniformly la-beled synthetic mRNA. Positions of 5´ and 3´ cleavage products are noted. Whereindicated, the siRNA was not 5´ phosphorylated and, in one case, ATP was not addedto the reconstitution reaction. (B) Ago2 or Ago2 mutants were assembled into RISCin vivo by cotransfection with siRNAs, followed by immunoaffinity purification or by invitro reconstitution, mixing affinity-purified proteins with single-stranded siRNAs. Thesemutants were tested for activity against a complementary mRNA substrate. 5´ and3´ cleavage products are as in (A). (C and D) Both mutant proteins were expressed at levels similar to wild-type Ago2 and bound siRNAs as readily. Ago2 (H634P) and Ago2(Q633R) behave similarly in this assay.

that RNAi protects against. This hypothesis is made even more poignant by consideringthe role of RNA-dependent RNA polymerases in RNAi, their functional relationship toviral replicases, and the possibility that the siRNAs themselves might first have served asprimers that enable such replicases to duplicate primordial genomes.

References and Notes1. G. J. Hannon, Nature 418, 244 (2002).2. A. Fire et al., Nature 391, 806 (1998).3. G. Hutvagner, P. D. Zamore, Curr. Opin. Genet. Dev. 12, 225 (2002).4. A. Hamilton, O. Voinnet, L. Chappell, D. Baulcombe, EMBO J. 21, 4671 (2002).5. E. Bernstein, A. A. Caudy, S. M. Hammond, G. J. Hannon, Nature 409, 363 (2001).6. Y. Lee et al., Nature 425, 415 (2003).7. S. M. Hammond, E. Bernstein, D. Beach, G. J. Hannon, Nature 404, 293 (2000).8. M. F. Mette, W. Aufsatz, J. van der Winden, M. A. Matzke, A. J. Matzke, EMBO J. 19,

5194 (2000).9. I. M. Hall et al., Science 297, 2232 (2002).10. T. Volpe et al., Science 297, 1833 (2002).11. M. Pal-Bhadra, U. Bhadra, J. A. Birchler, Mol. Cell 9, 315 (2002).12. P. H. Olsen, V. Ambros, Dev. Biol. 216, 671 (1999).13. D. P. Bartel, Cell 116, 281 (2004).14. T. Tuschl, P. D. Zamore, R. Lehmann, D. P. Bartel, P. A. Sharp, Genes Dev. 13, 3191

(1999).15. P. D. Zamore, T. Tuschl, P. A. Sharp, D. P. Bartel, Cell 101, 25 (2000).16. S. M. Elbashir, J. Martinez, A. Patkaniowska, W. Lendeckel, T. Tuschl, EMBO J. 20,

6877 (2001).17. J. Martinez, A. Patkaniowska, H. Urlaub, R. Luhrmann, T. Tuschl, Cell 110, 563

(2002).18. G. Hutvagner, P. D. Zamore, Science 297, 2056 (2002).19. S. M. Hammond, S. Boettcher, A. A. Caudy, R. Kobayashi, G. J. Hannon, Science

293, 1146 (2001).20. M. A. Carmell, G. J. Hannon, Nature Struct. Mol. Biol. 11, 214 (2004).21. L. Cerutti, N. Mian, A. Bateman, Trends Biochem. Sci 25, 481 (2000).22. H. Tabara et al., Cell 99, 123 (1999).23. M. A. Carmell, Z. Xuan, M. Q. Zhang, G. J. Hannon, Genes Dev. 16, 2733 (2002).24. A. A. Caudy, M. Myers, G. J. Hannon, S. M. Hammond, Genes Dev. 16, 2491 (2002).25. K. Okamura, A. Ishizuka, H. Siomi, M. C. Siomi, Genes Dev. 18, 1655 (2004).26. B. Zheng, A. A. Mills, A. Bradley, Nucleic Acids Res. 27, 2354 (1999).27. S. J. Conway, A. Kruzynska-Frejtag, P. L. Kneer, M. Machnicki, S. V. Koushik, Genesis

35, 1 (2003).28. W. Deng, H. Lin, Dev. Cell 2, 819 (2002).29. S. Kuramochi-Miyagawa et al., Development 131, 839 (2004).30. T. Sasaki, A. Shiohama, S. Minoshima, N. Shimizu, Genomics 82, 323 (2003).31. S. Yekta, I. H. Shih, D. P. Bartel, Science 304, 594 (2004).32. J. G. Doench, C. P. Petersen, P. A. Sharp, Genes Dev. 17, 438 (2003).33. M. Kiriakidou et al., Genes Dev. 18, 1165 (2004).34. A. Nykanen, B. Haley, P. D. Zamore, Cell 107, 309 (2001).35. J. W. Pham, J. L. Pellino, Y. S. Lee, R. W. Carthew, E. J. Sontheimer, Cell 117, 83

(2004).36. Y. Tomari et al., Cell 116, 831 (2004).37. J.-J. Song et al., Science 305, 1434 (2004). Published online 29 July 2004; 10.1126/

science.1102514.38. B. R. Chapados et al., J. Mol. Biol. 307, 541 (2001).39. W. Yang, T. A. Steitz, Structure 3, 131 (1995).40. D. S. Schwarz, Y. Tomari, P. D. Zamore, Curr. Biol. 14, 787 (2004).41. R. F. Ketting, T. H. Haverkamp, H. G. van Luenen, R. H. Plasterk, Cell 99, 133

(1999).42. T. Sijen, R. H. Plasterk, Nature 426, 310 (2003).43. E. Sarot, G. Payen-Groschene, A. Bucheton, A. Pelisson, Genetics 166, 1313

(2004).44. Materials and methods are available as supporting material on Science Online.45. The authors thank members of the Hannon lab for helpful discussions, Alea Mills for

advice on ES cell work and for providing the library of targeting constructs, Sang

Yong Kim for generating chimeras, Kathryn Anderson for insightful discussions andadvice, and Phil Sharp for providing the CXCR4 constructs. M.C. is supported by the U.S. Army Breast Cancer Research Program, F.V.R. by the Jane Coffin Childs Memorial Fund, and J.S. by a Bristol Myers Squibb predoctoral fellowship. S.M.H. is a General Motors Cancer Research Foundation Scholar. This work was supported inpart by grants from NIH (L.J. and G.J.H.).

Supporting Online Materialwww.sciencemag.org/cgi/content/full/1102513/DC1Materials and MethodsFigs. S1 to S9

8 July 2004; accepted 19 July 2004Published online 29 July 2004;10.1126/science.1102513Include this information when citing this paper.

Structural Basis for Double-Stranded

RNA Processing by Dicer

Ian J. MacRae,1,3 Kaihong Zhou,1,3 Fei Li,1 Adrian Repic,1 Angela

N. Brooks,1 W. Zacheus Cande,1 Paul D. Adams,4 Jennifer A.

Doudna1,2,3,4*

The specialized ribonuclease Dicer initiates RNA interference by cleaving dou-

ble-stranded RNA (dsRNA) substrates into small fragments about 25 nucleo-

tides in length. In the crystal structure of an intact Dicer enzyme, the PAZ do-

main, a module that binds the end of dsRNA, is separated from the two catalytic

ribonuclease III (RNase III) domains by a flat, positively charged surface.The 65

angstrom distance between the PAZ and RNase III domains matches the length

spanned by 25 base pairs of RNA. Thus, Dicer itself is a molecular ruler that

recognizes dsRNA and cleaves a specified distance from the helical end.

RNA interference (RNAi) is an ancient gene-silencing process that plays a fundamentalrole in diverse eukaryotic functions including viral defense (1), chromatin remodeling(2), genome rearrangement (3), developmental timing (4), brain morphogenesis (5),and stem cell maintenance (6). All RNAi pathways require the multidomain ribonucle-ase Dicer (7). Dicer first processes input dsRNA into small fragments called short in-terfering RNAs (siRNAs) (8), or microRNAs (miRNA) (9), which are the hallmark ofRNAi. Dicer then helps load its small RNA products into large multiprotein complexestermed RNAinduced silencing complexes (RISC) (10). RISC and RISC-like complex-es use the small RNAs as guides for the sequence-specific silencing of cognate genesthrough mRNA degradation (11), translational inhibition (12), and heterochromatin for-mation (13).

Dicer products are typically 21 to 25 nucleotides long, which is the ideal size for a genesilencing guide, because it is long enough to provide the sequence complexity required

1Department of Molecular and Cell Biology, 2Department of Chemistry, 3HowardHughes Medical Institute, University of California, Berkeley, CA 94720, USA. 4PhysicalBiosciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720,USA.

*To whom correspondence should be addressed: E-mail: [email protected]

to uniquely specify a single gene in a eukaryotic ge-nome. Several models have been proposed for howDicer generates RNA fragments of this specific size(14–16), but structural information is lacking. In aneffort to deepen our understanding of the initiationstep of RNAi, we determined the crystal structureof an intact and fully active Dicer enzyme.

Conservation of a highly active Dicer

in Giardia intestinalis. We identified an open read-ing frame in Giardia intestinalis that encodes thePAZ and tandem RNase III domains characteristicof Dicer (7), but lacks the N-terminal DExD/Hhelicase, C-terminal double-stranded RNA bind-ing domain (dsRBD), and extended interdomainregions associated with Dicer in higher eukaryotes(Fig. 1A). A recombinant form of this protein pos-sesses robust dicing activity in vitro (Fig. 1B). TheRNA fragments produced by Giardia Dicer are 25to 27 nucleotides long, which is similar to a class ofsmall RNAs associated with RNAi-mediated DNAelimination in Tetrahymena (17) and RNA-direct-ed DNA methylation in plants (18). dsRNA cleavageby Giardia Dicer is magnesium-dependent, althoughseveral other divalent cations including Mn2+, Ni2+,and Co2+ also support catalytic activity (19). Thepresence of discrete dicing intermediates sepa-rated by intervals of ~25 nucleotides indicates thatGiardia Dicer processes dsRNA from the helical endin a fashion similar to human Dicer (20). However, incontrast to human Dicer, Giardia Dicer has a low af-finity for its small RNA product (~1 μM) (19) anddisplays multiple turnover kinetics (20).

Structural overview. We determined the crystalstructure of the full-length Giardia Dicer at 3.3 Å resolution (table S1). The structure re-veals an elongated molecule that, when viewed from the front, takes on a shape resem-bling a hatchet; the RNase III domains form the blade and the PAZ domain makes upthe base of the handle (Fig. 2A). The PAZ domain is directly connected to the RNaseIIIa domain by a long α helix that runs through the handle of the molecule. This ‘‘con-nector’’ helix is encircled by the N-terminal residues of the protein, which form a plat-form domain composed of an antiparallel β sheet and three α helices. A large helicaldomain bridges the two RNase III domains and forms the back end of the blade. ViewingDicer from the side reveals a contiguous flat surface that extends along one face of themolecule.

Two–metal-ion mechanism of dsRNA cleavage. The two RNase III domains ofDicer sit adjacent to each other in the blade region and form an internal heterodimer thatis similar to the homodimeric structure of bacterial Rnase III (fig. S1). Although previousbacterial RNase III crystal structures revealed a single catalytic metal ion in each RNaseIII domain (21), subsequent studies implicated two metal ions in the hydrolysis of eachstrand of the dsRNA (22).

During our biochemical characterization of Giardia Dicer, we noticed that the en-zyme is potently inhibited by trivalent lanthanide cations such as Er3+ (19). Lanthanidesoften bind more tightly to cation binding sites than divalent cations do, a property pre-viously used to identify transient Mn2+ binding sites in proteins (23). Inspection of theanomalous difference electron density map from a crystal derivatized with ErCl

3re-

vealed a pair of Er3+ cations in the active site of each RNase III domain of Giardia Dicer

Fig. 1. Giardia encodes an ac-tive Dicer enzyme. (A) Schematicrepresentation of the primary sequence of human and GiardiaDicers. (B) Time course of in vi-tro Giardia Dicer dsRNA cleav-age assay. RNA product sizes were determined by comparisonwith RNase T1 and alkaline hy-drolysis (OH) sequencing lad-ders (lanes 1 and 2). Dicing re-quires the protein (Dcr) and Mg2+

(lanes 3 and 4).

(Fig. 2B). The prominent Er3+ metal (M1) in each domain resides between four strictlyconserved acidic residues, which make up the previously identified Mn2+ binding site ofbacterial RNase III (21). The second Er3+ binding site (M2) lies adjacent to the first, out-side of the acidic residue cluster. The distances between the two Er3+ metals in the RNaseIIIa and IIIb domains are ~4.2 Å and ~5.5 Å , respectively. These distances are similar tothose previously observed in the active site of RNase H (4.1 Å ) (24), avian sarcoma virus(ASV) integrase (3.6 Å ) (25), the restriction enzyme EcoRV (4.2 Å ) (26), and the groupI intron (3.9 Å ) (27), all of which are thought to use a two–metal-ion mechanism ofcatalysis. The 17.5 Å distance between the metal-ion pairs closely matches the width ofthe dsRNA major groove. We also observed Mn2+ in all M1 and some M2 sites in crystalsgrown in high concentrations of MnCl

2. Therefore, we propose that the Er3+ metals seen

in Giardia Dicer denote true catalytic metal-ion binding sites and that Giardia Dicer usesa two–metal-ion mechanism of catalysis. Given the high level of sequence conservationthroughout the RNase III family, it is likely that all RNase III enzymes, including bacte-rial RNase III and Drosha, contain similar catalytic metal-ion binding sites.

Structural features of the Dicer PAZ domain. The PAZ domain is an RNA bind-ing module found in Dicers and in the Argonaute family of proteins that are core com-ponents of RISC and other siRNA- and miRNA-containing complexes. Previous studiesof PAZ domains from several Argonaute proteins revealed a degenerate oligonucleotide/oligosaccharide-binding (OB) fold that specifically recognizes dsRNA ends containing a3´ two-base overhang (28–31). Superposition of the PAZ domains of Giardia Dicer andhuman Argonaute1 reveals that the two domains share the same overall fold and 3´ two-nucleotide RNA binding pocket (Fig. 3A).

The Dicer PAZ domain contains a large extended loop that is conserved amongDicer sequences and absent in Argonaute (fig. S1). The Dicer-specific loop dramaticallychanges the electrostatic potential and molecular surface surrounding the 3´ overhang-binding pocket relative to the Argonaute PAZ domain (Fig. 3B). The presence of manybasic amino acid residues in the extra loop could substantially affect the way the RNA isrecognized and perhaps handed off to other complexes by each family of proteins.

A model for siRNA formation. The structure of Giardia Dicer immediatelysuggests how Dicer enzymes specify siRNA length. Measuring from the active site of

Fig. 2. Crystal structure of Giardia Dicer. (A) Front and side view ribbon representa-tions of Dicer showing the N-terminal platform domain (blue), the PAZ domain (or-ange), the connector helix (red), the RNase IIIa domain (yellow), the RNase IIIb domain(green) and the RNase-bridging domain (gray). Disordered loops are drawn as dot-ted lines. (B) Close-up view of the Dicer catalytic sites; conserved acidic residues (sticks); erbium metal ions (purple); and erbium anomalous difference electron density map, contoured at 20σ (blue wire mesh). Dashed lines indicate distances describedin the text.

the RNase IIIa domain to the 3´ overhang-binding pocket in the PAZ domain gives adistance of ~65 Å (Fig. 4), which matches the length of 25 dsRNA base pairs. To pro-duce a likely model of a Dicer-dsRNA complex, the positions of the metal-ion pairsbound in each RNase III domain were used to anchor the two scissile phosphates ofan ideal A-form dsRNA helix into the RNase III active sites. This placement positionsthe twofold symmetry element of the dsRNA coincident with the pseudo twofold sym-metry axis relating the two RNase III domains, which is analogous to how restrictionenzymes typically bind dsDNA substrates (32). Bacterial RNase III has been proposed tobind dsRNA in a similar fashion (16, 33). Outside of the RNase III region, the modeleddsRNA extends along a flat surface formed by the platform domain. This surface containsa large positively charged region that could interact directly with the negatively chargedphosphodiester backbone of the modeled dsRNA helix. The 3´ end of the RNA duplexfalls directly into the 3´ overhang-binding pocket of the PAZ domain, and the 5´ end liesadjacent to the Dicer-specific PAZ domain loop. There are exactly 25 nucleotides be-

Fig. 3. Structural features of the Dicer PAZ domain. (A) Superposition of the Cα atoms ofPAZ domains from Giardia Dicer (orange) and human Argonaute1 (white). Amino acids forming the 3´ overhang-binding pocket are shown as sticks. (B) Electrostatic surfacerepresentation of the PAZ domains of Giardia Dicer and Argonaute1 (hAGO1). Asterisks denote 3´ overhang-binding pockets. The RNA in Argonaute1 PAZ structure is drawnas green sticks.

Fig. 4. A model for dsRNAprocessing by Dicer.Front and side views ofa surface representationof Giardia Dicer with mod-eled dsRNA. Red and bluerepresent acidic and basicprotein surface charge,respectively. Electrostaticsurface potentials do not include contributions frombound metal ions. Putativecatalytic metal ions areshown as green spheres.White arrows point to scis-sile phosphates. Asteriskdenotes PAZ domain 3´overhang-binding pocket.

tween the 3' end of the helix bound to the PAZ domain and the scissile phosphate in theRNase IIIa domain.

Thus, Dicer is a molecular ruler that measures and cleaves ~25 nucleotides fromthe end of a dsRNA. The length of the small RNAs produced by Dicer is set by thedistance between the PAZ and RNase III domains, which is largely a function of thelength of the connector helix. This model of dsRNA processing is consistent with theproposed architecture of human Dicer based on biochemical studies in which the RNaseIIIa and IIIb domains were shown to produce the siRNA 5´ and 3´ ends, respectively(16). Furthermore, closing the ends of a dsRNA substrate by hybridization or ligationgreatly diminished dicing activity (20, 34), which may explain why circular viral dsRNAis resistant to RNAi (35).

Giardia Dicer can support RNAi in fission yeast. Given that Giardia Dicer lackssome of the domains commonly associated with Dicer enzymes, most notably the N-terminal helicase, we wondered if the structure represents an intact Dicer or merely thecatalytic subunit of a larger complex required for complete Dicer function in vivo. To ad-dress this question, we introduced the Giardia Dicer gene into a strain of the fission yeastSchizosaccharomyces pombe that contains a deletion of its endogenous Dicer (dcrΔ).

Like most Dicer proteins, the S. pombe Dicer contains an N-terminal helicase do-main and a C-terminal dsRBD. The S. pombe dcrΔ strain is defective in RNAi and ishypersensitive to the microtubule-destabilizing drug thiobendazole (TBZ) because ofchromosome missegregation (36). Plasmid expression of S. pombe dcr1+ fully rescuedTBZ sensitivity of the dcrΔ cells. A partial functional rescue of TBZ sensitivity was alsoachieved by episomal expression of Giardia Dicer, indicating that Giardia Dicer cansuppress the chromosome segregation defect (Fig. 5A). Furthermore, Giardia Dicer re-stores silencing of centromeric regions that are aberrantly transcribed in the dcrΔ mutant(Fig. 5B). These results demonstrate that Giardia Dicer is sufficient to function as anintact Dicer in vivo.

A conserved architecture in Dicer enzymes. Considering the structural roleplayed by the connector helix that links the PAZ and RNase III domains (Fig. 2), wewondered whether larger Dicer proteins found in higher eukaryotes contain an analogoushelix. Sequence alignment of the region directly following the PAZ domain of severalevolutionarily diverse Dicer enzymes reveals a conserved pattern of hydrophobic andhydrophilic amino acids that is predicted to form a long α helix by secondary structur-al analysis (fig. S2). All Dicers contain a conserved proline about 11 amino acid res-idues from the predicted N terminus of the helix. In the crystal structure of GiardiaDicer, this proline induces a distinct kink that aids in directing the helix toward theRNase IIIa domain.

Fig. 5. Giardia Dicer supports RNAi in vivo.(A) Overexpression (OE) of Giardia Dicerrescues the TBZ sensitivity of the S. pombeDicer delete (dcrΔ). Growth was assayedby spotting 10-fold serial dilutions of cul-tures indicated. (B) Overexpression ofGiardiaDicerrestores transcriptional silenc-ing at centromeres (cen). Transcript levels were determined by semiquantitativereverse-transcriptase polymerase chainreaction. Actin (act) served as an internalcontrol. bp, base pair.

Most Dicer proteins contain a conserved region of ~100 amino acids termed ‘‘do-main of unknown function 283’’ (DUF283), which lies between the helicase and PAZdomains in the primary sequence. Low but consistent sequence homology between theN-terminal domain of Giardia Dicer and DUF283 (fig. S3) suggests that in the Dicers ofhigher eukaryotes, DUF283 forms a platform structure similar to that of Giardia Dicer.

The conserved Dicer architecture, together with the demonstration that Giardia Dicercan substitute for S. pombe Dicer in vivo, argues that the mechanism of Dicer-catalyzeddsRNA processing is conserved. Moreover, these results indicate that all Dicers evolvedfrom a common ancestral enzyme. Because Giardia is one of the most anciently divergedmembers of the eukaryotic kingdom, we may consider that the earliest eukaryotic organ-isms had a similar Dicer enzyme and therefore were capable of RNAi-like processes.It will be of evolutionary interest to determine the cellular function of Dicer and RNAiin Giardia.

The structure of Giardia Dicer also provides new insight into eukaryotic RNaseIII enzymes in general. This family of enzymes performs a range of specific cellularfunctions involving the cleavage of dsRNA [reviewed in (37)]. The structure of Dicer il-lustrates how the presence of RNA binding modules, like the PAZ and platform domains,can impart a specific function to the otherwise nonspecific double-stranded RNase ac-tivity of the RNase III dimer (38). This is likely to be the structural paradigm for alleukaryotic RNase III enzymes that have specific activities and cellular functions.

References and Notes1. D. Baulcombe, Trends Microbiol. 10, 306 (2002).2. T. A. Volpe et al., Science 297, 1833 (2002).3. K. Mochizuki, N. A. Fine, T. Fujisawa, M. A. Gorovsky, Cell 110, 689 (2002).4. A. Grishok et al., Cell 106, 23 (2001).5. A. J. Giraldez et al., Science 308, 833 (2005).6. S. D. Hatfield et al., Nature 435, 974 (2005).7. E. Bernstein, A. A. Caudy, S. M. Hammond, G. J. Hannon, Nature 409, 363 (2001).8. S. M. Elbashir, W. Lendeckel, T. Tuschl, Genes Dev. 15, 188 (2001).9. G. Hutvagner et al., Science 293, 834 (2001).10. Q. Liu et al., Science 301, 1921 (2003).11. S. M. Hammond, E. Bernstein, D. Beach, G. J. Hannon, Nature 404, 293 (2000).12. R. S. Pillai et al., Science 309, 1573 (2005).13. A. Verdel et al., Science 303, 672 (2004).14. M. A. Carmell, G. J. Hannon, Nat. Struct. Mol. Biol. 11, 214 (2004).15. P. D. Zamore, Mol. Cell 8, 1158 (2001).16. H. Zhang, F. A. Kolb, L. Jaskiewicz, E. Westhof, W. Filipowicz, Cell 118, 57 (2004).17. C. D. Malone, A. M. Anderson, J. A. Motl, C. H. Rexer, D. L. Chalker, Mol. Cell. Biol.

25, 9151 (2005).18. Z. Xie et al., PLoS Biol. 2, E104 (2004).19. I. J. MacRae, K. Zhou, J. A. Doudna, data not shown.20. H. Zhang, F. A. Kolb, V. Brondani, E. Billy, W. Filipowicz, EMBO J. 21, 5875 (2002).21. J. Blaszczyk et al., Structure (Camb) 9, 1225 (2001).22. W. Sun, A. Pertzev, A. W. Nicholson, Nucleic Acids Res. 33, 807 (2005).23. M. Sundaramoorthy, H. L. Youngs, M. H. Gold, T. L. Poulos, Biochemistry 44, 6463

(2005).24. M. Nowotny, S. A. Gaidamakov, R. J. Crouch, W. Yang, Cell 121, 1005 (2005).25. G. Bujacz et al., J. Biol. Chem. 272, 18161 (1997).26. I. B. Vipond, G. S. Baldwin, S. E. Halford, Biochemistry 34, 697 (1995).27. M. R. Stahley, S. A. Strobel, Science 309, 1587 (2005).28. J. B. Ma, K. Ye, D. J. Patel, Nature 429, 318 (2004).29. J. J. Song et al., Nat. Struct. Biol. 10, 1026 (2003).30. A. Lingel, B. Simon, E. Izaurralde, M. Sattler, Nature 426, 465 (2003).31. K. S. Yan et al., Nature 426, 468 (2003).32. A. K. Aggarwal, Curr. Opin. Struct. Biol. 5, 11 (1995).33. D. L. Akey, J. M. Berger, Protein Sci. 14, 2744 (2005).34. A. Repic, J. A. Doudna, unpublished data.35. J. Chang, P. Provost, J. M. Taylor, J. Virol. 77, 11910 (2003).

1Department of Cell Biology, 3Taplin Biological Mass Spectrometry Facility, HarvardMedical School, Boston, MA 02115, USA. 2Laboratory of Molecular Cell Biology,National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA.*To whom correspondence should be addressed. Email: [email protected],[email protected].

36. I. M. Hall, K. Noma, S. I. Grewal, Proc. Natl. Acad. Sci. U.S.A. 100, 193 (2003).37. D. Drider, C. Condon, J. Mol. Microbiol. Biotechnol. 8, 195 (2004).38. W. Sun, E. Jun, A. W. Nicholson, Biochemistry 40, 14976 (2001).39. We thank members of the Doudna and Berger labs for helpful discussions, A. Fischer

for work with tissue culture, and D. King for mass spectrometry analysis. We aregrateful to C. Ralston and J. Dickert for technical support on beam lines 8.2.1 and8.2.2 at the Advanced Light Source at the Lawrence Berkeley National Lab. I.J.M. is a Howard Hughes Medical Institute fellow of the Life Sciences Research Foundation.This work was supported in part by a grant from NIH (to J.A.D.). Dicer coordinates and structure factors have been deposited in the Protein Data Bank with accessioncode 2FFL.

Supporting Online Materialwww.sciencemag.org/cgi/content/full/311/5758/195/DC1Materials and MethodsFigs. S1 to S5Table S1

21 October 2005; accepted 7 December 200510.1126/science.1121638

Include this information when citing this paper.

RNAi-Mediated Targeting of

Heterochromatin by the RITS Complex

André Verdel,1 Songtao Jia,2 Scott Gerber,1,3

Tomoyasu Sugiyama,2 Steven Gygi,1,3 Shiv I. S. Grewal,2*

Danesh Moazed1*

RNA interference (RNAi) is a widespread silencing mechanism that acts at both

the posttranscriptional and transcriptional levels. Here, we describe the puri-

fication of an RNAi effector complex termed RITS (RNA-induced initiation of

transcriptional gene silencing) that is required for heterochromatin assembly in

fission yeast.The RITS complex contains Ago1 (the fission yeast Argonaute ho-

molog), Chp1 (a heterochromatin-associated chromodomain protein), and Tas3

(a novel protein). In addition, the complex contains small RNAs that require

the Dicer ribonuclease for their production.These small RNAs are homologous

to centromeric repeats and are required for the localization of RITS to hetero-

chromatic domains. The results suggest a mechanism for the role of the RNAi

machinery and small RNAs in targeting of heterochromatin complexes and epi-

genetic gene silencing at specific chromosomal loci.

The fission yeast Schizosaccharomyces pombe contains large stretches of heterochro-matin that are associated with telomeres, repetitive DNA elements surrounding centro-meres, and with the silent mating-type loci (1). Assembly of heterochromatin at theseloci involves an orchestrated array of chromatin modifications that lead to the recruit-

Science 311, 195-198 (2006)

ment of two chromodomain histone-binding proteins Swi6, a homolog of the Drosophilaand mammalian HP1 proteins, and Chp1 (2, 3). The RNAi pathway has also been im-plicated in regulation at the DNA and chromatin level in Arabidopsis (4–6), Drosophila(7), and Tetrahymena (8), and in heterochromatin assembly in S. pombe (9, 10).

RNAi silencing is triggered by doublestranded RNA (dsRNA), which is cleavedby the ribonuclease III (RNase III)–like enzyme Dicer to generate small RNA moleculesof ~22 nucleotides (nt) (11–13). These small interfering RNAs (siRNAs), load onto aneffector complex called RISC (RNA-induced silencing complex) that contains an Argo-naute/PIWI family protein and targets cognate mRNAs for inactivation (12–15).

Factors involved in the RNAi pathway in other organisms are required for hetero-chromatin formation in S. pombe. Deletion of any of these factors, such as Dicer (dcr1+),Argonaute (ago1+), and RNA-dependent RNA polymerase (rdp1+), disrupts hetero-chromatin assembly (9, 10). In support of a role for RNAi in heterochromatin assembly,both DNA strands of the S. pombe centromeric repeats are transcribed (9), and siRNAshave been identified that match the S. pombe centromeric repeats (16). Moreover, re-cent experiments suggest that artificial generation of dsRNA from a hairpin constructcan silence homologous sequences by heterochromatin formation in an RNAi-dependentmanner (17). Here, we address the key question of how small RNAs generated by theRNAi machinery initiate heterochromatin assembly in fission yeast.

To identify factors important for RNAi-mediated targeting of heterochromatincomplexes, we reasoned that such factor(s) would act in early steps in heterochroma-tin assembly and would be required for the establishment of heterochromatin-specifichistone modification patterns. The Chp1 protein binds to centromeric repeats and is re-quired for methylation of histone H3-K9 and for localization of Swi6 (3, 18). Moreover,the phenotypes displayed by chp1Δ strains are identical to RNAi mutants. To testwhether Chp1 provides a physical and functional link between RNAi and heterochroma-tin assembly, we used a tandem affinity purification procedure (TAP) and a TAP tag toidentify factors that interact with Chp1 (Fig. 1). Several protein species of about 65, 90,100, and 120 kD were specifically purified from the Chp1-TAP strain (Fig. 1A). Mass

Fig. 1. Purification ofChp1-TAP and identi-fication of associatedproteins. Extracts froma Chp1-TAP strain andan untagged controlstrain were purifiedby the TAP proce-dure and applied toa 4 to 12% poly-acrylamide gel, whichwas stained with col-loidal Coomassie blue(A). The bands in theChp1-TAP purificationwere excised from thegel and sequencedby tandem mass spectrometry (22). The identity of each band is based on multiple sequenced pep-tides and is indicated on the right. *Residual GST-TEV, the protease used for elutionfrom the first affinity column. (B) The Chp1-TAP protein was fully functional for silencingof a centromeric imr::ura4+ reporter gene as indicated by wild-type levels of growthon 5-FOA medium, which only allows growth when ura4+ is silenced. N/S, nonselec-tive medium. (C) Schematic diagram showing the subunits of the RITS complex andtheir conserved motifs. The chromodomain (ChD) in Chp1, the PAZ and PIWI domains in Ago1, and a region of sequence similarity between Tas3 and the mouse OTT (ovary testis transcribed) protein are indicated.

spectrometry of excised gel bands, as well as protein mixtures, identified the 120- and100-kD bands as Chp1, the 90-kD band as Ago1, and the 65-kD band as SPBC83.03c, apreviously uncharacterized protein (Fig. 1, A and C; table S1; figs. S1 and S2), which wenamed Tas3 (targeting complex subunit 3). The ratio of the 120- and 100-kD bands variesfrom experiment to experiment, which suggests that the 100-kD protein is a degradationproduct of Chp1.

To verify that Chp1, Ago1, and Tas3 are associated together in a complex, we con-structed an S. pombe strain that produced a fully functional Tas3-TAP protein (Fig. 2, Aand B). Affinity purification followed by mass spectrometry sequencing identified Ago1and Chp1 as Tas3-associated proteins (Fig. 2C, table S1). N- or C-terminally taggedAgo1 proteins were not functional in centromeric silencing and were not used for purifi-cation experiments. However, identical purification profiles of Chp1-TAP and Tas3-TAPsuggests that Chp1, Ago1, and Tas3 are associated together in a complex, which we havenamed RITS.

Chp1, as well as Ago1 and other components of the RNAi pathway, have previ-ously been shown to be required for the assembly of heterochromatin and silencing ofreporter genes inserted within heterochromatic domains (9, 10, 19, 20). A tas3 deletionstrain carrying the ura4+ reporter gene inserted at innermost (imr) or outermost (otr)

Fig. 2.Purificationof the RITScomplexby usinga Tas3-TAP strainand the requirement oftas3+ in silencing and methylation of H3-K9 and Swi6 localization. Western blot show-ing that (A) the Tas3-TAP and Chp1-TAP proteins areexpressed to similar levels and(B) growth assays showing that Tas3-TAP displays wild-type levels of silencing for a cen-tromeric imrIR::ura4+ reporter gene. (C) Tas3-TAP was purified, and silver-stained pro-tein bands were sequencedby tandemmass spectrometry. *GST-TEV. (D) In tas3Δcells,silencing of a ura4+ reporter gene inserted at the centromeric repeats (imr1R::ura4+ andotr1R::ura4+) is lost, but silencing of the same reporter gene at the silent mating-type in-terval (Kint2::ura4+) is unaffected. Loss of silencing in sir2Δ, chp1Δ, and ago1Δ is shownfor comparison. Loss of silencing results in loss of growth on counterselective 5-FOAmedium. (E) ChIP experiments showing that in tas3Δ cells methylation of histone H3-K9and localization of Swi6 to a ura4+ reporter gene inserted at otr1R and imr1R centro-meric repeats is abolished. In contrast, deletion of tas3+ has little or no effect on H3-K9methylation and Swi6 localization (Kin2::ura4+). ChIP analysis and quantification wereperformed as described previously (26). The ratios of ura4+ or cen signals to ura4DS/E-minigene signal present in the immunoprecipitated DNA(ChIP) and whole-cell extracts (WCE) were used to calculate fold enrichment shown underneath each lane.

centromeric repeats of chromosome 1 (imr1R::ura4+ and otr1R::ura4+, respectively)displayed a loss of silencing of both reporter genes (Fig. 2D) to an extent similar tothat of the deletion of sir2, chp1, or ago1 (Fig. 2D) (9, 10, 19, 21). Further, chromatinimmunoprecipitation (ChIP) showed that Tas3 was required for H3-K9 methylation andSwi6 localization of a ura4+ reporter gene inserted at each of the above loci (Fig. 2E).

As is the case for RNAi mutants (10), deletion of tas3+ had little or no effect on silenc-ing or localization of H3-K9 methylation and Swi6 to the ura4+ reporter gene insertedat the mat locus (Kint2::ura4+) (Fig. 2, D and E). The similarity in phenotypes displayedby tas3Δ, chp1Δ, and RNAi mutants underscores the importance of Tas3 interaction withChp1 and the role of the RITS complex in RNAi-mediated heterochromatin assembly.

Members of the Argonaute family of proteins constitute the core subunit ofRISC, which is associated with small RNA molecules that target it to specific mRNAs(12, 13). To determine whether the RITS complex is associated with small RNA mol-ecules, we subjected Chp1-TAP or control purifications to phenol-chloroform extrac-tion and precipitated the aqueous phase of the extraction containing any nucleic acid.The precipitated material was then labeled with [5´-32P]pCp and T4 RNA ligase (22). As

Fig. 3. Dicer-dependent association of RITS with siRNAs. (A)Small RNAs of ~22 to 25 nt copurify with Chp1-TAP. RNAs iso-lated from untagged control (–) and Chp1-TAP (+) strains were3´ end-labeled with [5´-32P]pCp and separated on 15% dena-turing urea polyacrylamide gel. Lane 1, [γ-32P]ATP–labeled RNA markers (Ambion);lanes 2 and 3, labeling of RNA from whole-cell extract (WCE) (~1/2500 of input); lanes 3 and 4, labeling of RNAs after purification. Bracket on the right side indicates the posi-tion of small RNAs specifically associated with Chp1-TAP. (B) Copurification of smallRNAs with Tas3-TAP. (C) No small RNAs are associated with RITS purified from dcr1Δcells. Parallel purifications were performed from an untagged (control, lane 1) strain as well as chp1-TAP, dcr1+ (lane 2) and chp1-TAP, dcr1Δ (lane 3) cells, and the associ-ated RNAs were [5´-32P]pCp labeled (compare lanes 2 and 3, bracket). (D) Northernblot showing that siRNAs associated with RITS hybridize to 32P-labeled probes cor-responding to centromeric repeat sequences. RNA from untagged control (lane1) andChp1-TAP cells (lane 2), purified as described in (B), was separated on a denaturinggel and electrotransferred to a nylon membrane (22). DNA oligonucleotides with se-quence complementary to the 12 heterochromatic siRNAs identified by Reinhart andBartel (16) were 5´ labeled with [γ-32P]ATP and used as probes for the Northern blot.(E) Southern blot showing that RITS contains siRNAs complementary to the outer cen-tromeric repeats (otr). dg (lanes 2 and 4) and dh (lane 3) repeats, actin (lane 5), andLTRs (lane 6) were amplified by polymerase chain reaction (PCR) from genomic DNA,separated on 1.1% agarose gel, and transferred to nylon membrane. 32P-labeled RITSsiRNAs, obtained by labeling RNAs as described in (A), were separated on a denatur-ing urea gel, eluted, and used as probes for the blot.

shown in Fig. 3A, Chp1-TAP is specifically associated with small RNA molecules rang-ing in size from ~22 to 25 nt. In contrast, the predominant RNA species prepared from awhole-cell extract (total RNA) are 70 to 100 nt in size, most likely representing transferRNA (tRNA) and 5S RNA (Fig. 3A). RNA species, mainly in the size range of abundanttRNAs, as well as a small amount of an RNA species of ~25 nt, were present in boththe untagged control and Chp1-TAP purification and represent nonspecific backgroundbinding (Fig. 3A, lanes 2 to 4). Similar results were obtained when the RITS complexwas purified from a strain producing Tas3-TAP (Fig. 3B).

siRNAsareproducedby the ribonuclease Dicer (12,13).Wepurified theRITS complexfrom a strain that carried a deletion of dcr1+, the only S. pombe gene that codes for Dicer.Deletion of dcr1+ resulted in a loss of small RNA species that specifically copurify withChp1-TAP but had no effect on the presence of nonspecific RNA species, which werealso present in the untagged control purification (Fig. 3C). These results indicated thatthe small RNA species specifically associated with RITS are siRNAs that are producedin a Dcr1-dependent manner.

Sequencing of small RNAs from S. pombe has identified a series of small RNAspecies that are complementary to the centromeric repeat sequences (16). These smallRNAs have been termed heterochromatic siRNAs and are clustered at two regions with-in the centromeric repeats, the dh repeats and a region immediately downstream of thedg repeats. Centromeric siRNAs have been proposed to function in sequence-specifictargeting of homologous DNA regions (i.e., centromeric repeats) for heterochromatinassembly. To determine whether siRNAs associated with RITS originate from cen-tromeric repeats, we first analyzed RITS-associated RNAs on a Northern blot probedwith a mixture of oligonucleotides derived from the centromeric repeats. These oligo-nucleotides were specifically designed to hybridize to siRNAs previously identified byReinhart and Bartel (16). The 32P-labeled oligonucleotide probes specifically hybridized

Fig. 4. The RNAi pathway is re-quired for localization of RITS toheterochromatin. (A) ChIP experi-ments showing that Tas3-TAP is lo-calized to centromeric heterochro-matin in an RNAi-dependent man-ner. Tas3-TAP is associated withura4+ inserted at the otr centromer-ic repeats (otr1::ura4+, left panels)and with native centromeric repeat sequences (cen, right panels) inwild-type (wt) but not ago1Δ, dcr1Δ, or rdp1Δ cells. The ura4DS/E-minigene at theendogenous euchromatic location is used as a control. (B) The RNAi pathway is re-quired for the localization of Chp1-(Flag)3 to centromeric heterochromatin. (C) Tas3 isrequired for the localization of Chp1-(Flag)3 to heterochromatin. Immunoprecipitations were carried out using a Flag-specific antibody from tas3+ and tas3Δ cells. (D) Tas3 isassociated with ura4+ inserted at the imr centromeric region (imr1::ura4+). WCE, whole-cell extract. Fold enrichment values are shown underneath each lane.

to RNA species of ~22 to 25 nt in size present in the Chp1-TAP purification but not withnonspecific RNAs present in the untagged control purification (Fig. 3D).

As a second test for the identities of the siRNAs associated with RITS, we la-beled RITS-associated siRNAs with [5´-32P]pCp, then gel purified and used them toprobe a Southern blot containing equal amounts of DNA fragments (ranging in size from300 to 700 base pairs) corresponding to the dg and dh centromeric repeats, the regiondownstream of dg repeats to which siRNAs map (designated dg-D), retrotransposon longterminal repeats (LTRs) that have been shown to mediate RNAi-dependent gene silenc-ing (17), and DNA fragments corresponding to actin and molecular size markers. Thelabeled siRNAs specifically hybridized to dg, dh, and dg-D centromeric sequences (Fig.3E). No hybridization was detected to LTR, actin, or DNA size markers (Fig. 3E). Our inability to detect hybridization of RITS-associated siRNAs with LTR sequences may bedue to a relatively lower abundance of LTR siRNAs compared with siRNAs that origi-nate from the centromeric repeats. Together, these experiments show that RITS is associ-ated with siRNAs that originate from processing of centromeric dsRNA transcripts.

We next used S. pombe strains that produced either Tas3-TAP or Chp1-Flag to deter-mine the in vivo chromatin localization of the RITS complex and the requirement for theRNAi pathway in its localization. It has previously been shown that Chp1 localizes to thecentromeric repeat regions and together with the Clr4 methyltransferase is required for H3-K9 methylation and Swi6 localization (3). ChIP experiments showed that Tas3-TAPis similarly localized to a ura4+ reporter gene inserted within in the otr centromeric repeat region (otr1::ura4+) and centromeric repeat sequences but not to the control mini-ura4(ura4DS/E) gene at the endogenous euchromatic location (Fig. 4). Tas3-TAP, like Chp1(18), is also localized to the imr centromeric repeats (Fig. 4D). Furthermore, deletion ofago1+, dcr1+, or rdp1+ abolished the association of Chp1-Flag and Tas3-TAP withotr1::ura4+, as well as with centromeric repeat sequences (Fig. 4, A and B). These results indicated that the RNAi pathway is required for association of the Chp1 and Tas3 sub-units of RITS with heterochromatic DNA regions. Our purification of the RITS complexfrom dcr1Δ cells showed that the protein subunits of the complex remained associatedtogether in the absence of siRNAs (fig. S4). The purification results, together with theChIP analysis, indicate that the “empty” RITS complex is inactive and can only associatewith its chromosomal target after it is programmed by siRNAs.

We further tested whether Tas3 was required for the localization of Chp1-Flagto each of the above regions. Deletion of tas3+ abolished the association of Chp1-Flagwith otr1::ura4+, as well as with native cen sequences (Fig. 4C). These results support the biochemical identification of Tas3 as an integral subunit of RITS and indicate that it plays an essential role in localizing the complex to heterochromatin.

Our analysis suggests a remarkably direct role for the RNAi machinery in hetero-chromatin assembly. By analogy to RISC complexes, which use small RNAs as guides to target specific mRNAs for degradation or translational repression, we proposethat RITS uses siRNAs to recognize and to bind to specific chromosome regions so as to initiate heterochromatic gene silencing (Fig. 5). Four lines of evidence support this view. First, RITS contains Ago1, the S. pombe homolog of the Argonaute family ofproteins, which form the common subunit of RISC complexes purified from different organisms and are thought to be directly responsible for target recognition (12). Second,RITS is associated with siRNAs that require Dcr1 for their formation and originatefrom heterochromatin repeat regions. Thus, this complex contains the expected specific-ity determinants, i.e., siRNAs, which in other systems have been shown to direct target recognition (14, 15, 23, 24). Third, at least two subunits of the RITS complex, Chp1and Tas3, are specifically associated with the expected heterochromatic DNA regions,which suggests that the complex localizes directly to its target DNA. Fourth, in additionto Ago1, RITS contains a chromodomain protein, Chp1, which is localized throughout heterochromatic DNA regions (18) (Fig. 4) and requires the methyltransferase Clr4 andhistone H3-K9 methylation for localization to chromatin (3, 18). Thus, RITS contains

both a subunit (Ago1) that binds to siRNAs and can function in RNA or DNA target-ing by sequence-specific pairing interaction and a subunit (Chp1) that associates withspecifically modified histones and may be involved in further stabilizing its associationwith chromatin (Fig. 5).

Mechanisms analogous to the RITS-mediated targeting of heterochromatin complex-es are likely to be conserved in other systems. For example, in Tetrahymena, genome-wide DNA elimination during macronucleus development requires an Argonaute familyprotein, Twi1, and a chromodomain protein, Pdd1, both of which are also required forH3-K9 methylation and accumulation of small RNAs corresponding to target sequences(8, 25). Similarly, in Drosophila repeat-induced transcriptional gene silencing requiresan Argonaute family protein, Piwi, and a chromodomain protein, Polycomb (7). Ourresults support the hypothesis that Argonaute proteins form the core subunit of a num-ber of different effector complexes that use sequence-specific recognition to target eitherRNA or DNA.

References and Notes1. S. I. S. Grewal, J. Cell. Physiol. 184, 311 (2000).2. S. I. S. Grewal, D. Moazed, Science 301, 798 (2003).3. J. F. Partridge, K. S. Scott, A. J. Bannister, T. Kouzarides, R. C. Allshire, Curr. Biol. 12,

1652 (2002).4. D. Zilberman, X. Cao, S. E. Jacobsen, Science 299, 716 (2003).5. M. Matzke, A. J. M. Matzke, J. M. Kooter, Science 293, 1080 (2001).6. F. E. Vaistij, L. Jones, D. C. Baulcombe, Plant Cell 14, 857 (2002).7. M. Pal-Bhadra, U. Bhadra, J. A. Birchler, Mol. Cell 9, 315 (2002).8. K. Mochizuki, N. A. Fine, T. Fujisawa, M. A. Gorovsky, Cell 110, 689 (2002).9. T. A. Volpe et al., Science 297, 1833 (2002).10. I. M. Hall et al., Science 297, 2232 (2002).11. A. Fire et al., Nature 391, 806 (1998).12. G. J. Hannon, Nature 418, 244 (2002).13. P. D. Zamore, Science 296, 1265 (2002).14. S. M. Hammond, S. Boettcher, A. A. Caudy, R. Kobayashi, G. J. Hannon, Science

293, 1146 (2001).15. G. Hutvágner, P. D. Zamore, Science 297, 2056 (2002).16. B. J. Reinhart, D. P. Bartel, Science 297, 1831 (2002).17. V. Schramke, R. Allshire, Science 301, 1069 (2003).18. J. F. Partridge, B. Borgstrom, R. C. Allshire, Genes Dev. 14, 783 (2000).19. G. Thon, J. Verhein-Hansen, Genetics 155, 551 (2000).20. C. L. Doe et al., Nucleic Acids Res. 26, 4222 (1998).

Fig. 5. A model for siRNA-dependent initiation of heterochromatin assembly by RITS. The RITS complex is programmedby Dcr1-produced siRNAs to target spe-cific chromosome regions by sequence-specific interactions involving either siR-NA-DNA or siRNA-nascent transcript (blue arrows) base pairing. Nuc, nucleo-some; red triangle, K9-methylation on theamino terminus of histone H3. See text for further discussion and references.

21. G. D. Shankaranarayana, M. R. Motamedi, D. Moazed, S. I. S. Grewal, Curr. Biol. 13,1240 (2003).

22. Materials and methods are available as supporting material on Science Online.23. D. S. Schwarz, G. Hutvágner, B. Haley, P. D. Zamore, Mol. Cell 10, 537 (2002).24. J. Martinez, A. Patkaniowska, H. Urlaub, R. Luhrmann, T. Tuschl, Cell 110, 563

(2002).25. S. D. Taverna, R. S. Coyne, C. D. Allis, Cell 110, 701 (2002).26. J. Nakayama, J. C. Rice, B. D. Strahl, C. D. Allis, S. I. S. Grewal, Science 292, 110

(2001).27. We thank M. Ohi, K. Gould, C. Hoffman, and D. Wolf for gifts of strains and plas-

mids; members of the Moazed, Grewal, and Reed laboratories for support andencouragement; R Ohi and El C. Ibrahim for advice; El C. Ibrahim and M. Wahifor comments on the manuscript; and C. Centrella for technical help. A.V. was supported by a postdoctoral fellowship from INSERM and is now a fellow of theHuman Frontier Science Programme. This work was supported by grants from theNIH (S.I.S.G. and D.M.) and a Carolyn and Peter S. Lynch Award in Cell Biology andPathology (D.M.). D.M. is a scholar of the Leukemia and Lymphoma Society.

Supporting Online Materialwww.sciencemag.org/cgi/content/full/1093686/DC1Materials and MethodsFigs. S1 to S4Tables S1 and S2

14 November 2003; accepted 5 December 2003Published online 2 January 2004;10.1126/science.1093686Include this information when citing this paper.

IntroductionThe technology of RNA interference emerged in its earliest form following a 1998 study inCaenorhabditis elegans1 and has since rapidlyevolved to its current form as a revolutionary toolfor studying gene function, biological pathways,and the physiology of disease. Work and refinementof the RNAi technology has exploded in recentyears. We now know the basic mechanism of theendogenous RNAi pathway,2,3 that the pathway ispresent in most eukaryotes,4 and how cellularmachinery can be harnessed to silence geneexpression without triggering the cell’s antiviralresponse mechanism.5 Further advances haveshown that siRNAs can be expressed from DNAvectors within the host cell, providing methods forlonger term silencing, inducible silencing, and aplasmid DNA format that can be replicated forunlimited supply (compared to synthetic siRNA). Inaddition, these vector-based RNAi platforms may beintegrated with viral delivery systems6 allowing theresearcher to perform gene knockdown in a myriadof cell lines. Studies of endogenous microRNAs(miRNAs) suggested that synthetic or expressedmiRNA mimics could be used to induce the RNAipathway rather than directly using the standard 21bp siRNA sequence. Short hairpin RNAs (shRNAs)7

are structurally related to miRNA and can beexpressed from pol II or pol III promoters.

DiscussionAs the portfolio of RNAi methods continues to expand,options become available for even the most complexsystems being studied. Until recently, synthetic siRNAwas the RNAi vehicle most broadly applicable to awide variety of systems and applications. However,obstacles for using synthetic siRNA include being anon-renewable resource, the transient nature ofsilencing, and the difficulty faced in transfectingprimary cells and non-dividing cell lines such asneurons, lymphocytes and macrophages. In addition,in vivo knockdown studies are particularly cumbersome.

For those facing the above hurdles, DNA vector-based shRNA methods provide the necessary

MISSION™ TRC shRNA Library: Next Generation RNA InterferenceBy Stephanie Uder, Henry George, and Betsy BoedekerSigma-Aldrich Corporation, St. Louis, MO, USA

sigma.com/rnai

solutions. shRNA expression vectors may bepropagated in Escherichia coli and thus provide anunlimited supply of DNA for transfection. In addition,such vectors provide selectable markers for stableshRNA expression and gene silencing. One of themost attractive features of plasmid-based systems isthe coupling of the technology to viral delivery systems.Vectors containing appropriate viral packagingsignals and regulatory elements may be used topackage the shRNA sequence into infectiousvirions. When appropriately pseudotyped, theseviral particles can transduce a broader spectrum ofcell lines and overcome issues faced in standardtransfection methods. Viral delivery systems havebeen extensively studied for gene therapy researchand have thus undergone numerous modificationsfor safety and use.

The lentiviral system, pseudotyped with the VSV-Genvelope protein, presents one of the mostattractive systems for viral packaging and deliveryof shRNA constructs. This is due to its broadtropism and receptor independent delivery, itsability to integrate into the genome for stable gene

5' -

3' - UU

Sense Strand

Antisense Strand

Sense StrandCCGGNNNNNNNNNNNNNNNNNNNNNCTCGAGNNNNNNNNNNNNNNNNNNNNNT T T T TGGCCNNNNNNNNNNNNNNNNNNNNNGAGCTCNNNNNNNNNNNNNNNNNNNNNAAAAA

Antisense StrandLoop

Loo

p

cppt

pLKO.1-puro7,091 bp

hPGK

puroR

pUC ori

RRE

SIN/3’ LTR

RSV/5' LTR

U6

f1 ori

ampR

(Ψ) Psi

Figure 1. The pLKO.1-puro Vector for transient or stable expression of shRNA.


INNOVATION @ WORK

silencing, and the fact that it does not require amitotic event for integration into the genome, whichextends its use to both dividing and non-dividingcell lines. The lentiviral system is also not known toelicit immune responses minimizing concerns of off-target effects and use in in vivo applications.

In an effort to help further the development anddistribution of tools for RNAi research, Sigma-Aldrich announced its membership and sponsorshipof The RNAi Consortium (TRC) in March 2005. TRCis a research collaboration between The BroadInstitute and renowned scientists and their

expanded to 150,000 clones targeting 15,000annotated human genes (MISSION TRC-Hs1.0) and15,000 annotated mouse genes (MISSION TRC-Mm1.0).

Approximately 72,600 clones targeting 10,500human and 5,300 mouse genes are currentlyavailable. The libraries include a broad range ofgene families, functional classes, and druggabletargets. MISSION shRNA constructs are designedusing a rules-based algorithm for efficientknockdown and to minimize off-target effects. Up to five shRNA sequences are individually clonedinto pLKO.1-puro (Figure 1) for broad coverage ofeach target gene and varying degrees of knockdown(Figure 2). The hairpin structure includes anintramolecular 21 bp stem and 6 base loop that isrecognized and cleaved by the enzyme Dicer uponexpression via the U6 (pol III) promoter in the hostcell. The resulting siRNA duplex then continues inthe RNAi pathway by association with the RNAi-induced Silencing Complex (RISC). The puromycinresistance marker is present for stable selection inmammalian cells while the ampicillin resistancemarker provides for plasmid propagation in E. coli.The constructs may be used for transient or stabletransfection of mammalian cells. In addition,pLKO.1-puro allows the generation of lentiviralparticles to infect a wide variety of cells, enablingstable long-term expression of the shRNA.

AcknowledgementsThe authors would like to thank David Root andcolleagues at the Broad Institute, Sheila Stewart ofWashington University, and all of TRC for theirscientific consultation and collaboration.

References1.Fire, A., et al., Potent and specific genetic interference by double-stranded

RNA in Caenorhabditis elegans. Nature, 391, 806-811 (1998).2.Bernstein, E., et al., Role for a bidentate ribonuclease in the initiation step

of RNA interference. Nature, 409, 363-366 (2001).3.Hammond, S.M., et al., An RNA-directed nuclease mediates post-

transcriptional gene silencing in Drosophila cells. Nature, 404, 293-296 (2000).4. Hannon, G.J., RNA Interference. Nature, 418, 244-251 (2002).5. Williams, B.R., PKR; a sentinel kinase for cellular stress. Oncogene, 18,

6112-6120 (1999).6. Stewart, S.A., et al., Lentivirus-delivered stable gene silencing by RNAi in

primary cells. RNA, 9, 493-501 (2003).7. Brummelkamp, T.R., et al., A system for stable expression of short

interfering RNAs in mammalian cells. Science, 296, 550-553 (2002).

MISSION is a trademark belonging to Sigma-Aldrich Co. and its affiliateSigma-Aldrich Biotechnology LP.The RNAi Consortium shRNA library is produced and distributed underlicense from the Massachusetts Institute of Technology.

Perc

enta

ge

Exp

ress

ion

Lev

el

1 2 3 4

100

110

90

80

70

60

50

40

30

20

10

0

72 hours 144 hours

5 Control

Silencing of MAPK1 Using MISSION shRNA Constructs

Quantitative RT-PCR Results Normalized to GAPDH

Figure 2. The Human Mitogen-Activated Protein Kinase 1 (MAPK 1, NM_138957)MISSION shRNA set (5 individual hairpins) was used to achieve gene silencingof MAPK 1.

laboratories from Harvard Medical School,Massachusetts Institute of Technology, TheWhitehead Institute, Dana Farber Cancer Institute,Massachusetts General Hospital, WashingtonUniversity, and Columbia University. Four otherleading life science research organizations aresponsoring members in addition to Sigma-Aldrich.The mission of TRC is to create comprehensive toolsfor genomics medicine, make them broadlyavailable to scientists worldwide, and to pioneerapplications of these tools to the study of disease.The consortium is creating a comprehensive libraryof RNAi reagents with an anticipated completiondate of March 2007. As a scientific collaboratorand commercial partner, Sigma-Aldrich is assistingthe development, manufacturing, and globaldistribution of TRC’s human and mouse lentiviralvector-based shRNA libraries (MISSION™ TRC). Thecollection is designed by the Broad Institute of MITand Harvard and is in the process of being

The future of

is here.www.aaas.org/future

Innovate!

INNOVATION @ WORK




with Sigma, the new leader in RNAiinnovate with our workflow solutions

sigma.com/rnai

Innovation is at the core of scientific advancement. By becoming theonly fully licensed provider of both siRNA and shRNA reagents forRNAi, we at Sigma are facilitating such advancement. Allow us toshow you how our extensive product offering can enable innovationat every step of your workflow.

Whether you are determining gene function, analyzing signaltransduction or screening for potential drug targets, why not discoverhow Sigma’s innovative approach can facilitate your breakthroughs.

• Principal collaborator andmember of The RNAiConsortium (TRC) for globaldistribution of pre-clonedMISSION™ shRNA libraries andfuture advancements in RNAi

• State-of-the-art siRNAmanufacturing capabilities at Sigma Proligo

• Partnerships with key patentholders providing access tomajor RNAi technologiesincluding lentiviral shRNAdelivery

• Comprehensive upstream anddownstream workflow solutions

Sigma is committed to help you innovate.

Create!

INNOVATION @ WORK




sigma.com/rnai

with Sigma, the new leader in RNAicreate your advantage

• Taking siRNA manufacturing toa new level by providing arapid turnaround, highthroughput and cost effectiveservice that caters to yoursiRNA needs

• MISSION™ TRC shRNA libraries,comprising 150,000 pre-clonedshRNA constructs targeting

15,000 human genes and 15,000 mouse genes

• Lentiviral shRNA delivery thatboasts flexibility of long andshort term silencing, 100%transduction efficiency andenables experimentation withdifficult to study cell types suchas non-dividing or primary cells

Faster siRNA manufacturing? 100% transduction efficiency of shRNAconstructs? Long and short term silencing?

Sigma has developed the most comprehensive array of cutting edgeproducts for every step of your RNAi experimental design – creatingfor you a real advantage.

So whether you are determining gene function, analyzing signaltransduction or screening for potential drug targets, why not discoverhow you can create your RNAi advantage.

RNAi Collection - Science · Science RNAi Collection Supported by Sigma-Aldrich RNA interference...

Documents

Transcript of RNAi Collection - Science · Science RNAi Collection Supported by Sigma-Aldrich RNA interference...