Downloaded from
`
`genesdev.cshlp.org
`
`
`
` on May 15, 2018 - Published by Cold Spring Harbor Laboratory Press
`
`
`
`RESEARCH COMMUNICATION
`
`RISC is a 5ⴕ
`phosphomonoester-producing
`RNA endonuclease
`Javier Martinez and Thomas Tuschl1
`Laboratory of RNA Molecular Biology, The Rockefeller
`University, New York, New York 10021, USA
`
`Gene silencing in the process of RNA interference is me-
`diated by a ribonucleoprotein complex referred to as
`RNA-induced silencing complex (RISC). Here we de-
`scribe the molecular mechanism of target RNA cleavage
`using affinity-purified minimal RISC from human cells.
`Cleavage proceeds via hydrolysis and the release of a
`3ⴕ-hydroxyl and a 5ⴕ-phosphate terminus. Substitution of
`the 2ⴕ-hydroxyl group at the cleavage site by 2ⴕ-deoxy
`had no significant effect, suggesting that product release
`and/or a conformational transition rather than a chemi-
`cal step is rate-limiting. Substitution by 2ⴕ-O-methyl at
`the cleavage site substantially reduced cleavage, which
`is presumably due to steric interference. Mutational
`analysis of the target RNA revealed that mismatches
`across from the 5ⴕ or the 3ⴕ end of the siRNA had little
`effect and that substrate RNAs as short as 15 nucleotides
`were cleaved by RISC.
`
`Supplemental material is available at http://www.genesdev.org.
`
`Received January 21, 2004; revised version accepted March
`29, 2004.
`
`RNAi is a conserved posttranscriptional gene silencing
`mechanism triggered by the appearance of double-
`stranded RNA (dsRNA) in most eukaryotic cells (Cerutti
`2003; Denli and Hannon 2003; Dykxhoorn et al. 2003).
`The dsRNA is first processed by Dicer into short du-
`plexes of 21–23-nt in length with characteristic 2-nt 3⬘-
`hydroxyl overhanging ends and 5⬘-phosphate termini, re-
`ferred to as siRNA duplexes (Hammond et al. 2000;
`Zamore et al. 2000; Bernstein et al. 2001; Elbashir et al.
`2001a). Subsequently, one strand of the siRNA duplex is
`incorporated into the RNA-induced silencing complex
`(RISC) to guide the sequence-specific cleavage of comple-
`mentary target or substrate mRNAs (Elbashir et al.
`2001a; Nykänen et al. 2001; Hutvágner and Zamore
`2002; Martinez et al. 2002; Liu et al. 2003; Schwarz et al.
`2003). The substrate RNA cleavage site has been mapped
`10 nt upstream of the nucleotide opposed to the 5⬘-most
`residue of the guide siRNA (Elbashir et al. 2001b).
`RISC has been partially purified from Drosophila me-
`lanogaster Schneider cells as a 500-kD ribonucleopro-
`tein complex that catalyses the complete degradation of
`a complementary RNA (Hammond et al. 2000, 2001).
`The complex contains Argonaute 2 (Ago2; Hammond et
`
`[Keywords: RNAi; endonuclease; 2⬘ modification]
`1Corresponding author.
`E-MAIL ttuschl@rockefeller.edu; FAX (212) 327-7652.
`Article published online ahead of print. Article and publication date are
`at http://www.genesdev.org/cgi/doi/10.1101/gad.1187904.
`
`al. 2001), the Vasa Intronic Gene product (VIG; Caudy et
`al. 2002), the Drosophila homolog of the Fragile X Men-
`tal Retardation Protein (FMRP), dFXR (Caudy et al.
`2002), and the Tudor Staphylococcal Nuclease, Tudor
`SN (Caudy et al. 2003). Tudor SN has been regarded as a
`potential candidate for the endonuclease component of
`RISC, because it contains five repeats of a staphylococ-
`cal/micrococcal nuclease domain. Chromatographic size
`fractionation using D. melanogaster embryonic extracts
`suggested a smaller, <230 kD, size for RISC (Nykänen et
`al. 2001). Human RISC, which has been affinity-purified
`from HeLa cell extracts, only showed an apparent mo-
`lecular mass of 160 kD, containing a single-stranded
`siRNA, 100 kD eIF2C1 and/or eIF2C2, both members of
`the Argonaute protein family, and a yet to be identified
`endonuclease (Martinez et al. 2002).
`Sequence and chemical variants of siRNAs have been
`previously used to study the specificity and structural
`requirements of the guide RNA embedded in RISC. The
`results are not readily summarized, but dependent on
`the number and positions of modifications, differences
`in gene silencing efficiency were observed. Substitution
`of 2⬘-hydroxyl by 2⬘-fluoro groups (Capodici et al. 2002;
`Braasch et al. 2003; Chiu and Rana 2003; Harborth et al.
`2003) appears to be extremely well tolerated, whereas
`the introduction of 2⬘-deoxy residues or the bulkier 2⬘-
`O-methyl and 2⬘-O-allyl group could be more disruptive
`(Amarzguioui et al. 2003; Chiu and Rana 2003; Czaud-
`erna et al. 2003; Holen et al. 2003). These studies also
`show that the sense siRNA strand is frequently more
`tolerant to chemical modifications than the guide anti-
`sense siRNA strand, indicating different requirements
`for assembly of RISC and targeting by RISC.
`Here we describe the mechanism by which the sub-
`strate RNA is recognized and cleaved by the minimal
`RISC that was prepared by stringent affinity-purification
`using biotinylated siRNAs and HeLa cell cytoplasmic
`extract. We determined kinetic parameters and sequence
`and structural constraints, showing that substrates as
`short as 15 nt are cleaved by RISC, and that cleavage
`releases a 5⬘-phosphomonoester and a 3⬘-hydroxyl group.
`
`Results and Discussion
`Human RISC was assembled by incubation of a 3⬘-bioti-
`nylated siRNA duplex in HeLa cell cytoplasmic extract,
`followed by biotin-streptavidin affinity chromatography
`(Martinez et al. 2002). We noticed that the potassium
`chloride concentration in the wash steps of the affinity
`column purification could be increased up to 2.5 M with-
`out loss of RISC activity. We refer to this surprisingly
`stable and highly purified complex as minimal RISC.
`The high-stringency wash removed a significant amount
`of unspecific cellular nucleases, and possibly some RISC-
`specific but partly dispensable proteins, which allowed
`us to use conventionally radiolabeled target RNAs that
`would otherwise be unstable in less purified RISC frac-
`tions. To determine the minimal length for target RNAs
`to be cleaved by RISC, we first tested a 5⬘ 32P-labeled
`21-nt RNA (S21) that was identical in sequence to the
`sense strand of the siRNA duplex used for formation of
`RISC. Encouraged by the observation that this substrate
`was effectively cleaved, we prepared a 21-nt substrate
`site-specifically labeled at the scissile bond located 10 nt
`
`GENES & DEVELOPMENT 18:975–980 © 2004 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/04; www.genesdev.org
`
`975
`
`Alnylam Exh. 1077
`
`

`

`
`
`Downloaded from genesdev.cshlp.org on May 15, 2018 - Published by Cold Spring Harbor Laboratory Press
`
`
`
`
`
`
`
`Martinez and Tuschl
`
`upstream of the nucleotide paired up with the 5⬘-most
`nucleotide of the guide siRNA (Elbashir et al. 2001a).
`The preparation of this 21-nt substrate and the RISC
`cleavage assay are schematized in Figure 1A.
`The cleavage reaction of S21 is expected to yield a 9-nt
`5⬘ cleavage product and a 12-nt 3⬘ cleavage product, but
`depending on the mechanism of target RNA cleavage,
`the radiolabeled phosphate might be found at either one
`of the two products. Comigration analysis indicated that
`the radiolabeled phosphate was present as a 5⬘-phosphate
`at the 12-nt 3⬘ cleavage product (Fig. 1B, lanes 2–4), in-
`dicating that the reaction proceeded via hydrolysis re-
`leasing cleavage products carrying 3⬘-hydroxyl and 5⬘-
`phosphate termini.
`Thin-layer chromatography of the RNase T2-treated
`12-nt cleavage product confirmed that the radiolabeled
`terminal nucleotide contained a 5⬘-phosphate based on
`comigration with synthetic adenosine 3⬘,5⬘-diphosphate
`(Fig. 1C).
`
`In order to visualize both of the cleavage products si-
`multaneously, we also 5⬘-32P-labeled the internally la-
`beled substrate. As expected, the radiolabeled 9-nt 5⬘
`cleavage product comigrated precisely with the indepen-
`dently 5⬘-32P-labeled 9-nt oligoribonucleotide (Fig. 1B,
`lanes 1,5), conclusively demonstrating that minimal
`RISC cleaves substrate RNA only at a single site. Release
`of the cleavage products in a more cellular environment
`would expose the cleavage fragments to further degrada-
`tion by unspecific nucleases, including 3⬘ to 5⬘ exonucle-
`ases that can degrade RNAs with free 3⬘-hydroxyl but
`not 3⬘-phosphate termini
`(Schoenberg and Cherno-
`kalskaya 1997).
`To determine the kinetic parameters of target RNA
`cleavage, initial velocities of S21 cleavage were deter-
`mined as a function of substrate concentrations varied in
`the low nanomolar range (Supplementary Fig. S1). Fitting
`the initial rates to the Michaelis-Menten equation, a
`maximum velocity Vmax of 0.017 ± 0.002 nM/min
`and a Michaelis constant KM of 1.1 ± 0.3 nM were
`obtained (Supplementary Fig. S1A). Examination
`of a second independent preparation of RISC
`yielded a Vmax of 0.43 ± 0.02 nM/min, and a Mi-
`chaelis constant KM of 2.3 ± 0.3 nM (Supplemen-
`tary Fig. S1C). The differences in Vmax are due to
`differences in the concentration of RISC in the dif-
`ferent preparations. Using the more active prepa-
`ration of RISC, the fit of initial rates at substrate
`concentrations above 5 nM intersected with the
`ordinate at a concentration of ∼0.4 nM, which rep-
`resents the concentration of active enzyme. The
`presence of this burst is consistent with product
`release being rate-limiting under our experimental
`conditions. It was not possible to measure the
`burst for the low-activity RISC preparation be-
`cause of its more than 10-fold lower activity.
`It is important to note that the cleavage reac-
`tions did not proceed to completion, not even at
`the lowest substrate concentration tested (0.02
`nM). This could be explained by the presence of
`substrate–RNA-binding complexes that are not as-
`sociated with a nuclease, or free single-stranded
`siRNA. Presumably, these copurified complexes
`compete irreversibly during the time course of
`the experiment for binding of the RNA sub-
`strate to active RISC. The IC50 concentration, at
`which a 2⬘-O-methyl oligoribonucleotide irrevers-
`ibly blocked our low- and high-activity RISC was
`determined to be 0.4 and 0.7 nM, respectively
`(Hutvágner et al. 2004; Meister et al. 2004). The
`IC50 concentrations are higher than the estimated
`concentration of RISC by burst analysis, suggest-
`ing competitive binding of the substrate by a non-
`catalytic affinity-copurified single-stranded anti-
`sense siRNA or siRNA–protein complex.
`For subsequent comparative analysis of modi-
`fied substrates, we used the low-activity RISC and
`1 nM constant substrate concentrations, and we
`stopped the reaction before plateau levels of sub-
`strate conversion were reached. Changes in the
`cleavage fraction therefore should reflect maxi-
`mum changes in the cleavage rate.
`To determine the minimal length of substrate
`structurally required for cleavage by RISC, we pre-
`pared truncated substrates removing 2-nt seg-
`ments from either the 3⬘ or the 5⬘ end (Fig. 2A).
`
`Figure 1. Target RNA is cleaved endonucleolytically producing 5⬘-phosphate
`and 3⬘-hydroxyl termini. (A) Preparation of site-specifically labeled substrates
`and cleavage assay. 5⬘-32P-labeled and 3⬘ aminolinker (L) protected 12-nt oli-
`goribonucleotide was ligated to nonphosphorylated 9-nt oligoribonucleotide
`using T4 RNA ligase. An aliquot of the ligation product was further 5⬘-32P-
`labeled using T4 polynucleotide kinase. The purified substrates were incubated
`with affinity-purified RISC programmed with single-stranded guide siRNA. (B)
`PhosphorImaging of cleavage reactions incubated fo r 2 h at 30°C, andresolved
`on a 20% denaturing polyacrylamide gel. 5⬘-32P-labeled 9- and 12-nt oligoribo-
`nucleotides were loaded as marker in lanes 1 and 2, respectively. The cleavage
`reactions with single- and double-labeled 21-nt substrate are loaded in lanes 4
`and 5, respectively. Lane 3 contains the 12-nt cleavage product isolated from a
`prior cleavage reaction. (C) Two-dimensional thin layer chromatography analy-
`sis of the ribonuclease T2-digested RISC-cleavage product. The oval depicts the
`unlabeled pAp as detected by UV shadowing. The radioactive signal comigrates
`with the 5⬘ 32pAp released by ribonuclease T2 digestion from the gel-purified
`12-nt cleavage product.
`
`976
`
`GENES & DEVELOPMENT
`
`

`

`
`
`Downloaded from genesdev.cshlp.org on May 15, 2018 - Published by Cold Spring Harbor Laboratory Press
`
`
`
`
`
`
`
`RISC target RNA cleavage mechanism
`
`point mutations introduced proximal to the cleavage site
`reduced cleavage 9.3-fold for S21–7 and 70-fold for S21–8
`relative to S21 (Fig. 3B, lanes 7,8). It is surprising to find
`that a mutation more proximal to the cleavage site (S21–
`7) had a less significant effect on cleavage than the mu-
`tation in S21–8 that was located 2 nt away from the
`cleavage site. Mutation and truncation analyses con-
`verge to a similar picture defining the minimal require-
`ments for substrate targeting by RISC, and emphasize
`the need for devoting special attention to the identifica-
`tion of unique 13-nt sequence segments positioned at the
`center segment of the targeting siRNA in siRNA gene-
`targeting experiments.
`Chemical synthesis of short RNA substrates offers
`unique advantages for site-specific introduction of
`single-atom changes, such as the replacement of a single
`2⬘-hydroxyl group by 2⬘-hydrogen or 2⬘-O-methyl. We
`prepared several substrates by substituting the 2⬘-ribose
`residue of S21 at G9, A10, or A11 by 2⬘-deoxy or 2⬘-O-
`methyl ribose (Fig. 4A). The cleaved phosphodiester
`bond is located between G9 and A10. Substitutions by
`2⬘-deoxyribose caused a 1.3-fold reduction in cleavage for
`S21-dG9, 1.5-fold for S21-dA10, and a 1.1-fold increase
`for S21-dA11, relative to S21 (Fig. 4B, lanes 2,4,6). Sub-
`stitutions by 2⬘-O-methylribose, however, caused a 73-
`fold reduction in cleavage for S21-meG9, but only a 1.3-
`fold increase for S21-meA10, and a 1.3-fold reduction for
`S21-meA11, relative to S21 (Fig. 4B, lanes 3,5,7). Con-
`sistent with the conversion of substrate observed in
`the single time point analysis (Fig. 4B), we did not ob-
`serve any significant differences in Vmax and KM rela-
`tive to unmodified S21, when we determined these pa-
`rameters for S21-dG9 (Vmax = 0.012 ± 0.001 nM min−1,
`KM = 0.9 ± 0.1 nM, Supplementary Fig. S1B) or S21-dA10
`
`Figure 3. Specificity of target RNA cleavage.
`(A) Sequences of
`modified substrate RNAs. (B) PhosphorImaging of 1 h at 30°C cleav-
`age reactions resolved on a 15% denaturing polyacrylamide gel. Sub-
`strate RNAs were radiolabeled at the 5⬘ terminus. The arrow indi-
`cates the 9-nt cleavage product, and the arrow with the dotted line
`indicates miscleavage guided by substrate S4 that contains a bulge
`over the cleavage site. The fraction of cleaved material is indicated
`at the bottom of the gel.
`
`GENES & DEVELOPMENT
`
`977
`
`Figure 2. Substrates as short as 15-nt are cleaved by RISC. (A) Se-
`quences of the truncated substrates. The 13-nt core sequence re-
`quired for RISC-guided cleavage is shaded. (B) PhosphorImaging of
`cleavage reactions incubated fo r 1 h at30°C and resolved on a 15%
`denaturing polyacrylamide gel. Substrate RNAs were radiolabeled
`only at the 5⬘ terminus. Arrows indicate the 7- and 9-nt 5⬘ cleavage
`product. The fraction of cleaved material is indicated at the bottom
`of the gel.
`
`Deletion of the 3⬘-most residues, which are unpaired in
`the enzyme–substrate complex, reduced cleavage by a
`marginal 1.1-fold in S19 compared with S21 (Fig. 2B).
`Subsequent truncations from the 3⬘ end reduced cleavage
`by 2.0-fold for S17, 5.2-fold for S15, and 13-fold for S13.
`Removal of 2 nt from the 5⬘ end of S17 reduced the re-
`action rate 2.2-fold compared with S21 and 1.1-fold com-
`pared with S17. These data suggest that pairing of the
`central 13 nt of the guide siRNA of RISC is required for
`constituting the active site, or for promoting a structural
`change to constitute the active site. SiRNA residues
`comprising segments of up to 4 nt at either the 5⬘ or 3⬘
`terminus can remain unpaired without causing a com-
`plete loss of cleavage activity. This observation is remi-
`niscent of studies questioning the specificity of siRNA-
`guided target RNA degradation, reporting off-target
`cleavage activity for some substrates that contained only
`11-nt segments of complementarity (Jackson et al. 2003).
`We therefore evaluated the importance of mismatches
`introduced into the S21 substrate (Fig. 3A). Mismatches
`were introduced by inverting segments of 3 or 4 nt, or by
`introducing point mutations. Four mismatches intro-
`duced at the 5⬘ end of the substrate (S21–2) reduced
`cleavage 1.3-fold, and two mismatches introduced at the
`3⬘ end (S21–6) reduced cleavage 3.2-fold, compared to S21
`(Fig. 3B, lanes 2,6). Mutation of the 4-nt segments closer
`to or overlapping the cleavage site (S21–3, 4, and 5) re-
`duced cleavage at least 70-fold relative to S21 (Fig. 3B,
`lanes 3–5). The substrate S21–4, which is mismatched
`across the cleavage site, appears to be miscleaved by 1 nt
`(Fig. 3B, lane 4, see dotted arrow). This is an interesting
`observation, as it suggests that naturally expressed mi-
`croRNAs, which pair imperfectly to their natural target
`mRNAs, may act partly by cleaving their target mRNAs
`and not only mediate translational regulation. Single-
`
`

`

`
`
`Downloaded from genesdev.cshlp.org on May 15, 2018 - Published by Cold Spring Harbor Laboratory Press
`
`
`
`
`
`
`
`Martinez and Tuschl
`
`Figure 4. Cleavage analysis of 2⬘- modified substrate RNAs. (A)
`Sequence and position of 2⬘-deoxy and 2⬘-O-methyl modified sub-
`strates. (B) PhosphorImaging of cleavage reactions incubated for 1 h
`at 30°C and resolved on a 15% denaturing polyacrylamide gel. Sub-
`strate RNAs were radiolabeled at the 5⬘ terminus. The arrow indi-
`cates the 9-nt cleavage product. The fraction of cleaved material is
`indicated at the bottom of the gel.
`
`(Vmax = 0.0088 ± 0.0002 nM min−1, KM = 0.9 ± 0.1 nM).
`Comparable relative rates were obtained using the high-
`activity RISC. Although individual 2⬘-deoxynucleotide
`substitutions in the substrate showed no significant per-
`turbation of cleavage by minimal RISC, a substrate en-
`tirely composed of 2⬘-deoxynucleotide was not cleaved
`at all (Supplementary Fig. S2).
`Modification of the ribose 2⬘ position affects the pKa of
`the 3⬘-hydroxyl, and therefore should also affect the rate
`of hydrolysis of the phosphodiester bond (Izatt et al.
`1965; Herschlag et al. 1993; Smith and Pace 1993;
`Yoshida et al. 2000); the pKa values reported for ribose
`and 2⬘-deoxyribose are 12.4 and 15, respectively. The less
`acidic 3⬘-hydroxyl of 2⬘-deoxyribose is expected to be a
`worse leaving group than the 3⬘-hydroxyl of regular ri-
`bose. However, no significant change in the rate of hy-
`drolysis occurred upon 2⬘-deoxy modification at the
`cleavage site, indicating that product release and/or a
`conformational change rather than the rate of chemical
`cleavage is limiting for catalysis by minimal RISC. It is
`conceivable that RNA helicases or RNP remodeling ac-
`tivities (Will and Lührmann 2001), which presumably
`were removed by our purification protocol, participate in
`such a transition and promote RISC activity in a more
`natural environment. In contrast to the 2⬘-deoxy modi-
`fication, introduction of the bulky, but chemically inert
`2⬘-O-methyl group at the cleavage site significantly re-
`duced the substrate cleavage, presumably due to steric
`interference with either a conformational transition or
`positioning catalytic residues or metal ions at the active
`site.
`The cleavage mechanism of RISC shares molecular
`features characteristic of dsDNA restriction enzymes
`(Pingoud and Jeltsch 2001) and RNase III enzymes (Dunn
`
`978
`
`GENES & DEVELOPMENT
`
`1982), and is unrelated to conventional RNase A-type
`endonucleases that cleave via 2⬘, 3⬘-cyclic phosphate in-
`termediates and leave a hydrolyzed 3⬘-phosphomonoes-
`ter and a free 5⬘-hydroxyl (Usher et al. 1970). Although
`these findings might suggest that Dicer RNase III, which
`is involved in generating siRNAs from dsRNA precur-
`sors, might be a good candidate nuclease for RISC, this is
`unlikely, because cytoplasmic extracts effectively de-
`pleted of Dicer were still able to reconstitute RISC (Mar-
`tinez et al. 2002). In addition, the 220-kD molecular
`mass of human Dicer (Provost et al. 2002; Zhang et al.
`2002) appears too large to be possibly present in our RISC
`preparation of an estimated apparent molecular mass of
`160 kD, that also has to account for a 100-kD Argonaute
`protein and a 21-nt single-stranded RNA.
`A nuclease related to micrococcal nuclease, Tudor SN,
`was recently reported in association with RISC, and the
`bacterial recombinantly expressed protein was shown to
`cleave DNA and RNA nonspecifically (Caudy et al.
`2003). Micrococcal nucleases degrade DNA and RNA,
`however, by generating 3⬘-phosphonucleotides and 3⬘-
`phosphodinucleotides (Reddi 1959; Alexander et al.
`1961). The initial cleavage by micrococcal nuclease is
`endonucleolytic, and is followed by exonucleolytic deg-
`radation of the resulting oligonucleotides (Alexander et
`al. 1961; Williams et al. 1961; Sulkowski and Laskowski
`1962). This argues against Tudor SN being the core en-
`donucleolytic component of RISC. This specific example
`illustrates the importance of defining the mechanism of
`RISC-endonucleolytic cleavage, and how mechanistic
`studies provide constraints for the critical evaluation of
`potential candidate nucleases that may soon emerge
`from genetic or biochemical studies aimed at uncovering
`this key RNAi component. The described biochemical
`assays will also guide future investigation to uncover the
`role of RISC and RISC-associated proteins.
`
`Materials and methods
`Oligonucleotide synthesis
`Oligoribonucleotides and modified oligoribonucleotides were chemically
`synthesized using 5⬘-silyl, 2⬘-ACE phosphoramidites, 5⬘-silyl, 2⬘-deoxy
`phosphoramidites, and 5⬘-silyl, 2⬘-O-methyl phosphoramidites (Dharma-
`con Research) on 0.2 µmol synthesis columns using a modified ABI 394
`synthesizer (Scaringe 2001). The aminolinker-modified 12-nt RNA was
`synthesized using a 3⬘ C6-aminolinker synthesis column (Dharmacon).
`The phosphate methyl group was removed by flushing the column with
`2 mL of 0.2 M 2-carbamoyl-2-cyanoethylene-1,1-dithiolate trihydrate in
`DMF/water (98:2 v/v) for 30 min at room temperature. The reagent was
`removed and the column rinsed with 10 mL water followed by 10 mL
`acetonitrile. Oligonucleotides were cleaved and eluted from the solid
`support by flushing with 1.9 mL of 40% aqueous methylamine over 2
`min, collected in a screw-cap vial, and incubated for 10 min at 55°C.
`Subsequently, base-treated oligonucleotides were dried down in an Ep-
`pendorf concentrator to remove methylamine and water. The residues
`were dissolved in sterile 2⬘-deprotection buffer (400 µL of 100 mM ac-
`etate-TEMED at pH 3.8 for a 0.2 µmole scale synthesis) and incubated for
`30 min at room temperature and another 30 min at 60°C to remove the
`2⬘ ACE group. Oligonucleotides were precipitated from the acetate-
`TEMED solution by adding 24 µL 5 M NaCl and 1.2 mL of absolute
`ethanol. Pellets were dissolved in 500 µL of water. DNA oligonucleotides
`were prepared using 0.2 µmole scale synthesis and standard DNA syn-
`thesis reagents (Proligo). Synthesis of 3⬘ C7-aminolinker-containing
`siRNAs and biotin conjugation were performed as described (Martinez
`et al. 2002). Annealing of siRNA duplexes was performed as described
`(Elbashir et al. 2002).
`
`Labeling of oligoribonucleotide substrates
`5⬘ labeling reactions contained 10 pmol oligonucleotide, 5 pmol ␥-32P-
`ATP (Amersham, 3000 Ci/mmol), 1 unit T4 polynucleotide kinase (New
`
`

`

`
`
`Downloaded from genesdev.cshlp.org on May 15, 2018 - Published by Cold Spring Harbor Laboratory Press
`
`
`
`
`
`
`
`England Biolabs), and 10 mM MgCl2/5 mM dithiothreitol/70 mM Tris-
`HCl (pH 7.6) in a final volume of 10 µL. The reaction mixture was in-
`cubated for 20 min at 37°C, followed by the addition of 1 µL 100 mM
`ATP and incubation for another 3 min at 37°C. The reaction was stopped
`by the addition of 1 volume urea stop solution (8 M urea, 0.05% w/v
`Bromophenol Blue, 50 mM EDTA at pH 8.0) and 1 µg carrier tRNA. The
`samples were heat-treated for 2 min at 90°C and separated on a 20%
`denaturing polyacrylamide gel. The radioactive product bands were ex-
`cised and eluted overnight at 4°C into 300 µL 0.3 M NaCl containing 1
`µg carrier tRNA. To optimize the recovery of the 9- and 12-nt oligoribo-
`nucleotides, gel elution was performed usin g 1 M Naacetate (pH 6.0)
`containing 1 µg of carrier tRNA. The oligonucleotides were precipitated
`by the addition of 3 volumes absolute ethanol, collected by centrifuga-
`tion, and dissolved in water.
`
`RNA ligation
`Site-specific internally labeled substrate RNA was prepared by incubat-
`ing 2 pmol 5⬘-32P-phosphorylated and linker-modified 12-nt RNA with
`20 pmol 9-nt RNA, and 40 units of T4 RNA ligase (Amersham) in 10 mM
`MgCl2/10 mM 2-mercaptoethanol/50 mM Tris-HCl (pH 7.6)/0.1 mg/mL
`acetylated bovine serum albumin (Ac-BS, Sigma)/0.2 mM ATP/15%
`DMSO in a final volume of 20 µL. The reaction mixture was incubated
`for 3 h at 37°C. The reaction was stopped by the addition of 1 volume
`urea stop solution and 1 µg of carrier tRNA. The samples were heat-
`treated for 2 min at 90°C and separated on a 20% denaturing polyacryl-
`amide gel. The ligation product was recovered from the gel as described
`above, and eluted in the presence of 50 ng of carrier tRNA.
`
`Affinity purification of RISC
`RISC was affinity-purified as described using the siRNA duplex contain-
`ing biotin at both of its 3⬘ ends (Martinez et al. 2002). The concentration
`of siRNA duplex was 2 nM during RISC assembly in HeLa cell cytoplas-
`mic extract. The streptavidin resin was washed with increasing concen-
`trations of KCl (0.1–2.5 M) after binding of RISC. The complex was
`cleaved from the resin by irradiating at 302 nm and recovered in 0.1 M
`KCl. The affinity-purified RISC only targets sense substrates for degra-
`dation, not antisense substrates, indicating completely asymmetric as-
`sembly of RISC on the antisense siRNA.
`
`Target RNA cleavage assay
`Cleavage assays were performed as described (Martinez et al. 2002), but
`adjusting the concentration of MgCl2 to 5 mM and omitting phospho-
`creatine and creatine phosphokinase. Cleavage by RISC is magnesium
`ion-dependent and plateaus between 2 and 10 mM MgCl2. The concen-
`tration of carrier tRNA in the target RNA substrates was adjusted to 10
`ng/µL. Cleavage assays were performed using 1 nM concentration of
`substrate, except for measuring kinetic constants, where the substrate
`concentration was then varied from 0.02 to 20 nM. Substrate RNA dilu-
`tions were heat-treated for 40 sec at 90°C and shortly chilled on ice prior
`to setting up the reactions. Aliquots were withdrawn at the indicated
`time points, and the reaction was stopped by the addition of 1 volume of
`urea-stop solution. To improve the gel resolution of the RNA cleavage
`products shown in Figure 1, a 100-fold excess of 12-nt RNA relative to
`substrate was added to the gel loading solution to act as competitor for
`the radiolabeled product. The gel loading solution was incubated for 2
`min at 90°C and then loaded on the gel.
`
`Ribonuclease T2 digestion and two-dimensional thin
`layer chromatography
`The radioactive 12-nt cleavage product shown in Figure 1A was excised
`from a 20% denaturing polyacrylamide gel, eluted and redissolved in 3.5
`µL of water, and digested with 2 units of ribonuclease T2 (Invitrogen) in
`50 mM sodium acetate (pH 4.5) fo r 1 h at 37°C.Samples was cospotted
`with 1 µg of nonradioactive adenosine 3⬘,5⬘-diphosphate (Sigma, catalog
`no. A-5763) on cellulose HPTLC plates (Merck, catalog no. 1.05787) and
`separated in the first dimension in isobutyric acid/ammonia 32%/water
`(577:38:385) for ∼8 h. The TLC plate was dried overnight at room tem-
`perature, and the separation in the second dimension was performed
`using tert-butanol/concentrated hydrochloric acid (37.6%)/water (14:3:3)
`for ∼8 h. The position of migration of the nonradioactive pAp was de-
`tected by UV-shadowing, and the position of the T2-digested radioactive
`product was determined by PhosphorImaging.
`
`RISC target RNA cleavage mechanism
`
`Acknowledgments
`We thank P.Y. Chen and J. Meyer for oligonucleotide synthesis and tech-
`nical assistance; R. Lührmann (Max-Planck-Institute for Biophysical
`Chemistry, Göttingen, Germany) for providing HeLa cytoplasmic ex-
`tracts; M. Konarska for advice, and H. Dormann, W. Fischle, and all
`members of our laboratory for discussions and comments on the manu-
`script. This work was supported by NIH grant R01 GM068476-01.
`
`Note added in proof
`
`While this manuscript was under revision, a study by Schwarz et al.
`(2004) also reported on the endonuclease mechanism of RISC using D.
`melanogaster lysates.
`The publication costs of this article were defrayed in part by payment
`of page charges. This article must therefore be hereby marked “adver-
`tisement” in accordance with 18 USC section 1734 solely to indicate this
`fact.
`
`References
`
`Alexander, M., Heppel, L.A., and Hurwitz, J. 1961. The purification and
`properties of micrococcal nuclease. J. Biol. Chem. 236: 3014–3019.
`Amarzguioui, M., Holen, T., Babaie, E., and Prydz, H. 2003. Tolerance for
`mutations and chemical modifications in a siRNA. Nucleic Acids
`Res. 31: 589–595.
`Bernstein, E., Caudy, A.A., Hammond, S.M., and Hannon, G.J. 2001. Role
`for a bidentate ribonuclease in the initiation step of RNA interfer-
`ence. Nature 409: 363–366.
`Braasch, D.A., Jensen, S., Liu, Y., Kaur, K., Arar, K., White, M.A., and
`Corey, D.R. 2003. RNA interference in mammalian cells by chemi-
`cally-modified RNA. Biochemistry 42: 7967–7975.
`Capodici, J., Kariko, K., and Weissman, D. 2002. Inhibition of HIV-1
`infection by small interfering RNA-mediated RNA interference. J.
`Immunol. 169: 5196–5201.
`Caudy, A.A., Myers, M., Hannon, G.J., and Hammond, S.M. 2002. Fragile
`X-related protein and VIG associate with the RNA interference ma-
`chinery. Genes & Dev. 16: 2491–2496.
`Caudy, A.A., Ketting, R.F., Hammond, S.M., Denli, A.M., Bathoorn,
`A.M., Tops, B.B., Silva, J.M., Myers, M.M., Hannon, G.J., and Plas-
`terk, R.H. 2003. A micrococcal nuclease homologue in RNAi effector
`complexes. Nature 425: 411–414.
`Cerutti, H. 2003. RNA interference: Traveling in the cell and gaining
`functions? Trends Genet. 19: 39–46.
`Chiu, Y.L. and Rana, T.M. 2003. siRNA function in RNAi: A chemical
`modification analysis. RNA 9: 1034–1048.
`Czauderna, F., Fechtner, M., Dames, S., Aygun, H., Klippel, A., Pronk,
`G.J., Giese, K., and Kaufmann, J. 2003. Structural variations and sta-
`bilising modifications of synthetic siRNAs in mammalian cells.
`Nucleic Acids Res. 31: 2705–2716.
`Denli, A.M. and Hannon, G.J. 2003. RNAi: An ever-growing puzzle.
`Trends Biochem. Sci. 28: 196–201.
`Dunn, J.J. 1982. Ribonuclease III. In The enzymes (ed. P.D. Boyer), Vol.
`15, part B, pp. 485–499. Academic Press, New York.
`Dykxhoorn, D.M., Novina, C.D., and Sharp, P.A. 2003. Killing the mes-
`senger: Short RNAs that silence gene expression. Nat. Rev. Mol. Cell
`Biol. 4: 457–467.
`Elbashir, S.M., Lendeckel, W., and Tuschl, T. 2001a. RNA interference is
`mediated by 21 and 22 nt RNAs. Genes & Dev. 15: 188–200.
`Elbashir, S.M., Martinez, J., Patkaniowska, A., Lendeckel, W., and
`Tuschl, T. 2001b. Functional anatomy of siRNAs for mediating effi-
`cient RNAi in Drosophila melanogaster embryo lysate. EMBO J. 20:
`6877–6888.
`Elbashir, S.M., Harborth, J., Weber, K., and Tuschl, T. 2002. Analysis of
`gene function in somatic mammalian cells using small interfering
`RNAs. Methods 26: 199–213.
`Hammond, S.M., Bernstein, E., Beach, D., and Hannon, G.J. 2000. An
`RNA-directed nuclease mediates post-transcriptional gene silencing
`in Drosophila cells. Nature 404: 293–296.
`Hammond, S.M., Boettcher, S., Caudy, A.A., Kobayashi, R., and Hannon,
`G.J. 2001. Argonaute2, a link between genetic and biochemical analy-
`ses of RNAi. Science 293: 1146–1150.
`Harborth, J., Elbashir, S.M., Vandenburgh, K., Manninga, H., Scaringe,
`
`GENES & DEVELOPMENT
`
`979
`
`

`

`
`
`Downloaded from genesdev.cshlp.org on May 15, 2018 - Published by Cold Spring Harbor Laboratory Press
`
`
`
`
`
`
`
`Martinez and Tuschl
`
`S.A., Weber, K., and Tuschl, T. 2003. Sequence, chemical, and struc-
`tural variation of small interfering RNAs and short hairpin RNAs and
`the effect on mammalian gene silencing. Antisense Nucleic Acid
`Drug Dev. 13.
`Herschlag, D., Eckstein, F., and Cech, T.R. 1993. The importance of being
`ribose at the cleavage site in the Tetrahymena ribozyme reaction.
`Biochemistry 32: 8312–8321.
`Holen, T., Amarzguioui, M., Babaie, E., and Prydz, H. 2003. Similar be-
`haviour of single-strand and double-strand siRNAs suggests they act
`through a common RNAi pathway. Nucleic Acids Res. 31: 2401–2407.
`Hutvágner, G. and Zamore, P.D. 2002. A microRNA in a multiple-turn-
`over RNAi enzyme complex. Science 297: 2056–2060.
`Hutvágner, G., Simard, M.J., Mello, C.C., and Zamore, P.D. 2004. Se-
`quence-specific inhibition of small RNA function. PLoS Biol. 2: E98.
`Izatt, R.M., Hansen, L.D., Rytting, J.H., and Christensen, J.J. 1965