siDirect 2.0: updated software for designing functional siRNA with reduced seed-dependent off-target effect
© Naito et al; licensee BioMed Central Ltd. 2009
Received: 03 August 2009
Accepted: 30 November 2009
Published: 30 November 2009
RNA interference (RNAi), mediated by 21-nucleotide (nt)-length small interfering RNAs (siRNAs), is a powerful tool not only for studying gene function but also for therapeutic applications. RNAi, requiring perfect complementarity between the siRNA guide strand and the target mRNA, was believed to be extremely specific. However, a recent growing body of evidence has suggested that siRNA could down-regulate unintended genes whose transcripts possess complementarity to the 7-nt siRNA seed region. This off-target gene silencing may often provide incongruous results obtained from knockdown experiments, leading to misinterpretation. Thus, an efficient algorithm for designing functional siRNAs with minimal off-target effect based on the mechanistic features is considered of value.
We present siDirect 2.0, an update of our web-based software siDirect, which provides functional and off-target minimized siRNA design for mammalian RNAi. The previous version of our software designed functional siRNAs by considering the relationship between siRNA sequence and RNAi activity, and provided them along with the enumeration of potential off-target gene candidates by using a fast and sensitive homology search algorithm. In the new version, the siRNA design algorithm is extensively updated to eliminate off-target effects by reflecting our recent finding that the capability of siRNA to induce off-target effect is highly correlated to the thermodynamic stability, or the melting temperature (Tm), of the seed-target duplex, which is formed between the nucleotides positioned at 2-8 from the 5' end of the siRNA guide strand and its target mRNA. Selection of siRNAs with lower seed-target duplex stabilities (benchmark Tm < 21.5°C) followed by the elimination of unrelated transcripts with nearly perfect match should minimize the off-target effects.
siDirect 2.0 provides functional, target-specific siRNA design with the updated algorithm which significantly reduces off-target silencing. When the candidate functional siRNAs could form seed-target duplexes with Tm values below 21.5°C, and their 19-nt regions spanning positions 2-20 of both strands have at least two mismatches to any other non-targeted transcripts, siDirect 2.0 can design at least one qualified siRNA for >94% of human mRNA sequences in RefSeq. siDirect 2.0 is available at http://siDirect2.RNAi.jp/.
RNA interference (RNAi) mediated by double-stranded RNA has become a powerful tool not only for studying gene functions, but also for therapeutic applications [1, 2]. In mammalian cells, RNAi is induced by small interfering RNA (siRNA), a duplex of 21-nucleotide (nt) RNAs containing 2-nt 3' overhangs. The siRNAs incorporated into cells are transferred to the RNAi effector complex called RNA-induced silencing complex (RISC) [3, 4]. RISC assembles on one of the two strands of siRNA duplex, and is activated upon the removal of the passenger strand [5–9]. The activated RISC is a ribonucleoprotein complex minimally consisting of the core protein Argonaute (Ago) and single-stranded siRNA, which acts as the guide to target complementary sequences within mRNAs [10–13]. The 5' end of the siRNA guide strand is anchored in the binding pocket of the Mid domain of Archaeoglobus fulgidus Ago-like protein [14, 15], and the 3' end is anchored to the PAZ domain of human  and Drosophila  Ago in the RISC complex. Thus, in the siRNA guide strand, 19 nucleotides positioned at 2-20 from 5' end may be responsible for target RNA recognition, leading to the silencing of gene expression by cleaving target mRNA [10–13]. Since RNAi is based on sequence recognition by the siRNA, it can give rise to the silencing of other genes with similar sequences. This phenomenon is referred to as an off-target effect, and the growing evidence from large-scale knockdown experiments indicates that the off-target silencing is induced by the base-pairing between the seed region at positions 2-8 from the 5' end of the RISC-loaded siRNA strand, and its complementary sequences in the 3' UTR of the unrelated mRNAs [18–23]. Although RNAi is now widely and routinely used as an experimental tool, the remaining fundamental concern is whether the target gene can be specifically silenced. Especially, accurate knowledge of RNAi specificity is critical for therapeutic technologies.
To avoid off-target effects, one approach may be to select the siRNA whose seed sequence is not complementary to any sequences in the 3' UTR of all non-targeted genes. However, this approach is problematic because random 7-nt sequence is predicted to appear in every 16,384 bp on average. In fact, we analyzed the human 3' UTR database and it proved impossible to select such siRNAs. That is, human siRNAs with the most infrequent 7-nt seed sequence still have seed-complementarities with 17 3' UTR sequences. Recently, we have revealed that the capability of siRNAs to induce seed-dependent off-target effect is highly correlated to the thermodynamic stability of the duplex formed between the seed region of siRNA guide strand and its target mRNA : the melting temperature (Tm) of the seed-target duplex showed strong positive correlation with the induction of seed-dependent off-target effects. The results suggested that the Tm of 21.5°C may serve as the benchmark, which discriminates the almost off-target-free seed sequences from the off-target-positive ones. Thus, selecting the siRNAs with low Tm of the seed-target duplex should minimize seed-dependent off-target silencing.
We have previously released highly effective, target-specific siRNA design software, siDirect , in which siRNA sequences were selected using our guidelines established through extensive experiments to clarify the relationship between siRNA sequences and RNAi activities . In order to exclude potential cross-hybridization candidates, siDirect used the rigorous homology search algorithm to select siRNA sequences that have at least three mismatches to any other non-targeted transcripts . In the updated software, siDirect 2.0, the siRNA design algorithm has been extensively updated to select off-target minimized siRNAs by considering the thermodynamic stability of the seed-target duplex. By using the default parameters, at least one functional siRNA could be designed for >94% of the human mRNA sequences in RefSeq release 30.
Selection of highly functional siRNAs
In the first step, highly functional siRNA sequences were selected using our algorithm  (Figure 1, Step 1). We have revealed that efficient RNAi could be induced by the siRNAs that satisfies the following three sequence conditions simultaneously: A/U at the 5' terminus of the guide strand; G/C at the 5' terminus of the passenger strand; at least 4 A/U residues in the 5' terminal 7 bp of the guide strand. In addition, G/C stretch longer than 9 bp should be absent . The experimental validation showed that 98% of the siRNAs predicted to be functional have reduced the target gene expression . The proportion of functional siRNA sequences selected by this algorithm is 14.7% of all human 23-mer sequences generated from RefSeq 30 (Figure 1A, see Step 1).
Reduction of seed-dependent off-target effects
Elimination of near-perfect matched genes
Several studies have indicated that the effect of single-base mismatches between the siRNA guide strand and the target mRNA varies, according to the positions of the mismatch and/or the sequence of siRNA [21, 29]. However, as shown in our previous report, it is obvious that even when the Tm value of the seed-target duplex is sufficiently low, the target gene silencing can still take place if the non-seed region is completely complementary . Therefore, in the third step, siRNAs that have near-perfect matches to any other non-targeted transcripts were eliminated. In siDirect 2.0, off-target searches are performed for 19-mer sequences at positions 2-20 of both strands of the siRNA duplex (Figure 1B, Step 3), because these 19 nucleotides are thought to be involved in target mRNA recognition. Since widely-used BLAST tends to overlook near-perfect match candidates frequently, we used our fast and sensitive algorithm . In addition, all of the near-perfect match hits are precomputed for all the functional human siRNAs to accelerate the computational performance. Precomputed results are stored in the memory engine of MySQL relational database management system. This makes it possible to return the list of siRNA candidates within a few seconds (Figure 2A). The output page includes the minimum number of mismatches against any near-perfect match candidates for each siRNA (Figure 2A). By clicking the individual siRNA in Figure 2A, a detailed list of candidate genes will appear (Figure 2B). By default, siRNA sequences that have at least two mismatches to any other non-targeted transcripts are selected.
Results and Discussion
We performed a genome-wide design of siRNAs for human mRNAs in RefSeq release 30 with the following parameters: 1) satisfying our functional siRNA design algorithm [7, 24], 2) Tm values at the seed-target duplex of both the guide and the passenger strands below 21.5°C, and 3) no off-target hits with less than two mismatches.
It is also desirable to select siRNA that contains as many mismatches as possible to any non-targeted mRNAs. In addition to the Tm value of below 21.5°C, siRNA sequences with at least two mismatches to any other non-targeted transcripts are selectable for 94.7% of human mRNAs (Figure 3B). However, if the siRNAs having near-perfect match hits with less than three mismatches, with their Tm of seed sequences below 21.5°C, are selected, one or more siRNA can be designed for only 77.2% of the human mRNAs (Figure 3B). When siRNAs with seed Tm below 15°C and 10°C were selected, siRNAs can be designed for only 47.0% and 18.5%, respectively (Figure 3B). Furthermore, the percentage of human mRNAs drops severely to 0.15% if the near-perfect match hits with less than four mismatches are filtered. Thus, siDirect 2.0 filters siRNAs with less than two mismatches by default to avoid severe reduction in the number of siRNA candidates.
We were unable to design functional, off-target minimized siRNAs for 5.3% of the RefSeq mRNAs using the default parameters. Typical examples of these mRNAs are the histone clusters (NM_003523, etc.) and ribosomal proteins (NM_002952, etc.), which are known to form multigene families. When designing siRNAs targeting such genes, users can manually investigate the detailed list of off-target gene candidates (Figure 2B) and select the siRNA that does not have off-target hits to unrelated transcripts.
We have extensively updated siDirect 2.0 based on our experimental knowledge, and provided a promising website for reducing siRNA off-target silencing. The website selects: 1) functional siRNAs that satisfy our guideline , 2) siRNAs with reduced seed-dependent off-target effects by considering the thermodynamic stability of the seed-target duplex, 3) siRNAs that do not hit any non-targeted genes with near-perfect matches. When the candidate functional siRNAs could form seed-target duplexes with Tm values below 21.5°C, and their 19-nt region spanning positions 2-20 of both strands have at least two mismatches to any other non-targeted transcripts, siDirect 2.0 can design at least one qualified siRNA for >94% of human mRNA sequences in RefSeq. This website should provide a wide scope of applications in RNAi studies.
Availability and requirements
Project name: siDirect
Project home page: http://siDirect2.RNAi.jp/
Operating system(s): Platform independent
Programming language: Perl
Any restrictions to use by non-academics: Contact license@RNAi.jp
This work was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan to YN, SM and KU-T. KU-T is a member of the Genome Network Project (MEXT).
- Boutros M, Ahringer J: The art and design of genetic screens: RNA interference. Nat Rev Genet 2008, 9: 554–566. 10.1038/nrg2364View ArticlePubMedGoogle Scholar
- Castanotto D, Rossi JJ: The promises and pitfalls of RNA-interference-based therapeutics. Nature 2009, 457: 426–433. 10.1038/nature07758PubMed CentralView ArticlePubMedGoogle Scholar
- Hutvagner G, Simard MJ: Argonaute proteins: key players in RNA silencing. Nat Rev Mol Cell Biol 2008, 9: 22–32. 10.1038/nrm2321View ArticlePubMedGoogle Scholar
- Jinek M, Doudna JA: A three-dimensional view of the molecular machinery of RNA interference. Nature 2009, 457: 405–412. 10.1038/nature07755View ArticlePubMedGoogle Scholar
- Schwarz DS, Hutvágner G, Du T, Xu Z, Aronin N, Zamore PD: Asymmetry in the assembly of the RNAi enzyme complex. Cell 2003, 115: 199–208. 10.1016/S0092-8674(03)00759-1View ArticlePubMedGoogle Scholar
- Khvorova A, Reynolds A, Jayasena SD: Functional siRNAs and miRNAs exhibit strand bias. Cell 2003, 115: 209–216. 10.1016/S0092-8674(03)00801-8View ArticlePubMedGoogle Scholar
- Ui-Tei K, Naito Y, Takahashi F, Haraguchi T, Ohki-Hamazaki H, Juni A, Ueda R, Saigo K: Guidelines for the selection of highly effective siRNA sequences for mammalian and chick RNA interference. Nucleic Acids Res 2004, 32: 936–948. 10.1093/nar/gkh247PubMed CentralView ArticlePubMedGoogle Scholar
- Matranga C, Tomari Y, Shin C, Bartel DP, Zamore PD: Passenger-strand cleavage facilitates assembly of siRNA into Ago2-containing RNAi enzyme complexes. Cell 2005, 123: 607–620. 10.1016/j.cell.2005.08.044View ArticlePubMedGoogle Scholar
- Rand TA, Petersen S, Du F, Wang X: Argonaute2 cleaves the anti-guide strand of siRNA during RISC activation. Cell 2005, 123: 621–629. 10.1016/j.cell.2005.10.020View ArticlePubMedGoogle Scholar
- Elbashir SM, Lendeckel W, Tuschl T: RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev 2001, 15: 188–200. 10.1101/gad.862301PubMed CentralView ArticlePubMedGoogle Scholar
- Meister G, Landthaler M, Patkaniowska A, Dorsett Y, Teng G, Tuschl T: Human Argonaute2 mediates RNA cleavage targeted by miRNAs and siRNAs. Mol Cell 2004, 15: 185–197. 10.1016/j.molcel.2004.07.007View ArticlePubMedGoogle Scholar
- Song JJ, Smith SK, Hannon GJ, Joshua-Tor L: Crystal structure of Argonaute and its implications for RISC slicer activity. Science 2004, 305: 1434–1437. 10.1126/science.1102514View ArticlePubMedGoogle Scholar
- Liu J, Carmell MA, Rivas FV, Marsden CG, Thomson JM, Song JJ, Hammond SM, Joshua-Tor L, Hannon GJ: Argonaute2 is the catalytic engine of mammalian RNAi. Science 2004, 305: 1437–1441. 10.1126/science.1102513View ArticlePubMedGoogle Scholar
- Parker JS, Roe SM, Barford D: Structural insights into mRNA recognition from a PIWI domain-siRNA guide complex. Nature 2005, 434: 663–666. 10.1038/nature03462PubMed CentralView ArticlePubMedGoogle Scholar
- Ma J-B, Yuan YR, Meister G, Pei Y, Tuschl T, Patel DJ: Structural basis for 5'-end-specific recognition of guide RNA by the A. fulgidus Piwi protein. Nature 2005, 434: 666–670. 10.1038/nature03514View ArticlePubMedGoogle Scholar
- Ma J-B, Ye K, Patel DJ: Structural basis for overhang-specific small interfering RNA recognition by the PAZ domain. Nature 2004, 429: 318–322. 10.1038/nature02519View ArticlePubMedGoogle Scholar
- Lingel A, Simon B, Izaurralde E, Sattler M: Nucleic acid 3'-end recognition by the Argonaute2 PAZ domain. Nat Struct Mol Biol 2004, 11: 576–577. 10.1038/nsmb777View ArticlePubMedGoogle Scholar
- Jackson AL, Bartz SR, Schelter J, Kobayashi SV, Burchard J, Mao M, Li B, Cavet G, Linsley PS: Expression profiling reveals off-target gene regulation by RNAi. Nat Biotechnol 2003, 21: 635–637. 10.1038/nbt831View ArticlePubMedGoogle Scholar
- Scacheri PC, Rozenblatt-Rosen O, Caplen NJ, Wolfsberg TG, Umayam L, Lee JC, Hughes CM, Shanmugam KS, Bhattacharjee A, Meyerson M, Collins FS: Short interfering RNAs can induce unexpected and divergent changes in the levels of untargeted proteins in mammalian cells. Proc Natl Acad Sci USA 2004, 101: 1892–1897. 10.1073/pnas.0308698100PubMed CentralView ArticlePubMedGoogle Scholar
- Lin X, Ruan X, Anderson MG, McDowell JA, Kroeger PE, Fesik SW, Shen Y: siRNA-mediated off-target gene silencing triggered by a 7 nt complementation. Nucleic Acids Res 2005, 33: 4527–4535. 10.1093/nar/gki762PubMed CentralView ArticlePubMedGoogle Scholar
- Birmingham A, Anderson EM, Reynolds A, Ilsley-Tyree D, Leake D, Fedorov Y, Baskerville S, Maksimova E, Robinson K, Karpilow J, Marshall WS, Khvorova A: 3' UTR seed matches, but not overall identity, are associated with RNAi off-targets. Nat Methods 2006, 3: 199–204. 10.1038/nmeth854View ArticlePubMedGoogle Scholar
- Jackson AL, Burchard J, Schelter J, Chau BN, Cleary M, Lim L, Linsley PS: Widespread siRNA "off-target" transcript silencing mediated by seed region sequence complementarity. RNA 2006, 12: 1179–1187. 10.1261/rna.25706PubMed CentralView ArticlePubMedGoogle Scholar
- Ui-Tei K, Naito Y, Nishi K, Juni A, Saigo K: Thermodynamic stability and Watson-Crick base pairing in the seed duplex are major determinants of the efficiency of the siRNA-based off-target effect. Nucleic Acids Res 2008, 36: 7100–7109. 10.1093/nar/gkn902PubMed CentralView ArticlePubMedGoogle Scholar
- Naito Y, Yamada T, Ui-Tei K, Morishita S, Saigo K: siDirect: highly effective, target-specific siRNA design software for mammalian RNA interference. Nucleic Acids Res 2004, 32: W124-W129. 10.1093/nar/gkh442PubMed CentralView ArticlePubMedGoogle Scholar
- Yamada T, Morishita S: Accelerated off-target search algorithm for siRNA. Bioinformatics 2005, 21: 1316–1324. 10.1093/bioinformatics/bti155View ArticlePubMedGoogle Scholar
- Naito Y, Saigo K, Ui-Tei K: Evaluation of published rational siRNA design algorithms using firefly luciferase gene as a reporter. In RNA interference research progress. Edited by: Lyland RT, Browning IB. New York: Nova Science Publishers; 2008:3–11.Google Scholar
- Freier SM, Kierzek R, Jaeger JA, Sugimoto N, Caruthers MH, Neilson T, Turner DH: Improved free-energy parameters for predictions of RNA duplex stability. Proc Natl Acad Sci USA 1986, 83: 9373–9377. 10.1073/pnas.83.24.9373PubMed CentralView ArticlePubMedGoogle Scholar
- Clark PR, Pober JS, Kluger MS: Knockdown of TNFR1 by the sense strand of an ICAM-1 siRNA: dissection of an off-target effect. Nucleic Acids Res 2008, 36: 1081–1097. 10.1093/nar/gkm630PubMed CentralView ArticlePubMedGoogle Scholar
- Du Q, Thonberg H, Wang J, Wahlestedt C, Liang Z: A systematic analysis of the silencing effects of an active siRNA at all single-nucleotide mismatched target sites. Nucleic Acids Res 2005, 33: 1671–1677. 10.1093/nar/gki312PubMed CentralView ArticlePubMedGoogle Scholar
- Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ: Basic local alignment search tool. J Mol Biol 1990, 215: 403–410.View ArticlePubMedGoogle Scholar
- Levenkova N, Gu Q, Rux JJ: Gene specific siRNA selector. Bioinformatics 2004, 20: 430–432. 10.1093/bioinformatics/btg437View ArticlePubMedGoogle Scholar
- Chalk AM, Wahlestedt C, Sonnhammer EL: Improved and automated prediction of effective siRNA. Biochem Biophys Res Commun 2004, 319: 264–274. 10.1016/j.bbrc.2004.04.181View ArticlePubMedGoogle Scholar
- Henschel A, Buchholz F, Habermann B: DEQOR: a web-based tool for the design and quality control of siRNAs. Nucleic Acids Res 2004, 32: W113-W120. 10.1093/nar/gkh408PubMed CentralView ArticlePubMedGoogle Scholar
- Yuan B, Latek R, Hossbach M, Tuschl T, Lewitter F: siRNA Selection Server: an automated siRNA oligonucleotide prediction server. Nucleic Acids Res 2004, 32: W130-W134. 10.1093/nar/gkh366PubMed CentralView ArticlePubMedGoogle Scholar
- Santoyo J, Vaquerizas JM, Dopazo J: Highly specific and accurate selection of siRNAs for high-throughput functional assays. Bioinformatics 2005, 21: 1376–1382. 10.1093/bioinformatics/bti196View ArticlePubMedGoogle Scholar
- Shah JK, Garner HR, White MA, Shames DS, Minna JD: sIR: siRNA Information Resource, a web-based tool for siRNA sequence design and analysis and an open access siRNA database. BMC Bioinformatics 2007, 8: 178. 10.1186/1471-2105-8-178PubMed CentralView ArticlePubMedGoogle Scholar
- Chalk AM, Sonnhammer EL: siRNA specificity searching incorporating mismatch tolerance data. Bioinformatics 2008, 24: 1316–1317. 10.1093/bioinformatics/btn121View ArticlePubMedGoogle Scholar
- Park YK, Park SM, Choi YC, Lee D, Won M, Kim YJ: AsiDesigner: exon-based siRNA design server considering alternative splicing. Nucleic Acids Res 2008, 36: W97-W103. 10.1093/nar/gkn280PubMed CentralView ArticlePubMedGoogle Scholar
- Gong W, Ren Y, Zhou H, Wang Y, Kang S, Li T: siDRM: an effective and generally applicable online siRNA design tool. Bioinformatics 2008, 24: 2405–2406. 10.1093/bioinformatics/btn442View ArticlePubMedGoogle Scholar
- Vert JP, Foveau N, Lajaunie C, Vandenbrouck Y: An accurate and interpretable model for siRNA efficacy prediction. BMC Bioinformatics 2006, 7: 520. 10.1186/1471-2105-7-520PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.