A role for human Dicer in pre-RISC loading of siRNAs

RNA interference is a powerful mechanism for sequence-specific inhibition of gene expression. It is widely known that small interfering RNAs (siRNAs) targeting the same region of a target-messenger RNA can have widely different efficacies. In efforts to better understand the siRNA features that influence knockdown efficiency, we analyzed siRNA interactions with a high-molecular weight complex in whole cell extracts prepared from two different cell lines. Using biochemical tools to study the nature of the complex, our results demonstrate that the primary siRNA-binding protein in the whole cell extracts is Dicer. We find that Dicer is capable of discriminating highly functional versus poorly functional siRNAs by recognizing the presence of 2-nt 3′ overhangs and the thermodynamic properties of 2–4 bp on both ends of effective siRNAs. Our results suggest a role for Dicer in pre-selection of effective siRNAs for handoff to Ago2. This initial selection is reflective of the overall silencing potential of an siRNA.     

Efficient prediction of siRNAs with siRNArules 1.0: an open-source JAVA approach to siRNA algorithms

RNAi interference and siRNA have become useful tools for investigation of gene function. However, the discovery that not all siRNA are equally efficient made necessary screens or design algorithms to obtain high activity siRNA candidates. Several algorithms have been published, but they remain inefficient, obscure, or commercially restricted. This article describes an open-source JAVA program that is surprisingly efficient at predicting active siRNAs (Pearson correlation coefficient r = 0.52, n = 526 siRNAs). Furthermore, this version 1.0 sets the stage for further improvement of the free code by the open-source community (http://sourceforge.net/).     

Designing highly active siRNAs for therapeutic applications

The discovery of RNA interference (RNAi) generated considerable interest in developing short interfering RNAs (siRNAs) for understanding basic biology and as the active agents in a new variety of therapeutics. Early studies showed that selecting an active siRNA was not as straightforward as simply picking a sequence on the target mRNA and synthesizing the siRNA complementary to that sequence. As interest in applying RNAi has increased, the methods for identifying active siRNA sequences have evolved from focusing on the simplicity of synthesis and purification, to identifying preferred target sequences and secondary structures, to predicting the thermodynamic stability of the siRNA. As more specific details of the RNAi mechanism have been defined, these have been incorporated into more complex siRNA selection algorithms, increasing the reliability of selecting active siRNAs against a single target. Ultimately, design of the best siRNA therapeutics will require design of the siRNA itself, in addition to design of the vehicle and other components necessary for it to function in vivo. In this minireview, we summarize the evolution of siRNA selection techniques with a particular focus on one issue of current importance to the field, how best to identify those siRNA sequences likely to have high activity. Approaches to designing active siRNAs through chemical and structural modifications will also be highlighted. As the understanding of how to control the activity and specificity of siRNAs improves, the potential utility of siRNAs as human therapeutics will concomitantly grow.     

More complete gene silencing by fewer siRNAs: transparent optimized design and biophysical signature

Highly accurate knockdown functional analyses based on RNA interference (RNAi) require the possible most complete hydrolysis of the targeted mRNA while avoiding the degradation of untargeted genes (off-target effects). This in turn requires significant improvements to target selection for two reasons. First, the average silencing activity of randomly selected siRNAs is as low as 62%. Second, applying more than five different siRNAs may lead to saturation of the RNA-induced silencing complex (RISC) and to the degradation of untargeted genes. Therefore, selecting a small number of highly active siRNAs is critical for maximizing knockdown and minimizing off-target effects. To satisfy these needs, a publicly available and transparent machine learning tool is presented that ranks all possible siRNAs for each targeted gene. Support vector machines (SVMs) with polynomial kernels and constrained optimization models select and utilize the most predictive effective combinations from 572 sequence, thermodynamic, accessibility and self-hairpin features over 2200 published siRNAs. This tool reaches an accuracy of 92.3% in cross-validation experiments. We fully present the underlying biophysical signature that involves free energy, accessibility and dinucleotide characteristics. We show that while complete silencing is possible at certain structured target sites, accessibility information improves the prediction of the 90% active siRNA target sites. Fast siRNA activity predictions can be performed on our web server at http://optirna.unl.edu/.     

Effect of asymmetric terminal structures of short RNA duplexes on the RNA interference activity and strand selection

Short interfering RNAs (siRNAs) are valuable reagents for sequence-specific inhibition of gene expression via the RNA interference (RNAi) pathway. Although it has been proposed that the relative thermodynamic stability at the 5'-ends of siRNAs plays a crucial role in siRNA strand selection, we demonstrate here that a character of the 2-nt 3'-overhang of siRNAs is the predominant determinant of which strand participates in the RNAi pathway. We show that siRNAs with a unilateral 2-nt 3'-overhang on the antisense strand are more effective than siRNAs with 3'-overhangs at both ends, due to preferential loading of the antisense strand into the RNA-induced silencing complex (RISC). Regardless of the relative thermodynamic stabilities at the ends of siRNAs, overhang-containing strands are predominantly selected as the guide strand; whereas, relative stability markedly influences opposite strand selection. Moreover, we show that sense strand modifications, such as deletions or DNA substitutions, of siRNAs with unilateral overhang on the antisense strand have no negative effect on the antisense strand selection, but may improve RNAi potency. Our findings provide useful guidelines for the design of potent siRNAs and contribute to understanding the crucial factors in determining strand selection in mammalian cells.     

A novel program to design siRNAs simultaneously effective to highly variable virus genomes

A major concern of antiviral therapy using small interfering RNAs (siRNAs) targeting RNA viral genome is high sequence diversity and mutation rate due to genetic instability. To overcome this problem, it is indispensable to design siRNAs targeting highly conserved regions. We thus designed CAPSID (Convenient Application Program for siRNA Design), a novel bioinformatics program to identify siRNAs targeting highly conserved regions within RNA viral genomes. From a set of input RNAs of diverse sequences, CAPSID rapidly searches conserved patterns and suggests highly potent siRNA candidates in a hierarchical manner. To validate the usefulness of this novel program, we investigated the antiviral potency of universal siRNA for various Human enterovirus B (HEB) serotypes. Assessment of antiviral efficacy using Hela cells, clearly demonstrates that HEB-specific siRNAs exhibit protective effects against all HEBs examined. These findings strongly indicate that CAPSID can be applied to select universal antiviral siRNAs against highly divergent viral genomes.     

A protocol for designing siRNAs with high functionality and specificity

Effective gene silencing by the RNA interference (RNAi) pathway requires a comprehensive understanding of the elements that influence small interfering RNA (siRNA) functionality and specificity. These include (i) sequence space restrictions that define the boundaries of siRNA targeting, (ii) structural and sequence features required for efficient siRNA performance, (iii) mechanisms that underlie nonspecific gene modulation and (iv) additional features specific to the intended use (i.e., inclusion of native sugar or base chemical modifications for increased stability or specificity, vector design, etc.). Attention to each of these factors enhances siRNA performance and heightens overall confidence in the output of RNAi-mediated functional genomic studies. Here, we provide a detailed protocol explaining the methodologies used for manual and web-based design of siRNAs.     

Improved and automated prediction of effective siRNA

Short interfering RNAs are used in functional genomics studies to knockdown a single gene in a reversible manner. The results of siRNA experiments are highly dependent on the choice of siRNA sequence. In order to evaluate siRNA design rules, we collected a database of 398 siRNAs of known efficacy from 92 genes. We used this database to evaluate previously proposed rules from smaller datasets, and to find a new set of rules that are optimal for the entire database. We also trained a regression tree with full cross-validation. It was however difficult to obtain the same precision as methods previously tested on small datasets from one or two genes. We show that those methods are overfitting as they work poorly on independent validation datasets from multiple genes. Our new design rules can predict siRNAs with efficacy >/= 50% in 91% of cases, and with efficacy >/=90% in 52% of cases, which is more than a twofold improvement over random selection. Software for designing siRNAs is available online via a web server at or as a standalone version for high-throughput applications.     

Rational siRNA design for RNA interference

Short-interfering RNAs suppress gene expression through a highly regulated enzyme-mediated process called RNA interference (RNAi). RNAi involves multiple RNA-protein interactions characterized by four major steps: assembly of siRNA with the RNA-induced silencing complex (RISC), activation of the RISC, target recognition and target cleavage. These interactions may bias strand selection during siRNA-RISC assembly and activation, and contribute to the overall efficiency of RNAi. To identify siRNA-specific features likely to contribute to efficient processing at each step, we performed a systematic analysis of 180 siRNAs targeting the mRNA of two genes. Eight characteristics associated with siRNA functionality were identified: low G/C content, a bias towards low internal stability at the sense strand 3'-terminus, lack of inverted repeats, and sense strand base preferences (positions 3, 10, 13 and 19). Further analyses revealed that application of an algorithm incorporating all eight criteria significantly improves potent siRNA selection. This highlights the utility of rational design for selecting potent siRNAs and facilitating functional gene knockdown studies.     

In Silico Target-Specific siRNA Design Based on Domain Transfer in Heterogeneous Data

RNA interference via exogenous small interference RNAs (siRNA) is a powerful tool in gene function study and disease treatment. Designing efficient and specific siRNA on target gene remains the key issue in RNAi. Although various in silico models have been proposed for rational siRNA design, most of them focus on the efficiencies of selected siRNAs, while limited effort has been made to improve their specificities targeted on specific mRNAs, which is related to reducing off-target effects (OTEs) in RNAi. In our study, we propose for the first time that the enhancement of target specificity of siRNA design can be achieved computationally by domain transfer in heterogeneous data sources from different siRNA targets. A transfer learning based method i.e., heterogeneous regression (HEGS) is presented for target-specific siRNA efficacy modeling and feature selection. Based on the model, (1) the target regression model can be built by extracting information from related data in other targets/experiments, thus increasing the target specificity in siRNA design with the help of information from siRNAs binding to other homologous genes, and (2) the potential features correlated to the current siRNA design can be identified even when there is lack of experimental validated siRNA affinity data on this target. In summary, our findings present useful instructions for a better target-specific siRNA design, with potential applications in genome-wide high-throughput screening of effective siRNA, and will provide further insights on the mechanism of RNAi.     

On the art of identifying effective and specific siRNAs

Small interfering RNAs (siRNAs) have been widely exploited for sequence-specific gene knockdown, predominantly to investigate gene function in cultured vertebrate cells, and also hold promise as therapeutic agents. Because not all siRNAs that are cognate to a given target mRNA are equally effective, computational tools have been developed based on experimental data to increase the likelihood of selecting effective siRNAs. Furthermore, because target-complementary siRNAs can also target other mRNAs containing sequence segments that are partially complementary to the siRNA, most computational tools include ways to reduce potential off-target effects in the siRNA selection process. Though these methods facilitate selection of functional siRNAs, they do not yet alleviate the need for experimental validation. This perspective provides a practical guide based on current wisdom for selecting siRNAs.     

Defining the optimal parameters for hairpin-based knockdown constructs

Induction of gene silencing using intracellularly expressed silencing triggers has been explored for large-scale loss-of-function screening, creation of knockdown cell lines or knockdown animals, and disease intervention. In all of these applications, the use of highly potent silencing constructs can maximize the possibility of obtaining target knockdown and thereby is intrinsically important for the chance of success. Several attempts have been made to improve the potency of a silencing construct. Results published in high profile journals such as Nature Biotechnology and Nature Genetics suggest that shRNAs with a 29-nucleotide (nt) stem is much more potent than shRNAs with a 19-nt stem, and miR30-based silencing constructs are much more potent than shRNA-based constructs. In this study, we systematically investigated several parameters, including the use of shRNA- or miR30-based scaffolds, the length of shRNA, and the selection of shRNA sequences for their impact on the knockdown efficiency of a silencing construct. Our studies revealed that the optimal configurations for a potent silencing trigger could be an shRNA with a 19-nt stem and a 9-nt loop. By comparing properties that favor the functional shRNAs and siRNAs using a set of 190 shRNAs against 19 targets and 360 siRNAs against four targets, we found that the functional shRNAs and siRNAs displayed similar but not identical nucleotide preferences. Based on the characteristic nucleotide preferences in the functional versus the nonfunctional shRNAs, we developed a computer program that outperforms an advanced siRNA selection algorithm for the enrichment of highly functional shRNAs.     

Filtering of ineffective siRNAs and improved siRNA design tool

MOTIVATION: Short interfering RNAs (siRNAs) can be used to suppress gene expression and possess many potential applications in therapy, but how to design an effective siRNA is still not clear. Based on the MPI (Max-Planck-Institute) basic principles, a number of siRNA design tools have been developed recently. The set of candidates reported by these tools is usually large and often contains ineffective siRNAs. In view of this, we initiate the study of filtering ineffective siRNAs. RESULTS: The contribution of this paper is 2-fold. First, we propose a fair scheme to compare existing design tools based on real data in the literature. Second, we attempt to improve the MPI principles and existing tools by an algorithm that can filter ineffective siRNAs. The algorithm is based on some new observations on the secondary structure, which we have verified by AI techniques (decision trees and support vector machines). We have tested our algorithm together with the MPI principles and the existing tools. The results show that our filtering algorithm is effective. AVAILABILITY: The siRNA design software tool can be found in the website http://www.cs.hku.hk/~sirna/ CONTACT: smyiu@cs.hku.hk     

Whole-Genome Thermodynamic Analysis Reduces siRNA Off-Target Effects

Small interfering RNAs (siRNAs) are important tools for knocking down targeted genes, and have been widely applied to biological and biomedical research. To design siRNAs, two important aspects must be considered: the potency in knocking down target genes and the off-target effect on any nontarget genes. Although many studies have produced useful tools to design potent siRNAs, off-target prevention has mostly been delegated to sequence-level alignment tools such as BLAST. We hypothesize that whole-genome thermodynamic analysis can identify potential off-targets with higher precision and help us avoid siRNAs that may have strong off-target effects. To validate this hypothesis, two siRNA sets were designed to target three human genes IDH1, ITPR2 and TRIM28. They were selected from the output of two popular siRNA design tools, siDirect and siDesign. Both siRNA design tools have incorporated sequence-level screening to avoid off-targets, thus their output is believed to be optimal. However, one of the sets we tested has off-target genes predicted by Picky, a whole-genome thermodynamic analysis tool. Picky can identify off-target genes that may hybridize to a siRNA within a user-specified melting temperature range. Our experiments validated that some off-target genes predicted by Picky can indeed be inhibited by siRNAs. Similar experiments were performed using commercially available siRNAs and a few off-target genes were also found to be inhibited as predicted by Picky. In summary, we demonstrate that whole-genome thermodynamic analysis can identify off-target genes that are missed in sequence-level screening. Because Picky prediction is deterministic according to thermodynamics, if a siRNA candidate has no Picky predicted off-targets, it is unlikely to cause off-target effects. Therefore, we recommend including Picky as an additional screening step in siRNA design.     

Improved serum stability and biophysical properties of siRNAs following chemical modifications

Small interfering RNAs (siRNAs) which mediate sequence-specific gene silencing through RNA interference have been harnessed for therapeutic applications. Here, we found combinations of modest 2′-ribose siRNA modifications that permit effective gene silencing and improve siRNA stability in human serum. Introduction of as few as two different or identical 2′-ribose modifications at a specific position of either strand, which has been previously suggested to suppress unintended off-target effects, enhances the thermal and thermodynamic stability of siRNA duplexes, and does not alter the overall A-form helical geometry that is required for silencing activity. Our results provide a useful template for the design for therapeutic siRNAs. Ik Sang Cho and Jung Kim contributed equally to this work.     

Solid-phase synthesis of modified RNAs containing amide-linked oligoribonucleosides at their 3'-end and their application to siRNA

siRNAs against luciferase mRNA were modified with amide-linked oligoribonucleosides (amide-linked RNA) at their 3 '-overhangs. Tm values of the modified siRNAs increased compared with that of the unmodified siRNA. These results indicate that the modified overhangs increase the thermodynamic stability of the siRNAs. The modified overhangs improved stability of siRNAs against degradation by nuclease S1 and 50% mouse plasma. Furthermore the modified siRNAs reduced the target gene expression in a similar manner to the unmodified siRNA in cultured cells. These results suggest that the overhang modifications are tolerated for the siRNA activity.