A Brief Historical Perspective on siRNA Design

Ciaran Faherty, Ph.D.

When designing a functional siRNA, major problems included the disparity between predicted and actual silencing activity and the specificity of the siRNA for its target.  The main challenge to designing these small regulatory sequences was identifying appropriate guidelines that were well-grounded on antecedent biological and biochemical principles, some relevant to functional silencing (e.g. G/C content, thermodynamic bias) and others that impeded non-productive activity (e.g., chemical modifications to block non-specific or off-target events).  Thus, the discovery that RNA interference (RNAi) proceeded via a complex endogenous multi-step mechanism was critical to the development of design algorithms.  This meant that a systematic approach could be applied to the design of siRNAs against a single gene or an entire genome.   In the following paragraphs we highlight some of the major contributions that defined practical siRNA design principles. .   
 

Figure1. Different Features of siRNA Design. Applied together a number of criteria are shown that markedly enhanced function and specificity of siRNA design. Blue shade box represents a hexamer sequence; gray shaded box represents 3’ termini of the sense strand where weaker base pairing facilitates enhanced guide strand uptake by RISC. Adapted from references.3, 4

   
One of the first features shown to be essential for siRNA functionality came from the analysis of microRNAs.   Detailed studies of the thermodynamic signatures of these endogenous hairpin substrates of the RNAi pathway revealed a consistent energy profile. In particular, differential instability of the mature microRNA termini and relatively relaxed structures at the center of the duplex were found to be hallmarks of the microRNAs derived from multiple species. Similar analysis of large collections of functional siRNAs uncovered comparable profiles, and subsequent  models built upon these observations proposed that instability on the 5’ terminus of the antisense strand played a role in duplex unwinding and strand selection, while “breathing” in the interior regions was important for adopting conformations amenable to RISC-mediated mRNA cleavage.^{1, 2}  Thus, emulating nature’s own design principles provided a roadmap for a systematic approach to design functional genomic tools; siRNAs.  Understanding and applying these principles proved fruitful in ensuring design of functional siRNAs.  However, to fully translate these initial observations into an efficient and accurate method for predicting the function of a siRNA required yet additional information.                     

Emboldened by the findings described above Reynolds et al. set about to systematically analyze the function of a test set of siRNAs that “walked” in two-base increments across specific gene targets.³  These siRNAs exhibited a diverse range of silencing abilities, demonstrating that a mere two-base shift along the target sequence could dramatically affect siRNA function.  The task was then to delineate those biophysical or thermodynamic properties that would distinguish the functional and nonfunctional siRNA sequences.  This approach led to several criteria that enhanced siRNA function (summarized in Figure 1).  Applied singularly, each criterion exhibited only incremental improvements but when combined into a single algorithm, these parameters markedly enhanced the ability to predict functionality and established the first rational approach to siRNA design.  With this cornerstone for reproducibly predicting functional siRNAs in place, other criteria such as filters for known toxic motifs and interferon responsive elements, could be incorporated into rational design strategies.                  

While applying core principles of the endogenous microRNA mechanism helped define parameters for the design of functional siRNA it also became apparent that sequence-specific off-target effects were also an inherent consequence of this regulatory pathway.  These effects were perpetuated by what is commonly referred to as the seed region (positions 2-7 or 2-8 of the targeting strand).  While earlier studies intimated a casual relationship between microRNA seed regions and siRNA-dependant off-target effects, Birmingham et al. rigorously tested and demonstrated a significant association.⁴  Anderson et al. further validated the role of matches to known seeds and correlated the frequency of these matches in the 3’UTR of gene targets to the specificity of a given siRNA.  Armed with this information, our siRNA design algorithm now filters sequences with undesirable seed regions to enhance specificity and minimize off-target effects and false positive phenotypes.

Clearly, siRNA design has benefited from elucidation of the endogenous RNAi biology and led to the development of powerful algorithms with unprecedented predictive power for enhanced function, potency and specificity.5, 6  Thus, the early hurdles associated with selecting highly functional, specific siRNAs have largely been addressed allowing us to focus our efforts on issues of delivery and sustained knockdown.  To address these latter challenges, stay tuned to future m360 notes which will focus on novel breakthroughs, including SMARTvector technology and Accell siRNA.

Take advantage of the full complement of up-to-date advances in pre-designed siRNA, or design your own siRNA using some of the critical features offered in our siDESIGN Center.

References

1. Schwarz et al. Asymmetry in the assembly of the RNAi enzyme complex. Cell. 2003. 115. p.199-208.
2. Khvorova et al. Functional siRNAs and miRNAs exhibit strand bias. Cell, 2003. 115: p.209-216.
3. Reynolds et al. Rational siRNA design for RNA interference. Nature Biotechnology, 2004. 22: p.326-330.
4. Birmingham et al. 3’UTR seed matches, but not overall identity, are associated with RNAi off-targets. Nature Methods. 2006. 3: p.199-204.
5. Anderson et al. Experimental Validation of the Importance of Seed Frequency to siRNA Specificity. 2008. RNA. In press
6. Birmingham et al. A protocol for designing siRNAs with high functionality and specificity. Nature Protocols. 2007. 2: p.2068-2077.