|
True Hits vs. False Positives
Kirk M. Brown, PhD
As a Field Application Scientist at Dharmacon, I present technological seminars on diverse topics surrounding the phenomenon of RNAi. Recently during an active discussion of siRNA-mediated off-target silencing, a question was posed to me concerning how one might go about discerning the difference between a true hit and a false positive result. However, before delving into this question, we must first gain an appreciation for the mechanism of off-target silencing.
One critical aspect for the success of an RNAi experiment lies in the specificity of the siRNA. Recent research has shown that siRNAs may not be as specific as initially believed. Jackson et al. performed a microarray analysis using eight different siRNAs targeting the same gene individually. A significant difference was observed in the number and type of genes down-regulated across the genome for each siRNA, given virtually equivalent levels of mRNA and protein knockdown [1]. This result was an eye-opening visual display of the effects of off-targeting by siRNAs. When the presence of an individual siRNA regulates the expression of multiple genes aside from the target gene of interest, enormous difficulty arises in ascribing the resulting phenotype exclusively to silencing of the gene of interest. A second study by Haley and Zamore uncovered a surprising tolerance for mismatches between the siRNA and the target, up to five bases at the 5' end and eight bases at the 3' end, while retaining some level of target knockdown from the duplex [2]. Finally, an example of a phenotypic outcome of off-targets was revealed in a screen designed to identify novel regulators of the hypoxia-inducible factor-1 (HIF-1) pathway [3]. The researchers determined that all three of the 'top hits' were a direct result of off-targeting and therefore, false positives. Further analysis showed a seven nucleotide motif within the siRNA complementary to HIF-1a was shared amongst the off-targeted genes. Moreover, the position and copy number of complementary sites within the transcript was critical in the production of off-targets. With the propensity to generate off-targets and in turn, false positives, it was imperative to uncover the means by which siRNA-mediated off-targets are produced.
One must look to the field of microRNA (miRNA) to gain insight into the mechanism of siRNA-mediated off-targeting. It had been shown that the targeting and regulation of numerous transcripts by miRNAs strongly correlates with the binding of the seed region, bases 2-7 of the targeting or guide strand, with the 3'UTR of targeted genes [4]. Recently, scientists at Dharmacon have discovered a similar mechanism for siRNA-mediated off-target events. Here, off-targeted genes identified in an experiment assessing the signature of siRNAs designed to twelve specific genes were compared directly to a control set of genes lacking off-target silencing. Again, a strong correlation was found between the seed region of each siRNA and the incidence of matches in the 3'UTRs of the off-targeted transcripts [5].
As intimidating as this all sounds, there are solutions to the problem of siRNA-mediated off-targeting which should greatly reduce your chances of producing a false positive result.
1) Functionality plays a critical role. Therefore, begin with rationally designed siRNA.
2) Concentration is also important. Excess siRNA has been shown to trigger non-specific responses including the interferon response and apoptosis. Therefore, the more potent the siRNA, the more likely one can titrate to a safe and effective concentration.
3) Multiplicity of the reagent and reduced overall concentration of individual siRNAs can be accomplished by utilizing pooled siRNAs in your experimental design (e.g. pre-designed SMARTpool silencing reagents). Using multiple siRNAs targeting different regions of the same gene mimics the natural process and increases the likelihood of knocking down the specific gene of interest while simultaneously reducing the effective concentration of each individual siRNA and the opportunity for off-target gene silencing.
4) Select positive, negative and transfection controls. This will provide a baseline and background for your experiment.
5) Finally, and most importantly, monitor your results using multi-parametric assays, especially when relying on phenotypic or protein-based assays that represent indirect measures of mRNA knockdown. It is unlikely that an off-target gene knockdown will meet three of four different experimental criteria thus eliminating false positives.
Problems using RNAi, in my opinion, stem from exploiting an evolutionarily conserved biological pathway whose intricacies and complexities are not fully understood. However, rigorous investigation is currently on-going, providing new data that impacts and updates the ever-changing model for the RNAi mechanism. Thus, to gain the greatest confidence in your results in spite of the complexity of the RNAi mechanism, it is imperative to adhere to the aforementioned guidelines when embarking on an experimental journey with RNAi.
References
1. Jackson, A.L., et al. Nature Biotechnology, 2003. 21(6): p. 635-637.
2. Haley, B. and Zamore, P. Nat. Struct. and Mol. Biol., 2004. 11(7): p. 599-606.
3. Lin, X., et al. Nucleic Acid Research, 2005. 33(14): p. 4527-4535.
4. Lim, L.P., et al. Nature, 2005. 433(7027): p. 769-773.
5. Birmingham, A., et al. Nature Methods, 2006. 3(3): p. 199-204.
|
|
|
|
|
