Whole-genome thermodynamic analysis reduces siRNA off-target effects

Abstract

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.

Figure 1. Workflow for designing, checking and selection of testing siRNAs.

In STEP 1, siRNA candidates are designed using siDesign and siDirect with their sequence-level screening turned on. Commercial siRNAs are designed by the vendors using their proprietary design pipelines which presumably also include sequence-level screening. In STEP 2, all siRNA candidates are examined by Picky to discover any potential off-target genes. Testing siRNAs are separated into the good and bad sets according to their siRNA score, i.e., the identified minimum difference between target and nontarget hybridization temperatures. Off-target genes for commercial siRNAs are similarly predicted by Picky. In STEP 3, the anti-sense strand secondary structures of the top 60 good and bad siRNA candidates are checked by the Mfold server, and 6 good and 7 bad siRNAs are selected for the validation experiments. All 18 commercial siRNAs are used in the experiments.

Figure 2. The secondary structures of a few representative siRNAs.

Examples of siRNAs with active secondary structures and accessible 5′- and 3′-ends: (A) ITPR2-G-design ΔG = 1.40 kcal/mol and (B) ITPR2-B-design ΔG = 1.60 kcal/mol. Commercial siRNAs with unfavourable secondary structures and thus reduced potency: (C) TRIM28-1 (IDT) ΔG = −0.50 kcal/mol and (D) TRIM28-3 (Sigma) ΔG = −6.60 kcal/mol. An interesting exception of a siRNA with active secondary structure but reduced potency: (E) ITPR2-1 (IDT) ΔG = 0.40 kcal/mol. All secondary structures and the Gibbs free energy (ΔG) values are obtained from the Mfold server.

Figure 3. siRNAs targeting IDH1.

(A) The relative expression levels of the target gene: IDH1. Two good siRNAs (-G-) and three bad siRNAs (-B-) are designed to target the IDH1 gene; all of them decrease the target gene expression by about 50%. (B–E) The relative expression levels of off-target genes predicted by Picky. Different subfigures correspond to different off-target genes that are associated with different siRNAs respectively. All expression levels are calibrated to the negative transfection controls (Neg: dark grey bars). The light grey bars correspond to treatments with siRNAs except the one with off-target effect predicted by Picky. A test was performed for significant reduction of gene expression by the siRNA with predicted off-target effect compared with other siRNAs. If p-value <0.05, the treatment using siRNA with the predicted off-target effect is red and labelled by *. Otherwise, it is pink. Remarkably, MAPK10, the predicted off-target gene of IDH1-B-design-3, has about 40% reduction in expression level by IDH1-B-deisgn-3.

Figure 4. siRNAs targeting ITPR2.

(A) The relative expression levels of the target gene: ITPR2. Two good siRNAs (-G-) and two bad siRNAs (-B-) are designed to target the ITPR2 gene. Most of them can reduce the target gene expression level to below 50%. (B–F) The relative expression levels of off-target genes predicted by Picky. Different subfigures correspond to different off-target genes that are associated with different siRNAs respectively. All expression levels are calibrated to the negative transfection controls (Neg: dark grey bars). The light grey bars correspond to treatments with siRNAs except the one with off-target effect predicted by Picky. A test was performed for significant reduction of gene expression by the siRNA with predicted off-target effect compared with other siRNAs. If p-value <0.05, the treatment using siRNA with the predicted off-target effect is red and labelled by *. Otherwise, it is pink. Remarkably, the expression level of UBE2R2 was reduced by 60% by ITPR2-B-design.

Figure 5. siRNAs targeting TRIM28.

(A) The relative expression levels of the target gene: TRIM28. Two good siRNAs (-G-) and two bad siRNAs (-B-) are designed to target the TRIM28 gene. All of them reduce the target gene expression level. (B–D) The relative expression levels of off-target genes predicted by Picky. Different subfigures correspond to different off-target genes that are associated with different siRNAs respectively. All expression levels are calibrated to the negative transfection controls (Neg: dark grey bars). The light grey bars correspond to treatments with siRNAs except the one with off-target effect predicted by Picky. A test was performed for significant reduction of gene expression by the siRNA with predicted off-target effect compared with other siRNAs. If p-value <0.05, the treatment using siRNA with the predicted off-target effect is red and labelled by *. Otherwise, it is pink.

Figure 6. Commercial siRNAs.

(A–C) 18 commercial siRNAs are purchased from Sigma (MISSION siRNA) and IDT (TriFECTa Kit) to target the same genes IDH1, ITPR2 and TRIM28. The commercial siRNA potency on target genes is mostly comparable to that of our siRNAs. (D–H) The relative expression levels of predicted off-target genes in different Sigma siRNA treated cells (IDT siRNAs are not included in off-target analyses). All expression levels are calibrated to the negative transfection controls (Neg: dark grey bars). The light grey bars correspond to treatments with siRNAs except the one with off-target effect predicted by Picky. A test was performed for significant reduction of gene expression by the siRNA with predicted off-target effect compared with other siRNAs. If p-value <0.05, the treatment using siRNA with the predicted off-target effect is red and labelled by *. Otherwise, it is pink. The SOGA1 off-target gene is significantly reduced by TRIM28-2 (Sigma).