Guidelines for the selection of highly effective siRNA sequences for mammalian and chick RNA interference

Abstract

In the present study, the relationship between short interfering RNA (siRNA) sequence and RNA interference (RNAi) effect was extensively analyzed using 62 targets of four exogenous and two endogenous genes and three mammalian and Drosophila cells. We present the rules that may govern siRNA sequence preference and in accordance with which highly effective siRNAs essential for systematic mammalian functional genomics can be readily designed. These rules indicate that siRNAs which simultaneously satisfy all four of the following sequence conditions are capable of inducing highly effective gene silencing in mammalian cells: (i) A/U at the 5' end of the antisense strand; (ii) G/C at the 5' end of the sense strand; (iii) at least five A/U residues in the 5' terminal one-third of the antisense strand; and (iv) the absence of any GC stretch of more than 9 nt in length. siRNAs opposite in features with respect to the first three conditions give rise to little or no gene silencing in mammalian cells. Essentially the same rules for siRNA sequence preference were found applicable to DNA-based RNAi in mammalian cells and in ovo RNAi using chick embryos. In contrast to mammalian and chick cells, little siRNA sequence preference could be detected in Drosophila in vivo RNAi.

Figure 1

Location of 16 cognate siRNAs for silencing of the firefly luc gene.

Figure 2

Relationship between luc siRNA sequence and induced luc gene silencing (RNAi) activity. AS and SS, respectively, in the upper margin, indicate siRNA ends with the 5′ antisense strand and 5′ sense strand ends. (A) Classification of 16 luc siRNAs. siRNA-dependent reduction in firefly luciferase activity in three mammalian (CHO-K1, HeLa and E14TG2a) and Drosophila (S2) cells was examined using 50 nM of 16 siRNAs, a–p, shown in Figure 1. Sixteen siRNAs were aligned according to their RNAi activity in mammalian cells from top to bottom. siRNAs were classified into three groups depending on RNAi or reduction in relative luciferase activity. (B) RNAi activity caused by siRNAs designed using our sequence preference rules. Using the rules, 15 class Ia and five class III siRNAs were designed and their capability to bring about RNAi in CHO-K1, HeLa, E14TG2a and S2 cells was examined. The siRNA number indicates the nucleotide position within the luc coding region, corresponding to the 3′ end of the siRNA AS. The concentration of siRNA was 50 nM and RNAi effects were observed 24 h after transfection. Data obtained from 2–4 experiments were averaged and are shown. Thin vertical lines indicate the average of three mammalian cells. The 7 bp terminal region with the 5′ AS end is boxed. While A/U and G/C in the boxed region are colored in red and blue, respectively, highly conserved, 5′-terminal bases are shown on red (A/U) or blue (G/C) backgrounds. Note that, in class Ia and highly effective siRNAs, the 5′ AS and SS ends, respectively, are almost exclusively A/U and G/C.

Figure 3

Highly effective silencing of endogenous genes by class Ia siRNAs (A and B), RNAi caused by an non-cognate siRNA (C) and class Ia siRNA-dependent RNAi in chick embryos (D). The sequences of siRNAs examined here are depicted in the right margin (see Fig. 2 for coloration and box). (A) Silencing of vimentin, a human endogenous gene, by class Ia and class III siRNAs. Ten class Ia (VIM-270, -368, -596, -812, -857, -1097, -1128, -1148, -1235 and -1298) and three class III siRNAs (VIM-35, -155 and -491) were designed and their RNAi activity was examined in HeLa cells subjected to three cycles of 50 nM siRNA transfection. On day 3, cells were stained for vimentin (target) and Yes (control). Vimentin and Yes protein signals are colored in yellow and green, respectively. Note that all class Ia but not class III siRNAs significantly reduced vimentin signals. (B) Effects of siRNA transfection on the expression of Oct 4, a mouse endogenous gene. E14TG2a (mouse ES) cells were transfected with class Ia (Oct-670, -797 and -821) or class III (Oct-161 and -566). The four pictures on the left are phase contrast photographs. Note that class Ia siRNAs (Oct-670 and -797) induced a flat morphology typical of trophectoderm cells. Class Ia siRNA-specific degradation of Oct 4 mRNA visualized with RT–PCR is also shown. Gapd was used as a control. (C) ECFP RNAi caused by an non-cognate EGFP siRNA. EGFP-441 is a class Ia EGFP siRNA but not identical in sequence to ECFP-441, a class II ECFP siRNA possessing G at the 5′ AS end. While ECFP-441 did not abolish ECFP signals 2 days after transfection (on day 2 of transfection), non-cognate EGFP-441 virtually completely eliminated ECFP signals. Note that RNAi effects 1 day after transfection (on day 1) indicate that EGFP is a better target for EGFP-441 than ECFP. Blue and red arrows indicate that EGFP and ECFP mRNA possess a base substitution at the position corresponding to the 5′ AS end of EGFP-441 or ECFP-441. (D) In ovo RNAi in chick embryo. EGFP and DsRed expression plasmids were co-electroporated into chick spinal cord with class Ia siRNAs (EGFP-416, -441 and DsRed-399) or class III siRNAs (EGCFP-666 and DsRed-383). Class Ia siRNAs significantly abolished the targeted fluorescence, while class III siRNAs did not. Expression of EGFP and DsRed is colored in yellow and red, respectively. Dotted lines show left (L) and right (R) portions of the spinal cords. The 5′ ends of SS and AS, and the 5′ region of the AS (7 bp) are boxed and colored as described in Figure 2.

Figure 4

Dose dependency of RNAi effects in CHO-K1 and S2 cells. The shaded area is the region bounded by two lines, intersecting, respectively, with the horizontal axis at 0.5 and 5 and the 50% line of luc activity at 0.05 and 0.5. The thick vertical bar at the right of each panel indicates the region with >77% reduction in luc activity. (A) Change in luc gene silencing activity with siRNAs ranging from 0.005 to 50 nM in CHO-K1 (left) and S2 (right) cells. siRNAs a–p in Figure 1 are grouped into three classes, I (open circles), II (open triangles) and III (closed circles). (B) Direct comparisons of RNAi activity curves in S2 (open circles) and CHO-K1 cells (filled circles). The sequences of corresponding or similar siRNAs are schematically shown in the lower margin. siRNAs i, e, j, m and h behave the same way as their corresponding examples above. Filled circles, G/C; open circles, A/U. The 7 bp duplex region containing the 5′ end is boxed. Eleven pictures are positioned based on siRNA classification and induced RNAi activity levels.

Figure 5

GC content distribution of highly effective class Ia siRNAs. The distribution of the GC content of 31 highly effective class Ia siRNAs shown in Figures 2 and 3 is presented. Position 1 corresponds to the siRNA duplex end including the 5′ AS end. The 5′ ends of AS and SS of all class Ia siRNAs causing highly effective RNAi in mammalian cells were A/U and G/C, respectively. The average GC content in the region 2–7 was 19%, while that of the region 8–18 was 52%.

Figure 6

Comparison of siRNA-based RNAi and DNA-based RNAi in HeLa cells. The predicted sequences of hairpin-type transcripts are shown on the left, while induced RNAi activity (reduction in relative luciferase activity) is shown by open boxes on the right. Stippled boxes indicate relative luciferase activity reduction due to cognate siRNA in HeLa cells. On the left, predicted AS are shaded. Note that the sequence preference of DNA-based RNAi is essentially identical to that of siRNA-based RNAi. Data obtained from 2–4 experiments were averaged and are shown.

Figure 7

Internal stability profiles of siRNAs. Thermodynamic profiles of highly effective 32 siRNAs (A), and 16 luc siRNAs shown in Figure 1 are composed of three groups giving rise to highly effective, intermediate and ineffective RNAi in mammalian cells (B–D). The vertical bars indicated in (A) show the standard deviation of 32 highly effective siRNAs. Thick and open vertical bars, respectively, indicate free energy change ranges at position 1 of highly effective and ineffective siRNAs.

Figure 8

Possible models of siRNA-based RNAi in mammalian cells. The rules for siRNA sequence preference are schematically shown in (A). siRNA unwinding might be effectively initiated from the AU-rich AS end in the case of class Ia siRNA, lacking a long GC stretch. On the other hand, siRNA duplex unwinding might be suppressed from the GC-rich class III AS end. G/C at the 5′ SS end of class Ia and the 5′ AS end of class III siRNAs might provide a site for binding of an unidentified protein possibly suppressing siRNA unwinding (B). Alternatively, A/U at the 5′ SS end of class III and the 5′ AS end of class Ia siRNAs might serve as a binding site for putative unwinding stimulation factors other than helicase. A long GC stretch such as that found in siRNA n (see Fig. 4B) might prevent the elongation of siRNA duplex denaturation from the AS end (C). A/U at the 5′ AS and SS ends and their counterparts in the SS and AS, respectively, are shown as hatched circles; G or C, closed circles. Terminal AU-rich or GC-rich regions are boxed. Open arrows indicate the direction of siRNA unwinding due to a hypothetical siRNA helicase.