Analysis of siRNA specificity on targets with double-nucleotide mismatches

Abstract

Although RNA interference as a tool for gene knockdown is a great promise for future applications, the specificity of small interfering RNA (siRNA)-mediated gene silencing needs to be thoroughly investigated. Most research regarding siRNA specificity has involved analysis of affected off-target genes instead of exploring the specificity of the siRNA itself. In this study we have developed an efficient method for generating a siRNA target library by combining a siRNA target validation vector with a nucleotide oligomix. We have used this library to perform an analysis of the silencing effects of a functional siRNA towards its target site with double-nucleotide mismatches. The results indicated that not only the positions of the mismatched base pair have an impact on silencing efficiency but also the identity of the mismatched nucleotide. Our data strengthen earlier observations of widespread siRNA off-target effects and shows that approximately 35% of the double-mutated target sites still causes knockdown efficiency of >50%. We also provide evidence that there may be substantial differences in knockdown efficiency depending on whether the mutations are positioned within the siRNA itself or in the corresponding target site.

Figure 1.

The oligomix containing all 1539 sequence variations of the siCD46 target site were PCR amplified with flanking primers resulting in a 85-bp long DNA molecule. The target sequence in this figure represents the wild-type version.

Figure 2.

Color-coded expression chart visualizing positional and regional trends for siCD46 knockdown efficiency and tolerance towards double-mutations in its specific target site. The vertical axis represents the position of the first mutation and the horizontal axis represents the position of the second mutation. The cross-section of each pair illustrates the remaining luciferase expression of each clone. Seven hundred and nine out of 1539 different combinations were identified and analyzed. These clones are each represented in the chart with a color corresponding to the remaining expression activity after targeting with siCD46. Dark green represents <10% remaining expression activity, lime green: <20%, light green: <30%, yellow: 30–40%, pink: >40%, red: >50% and black: >80% remaining activity. Light gray represents wild type target sequence and white nonidentified clones.

Figure 3.

Comparative study of mutations occurring in the siRNA antisense strand or in the corresponding positions of its target site. Light gray bars: siQ46-9T_17A and si-9T_17A, medium gray bars: siQ46-3A_14C and si-3A-14C, dark gray bars: siQ46-2C_8A and si-2C_8A, white bar: siQ46-wt co-transfected with unrelated siRNA. T—mutation of siRNA target, S—mutation of siRNA antisense strand. The bars represent remaining expression levels of firefly luciferase activity normalized to renilla luciferase. All values are also normalized to each reporter construct's own ability to express luciferase. *P < 0.05 and **P < 0.01, respectively are significant differences of luciferase expression between cells transfected with either mutated siQuant target constructs or mutated siRNAs.