Modulation of the RNA Interference Activity Using Central Mismatched siRNAs and Acyclic Threoninol Nucleic Acids (aTNA) Units

Abstract

The understanding of the mechanisms behind nucleotide recognition by Argonaute 2, core protein of the RNA-induced silencing complex, is a key aspect in the optimization of small interfering RNAs (siRNAs) activity. To date, great efforts have been focused on the modification of certain regions of siRNA, such as the 3'/5'-termini and the seed region. Only a few reports have described the roles of central positions flanking the cleavage site during the silence process. In this study, we investigate the potential correlations between the thermodynamic and silencing properties of siRNA molecules carrying, at internal positions, an acyclic L-threoninol nucleic acid (aTNA) modification. Depending on position, the silencing is weakened or impaired. Furthermore, we evaluate the contribution of mismatches facing either a natural nucleotide or an aTNA modification to the siRNA potency. The position 11 of the antisense strand is more permissive to mismatches and aTNA modification, in respect to the position 10. Additionally, comparing the ON-/OFF-target silencing of central mismatched siRNAs with 5'-terminal modified siRNA, we concluded: (i) central perturbation of duplex pairing features weights more on potency rather than silencing asymmetry; (ii) complete bias for the ON-target silencing can be achieved with single L-threoninol modification near the 5'-end of the sense strand.

Figure 1

Chemical structures of flexible acyclic derivatives of nucleosides used in RNA interference experiments. Threoninol and serinol nucleic acids may be formed by D or L stereoisomers.

Figure 2

Schematic representation of a 21-mer siRNA duplex. The upper strand is the sense (SS), the lower strand is the antisense (AS) that guides the cleavage of the cognate mRNA. Only one of the two strands is retained into the Ago2 protein (the antisense), the sense strand is cut and degraded.

Figure 3

(A) IC₅₀ assessments of siRNA molecules. To achieve the IC₅₀ values, HeLa cells were co-transfected with psiCHECK2 (AS) reporter and decreasing amounts (1 nM, 0.3 nM, 60 pM, 16 pM, 8 pM and 2 pM) of siRNA molecules. Luminescence was evaluated 24 h after transfection. n = 3 ± SD; (B) Plot of siRNAs melting temperature used in luciferase study. n = 3 ± SD.

Figure 4

Silencing activities of antisense strand (AS) and sense strand (SS) of mismatched siRNAs at position 9 (left panel) and position 10 (right panel) bearing either natural uridine (wt) or L-threoninol-thymine (T^L). To assess the on-target and off-target effects, 1 nM of the indicated siRNAs were co-transfected with psiCHECK2 reporters (AS or SS, SSU9, SSC9, SSG9, SSU10, SSC10 and SSG10, respectively) in HeLa cells. Mock transfection was set as 100%. n = 3 ± SD.

Figure 5

Renilla mRNA reduction in MEF^wt and MEF^Ago2−/− cells. 1 nM of unmodified (wt) and modified (T10A10 and T11A9) siRNAs were co-transfected with psiCHECK2 (AS) reporter. After 24 h, cells were harvested for RNA extraction and qRT-PCR measurement. n = 2 ± SD.

Figure 6

Plot of antisense ss-siRNAs activities of natural (ASWT) and T^L-modified ss-siRNAs (AST10 and AST11). Variable amounts of siRNAs (100 nM, 50 nM, 10 nM and 5 nM) were co-transfected with psiCHECK2 (AS) reporter in HeLa cells. 24 h post-transfection the luminescence were measured. n = 3 ± SD.

Figure 7

(left panel) ON-/OFF-target silencing of L-threoninol modified siRNA. Unmodified (wt) and modified at position 2 of the sense strand (wtT2) siRNAs, at final concentration of 1 nM, were co-transfected with psiCHECK2 sensors (AS and SS) in HeLa cells. Mock transfection was set as 100%. n = 3 ± SD. (right panel) Silencing activities of unmodified (SSWT) and position 2 T^L-modified (SST2) sense ss-siRNA in HeLa cells. Luminescence were determined 24 h after the co-trasfection of decreasing concentrations of ss-siRNAs (100 nM, 50 nM, 30 nM and 5 nM) with psiCHECK2 (SS) reporter. Mock trasfection was set as 100%. n = 3 ± SD.