Chemical modifications on the passenger strand impair cleavage of a target RNA. (A) Graphical representation of short interfering RNAs (siRNAs) containing unmodified or chemically modified passenger strands. Passenger strands are shown in grey and guide strands in black in this report. 2′-O-methyl modifications are depicted as circles and phosphorothioates as triangles. (B) Phosphorimaging of cleavage reactions using siRNAs depicted in (A) resolved on a 6% denaturing polyacrylamide gel. Arrows point to the RNA substrate and the labelled 5′-cleavage product. The percentage of cleaved target RNA is indicated at the bottom of the gel. (C) Graphical representation of siRNAs containing 4S-U substitutions (indicated as bars) at different positions on the passenger strand. (D) Phosphorimaging of cleavage reactions using siRNAs depicted in (C). (E) Graphical representation of siRNAs containing 4-nt mismatches at different positions on the passenger strand. (F) Phosphorimaging of cleavage reactions using siRNAs depicted in (E).

The passenger strand is cleaved during RNA-induced silencing complex assembly. (A) Graphical representation of short interfering RNAs (siRNAs) composed of unmodified guide strands and unmodified or chemically modified passenger strands. In all cases, the passenger strands were 5′ phospho-radiolabelled (indicated by ‘³²P'). The dotted line depicts the position where cleavage is predicted to take place on the passenger strand, that is, between guanosine-9 (G₉) and adenosine-10 (A₁₀). The expected 9-nt cleavage product is indicated. (B) Graphical representation of an siRNA formed by a 5′-phospho-radiolabelled, unmodified passenger strand (for the sequence, see ‘mismatch 9–12' in the legend of Fig 1E) and guide strands, the sequence of which either compensates for the 4-nt mutation in the passenger strand or leads to a central 4-nt mismatch. The dotted line depicts the predicted cleavage position. (C) Graphical representation of a blunt siRNA formed by an unmodified guide strand and a 5′-phospho-radiolabelled, unmodified passenger strand, in which the sequence has been shifted 2 nt towards the 5′ end. The dotted line depicts the predicted cleavage position, which has also shifted by 2 nt. The expected 11-nt cleavage product is indicated. (D) Phosphorimaging analysis of a time-course cleavage reaction resolved in a 15% denaturing gel electrophoresis. The region of the gel corresponding to sizes between 8 and 12 nt has been enhanced for an optimal visualization of the cleavage products. The picture of the whole gel is depicted in supplementary Fig S4 online.

Functional RNA-induced silencing complex assembles on a short interfering RNA in which an otherwise non-cleavable passenger strand is nicked at the putative cleavage site. (A) Graphical representation of short interfering RNAs (siRNAs) formed by one (21 nt) or two passenger strands (x+y) of different lengths leading to a nicked passenger strand of 21 nt. Numbers on the right refer to the length of the strands in nucleotides. The asterisk indicates a 2′-O-methyl ribose at G₉. (B) Phosphorimaging analysis of a cleavage reaction resolved in a 6% denaturing gel electrophoresis showing the efficiency of target RNA cleavage as a function of the position of a nick on the passenger strand and the presence or absence of a 2′-O-methyl ribose at the cleavage position. (C) Graphical representation of siRNAs formed by two passenger strands (12 nt+9 nt), in which the 12-nt fragment has been 5′ phospho-radiolabelled (indicated by ‘³²P'), and either left unmodified (duplex 12+9) or modified with a 2′-O-methyl ribose at position G₉ (duplex 12^*+9). Numbers on the right refer to the length of the strands in nucleotides. The dotted line depicts the position where cleavage is predicted to take place on the passenger strand, that is, between G₉ and A₁₀. The expected 9-nt cleavage product is indicated. (D) Phosphorimaging analysis of a time-course cleavage reaction resolved in a 15% denaturing gel electrophoresis. A phosphorylated, 9-nt RNA oligonucleotide was used as a marker. Note the different pattern between the cleavable and non-cleavable passenger strand. Whereas unspecific 3′-to-5′ exonucleolytic degradation of the non-cleavable passenger strand leads to the steady accumulation of <12 nt species during the course of the reaction, the cleavable passenger strand, which is subjected to a similar degradation, shows a clear enrichment of the 9-nt species. Arrows indicate the time point at which the quantification in (E) was performed. (E) Quantification of the phosphorimaging analysis in (D). Bars represent the relative intensity of the bands, the sizes of which are indicated along the x axis. The 9-nt cleavage product accumulates only when using the unmodified, cleavable 12-nt fragment of the passenger strand.

In vitro and in vivo analysis of a putative bypass mechanism for RNA-induced silencing complex assembly. (A) Graphical representation of short interfering RNA (siRNAs) containing unmodified or chemically modified passenger strands. (B) Phosphorimaging analysis of a cleavage reaction resolved in a 6% denaturing gel electrophoresis. (C) Same as (A). (D) Luciferase activities measured after 6 or 24 h after transfecting HeLa cells with siRNAs depicted in (C). (E) Same as (A). (F) Western blot analysis of a knockdown experiment in HeLa cells 24 h after transfecting HeLa cells with siRNAs depicted in (E).