Table 1.
Summary of 454 sequencing.
Figure 1.
Dot blot and 454 profile of vsiRNAs.
Short RNA fraction was purified from in vitro transcribed virus infected plants, dephosphorylated, 5′ labelled and hybridised to membranes containing 21mer oligonucleotides in a one nucleotide sliding window complementary to the plus strand of CymRSV in positions 2650–3169 (A–D) or 4300–4505 (G). Tissues of wild type N. benthamiana infected with either wild type CymRSV (A and G) or with p19 deficient mutant CymRSV (B) and either transgenic N. benthamiana with reduced RDR6 expression (C) or wild type N. clevelandii (D) infected with wild type CymRSV were used for RNA extraction. Signal intensity of dots on panel A was quantified and plotted along the viral genome (E) and compared with the read frequency for the same region obtained through 454 sequencing (F). Signal intensity of dots on panel G was quantified and plotted along the viral genome (H) and compared with the read frequency for the same region obtained through 454 sequencing (I). The layout of the membranes in panel A–D is the same since the same membrane was used after stripping. The position of the most 5′ oligonucleotides is indicated as 1 at the top left corner. The membranes contain two kinds of negative controls: two dots contain only water (W) and two non-specific oligonucleotides are M13 forward (F) and reverse (R). We monitored the specificity of hybridisation by dotting shorter oligonucleotides: an abundant probe (Cym12) was shortened from the 5′ end by 1, 2, 3, 4 and 5 nucleotides (Cym12-1 indicates the 20mer probe followed by the shorter probes). The same approach is used for another abundant probe Cym87.
Figure 2.
Terminator digest of synthetic siRNAs and vsiRNAs.
A 19mer in vitro synthesised and phosphorylated siRNA was digested with Terminator™ 5′-Phosphate-Dependent Exonuclease in decreasing concentration without (left panel) or after denaturation (right panel)(A). The sense strand of the phosphorylated 19mer RNA oligonucleotide was digested in the highest concentration as a control for phosphorylation. The efficiency of the digestion was monitored by northern blot assay using a probe complementary to the sense strand of the synthetic siRNA. Total RNA from CymRSV infected N. benthamiana plants was mixed with the 19mer in vitro synthesised and phosphorylated siRNA and digested with Terminator™ 5′-Phosphate-Dependent Exonuclease in decreasing concentration without (left panel) or after denaturation (right panel) (B). The sense strand of the phosphorylated 19mer RNA was also mixed with total RNA from CymRSV infected N. benthamiana plants and digested in the highest concentration. The efficiency of the digestion was monitored by northern blot assay using probes complementary to the minus strand of the virus (Cym mix, a pool of five probes or Cym 3025), miR159, U6 snRNA or the sense strand of synthetic siRNA, respectively.
Figure 3.
Terminator digest of imperfect duplexes.
In vitro synthesised and phosphorylated imperfect RNA duplexes (A) were digested with Terminator™ 5′-Phosphate-Dependent Exonuclease at the concentration of 2 µM without (left panel) or after denaturation (right panel) (B). The efficiency of the digestion was monitored by northern blot assay using a probe complementary to the sense strand of the imperfect duplexes, which was identical in all duplexes. 2mm, 3mm and 4mm siRNAs contained 2, 3 or 4 mismatches, respectively. 4mm mid siRNA also contained 4 mismatches but at a more central location. The imperfect duplexes were digested at a decreasing concentration without denaturation (C). As a control, the perfect duplex was digested at the lowest concentration (0.002 µM).
Table 2.
Summary of Solexa sequencing.
Figure 4.
Solexa and dot plot profiles of vsiRNAs.
vsiRNA sequences obtained by Solexa sequencing were mapped to the CymRSV genome and normalised abundances were plotted for each sample (A–D). Short RNAs were sequenced from wild type (CymRSV) or silencing protein disabled (Cym19stop) in vitro transcript infected N. benthamiana following two different protocols. The 5′ adapter was either directly ligated to the short RNAs (PHOS) or first depohosphorylated and then re-phosphorylated before adapter ligation (CIP). Profiles obtained by dot blot and Solexa sequencing were compared for regions 2650–3169 (E) and 4300–4505 (F). Please note that the dot blot profiles are the same as those shown in Figure 1E and H.
Figure 5.
Functional analysis of vsiRNAs in planta.
A: Schematic illustration of the GFP sensor sequences and their origins in the viral genomes. The GFP and viral coding sequences are indicated by coloured open boxes, and non coding regions are indicated by thick black lines. Viral target sequences, CymRSV: A to K and Photos Latent Virus (PoLV) [24] as a control, inserted downstream to the GFP ORF are scaled and positioned along the viral genomes. B: Length and position of viral sensor sequences are listed, and regions that contain hotspots are shown in bold. C: Northern blot analysis of viral sensor mRNAs agroinfiltrated into Cym19stop-infected N. benthamiana plants. RNA extracts at 3 days after agroinfiltration were analysed with a 32P-labeled RNA probe raised against the 3′-terminal part of the GFP ORF. PoLV indicates the negative uncleaved control, while the CymRSV viral sequences are indicated by the capital letters. The letter followed by “-” indicates the negative orientation of the sensor sequence, rRNA was used as loading control.