Distinct extremely abundant siRNAs associated with cosuppression in petunia

Abstract

Cosuppression is a classical form of eukaryotic post-transcriptional gene silencing. It was first reported in transgenic petunia, where a sense transgene meant to overexpress the host Chalcone Synthase-A (CHS-A) gene caused the degradation of the homologous transcripts and the loss of flower pigmentation. In this work, we used deep sequencing technology to characterize in detail the small RNA population generated from the CHS-A sequence in cosuppressed transgenic petunia. Unexpectedly, two distinct small interfering RNAs (siRNAs) were found to vastly predominate. Our demonstration that they guide prominent cleavage events in CHS-A mRNA provides compelling and previously lacking evidence of a causative association between induction of individual siRNAs and an example of cosuppression. The preferential accumulation of these siRNAs provides new insights about sense cosuppression that may apply to other natural and engineered RNA silencing events.

FIGURE 1.

Number of CHS-A siRNAs in WT versus silenced petals. TPM indicates transcripts per million.

FIGURE 2.

Position and abundance of CHS-A gene-matching siRNAs in silenced flowers (A) compared with WT flowers (B). Upper chart: sense strand; lower chart: antisense strand. Nucleotide positions and annotations are relative to the CHS-A gene sequence (GenBank X14591.1).

FIGURE 3.

(A) Cleavage site mapping around the extremely abundant siRNAs in CHS-A mRNA in silenced flowers. Numbers of RLM 5′ RACE clones are reported for each nucleotide position (genomic coordinates are reported for convenience). Although nonquantitative, one of the predicted cleavage sites predominated and the other was prominent. (B) Abundance distribution and phasing of small RNAs relative to the CHS-A coding sequence in silenced flowers. Top: total number of small RNA reads matching to CHS-A coding sequence in each position (genomic coordinates are reported for convenience). The most abundant siRNAs were not included in the analysis to avoid a spurious phasing signal. The approximate position of the major cleavage site is shown as a vertical gray line. TPM, transcripts per million. Bottom: phasing signals from small RNAs. Phasing scores are computed according to Howell et al. (2007) with modifications as described in Materials and Methods. Scores are assigned to the intermediate interval position in the phasing plot. Gridlines correspond to 21-nt intervals. Phase drift is illustrated by two regions where the best phase is shifted with respect to the major cleavage site by 1 and 2 nt, respectively. Phasing was not observed in WT flowers (data not shows).

FIGURE 4.

Position and abundance of CHS-J gene-matching siRNAs in silenced flowers. Upper chart: sense strand; lower chart: antisense strand. Nucleotide positions and annotation are relative to the CHS-J genomic sequence (GenBank X14597.1). Eighty-two percent of the siRNAs, including the most abundant one (phy-siR3), were unique to CHS-J among all CHS sequences available from petunia. In the WT, matching of siRNAs to CHS-J was negligible (data not shown).