The molecular mechanisms of plant gene silencing are beginning to be understood, and significant progress has been made in this area during the past year. In the April issue of Current Opinion in Plant Biology (vol. 5 (2): 146-50), Peter M Waterhouse and Ming-Bo Wang (CSIRO Plant Industry, Australia) discuss the current status of our knowledge about gene silencing in plants, focusing on the role of double stranded RNA (dsRNA) in this process. The authors also discuss potential applications of dsRNA-mediated gene silencing for functional genomics studies and crop improvement.
Gene silencing may occur at transcriptional (TGS) and post-transcriptional (PTGS) levels. There is evidence for a role of short dsRNA molecules, 21-25 nucleotides in length, in both processes. In TGS, cytosine residues in promoter sequences are methylated, which is thought to inactivate the promoter by hindering interactions with transcription factors or triggering formation of heterochromatin in the promoter region. In PTGS, dsRNA is thought to act as a trigger for sequence-specific TNA degradation. Thus, two major processes involved in gene silencing in plants are RNA-directed RNA degradation and RNA-directed DNA methylation (RdDM).
The authors note recent studies that have identified proteins required for gene silencing. These include the identification of an RNA helicase, SILENCING DEFECTIVE3 (SDE3), associated with PTGS. A chromomethylase, CMT3, has also been reported that is required for non-CG methylation and is thought to be involved in RdRM.
First reported in 1990, PTGS was initially termed “cosuppression” after studies in petunia noted that transformation with a transgene directing over-expression of chalcone synthase resulted in silencing of both the transgene and the endogenous gene in some lines. Later studies utilized transgenic tobacco plants expressing different versions of a potato virus Y (PVY) gene; only tobacco lines that expressed the PVY gene in both sense and antisense orientation, but not those expressing only one form, showed resistance to PVY viral infection. Another study utilized transgenic lines expressing an inverted-repeat form of beta-glucuronidase (GUS), in which the transgene transcript formed a hairpin structure. This hairpin RNA structure was capable of silencing resident GUS transgenes. An additional study used transgenes encoding inverted-repeat forms of 1-aminocyclopropane-1-carboxylate (ACC) oxidase to silence endogenous ACC oxidase in tomato. Taken together, these studies indicate that dsRNA is involved in gene silencing in plants. In animal systems, 21-25 nt dsRNA molecules active in RNA interference (RNAi) have been shown to be derived from dsRNA by a nuclease called Dicer. The dsRNA molecules guide the nuclease complex to cognate single-stranded RNA.
Recently, dsRNA molecules have been shown to induce sequence-specific RdDM in plants. This has been studied using plants transformed with potato spindle tuber viroid sequences. When these transgenic lines were subjected to replication of the potato spindle viroid, researchers observed methylation of the resident transgene. Subsequent studies have shown that hypermethylation of homologous nuclear transgene sequences can be triggered by potyviruses or potexviruses, viral satellite RNA or inverted-repeat transgene-derived RNAs. The authors suggest that the involvement of dsRNA in both PTGS and RdDM provides a possible link between PTGS and TGS.
The authors describe applications in which transgenes hairpin RNA structures have been or may be used to silence endogenous or transgene targets. Some studies have indicated that silencing is more efficient if a spliceable intron is included in the hpRNA transgene, though the mechanism for this is not yet clear. Silencing efficiency also seems to be increased by using sequences at least 300 nucleotides in length to target hpRNA constructs.
Silencing of multiple endogenous genes by hpRNA transgenes has been performed successfully, a strategy that may be useful for crop improvement. For example, silencing of cotton desaturase genes has been used to alter the fatty acid contents of cottonseed oil. The authors note that silencing of two or more genes may be possible by using one hpRNA construct containing multiple genes or by transforming plants with several hpRNA constructs. In other applications, a single member of a multigene family may be suppressed by using less conserved regions such as the 5’ or 3’ untranslated region as a target for hpRNA-induced silencing. A target region of sufficient size must be available for specific inactivation, in this case. The authors point out that although tissue- or cell-specific silencing is desirable in many applications, this may be difficult to achieve using hpRNA-induced silencing, as cell-to-cell and long-distance spread of PTGS has been reported.
Another application of hpRNA-induced silencing may be the development of virus-resistant plants. Recent studies have further that, when compared to conventional sense or antisense viral transgene approaches (which often require multiple copies of the transgene), virus-derived hpRNAs are more effective in conferring resistance to viral pathogens. A single copy hpRNA has been shown to be sufficient to protect the plant against viral attacks. In addition, this approach may lessen the risk of recombination between transgene RNA and viral RNA, if viral sequences from non-coding regions are used.
Many applications for hpRNA transgene-induced silencing exist in the field of functional genomics. Existing functional genomics approaches generate loss-of-function mutations using T-DNA or transposon insertions, or induction of sequence variations with chemical or radiation mutagenesis. The authors note that drawbacks to such approaches include the untargeted nature of the mutagenesis, the time required to saturate the genome, and the difficult of linking loss-of-function phenotypes to individual genes. In addition, homozygous lethal mutations will not be recoverable. The authors state that hpRNA-induced gene silencing avoids most of these drawbacks.
The authors describe a vector that has been developed for use in high-throughput studies to generate intron-containing hpRNA constructs. The vector, called pHELLSGATE, contains two pairs of the recombination sites attP1 and attP2 used in the GATEWAY™ system developed by Invitrogen. The recombination sites are separated by an intron and are arranged in inverse orientation. PCR products with attP1 and attP2 sites at either end may be directionally recombined into pHELLSGATE. This step removes a lethal ccdB cassette within the vector, facilitating selection. However, the authors point out that this approach is at present applicable only to the limited number of plant species that are amenable to transformation by Agrobacterium tumefaciens, such as Arabidopsis and japonica rice.
In plants not amenable to Agrobacterium-mediated transformation, virus-induced gene silencing (VIGS) may prove to be an effective tool for a high-throughput functional analysis. VIGS does not require plant transformation to silence a gene and phenotypes can be identified 1-3 weeks after virus inoculation. However, at present, VIGS systems are only available for a small number of viruses that have a narrow range of dicot host ranges. Other VIGS systems, such as one based on tobacco rattle virus, are under development and may bypass this limitation in the future.
Viral RNAs and hpRNAs can cause hypermethylation of homologous promoter sequences, thereby conferring transcriptional silencing of transgenes. The authors point out that it is not yet clear whether endogenous genes can be reliably suppressed by RdDM of their promoter sequences. If viral RdDM is perfected and de novo methylation is retained in the next generation without viral infection, the authors note that this approach may enable manipulation of heritable traits without transgenesis. Alternatively, hpRNAs may be used that include promoter and exon sequences, eliciting both PTGS and TGS.
The authors conclude that RNA-mediated PTGS and RdDM may be used for high throughput reverse genetics approaches to determining gene function. They describe the tools that have been developed to enable such efforts, and predict that PTGS and RdDM techniques will revolutionize investigative plant biology.
Click here to view the abstract.