Feature: RNA interference goes mobile

In a now-classic experiment in tobacco and rice plant models, reported in 1998, Peter Waterhouse, Ming-Bo Wang and several CSIRO Plant Industry colleagues confirmed their hunch that plant cells harbour molecular traps armed to detect and destroy invading viruses and silence genes.

In the first demonstration of synthetic RNA interference in plants, they crossed two lines of tobacco containing transgenes coding for complementary strands of a double-stranded RNA sequence copied from Potato virus Y.

The parental lines were susceptible to infection, but 25 per cent of their progeny were completely resistant. They also introduced a transgene into rice encoding a self-complementary RNA that caused potent silencing of the target gene.

The double-stranded, or “hairpin,” RNA molecule that formed in the transgenic plants had armed the plant’s cells to destroy the single-stranded RNA template that the virus requires to replicate in the host plant’s cells. In the rice experiment, the hairpin RNA molecule had induced the degradation of the messenger RNA template for assembling the encoded protein.

US scientists Andy Fire and Craig Mello were the early birds in publishing double-stranded RNA-mediated gene-silencing (RNA interference or RNAi) in the worm, C. elegans, in the same year, an achievement that saw them awarded the 2006 Nobel Prize for Medicine.

A year later, Waterhouse and Wang were awarded the Prime Minister’s Prize for Science as the first researchers to discover RNA interference in plants and for their contributions to developing a set of RNAi technologies that are now widely used in basic and biotechnology research around the world.

Dr Ming-Bo Wang is still at CSIRO Plant Industry, exploring the role of small RNA molecules in the perpetual war between plants and their pathogens. In 2011, he and Dr Neil Smith – another member from the original gene silencing team – finally explained the mystery of how viral satellite RNAs induce plant disease symptoms.

Satellite RNAs are small single-stranded RNA molecules that replicate by hitching a ride into plant cells within the capsids of plant viruses. Lately, Wang and his colleagues have been exploring the role of sRNAs in resistance to fungus diseases, the last and most daunting frontier in plant pathology.

Now a professor at the University of Sydney, Peter Waterhouse is developing small-RNA (sRNA) tools to improve crop plant architectures and induce disease resistance. He also has maintained a strong interest in still unexplained aspect of RNA interference in plants: the rapid, systemic spread of virus resistance from the original point of infection, resulting in the entire plant acquiring resistance to subsequent infection, not only by the original virus, but by related strains.

---PB---

Cross protection

Rewind 72 years, to an experimental plot of tobacco plants in a field in then-rural Arlington County, Virginia, where the hulking Pentagon building, now stands, housing the United States Department of Defense.

In 1929 the US Department of Agriculture plant pathologist Harold McKinney inoculated a small plot of tobacco plants with a mild strain of Tobacco Mosaic Virus (TMV). When McKinney challenged the inoculated plants with a more virulent TMV strain, they turned out to be fully resistant to infection.

The phenomenon, called cross protection, found practical application in tobacco and other horticultural crops like citrus, melons, papaws and grapes, protecting these crops against their viral pathogens.

Explaining how a point inoculation induced powerful, systemic protection that made the plants impervious to attack by all TMV strains, was beyond the science of the day – and continues to challenge latter-day science.

The nature of the silencing signal, and its targets, remain unclear. However, a review paper published last year by Waterhouse and Wang – along with Wang’s CSIRO colleagues Dacheng Liang, Jean Finnegan, and Liz Dennis – in Frontiers in Biology, a new English language journal published in China, summarised a decade of theory and experimentation. Of particular interest was a recent finding by Burnie Carroll of University of Queensland and his co-workers, which proposed a working model of mobile silencing.

Wang and co-authors noted that plants have evolved three modes of RNA-mediated gene silencing:

• The microRNA pathway
• 
  
• The small interfering RNA (siRNA)-mediated post-transcriptional gene-silencing pathway (PTGS)
• 
  
• The siRNA-mediated transcriptional gene silencing (TGS) pathway, where siRNA molecules silence genes by inducing methylation – otherwise known as RNA-directed DNA methylation (RdDM)

Silencing can occur over long distances, resulting in systemic silencing, or spread over short distances between adjacent cells, indicating at least two distinct mechanisms are involved.

With systemic, or long-distance mobile silencing, gene silencing initiated in a few cells, or in a particular tissue, spreads throughout the plant in a pattern resembling the movement of dye via the phloem system.

The signal moves longitudinally, in both directions, passing through the perforated ends of the phloem sieve tubes. En route, it propagates laterally through the phloem cell walls via microscopic holes call plasmodesmata, through which cytoplasmic interconnections allow cells to exchange ions and molecules.

Silencing experiments with hairpin transgenes targeting the messenger RNAs of endogenous genes like the photosynthesis gene RuBisCo result in silencing within a zone 10 to 15 cells deep around the phloem vasculature. A major puzzle is that mobile silencing seems to work only for transgenes, not for endogenous genes.

Small interfering RNAs come in three size classes: 21; 22; and 24 nucleotides (nt). Wang and his co-authors suggest that all size classes can move from cell to cell, and over long distances, and enter cells to induce either systemic or short-distance silencing. The 21- and 22nt siRNAs direct RNA-induced silencing complexes (RISCs) to degrade the messenger RNAs from the target gene, to effect both long- and short-distance silencing.

But 21-22nt siRNAs operate only within a peripheral zone 10-15 cells deep, around the vasculature; by themselves, they are incapable of driving systemic silencing, because they are not self-amplifying – their activity is restricted to the cytosol, where they degrade mRNAs from the target gene after they have exited the cell nucleus.

---PB---

Nuclear interaction

In contrast, 24nt siRNAs enter the cell nucleus and interact with a long RNA transcript from the target gene, inducing cleavage of the long RNA into shorter fragments. The shorter fragments now serve as templates for RNA-dependent RNA polymerase to synthesise double-stranded RNAs corresponding to the target region of the gene.

The Dicer-like enzymes (DCL) found in all plants then “dice” the double-stranded RNAs, yielding 21, to 22nt siRNAs, that in turn direct cleavage of the target mRNA, resulting in systemic silencing. The DCL enzymes can also dice the double-stranded RNA to 24nt siRNAs. Both primary and secondary 21, 22 and 24nt siRNAs can move between cells, via the plasmodesmata, or to remote tissues via the phloem, to repeat the cycle.

So why does the silencing system work for transgenes, but not for endogenous genes? Wang and co-authors believe it is because only transgenes that produce a long nuclear RNA transcript are susceptible to 24nt siRNA-mediated long-distance silencing. Most endogenous genes do not produce such transcripts, and are thus resistant to systemic silencing.

¬¬If their proposed model is correct, there are important practical implications for developing transgenic plants. Plant breeders would like to construct siRNAs to systemically silence endogenous genes in species that that are difficult to transform; either they are not amenable to transformation by the standard gene-delivery vector Agrobacterium tumifaciens, or transformed cells are difficult to regenerate into plantlets in tissue culture.

Plants with a prolonged juvenile period, like many fruit trees, are also problematic as it takes a long time for transgenic plants to reach maturity for phenotypic analysis and for producing progeny.

Transgenes appear to mediate silencing because they are usually designed to be constitutively expressed – i.e. “always on” – in every tissue. But endogenous genes are usually “on” in some tissues, and “off” in others.

Only a minority are likely to be constitutively expressed across all tissues. The model predicts that endogenous genes not producing nuclear transcripts – probably the majority – won’t be amenable to systemic silencing.

---PB---

So, naturalists observe, a flea Hath smaller fleas that on him prey; And these have smaller still to bit 'em; And so proceed ad infinitum. - Jonathan Swift

It seems that Jonathan Swift’s classic 1733 poem on parasitism might be true even for viruses, supposedly the ultimate ‘parasites.’

Many plant viruses, and at least one human virus – the human hepatitis B virus – carry their own genetic parasites: small, viroid-like RNA molecules called satellite RNAs, ranging in length from 220 to 1700 nucleotides. The satellite RNA is packaged with the virus’s own genome into the capsids of newly minted virions. Without its virus host, the satellite RNA cannot replicate.

In May last year, Ming-Bo Wang and CSIRO Plant Industry colleagues Neil Smith and Andrew Eamens published a paper in the open-access journal Public Library of Science Pathogens, describing the results of their study of the Y-satellite RNA of cucumber mosaic virus. CMV–assisted Y-satellite infection causes leaf yellowing in several species of Nicotiana, including commercial tobacco plants (N. tabacum).

Most satellite RNAs do not code for proteins, so the mechanism by which they induce disease symptoms has been a long-standing question. The answer, it seems, is by silencing plant genes via RNA interference.

Previous studies had shown that the virulence of satellite RNAs is determined at the nucleotide level: changing one or two nucleotides drastically alters virulence, and the host-specificity of the RNA. A subsequent study showed that satellite RNA replication results in the accumulation in cells of high levels of siRNAs derived from the satellite sequence.

From these early studies Wang and colleagues proposed a siRNA-mediated viral disease model in 2004 and provided some preliminary evidence in support of the model. SiRNAs direct RNA silencing in plants through sequence specific cleavage of messenger RNAs, or by inducing methylation of target genes. Do satellite siRNAs work the same way?

The CMV Y-satellite RNA is a single-stranded, 369nt molecule. Neil Smith searched the database of Nicotiana genes for sequences complementary to the segments of the Y-satellite’s sequence, and found a 24nt sequence – the typical length of a siRNA – that mapped to the CHLI gene in N. tabacum, which is involved in chlorophyll synthesis.

Without green chlorophyll to capture light, the plant’s tissues turn yellow, and its vigour is impaired. Smith and Wang showed that Y-satellite RNA disease symptoms could be prevented by expressing a natural or synthetic variant of the CHLI gene with a slightly different DNA sequence that confers resistance to silencing.

They believe silencing-resistant transgenes, designed to be resistant to the siRNAs contained in satellite RNAs, could provide an alternate strategy for preventing plant virus diseases.