Short strands of piwi-interacting RNA may detect foreign invaders by determining whether a gene has ever been turned on in an organism's past.
As scientists have added to a growing list of types of RNA molecules with roles that go beyond conveying the genetic code, they have found the short strands known as Piwi-interacting RNAs (piRNAs) particularly perplexing. New work from Howard Hughes Medical Institute (HHMI) scientists suggests those abundant molecules may be part of the cell's search engine, capable of querying the entire history of a cell's genetic past.
Organisms contain thousands of piRNA molecules, strands of 26 to 31 nucleotides encoded all over the genome. In two studies published online June 25, 2012, in the journal Cell, HHMI investigator Craig Mello of the University of Massachusetts Medical School has discovered that piRNAs may be responsible for detecting foreign RNA—such as that carried by viruses—relying on a complex search mechanism to reveal whether an invader is foreign based on prior gene activity.
This is really remarkable. It implies that an organism has a memory of all the previous gene sequences it’s ever expressed before.
Craig C. Mello
"piRNAs are found in all animals and some of their functions in some organisms have been explained," says Mello. "But overall they've been a very mysterious category of molecule."
Some piRNAs have sequences that match up identically to genes elsewhere in the genome, suggesting that they bind and regulate those genes. But most have no perfect genetic match. Mello and his team focused their attention on the more puzzling piRNAs—those that had no obvious targets.
In the worm Caenorhabditis elegans, the scientists unexpectedly found that foreign genes that they inserted into the genome were sometimes silenced and sometimes not. When they genetically modified the worm to lack the Piwi protein, the silencing no longer worked.
When the researchers probed which sequences piRNAs tended to shut down, they found that if a cell has ever turned on a gene in the past, the piRNA system will recognize it as a "self" gene and allow it to be expressed. But if it hasn't been active in the organism before, the piRNA will set the silencing mechanism into action so it remains off.
The silencing or lack of silencing is permanent, they found. If the piRNA doesn't silence a gene the first time it encounters it, it won't ever silence it. And if it silences it once, then every time that gene appears in the future, the system will turn it off.
"This is really remarkable," says Mello. "It implies that an organism has a memory of all the previous gene sequences it's ever expressed before."
The researchers think that the snippets of piRNA do not hold the memory in their sequences. Rather, two other small RNA pathways are thought to provide epigenetic memories of "self" and "non-self" RNA. Mello says piRNAs likely allow mismatched pairing as they scan, so that virtually they can potentially recognize all sequences that have been expressed. Silencing occurs only when a sequence has not been seen before.
While people have hypothesized that foreign RNA is recognized by cells as foreign based on a particular feature of the molecule—like a structural element or chemical tag—the new results suggest that the recognition may be sequence based.
That's not all Mello's lab discovered about piRNAs. Not only did the gene silencing pattern that they establish persist throughout an organism's life, but the memory was passed very stably between generations.
"These small RNAs are present in the germline at all stages and are transmitted to both the egg and the sperm," says Mello.
When genetically identical animals exhibit opposite and heritable phenotypes, the mechanism of inheritance is dubbed epigenetic. In this case, the inducers of epigenetic silencing are piRNAs, so with that in mind, Mello coined the term RNAe, for RNA-induced epigenetic silencing.
A last highlight of the team's findings was that although an organism retains its pattern of piRNA silencing throughout its lifetime, each individual establishes its own pattern. While one may silence a gene everytime it encounters it, another may allow its expression.
"It's interesting that the animal is a little bit lenient," says Mello. "Maybe we don't necessarily want to shut off everything that we haven't seen before. Maybe there is some adaptive value to this type of variation."
More questions than answers remain about RNAe: What are all its targets? How exactly is the memory of past gene expression stored? And does the system react to changes in the environment, allowing more or fewer silenced genes to express in times of stress?
"So far, these small RNA systems are turning out to be really remarkable," says Mello. "But there's lots more to nail down."