Box 1. How does it work?
From the following article:
Trisha Gura
Nature 404, 804-808(20 April 2000)
doi:10.1038/35009245
Although the mechanisms of gene silencing are far from completely understood, the working hypothesis goes like this: the initial trigger is the presence in the host's cells of an aberrant RNA. This could be a double-stranded RNA, a shortened RNA that lacks its 'cap' or 'tail', or a conventional RNA that is present in unusually large quantities — all of which can indicate that a virus is on the attack.
The host organism's response is to call on enzymes that slice and dice the offending RNA into pieces around 25 nucleotides long. At some stage — either before or after the formation of these fragments — the rogue RNA is copied many times over, to amplify the alarm signal. The fragments then spread throughout the host. Antisense strands, complementary to the target mRNA, bind to the target and prompt other enzymes to disable it.
Several recent discoveries have contributed to this hypothesis. The sequence of the qde-1 gene, for instance, involved in gene silencing in fungi, suggests that it encodes an enzyme belonging to a family called RNA-dependent RNA polymerases24. This enzyme may amplify the rogue RNA.
The fact that the triggering RNA is cut into small fragments was revealed by work in plants and Drosophila cell cultures. Last year, David Baulcombe's group at the John Innes Centre in Norwich found that transgenic tobacco plants undergoing co-suppression induced by potato X virus make 25-nucleotide-long fragments that correspond to the target of gene silencing25.
Meanwhile, Greg Hannon of the Cold Spring Harbor Laboratory on Long Island, New York, has been inducing RNAi in Drosophila cells, pulverizing them, and then screening different biochemical extracts for gene-silencing activity. The active fractions again contained discrete RNAs, around 25 nucleotides long, corresponding to sequences from the silenced gene26.
Most recently, Phillip Zamore of the University of Massachusetts Medical School in Worcester and his colleagues, working with a similar in vitro system, have shown that these RNA fragments are between 21 and 23 nucleotides long. These seem to spell doom for the target mRNA by directing enzymes to cut it into pieces27.