The evolutionary origin of RNA interference is obscure, but I favour (if only for explanatory reasons) one theory in particular, namely that it is an ancient defence against viruses that has been adapted to meet other additional demands. Let me explain.
Although all 'higher' lifeforms exclusively use DNA to store their genomes, viruses can use either DNA or RNA (or sometimes both!). In fact, most use RNA, in either a single- or a double-stranded form. The hardest part of fighting viruses is to find them - they are fundamentally just nucleic acid surrounded by a protein coat. The protein coat is usually their weakness, as it is something decidedly foreign that the immune system can recognise and attack. However, if a virus should inject its nucleic acid into a cell, the cell has almost no way to recognise the RNA or DNA as foreign - why would it? RNA is RNA!
Usually what has to happen is that a selection of the virus' (foreign) proteins are displayed by the routine sampling process common to almost all cells - the protein fragments are attached to the MHC class I molecule, which strongly interacts with surveillance cells of the immune system. If a "non-self" molecule is detected, the immune cell will politely ask its infected counterpart to self-destruct (apoptosis) or, failing that, simply destroy the cell itself. (Viruses which disable the MHC I sampling process are dealt with by natural killer cells, as explained here.)
This method usually works well enough (or we'd all be dead), but it is frustratingly indirect - a cell has to be long on the road to helping the virus replicate before it can be stopped. Is there any hope at all of identifying viral RNA before it does its damage? Well, although RNA is RNA is RNA (etc.) there is one form of viral RNA that can be reliably distinguished from our own native RNA - double-stranded RNA, which is never normally produced by our cells.
Just as you'd expect, our cells can not only recognise but also actively destroy any hint of double-stranded RNA (ds-RNA). A protein with the wonderful name of Dicer lives up to its name by slicing any identified ds-RNA into useless fragments, thereby preventing the RNA from being translated into a protein. Each of these RNA fragments is very small, by RNA standards, and they are collectively referred to as small interfering RNAs (siRNA).
Furthermore, the process is vindictive. Once the ds-RNA has been cleaved, the bits bind a protein complex, called RISC, which stands for RNA-induced silencing complex. One of the two RNA strands is eliminated, but the other, still bound to RISC, acts as bait for other copies of the RNA molecule, attaching to them by the familiar base pairing rules common to all life (A to T, C to G). If this occurs, RISC cuts the newly-bound mRNA strand, but retains the other fragment in order to sleuth out more of these foreign RNA molecules.
In this way, a cell may be purged of viral RNA. But this is only the tip of the proverbial iceberg when it comes to the power of microRNA, and it wasn't at all how miRNA was discovered. We'll cover that in the next post.