long non-coding RNA
/ long non-KO-ding AR-en-ay /
We are used to the idea that an RNA is a temporary copy of a gene on its way to becoming a protein. But the cell makes a whole population of long RNAs that are never translated — they carry no protein recipe at all, yet many of them do real jobs. These are the long non-coding RNAs, and they are one of the most surprising parts of the modern genome.
By definition a long non-coding RNA is an RNA longer than about 200 nucleotides that does not code for a protein (the cutoff at 200 just separates them from the small regulatory RNAs like microRNAs). They are made by the same RNA polymerase that makes mRNA, and many are capped, spliced, and tailed just like mRNA — they simply lack a meaningful open reading frame. What they do varies enormously. Some act as scaffolds, holding several proteins together in one place. Some act as decoys, soaking up a microRNA or a protein so it cannot act elsewhere. Some guide chromatin-modifying machines to specific genes. And some work by the very act of being transcribed, the passage of the polymerase itself changing the local DNA.
The most famous example shows how powerful a long non-coding RNA can be: in female mammals, one of the two X chromosomes is shut down, and the trigger is a long non-coding RNA called Xist that physically coats the chromosome it is made from and recruits silencing machinery. Long non-coding RNAs are a major reason the field now says the genome is pervasively transcribed and that much non-coding RNA is functional — though honesty requires noting that for many lncRNAs a clear function has not yet been proven.
Xist is a long non-coding RNA, thousands of bases long, that is spread along one X chromosome in a female cell. It never makes a protein; instead the RNA itself coats the chromosome and recruits machinery that silences nearly all of its genes — the basis of X-inactivation.
A long RNA that codes for no protein can still silence an entire chromosome.
Be cautious: the genome makes huge numbers of long non-coding RNAs, but having a transcript is not proof of a function. Some are biologically important like Xist; others may be incidental transcriptional noise. Demonstrating that a given lncRNA actually does something is hard, ongoing work.