from The Economist
A newly recognised class of genes really does matter
GENETICS often progresses by breaking things. Early experiments used naturally broken genes—mutations—to work out the basic rules. Then geneticists found out how to induce mutations with radiation and chemicals. That gave them more material to work with, but the mutations still appeared at random. Then the “knockout mouse” was invented, in which mice have specified genes rendered inoperable, thus revealing their functions. Now, as they write in eLife, John Rinn of Harvard University and his colleagues have used the idea of knockouts to prove that certain bits of DNA once regarded as junk are proper genes after all.
The DNA in question is propagated by being copied—as are all genes—into a similar molecule called RNA. In the case of conventional genes this RNA is then “translated”, in a subcellular machine called a ribosome, into a protein. Dr Rinn’s RNA—a type known as long intergenic non-coding RNA, or lincRNA—is not so translated. Geneticists used to assume that this meant it was useless. Now they know different: it actually goes off to do jobs uniquely suited to the chemistry of RNA itself. How important those jobs are, though, has been much debated. Dr Rinn has just demonstrated, by knocking out 18 lincRNAs, one at a time, and observing what happened to the mice in question, that they can be very important indeed.
In three cases the answer to the question “what happens when you knock out this gene?” was: no mouse. Knockouts work by mating males and females that each have one functional and one non-functional version of the gene under investigation (each having been derived from a different parent). A quarter of their offspring will thus have two non-functional copies, and whatever job the gene does will therefore not get done. If this job is vital, the embryo will die in utero and no mouse pup will be born. That was what usually happened in the case of genes called linc-Foxf1a, linc-Hoxd3 and linc-Sox2.
In two other cases, the embryos did come to term, but they were a lot smaller than they should have been. The guilty genes here are called linc-Pint and linc-Brn1b. LincRNA, then, really does matter.
Dr Rinn and his colleagues followed up one of their newly discovered examples in detail. Linc-Brn1b, they found, is involved in brain formation. Without it, the layers of the cortex do not develop properly. These knockouts therefore, in Dr Rinn’s view, do indeed knock out once and for all the idea that lincRNAs are not proper genes.