More uses for "junk" DNA: developmental

Sex and junk. Two subjects of endless fascination for geneticists. Today: Junk.

Because there's been a new discovery that turns the notion of junk upside down.


Junk DNA seems to make up a high proportion of a typical genome: that is, we don't know what huge strings of coding information are for, but we don't think they do anything - simply because we haven't yet found functionality for them.

Wikipedia lists a few hypotheses for "junk DNA":
- a "reservoir of sequences" from which new functionality can emerge - in effect making it an important evolutionary nexus;
- useful space to more easily enable formation of necessary products such as enzymes (one of the least likely options, to my mind);
- undiscovered regulatory functions to control the expression of genes, particularly at developmental stages such as embryos;
- some junk may have regulatory layers for meaningful genetic programming (I'm not fully on top of this idea; I would have lumped it in with the one above);
- extrapolation from some genomes suggests eukaryotes (organisms with cellular nuclei) may "require a minimum of non-coding DNA" - but that minimum is said to be about 5%, whereas the human genome contains 95% junk.

Further possibilities include:
- evolutionary redundancy - for example, accidents in coding that are not deleterious and so remain within the DNA;
- viral insertions into germ-line cells, which are subsequently propagated into progeny.

That last is quite significant, as it has in recent times been demonstrated that a significant proportion of many genomes are viral in origin. This in itself is a significant vehicle for evolution, as viral DNA has often contributed useful functionality to the host.

But 95% junk DNA in the human genome? Intuition suggests much of that is of unknown functionality rather than junk per se.


Now Princeton University (datelined 20 May) reports an answer for at least some of the "junk". And it's close to the developmental concept above.


In Oxytricha, a single-celled pond-dwelling organism, what was considered junk has been found to trigger complex regroupings of DNA strands, during the development phase of the organism. Once the functionality has been completed, the gene's products are disposed of, and the genome trimmed down again.

It sounds like the process is so ephemeral that it is easily missed, especially since the evidence is subsequently destroyed. But the genes once labelled junk actually control the assemblage of other genes.

It sounds analogous to an assembly hierarchy: instructions beget miniature machines, which produce outputs and are then disassembled again.

To complicate matters, the researcher says "the instruction set comes in the form of RNA that is passed briefly from parent to offspring and these maternal RNAs provide templates for the rearrangement process". This might be analogous to a mother feeding an embryo necessary nutrients to enable growth, but this is a step beyond - it's at the DNA level.

One final thought. Much junk DNA consists of apparent stutters. That is, coding sequences that repeat ad nauseum, reverse, and otherwise follow semi-meaningless patterns. Some of that is viral in origin - which is not to say that it is or isn't currently useful. I expect that we will reclassify much junk as meaningful. As for the rest, will there be any DNA that is completely useless? On a macro level, unused functionality tends to atrophy and disappear over evolutionary time - unless it finds new functionality first, which happens surprisingly often. It's hard to say that the same principle of atrophy applies at a DNA level (the question is, how much cost there is to carrying an incremental amount of freeloading DNA) - but it certainly makes sense that mutation can result in new-found functionality at the DNA level. Mutations, after all, are akin to random scientific experiments: if something beneficial arises, it stands a reasonable chance of surviving.