livescience.com — Scientists have discovered mystery snippets of mammal DNA that have survived eons of evolution and yet have no apparent purpose. The finding reveals just how much we don't know about the secrets hidden in our genome and that of other animals.
Oct 9, 2008 View in Crawl 4
fordiOct 10, 2008
Feh. That's bollocks. Everyone knows, if a university had written the genetic code, it'd have been MIT.
gordonjOct 10, 2008
It is very doubtful that these regions serve no purpose, as the default result of that is degradation of the region by mutations. It seems unlikely to me that these regions are protein coding either. Genes require certain architectural features, and I think it would be quite obvious that these regions were genes based on that, unless they represent a totally new class of genes that don't operate in the same manner as all other genes we know about. As you mention, there are other transcribed regions that don't code for protein. My guess would be that this is more likely. Non-coding RNAs usually have very specific functions which are totally reliant on their structures which require pairing of different regions of the RNA, and therefore are well conserved by these structural limitations. Certainly, new classes of RNA that play important roles in genetic control organisms are being discovered all the time, and this is what I would guess these regions are. Of course in biology there is also always the chance that it is something completely different to anything we have encountered before, however, I think it is safe to say that these regions are functional.
jitmasterOct 10, 2008
* Copyright(c) Tyrell Corporation* All rights reserved.
protodonOct 10, 2008
or I'm in a molecular genetics class right now...People just seem to think that everything must have a purpose. The word vestigial comes to mind. There's a lot of extra s**t in our DNA that we don't use anymore. The genes are turned off but from an evolutionary standpoint they don't affect anything whether they are there or not so they remain.
gordonjOct 10, 2008
@protodonThe whole point of the paper is using comparisons between the genomes of mammals which have varying levels of evolutionary distance to find regions that are WAY more similar than they should be if they were just non-functional DNA. If you're in a molecular genetics class then you probably know that coding regions (specifically exons) are expected to evolve at a slower rate than non-functional DNA due to purifying selection (because mutations that change proteins are usually detrimental) and are weeded out of a population by selection. Introns are much less conserved between species (if at all). There is a lot of "junk" and probably non-functional DNA in mammalian genomes, but that junk isn't conserved between the sequences at all compared to genes, and the non-coding regions studied here. I recommend you read the paper. The regions they are describing are most likely upstream regulatory elements. Most of them are located close to genes, and their analysis also pulled out nearly every characterised mammalian regulatory cis-element. The regions they knocked out in mice were also well characterised cis-elements regulating genes with a known phenotype if their expression was knocked out or majorly disrupted."The genes are turned off but from an evolutionary standpoint they don't affect anything whether they are there or not so they remain."If a gene is completely switched off, and doesn't affect the organism at all, then the selection pressure maintaining that gene will be relaxed and the DNA will accumulate mutations at the same rate as non-functional DNA. If genes like that really existed, then all it would take is one stop codon mutation and the fate of the gene would be sealed as a pseudogene which would then just degrade through random mutation.Here's the paper:<a class="user" href="http://genome.cshlp.org/cgi/reprint/gr.080184.108v1">http://genome.cshlp.org/cgi/reprint/gr.080184.108v ...</a>
gordonjOct 11, 2008
In the paper the regions they knocked out were all well characterised cis-regulatory elements for genes with known knockout phenotypes, but there was no noticeable effect. They mention several possible reasons for this like redundancy in the transcriptional regulatory network or that the phenotype may be environment specific. I have no doubt that these regions are functional based on the sequence conservation and the fact that this analysis also pulled out almost every characterised mammalian cis-regulatory element (which only made up a small part of the dataset).