Scientists Catch Jumping Genes Rewiring Genomes | Hacker News Reader

Scientists Catch Jumping Genes Rewiring Genomes

quantamagazine.org

73 pointsnsoonhui8 comments

>For more than a decade, Feschotte has pointed to transposons as the ultimate innovators in eukaryotic genomes.

To be clear, there is a long history of scientific theories and studies on how transposable elements are key to eukaryotic genome evolution, regulatory network formation and genome plasticity, predating the review linked in that sentence - in fact, dating all the way back to Barbara McClintock. I would call this an under-studied area of genomics - I think there are still quite a few Nobel prizes to be made in transposon-mediated evolution.

The new Feschotte article (https://science.sciencemag.org/content/371/6531/eabc6405) demonstrates a transcription factor in bats and uses a reporter assay to link it to signatures of transposon activity, and also uses comparative genomics to identify lineage-specific signatures of transposon activity in tetrapods. This is a big step forward, especially in using reporter assays to link TE activity to TF networks in mammals, but I think the next step would be to do both the in silico and the in vitro experiment together (use a reporter assay to prove that lineage-specific regulatory network patterns are linked to transposon activity).

also transposons certainly exist in prokarya too, and probably drive innovation there as well, possibly assisting horizontal gene transfer. When I was in the synthetic genome lab at the venter institute, one of the first things I did was win a bet about the orientation of a transposable element in the syn1.0 genome that was eventually removed from the genome (the bet was not important enough to merit me being on the paper)

Why was the TE removed from the syn1.0 genome?

Causes instability when you transfer into yeast as your amplification medium; deleting the TE boosts the correct colony ratio from 50% to nearly 100%. I also identified the gene responsible for the phenomenon: it's RAD42. So if you're ever cloning new genomes that don't have TEs removed yet you might consider using RAD42- yeast.

> Pax6 is only one of thousands of genes encoding transcription factors

Last time I checked humans only had ~1800 TFs. I don’t know about “thousands”. Maybe if we count all TFs in all organisms excluding homologs...

> While geneticists have made leaps in understanding how genes with relatively simple, direct functions could have evolved, explanations for transcription factors have largely eluded scientists.

Is it that much different though? TFs bind short DNA sequences throughout the genome. These sequences don’t have to be exact, some variation is allowed. Now, random mutations can change the DNA and enable TF binding, thus increasing the fitness(since TF increasing or decreasing expression of a gene can give some advantage).

A transcription factor gene regulatory network is much more likely to arise through gene duplication and upstream transposon insertions than by random mutation - especially in plants and animals, which have long generation times, low population sizes, and proofreading processes to suppress point mutations, while also facilitating gene duplication via sexual recombination. Eukaryotes have also evolved chromatin accessibility restrictions to control not just expression but evolvability of DNA regions. Point mutations on their own don't seem like a plausible mechanism for the level of plasticity exhibited in plant and animal genomes.

A bit of an aside, but I'm wondering if anyone is applying some of these ideas to genetic algorithms?

Leonid Zamdborg et al published a good paper on this a few years ago and it has a very good overview of related previous work. All the referenced papers in the background and related works section are all worth reading:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4353400/

Reply on news.ycombinator.com