Transposable elements, nicknamed ‘jumping genes’, refer to stretches of DNA that are able to move from one part of the genome to another. Retrotransposons belong to this family of mobile genetic elements but also require the transcription of RNA to DNA in order to move or ‘transpose’. LINEs (long interspersed elements) are one of several subtypes of retrotransposons. In addition to transposing themselves, they are able to mobilize sections of transcribed DNA that lack an associated LINE sequence. This results in retrocopied genes, which have proven to be important for the evolution of new genes. Adam Ewing and colleagues in the Haussler lab explore the diversity of these retrocopied genes, uncovering new insights into their frequency in the human genome and their occurrence in cancer. Their recent publication in Genome Biology explores gene retrocopy insertion polymorphisms (GRIPs).
What does your article set out to investigate?
When most genes are expressed, the transcripts are processed (spliced, capped, and polyadenylated), exported to the cytoplasm, and translated at a ribosome. One alternative scenario is where the processed transcripts are instead picked up by the ribonuclear particles that mediate retrotransposition, and inserted back into the genome as a reverse-transcribed copy at some distal location. This was demonstrated to happen via the same mechanism LINE-1 retrotransposons use to copy themselves about 13 years ago by Thierry Heidmann’s group (Nat Genet. 2000 24:363-367). These are often referred to as processed pseudogenes, but since the term ‘pseudogene’ has a functional implication (i.e. lack of function) we stick to calling them ‘gene retrocopies’.
In the study published by Genome Biology we set out to explore the inter-individual diversity in gene retrocopy content by identifying gene retrocopies present in some individuals and not in others, termed ‘gene retrocopy insertion polymorphisms’ or ‘GRIPs’, focusing on those not in the reference genome.
What was your motivation for studying GRIPs?
I’ve been interested in transposable element (TE) insertion polymorphisms for several years. This is a natural extension of prior work on TE polymorphisms and seemed to be an interesting and largely unexplored area that was now tractable using paired-end sequence data.
What do you consider to be the most interesting findings of your study?
I think there’s two main interesting points in here. The first is the finding that there are somatic gene retrocopy insertions in some cancers. Using data from the Cancer Genome Atlas allowed a comparison of whole genome sequence data between patient-matched tumor and normal tissue. Given that gene copy number amplifications of oncogenes are a driving force behind some cancers, novel gene retrocopy insertions could provide a means to increase copy number and expression if the insertions occur in a highly transcribed region.
The second major finding is the estimation of a rate of novel processed pseudogene formation in the human genome. Given that our sensitivity is likely not perfect (based on simulations presented in the paper), this rate of roughly one new insertion for every 6,000 individuals should be taken as a lower bound and could vary slightly with genetic background.
Do you think that, until now, GRIPs have been an underappreciated form of genetic variation?
There have been a number of prior studies identifying what we call GRIPs in flies, mice, and humans but in the large majority of these cases the actual insertion sites were not identified, precluding some of the population-based results in our paper. I should also highlight a very recent paper in PLOS Genetics by Daniel Schrider and colleagues that came out while our paper was in review (PLoS Genetics 2013 9:e1003242). Taken together I think this indicates a recent increase in appreciation for this sort of phenomenon and I think our paper adds significantly to what is known about the diversity between individuals due to GRIPs.
Do you have plans to continue the GRIPs project?
As part of a larger ongoing effort to characterise structural variants in cancers, yes. Our paper clearly leaves many questions about the functional impact of GRIPs wide open for other groups to pick up on, and I’d be interested to see what happens.
You have made your GRIP mapping tool, GRIPper, available for public use. Are there any datasets that you would like to see it applied to?
Obviously it’s an application-specific tool. I think its application to population sequencing of other species could be interesting, and continued application to paired tumor and normal whole genome sequence data may yield further examples of gene retrocopy insertions in cancer.
Questions from Naomi Attar (@naomiattar), Senior Editor for Genome Biology.