There’s a lot that can happen before the information inside a gene gets translated into a protein. For example, RNA transcripts that are derived from sequences that are located far away from each other on the genome can be fused in a process called trans-splicing. Trans-splicing has been observed in several eukaryotic species, including Hydra’s and Nematodes. During trans-splicing a common spliced leader (SL) sequence is fused to mRNAs. Trans-splicing has several functions, ranging from providing 5′ caps and start codons to resolving multicistronic transcripts into monocistronic mRNAs. In a recent paper published in Molecular Biology and Evolution, Douris and Telford report that trans-splicing also occurs in amphipods, a shrimp-like order of the arthropods (the exoskeletal branch of life that includes insects, crustaceans and the like)¹.
Hyalella Azteca, a close relative of the amphipod that was used in this study (wich is Parhyale Hawaiensis). Picture couresy of Wikipedia.
This observation prompted them to delve further in the evolutionary history of trans-splicing. By looking for commonly spliced leader sequences in transcripts from EST databases, they predicted whether trans-splicing occurred in a diverse set of species. They discovered that the closest relatives of amphipods didn’t have trans-splicing, but that it did occur in distantly related species like sponges and comb jellies! Phylogeneticists refer to such a fragmented distribution as “patchy”. Patchiness can be explained by two different scenarios:
- Trans-splicing evolved repeatedly in different species
- Trans-splicing was present in the ancestor of metazoa. It was lost in some descendants and retained in others..
Douris and Telford reconstructed these evolutionary scenario’s. The found that if scenario 2 would be true, trans-splicing would need to be independently lost in 17-20 lineages. In contrast, only 10 independent instances of gains and losses are required if we assume that the common ancestor of metazoa did not trans-splice its transcripts. Therefore, the most parsimonious reconstruction, points at scenario 1.
Two different scenario's for the origin of trans-splicing in some metazoans. Black lineages denote a gain of trans-splicing, white lineages a loss. In hatched lineages, loss and gain is equally parsimonious. Picture adapted from paper.
How could it be that trans-splicing evolve repeatedly in different lineages? A closer inspection of the mechanism of trans-splicing clarifies this issue. The molecular machinery involved in trans-splicing largely overlaps with that of conventional splicing: the only difference is that the U1 snRNP does not participate in trans-splicing, its role has been taken over by an snRNP derived from the SL precursor RNA. Experiments have shown that it is possible to convert the U1 snRNA into a SL precursor, by adding a splice donor site and a few mutations². What this implies is that it is relatively easy to evolve trans-splicing: small alterations to existing snRNAs are enough to evolve RNAs which are capable of trans-splicing. Furthermore, researchers have introduced SL precursor RNAs into cells derived from organisms that normally don’t trans-splice, and showed that trans-splicing was induced³! This means that when trans-splicing evolves, it quickly becomes functional, since a large part of the required machinery is already in place.
Taken together, these observations make the repeated evolution of trans-splicing a likely scenario.
However, as the authors rightly note, absence of evidence is no evidence of absence: the analyzed EST datasets might simply be too small to detect trans-splicing in certain species, whereas other species have not been sampled at all. Our perception of the evolution of trans-splicing might change, when more data becomes available.
1. Douris V, Telford MJ, & Averof M (2009). Evidence for multiple independent origins of trans-splicing in Metazoa. Molecular biology and evolution PMID: 19942614
2. Hannon GJ, Maroney PA, Yu YT, Hannon GE, & Nilsen TW (1992). Interaction of U6 snRNA with a sequence required for function of the nematode SL RNA in trans-splicing. Science (New York, N.Y.), 258 (5089), 1775-80 PMID: 1465612
3. Bruzik JP, & Maniatis T (1992). Spliced leader RNAs from lower eukaryotes are trans-spliced in mammalian cells. Nature, 360 (6405), 692-5 PMID: 1465136
You might also like:
I’m never particularly convinced of the ‘genes-lost’ arguement (unless it’s particularly obvious), possibly because in bacteria at least there’s a tendancy for missing genes to leave some kind of scar. I’m a little less certain with eukaryotes, but couldn’t they just coil up unwanted DNA and keep it safe with the huge numbers of other untranscribed genes?
On a slightly unrelated note, that is a beautiful picture.
If gene sequences become dysfunctional in eukaryotic species, chances are big that they’ll be retained for a time as pseudogenes . So you’re right that missing genes tend to leave these ‘scars’ (I prefer to think of them as some kind of gene fossils!). Eukaryotic genomes tend to be even more tolerant for ‘unfunctional’ sequences than prokaryotic ones! I’m curious as to how these prokaryotic scars look like? Can you deduce anything from the remaining, surrounding sequences (because the turn-over is so high!)?
Thans for the compliment, I actually spend about half the time of writing a blog-post looking for nice pictures to go with them ;).