Extensive analyses

“In phylogeny, it is important to compare conserved sequences”, says Professor Motte. And yet this is not the case of the domain rich in serine/arginine dipeptides, known as SR domain. But besides this domain, SR proteins have an RRM domain (which stands for RNA recognition motif), which is a RNA binding domain. Contrary to the SR domain, the former has the necessary “phylogenetic qualities”. It is therefore the RRM domain that the researchers chose in order to study the evolutionary history of the SR proteins. “It is a small domain of 80-90 amino acids, but it is the only really useful one from a phylogenetic point of view”, continues Patrick Motte.

Consequently, the researchers’ (Denis Baurain, Sophie Califice, Marc Hanikenne and Patrick Motte) mission consisted, no less, of identifying and studying all the RRM proteins within the proteomes of more than 200 organisms! Both in prokaryotes (bacteria and archaea) and eukaryotes (plants, animals, fungi and protists)... “We therefore found ourselves with more than 12 000 different sequences”, Patrick Motte specifies. “This was very difficult to manage but essential to retrace the evolution of these proteins. The majority of the other studies on the subject only related to several hundred sequences at most”, continues the scientist.  

In order to be able to analyse this vast amount of data, the researchers had to develop complicated techniques. Among other things, Professor Denis Baurain elaborated new algorithms in order to complement the computer programs already available for such analyses. “The analyses required months of calculations on a super computer and a vast amount of interpretation work”, explain the two scientists. “We had two approaches: automatic on the one hand, with the use of computers for data preprocessing, and manual on the other hand, with a meticulous examination of the results obtained”, continues Denis Baurain. Hence, the gigantic phylogenetic trees created by the super computer were then scrutinised by the researchers looking for SR proteins. Besides this double approach, the researchers also used various phylogenetic models to support their results. It was indeed a challenge to work on such a short sequence as the RRM domain. “As a result, we increased the approaches because they each came up with rather uncertain statistical results. But since all of them converged towards the same tendency, this reinforced the validity of our conclusions”, Denis Baurain explains. 

SR proteins, an old innovation!

The results obtained measured up to the time and work the researchers put into this study, which began in 2007. Indeed, it revealed several important conclusions that were published in the Plant Physiology (1) journal. The analyses revealed that SR proteins are present in a great many eukaryotic organisms. Better still, the subfamilies of these proteins are found within different groups of eukaryotes. “This means that these subfamilies were already present in an ancestral eukaryotic form. Because when we find the same family or subfamily of proteins in organisms as distant as plants and animals, either this is a case of convergence or they have a common origin, as confirmed by our work on SR proteins”, Patrick Motte emphasises. Thus, an ancestral sequence has evolved and led to different families and subfamilies of SR proteins that were already present in the last eukaryotic common ancestor (LECA). Besides the interest of this discovery in helping to understand the history of SR proteins, this also means that LECA already possessed highly complex RNA maturation machinery and was far from the simple organism that we might have imagined. “In general, we often think that if an organism is simple, it must be ancient (and vice versa) but this isn’t always the case”, Patrick Motte tells us. “Evolution doesn’t always progress towards complexity”, Denis Baurain continues. “Take baker’s yeast for instance. This organism has no SR proteins. Yeast has evolved by simplifying its RNA maturation machinery compared with that of LECA and that of its current cousins”, Denis Baurain specifies.

(1) Sophie Califice, Denis Baurain, Marc Hanikenne, and Patrick Motte. A Single Ancient Origin for Prototypical Serine/Arginine-Rich Splicing Factors. Plant Physiol. 2012 February; 158(2): 546–560.