This paper is an excellent demonstration of the power of phylogenomics for the discovery of genes involved in traits of interest.
The authors report a larger scale genome comparison between symbiotic (arbuscular mycorrhiza forming) and non-symbiotic plant groups. They identify gene loss in plant species that go back to a minimum of four independent loss-of-symbiosis events; one in the Brassicales, one in the Caryophyllales (Amaranthaceae), one in the Laminales (Orobanchaceae) and one in the Fabales (Lupinus).
They performed an impressive phylogenomic analysis and identified a list of 300 Medicago genes that are present in most of the analyzed species but absent in all non-symbiotic Brassicaceae. Upon filtering the list further, by including paraphyletic non-symbiotic species, they arrived at a list of around 100 genes that were consistently absent in the non-mycorrhizal species. Lupinus as a plant that lost arbuscular mycorrhiza but maintained root nodule symbiosis was very informative because common symbiosis genes should be maintained in this genus.
The results are consistent with an evolutionary scenario in which each of the independent loss-of-symbiosis events, for which the loss of a single gene function was sufficient, was followed by a subsequent larger scale gene erosion that consistently removed the same orthologous genes in the four different clades.
This very interesting and largely unexpected observation reveals two opposing evolutionary forces that decide over the prevalence of this 'symbiosis-associated' gene set. On the one hand, the existing symbiosis leads to a successful maintenance of symbiosis genes. On the other hand, a yet unidentified force resulted in a consistent pattern of larger scale gene loss after each independent loss-of-symbiosis event. The forces behind this erosion must have acted either very quickly, before each of the non-symbiotic clades diversified from their respective common ancestor, or they independently led to consistent gene loss patterns after speciation.
Because symbiosis-related genes are overrepresented in the eroded gene set, it is likely that a large proportion, if not all of them, are of specific functional relevance in arbuscular mycorrhization (AM). Therefore this study is of major importance not only from an evolutionary perspective, but also because it demonstrates a novel strategy to identify candidate genes involved in AM symbiosis.
By Martin Parniske, F1000 Plant Biology, Biocenter University of Munich (LMU), Martinsried, Germany.