Phylogenomics reveals surprising sets of essential and dispensable clades of MIKC(c)-group MADS-box genes in flowering plants.

MIKC(C)-group MADS-box genes are involved in the control of many developmental processes in flowering plants. All of these genes are members of one of 17 clades that had already been established in the most recent common ancestor (MRCA) of extant angiosperms. These clades trace back to 11 seed plant-specific superclades that were present in the MRCA of extant seed plants. Due to their important role in plant development and evolution, the origin of the clades of MIKC(C)-group genes has been studied in great detail. In contrast, whether any of these ancestral clades has ever been lost completely in any species has not been investigated so far. Here, we determined the presence of these clades by BLAST, PSI-BLAST, and Hidden Markov Model searches and by phylogenetic methods in the whole genomes of 27 flowering plants. Our data suggest that there are only three superclades of which all members have been lost in at least one of the investigated flowering plant species, and only few additional losses of angiosperm-specific MIKC(C)-group gene clades could be identified. Remarkably, for one seed plant superclade (TM8-like genes) and one angiosperm clade (FLC-like genes), multiple losses were identified, suggesting that the function of these genes is dispensable or that gene loss might have even been adaptive. The clades of MIKC(C)-group genes that have never been wiped out in any of the investigated species comprises, in addition to the expected floral organ identity genes, also TM3-like (SOC1-like), StMADS11-like (SVP-like), AGL17-like and GGM13-like (Bsister) genes, suggesting that these genes are more important for angiosperm development and evolution than has previously been appreciated.