Morphological innovations in Angiosperms : evolution of floral symmetry in basal eudicots – FLOSYM

Shifts in floral symmetry have occurred several times independently during Angiosperm evolution. It appears that bilateral symmetry (or zygomorphy) is most probably an adaptive character that co-evolved with specialized pollinators. It is remarkable that this floral trait characterises some of the most successful groups (as evaluated by the number of species per family – see for example legumes, orchids). As such, it can be considered as a key morphological innovation. Remarkable progress in understanding the mechanisms underlying floral symmetry has been achieved in the model species Antirrhinum majus. Petal type and abortion of the dorsal stamen were found to be controlled by the interplay of four genes belonging to two families of transcription factors. The first two genes characterised, Cycloidea (Cyc) and Dichotoma (Dich), are expressed in the dorsal domain of the early floral meristem and are responsible for the dorsal identity of petals and for stamen abortion. They are paralogous genes that belong to the class II TCP transcription factor gene family. A third gene, Divaricata (Div), is expressed all over the meristem and is responsible for the ventral identity of petals. This can be deduced from the observation of the flowers of cyc-dich double mutants, which are actinomorphic. The dorsal effect of Cyc and Dich appears largely mediated by a fourth gene, Radialis (Rad). Rad is also expressed in the dorsal domain of floral meristems, where it acts antagonistically with Div. So far, all evo-devo studies of floral symmetry have focused on Cyc-like genes. Changes in gene expression patterns have been found to be correlated with changes in floral symmetry in several members of the tribe Anthirrhineae, and more recently in the unrelated legume species Cadia purpurea. The role of a Cyc-like gene in floral symmetry was recently demonstrated in the model legume species Lotus japonicus. These results indicate that the setting up of zygomorphy, which has evolved independently in Asterids (A. majus) and Rosids (L. japonicus), relies at least partly, on a homologous molecular pathway. They also emphasize the relevance of studying the molecular evolution of Cyc-like genes, and searching for their role in floral development and the evolution of zygomorphy at the level of angiosperms. Diversity and evolution of Cyc-like genes have been studied mainly in core eudicot families, where zygomorphy has evolved recurrently. The most recent phylogenetic analysis of Cyc-like genes in core eudicots revealed the existence of three homology groups, each one being characterised by one Arabidopsis gene. The phylogeny suggests two successive duplications, probably at the base of the core eudicots. Several research groups in the UK and USA are involved in the analysis of Cyc-like genes in the core eudicots, where zygomorphy is widespread. In basal angiosperms and basal eudicots, where zygomorphy is quite rare, to our knowledge, very few groups are at work yet. The group of the coordinator is one of these, and has recently shown that Cyc-like genes have undergone a duplication in the Papaveraceae, independently from those that have taken place in core eudicots. Moreover, Cyc-like gene expression in the flowers and inflorescences of three species of Papaveraceae suggests an ancestral expression pattern at junction between organs, and a species specific expression during flower development, possibly related to flower symmetry. Using a typical evo-devo approach, our project aims to explore the molecular bases and architectural context underlying the evolution of floral zygomorphy in basal eudicots compared to core eudicots. We will use a multidisciplinary approach, combining (i) comparative evolutionary analysis of floral traits potentially associated with symmetry shifts in basal and core eudicots, (ii) molecular evolution analysis of Cyc-like genes in basal eudicots, with a special interest in the history of duplications, and (iii) analysis of patterns of Cyc-like gene expression as related to floral development in selected taxa chosen among basal eudicots. The expected results will shed light on the evolution of floral form in eudicots, and the diversification of a family of transcription factor gene implicated in developmental processes. Our results will settle the bases for future functional analyses of Cyc-like genes in non-model basal eudicot species.