Distyly is a polymorphism that promotes cross-pollination where two floral morphs called ‘pin' and ‘thrum' display reciprocal anthers-stigma height and co-occur within a species. It has evolved convergently across at least 34 flowering plant families and is generally coupled with self- and intramorph- incompatibility [1] (Fig. 1). Distyly has been reported early in Oleaceae and occurs in the four lineages Myxopyreae, Forsythieae, Jasmineae, and Oleeae. In Oleaceae, distyly and self-incompatibility are controlled by a hemizygous supergene termed ‘S-locus' - present in a single copy in the thrum morph [2,3]. Phylogenetic evidence supports the inheritance of the S-locus in Oleaceae from the common ancestor of Myxopyreae, Jasmineae and Oleeae. In these three lineages, two regulator genes are present in the S-locus, one from the brassinosteroid pathway called BZR1-S and another called GA2ox-S from the gibberellin pathway. In non-distylous Oleeae, self-incompatibility is also controlled by the S-locus where only the candidate gene for self-incompatibility GA2ox-S is functional (Fig. 1) [2,3]. However, no ortholog of GA2ox-S was found in Forsythieae although reported self-incompatible. Their homologous region of the S-locus (bearing the BZR1-S gene) would be heterozygous (not hemizygous) according to preliminary results, suggesting a transitory loss of the S-locus in Forsythieae. While a recent study in Forsythia suspensa has shown that the differential style length between the two floral morphs would be controlled by brassinosteroids [4] (Fig. 1), the gene cascade determining self-incompatibility in this taxa remains unknown.

 Here, we aim at identifying the molecular mechanisms underlying distyly and self-incompatibility in Forsythia by transcriptomics (Fig. 1). Six ornamental cultivars (of F. x intermedia or its parent F. viridissima) from three phenotypic categories were sampled: self-incompatible thrum, self-incompatible pin, and self-compatible pin (Fig. 2A). In order to determine the time window where distyly and self-incompatibility take place, samples were collected over three developmental stages (Fig. 2B). Organ-specific samples with stamens or gynoecia were obtained following flower dissection and their transcriptomes were generated with a short reads RNAseq approach (Fig. 2C). Statistical modelling of the gene expression will help identifying differentially expressed genes between fertile floral organs of the two morphs and between self-compatible and self-incompatible individuals of the pin morph. These genes may respectively account for the phenotype of self-incompatibility and distyly. Ultimately, this work will contribute to the understanding of the evolution of the mating system in Oleaceae, in which distyly may have been lost several times and secondarily reacquired in Forsythieae.