Evo-Devo Of CYC-Like Genes Controlling Floral Symmetry In Some Typical Species Of Plantaginaceae And Gesneriaceae, With Analysis On The Phylogenetic Placement Of Rehmannia And Triaenophora Based On Molecular Evidence

The developmental pathway of floral symmetry has been revealed at molecular developmental level in model species Antirrhinum majus. However, research in how modifications of development lead to the transformation among different types of floral symmetry during evolution is a vast unexplored field still at the molecular developmental level. Therefore, we here select some typical species from Plantaginaceae and Gesneriaceae to carry out comparative studies on the expression pattern and regulation mechanism of CYC-like genes controlling floral symmetry. The expression pattern of RgCYC in Rehmannia might reflect a new mechanism underlying the evolutionary change from pentamerous to tetramerous flowers in Lamiales s.l.. First, we found that the loss of the adaxial stamen in R. glutinosa is associated with the enhanced expression of RgCYC at its expected site, which is likely connected with specific amino acid replacements in its transcriptional factor. Second, the RgCYC expression expanded to lateral stamens but restricted in their marginal area contributes to the formation of the didynamous androecium in R. glutinosa. The different expression patterns between McCYC and RgCYC in space and timing are probably responsible for the remarkably different fates of the lateral stamens between M. confertiflora and R. glutinosa. Third, the differences of the adaxial petals between A. majus and R. glutinosa under similar expression patterns of CYC and RgCYC are probably because of the single RgCYC gene controlling the zygomorphy in R. glutinosa, or changes in downstream genes involved in cell division and expansion. It is a very interesting field to explore the developmental pathway and molecular mechanism of the morphological transformation from zygomorphy to secondary actinomorphy. By using the mTAIL-PCR method,we have first isolated two CYC-like genes, SiCYC1A and SiCYC1B, from zygomorphic and actinomorphic cultivars of Saintpaulia ionantha in Gesneriaceae. The two genes SiCYC1A and SiCYC1B from the zygomorphic cultivar both contain the whole regulation domain, i.e. TCP and R domain. Therefore, they should be functional in the floral symmetry establishment, homologous with CYC in A. majus. Unexpectedly, the two genes from the actinomorphic cultivar are absolutely identical to those of the zygomorphic cultivar in DNA sequence, respectively. Based on comparative analysis of the molecular pathways responsible for the morphological transformation from zygomorphy to actinomorphy, we suggest that the two closely related genes SiCYC1A and SiCYC1B might be regulated by a common upstream regulator, in which its change would result in silence of both SiCYC1A and SiCYC1B in controlling the adaxial and lateral organs of a flower. Investigation on the floral development in Triaenophora shows that the initiationsequence of the calyx and corolla lobes fits into the common pattern of Digitaleae and Veroniceae. Moreover, the delay in corolla early growth and aestivation is also intermediate between Digitaleae and Veroniceae. However, only Rehmannia and Triaenophora in Digitaleae have four stamens with a gap at the expected site of the fifth stamen. Furthermore, the phylogenetic analysis based on ITS (internal transcribed spacer) and cpDNA trnL-F sequences shows a close phylogenetic relationship between Rehmannia and Triaenophora that are sister to each other. It is unexpected that Rehmannia and Triaenophora are not nested within the clade that contains Digitalis,Veronica and Plantago as representatives of the newly defined Plantaginaceae. On the contrary, Rehmannia and Triaenophora, as an independent clade, are sister to the clade of Paulownia and newly defined Phrymaceae. The evidence from DNA sequences does not support the monophyly of Digitaleae, in which Rehmannia and Triaenophora are traditionally placed in close to Digitalis. It is necessary to reconsider the phylogenetic placement of Rehmannia and Triaenophora.