Dysfunction of CYC2g is responsible for the evolutionary shift from radiate to disciform flowerheads in the Chrysanthemum group (Asteraceae: Anthemideae)

Abstract

Evolutionary shifts among radiate, disciform and discoid flowerheads have occurred repeatedly in a number of major lineages across the Asteraceae phylogeny; such transitions may also appear within evolutionarily young groups. Although several studies have demonstrated that CYC2 genes partake in regulating floral morphogenesis in Asteraceae, the evolution of capitulum forms within a recently diverging lineage has remained poorly understood. Here, we study the molecular regulation of the shift from a radiate to a disciform capitulum within the Chrysanthemum group. This is a recently radiating group mainly comprising two genera, Chrysanthemum and Ajania, that are phylogenetically intermingled but distinct in flowerhead morphology: Chrysanthemum spp. with radiate capitula and Ajania spp. with disciform capitula. We found that the morphogenesis of zygomorphy in the marginal floret in Ajania was disrupted soon after floral primordium emergence; CYC2g, one of the CYC2 copies that was expressed prominently in the ray floret of Chrysanthemum was not expressed in flowerheads of Ajania. Weakening the expression of ClCYC2g in Chrysanthemum lavandulifolium led to the gradual transition of a ray flower toward the disc-like form. Molecular evolutionary analyses indicated that the disciform capitulum might have evolved only once, approximately 8 Mya, arising from dysfunction of the CYC2g orthologs. A 20-nt deletion, including a putative TATA-box of the Ajania-type CYC2g promoter, appeared to inhibit the expression of the gene. Considering the divergent habitats of Chrysanthemum and Ajania, we propose that the shift from radiate to disciform capitulum must have been related to changes in pollination strategies under selective pressure.

Clarifying Recent Adaptive Diversification of the Chrysanthemum-Group on the Basis of an Updated Multilocus Phylogeny of Subtribe Artemisiinae (Asteraceae: Anthemideae)

“…This suggests that capitulum polymorphism had long existed in the ancestral populations of the Chrysanthemum-group. Comparing the present multilocus coalescent tree and the tree based on CYC2g sequences of Chrysanthemum-group (Shen et al, 2021), we found a high consensus of topologies, implying that the evolution of this group is linked to the alteration of flowerhead morphology (Supplementary Figure 7). Within Chrysanthemum-group, low pairwise niche overlaps were found between subclades, although they diverged rather recently (Figures 4, 5).…”

“…This could be due to incomplete lineage sorting of genes analyzed here. We found that the sequences of their CYC2g genes were highly similar to those of A. parviflora (Shen et al, 2021; Supplementary Figure 7) which is a member distributed in the eastern part of the range of Ajania. However, the mix-up of three Ajania species with Chrysanthemum is more likely due to secondary contacts as we found all three species are polyploids with high ploidy levels from 2n = 6x to 8x and to 10x, suggesting complicated origins probably involving parentages from Ch.…”

“…Although functional assays of chrysanthemum ClCYC2g and sunflower HaCYC2c showed that downregulating the expression of these genes led to the appearance of dorsal petals and stamens in ray florets, we assume that the developmental process of the discoid capitulum might be completely different from that of the radiate/disciform due to lacking a module responsible for the development of zygomorphic or asymmetric marginal female flowers, namely, a marginality identity module ( Pozner et al, 2012 ; Zhao et al, 2016 ; Elomaa et al, 2018 ; Zoulias et al, 2019 ; Shen et al, 2021 ). The finding of our previous study that the CYC2g gene was lost in the discoid species Stilpnolepis centiflora ( Chen et al, 2018 ; Shen et al, 2021 ) may be a hint for this hypothesis. Certainly, to test this hypothesis, more detailed molecular developmental studies are required.…”

“…Later molecular genetic studies indicated that the CYC2-mediated module participates in inflorescence repatterning ( Broholm et al, 2008 ; Chapman et al, 2008 , 2012 ; Kim et al, 2008 ; Tähtiharju et al, 2012 ; Chen et al, 2018 ), and the rise of rayless capitula is always linked to a reduction in the expression or gene loss of one or more CYC2 members ( Chen et al, 2018 ). Our recent analyses demonstrated that dysfunction of CYC2g led to a shift from the radiate flowerhead to the disciform head in Chrysanthemum -group ( Shen et al, 2021 ). Therefore, the radiate should be the ancestral state of the capitulum in subtribe Artemisiinae, despite Stilpnolepis centiflora at its basal position possessing discoid flowerheads (probably an autoapomorphic state).…”

“…Considering their distinct capitulum morphologies, we have postulated that these two genera must have experienced adaptive divergence associated with differential environmental conditions ( Chen et al, 2020 ). Our recent developmental genetic study revealed that a disciform capitulum may have evolved from a radiated type owing to the dysfunction of a key ray-flower regulator, CYCLOIDEA2g ( Chen et al, 2018 ; Shen et al, 2021 ). To date, however, no molecular phylogeny has resolved Chrysanthemum and Ajania each as monophyletic ( Kondo et al, 2003 ; Masuda et al, 2009 ; Zhao et al, 2010a , b ; Liu et al, 2012b ; Huang et al, 2017 ).…”