Format: ORAL

Theresa A. Mustacchio1, Amila Radoncic1, YolandSavriama2, Jingjing Tong3, Dianella G. Howarth1

1 Department of Biological Sciences, St. John’s University, New York, USA 2 Department Evolutionary Genetics, Max-Planck Institute for Evolutionary Biology, 24306 Plön, Germany 3 Department of Plant and Soil Sciences, University of Delaware, Delaware, USA

Shifts between radial symmetry (actinomorphy) and bilateral symmetry (zygomorphy) in flowers have occurred multiple times independently within angiosperms. These morphological shifts are commonly associated with increased pollinator specialization and speciation rates. The CYCLOIDEA (CYC)-like genes have been shown to shift their expression dorsally independently in nearly all shifts to bilaterally symmetrical flowers examined. Three core eudicot clades of CYC-like genes have been identified: CYC1, CYC2 and CYC3, with only CYC2-like genes being shown to have a role in floral symmetry. We use the non-model plant, Fedia graciliflora (Caprifoliaceae) to examine the effect of CYC2 as well as CYC3 paralogs on floral shape and symmetry. F. graciliflora have strongly bilaterally symmetrical flowers with morphologically distinct dorsal (upper), lateral, and ventral (lower) petals and only two functional stamens. Fedia is nested within the Valerianella, a group with pseudo-radially symmetric flowers with three functional stamens, providing a model for studying the evolution of strong bilateral symmetry and teasing apart the effects from each paralog. Here we used Virus-Induced Gene Silencing (VIGS) to knockdown each of the paralogs separately: FgCYC2A, FgCYC2B, FgCYC3A, and FgCYC3B. The down-regulated plants were assayed using geometric morphometrics combined with tools of multivariate statistical shape analysis to quantify shape changes, using a combination of landmarks and semilandmarks to precisely capture in high detail the corolla outline. Using canonical variate analysis (CVA), we are able to statistically model the morphological effect of lower expression of each paralog. These analyses indicate that each paralog plays a slightly different role in patterning floral symmetry, both in location and size of petal lobes. Additionally, here we show evidence for the first time that CYC3, despite having lower expression in general, also patterns corolla shape.