PLoS ONE (Jan 2010)

Tinkering with the C-function: a molecular frame for the selection of double flowers in cultivated roses.

  • Annick Dubois,
  • Olivier Raymond,
  • Marion Maene,
  • Sylvie Baudino,
  • Nicolas B Langlade,
  • Véronique Boltz,
  • Philippe Vergne,
  • Mohammed Bendahmane

DOI
https://doi.org/10.1371/journal.pone.0009288
Journal volume & issue
Vol. 5, no. 2
p. e9288

Abstract

Read online

BACKGROUND: Roses have been cultivated for centuries and a number of varieties have been selected based on flower traits such as petal form, color, and number. Wild-type roses have five petals (simple flowers), whereas high numbers of petals (double flowers) are typical attributes of most of the cultivated roses. Here, we investigated the molecular mechanisms that could have been selected to control petal number in roses. METHODOLOGY/PRINCIPAL FINDINGS: We have analyzed the expression of several candidate genes known to be involved in floral organ identity determination in roses from similar genetic backgrounds but exhibiting contrasting petal numbers per flower. We show that the rose ortholog of AGAMOUS (RhAG) is differentially expressed in double flowers as compared to simple flowers. In situ hybridization experiments confirm the differential expression of RhAG and demonstrate that in the double-flower roses, the expression domain of RhAG is restricted toward the center of the flower. Conversely, in simple-flower roses, RhAG expression domain is wider. We further show that the border of RhAG expression domain is labile, which allows the selection of rose flowers with increased petal number. Double-flower roses were selected independently in the two major regions for domestication, China and the peri-Mediterranean areas. Comparison of RhAG expression in the wild-type ancestors of cultivated roses and their descendants both in the European and Chinese lineages corroborates the correlation between the degree of restriction of RhAG expression domain and the number of petals. Our data suggests that a restriction of RhAG expression domain is the basis for selection of double flowers in both the Chinese and peri-Mediterranean centers of domestication. CONCLUSIONS/SIGNIFICANCE: We demonstrate that a shift in RhAG expression domain boundary occurred in rose hybrids, causing double-flower phenotype. This molecular event was selected independently during rose domestication in Europe/Middle East and in China.