Roses are the most economically important horticultural species because of their use in fragrance industry and in horticulture. Modern roses are the results of a long selection process lying on the vase life, the pest and disease resistance, the shape and the colour of the flower. Fragrance of roses for the cut flower market was not selected during this process. But nowdays, fragrance is of great interest for the breeders. Perfume of roses is a complex mixture of volatile compounds, mainly monoterpenes. The aim of my work was to characterize the petal of roses as a secretory organ and to isolate and characterize two genes involved in the biosynthetic pathway of monoterpenes. For this purpose, we studied the petal at the ultrastructural level. Both epidermis of the petal are able to synthezize and emit volatile compounds. We identified some structural features putatively involved in the biosynthesis of monoterpenes. A comparative study was made between scented and non-scented varieties but no major differences were observed. Two enzymes involved in the biosynthetic pathway of monoterpenes were studied : RhDXR, involved in IPP biosynthesis and the homodimeric form of RhGPPS, involved in the production of GPP, precursor of monoterpenes. In an in vitro assay, no enzymatic activity was detected for the recombiant RhGPPS. Expression of the corresponding gene was studied by RT-PCR. Its expression is constitutive and not related to biosynthesis of monoterpenes in different rose varieties. RhDXR was characterized as a functional enzyme. The expression of RhDXR was studied by semi-quantitative RT-PCR. By fusions with GFP, we showed that the protein is targeted to the plastids of rose petals. The effect of the overexpression of this gene in transgenic tobaccos is currently under study, with the aim to evaluate the ability of plants to produce more volatiles compounds.
