||Lily (Liliumspp.) is among the most traditional and beloved ornamental flowers worldwide. The genus Lilium comprises almost one hundred species, among which is the primary subject of our research, described in this thesis, the species Lilium longiflorum (Thunb.), known as trumpet lily or Easter lily.Despite the great economic importance of ornamental lily species, little is known about its biology at the molecular level so far. In a time when two genomes are fully sequenced, Arabidopsis thaliana and Oryza sativa , only a few genes have been characterized in Lilium spp. yet. Possible reasons for this are discussed throughout this thesis.This work intends to be a contribution to bridging the fundamental research concerning transcription factors involved in development of flower morphology in model species and the applied objectives of molecular breeding for manipulating flower morphology, endeavouring to create new cultivars with specific and novel features, more specifically in Lilium spp.The ABC model for floral development was proposed more than 10 years ago and since then many studies have been performed in model species, such as Arabidopsis thaliana , Antirrhinum majus , petunia and many other species in order to confirm this model. This investigation has led to additional information on flower development and to more complex molecular models.In the first chapter of this thesis, notions of molecular floral development, the difficulties of working with molecular biology of lily, the state-of-the-art in lily transformation are introduced, as well as general overviews of transcription factors, MADS-box genes, the ABCDE model for flower development and functional characterization of genes in heterologous systems. These concepts will guide the reader throughout the work we present here.AGAMOUS( AG ) is the only C type gene found in Arabidopsis and it is responsible for stamen and carpel development as well as floral determinacy. In the second chapter, we describe the isolation of LLAG1 , a putative AG orthologue from lily ( L. longiflorum ) by screening a cDNA library derived from developing floral buds. The deduced amino acid sequence of LLAG1 revealed the MIKC structure and a high homology in the MADS-box among AG and other orthologues. Phylogenetic analysis indicated close relationship between LLAG1 and AG orthologues from monocot species. Spatial expression data showed LLAG1 transcripts exclusively in stamens and carpels, constituting the C domain of the ABC model. Functional analysis was carried out in Arabidopsis by overexpression of LLAG1 driven by the CaMV 35S promoter. Transformed plants showed homeotic changes in the two outer floral whorls with some plants having the second whorl completely converted into stamens. Altogether, these data indicate a functional relationship between LLAG1 and AG .( SEP3 ) is a MADS-box homeotic gene possibly determining the E function in the ABCDE model. This function is essential for proper development of petals, stamens and carpels. In order to gain further information on lily ( Lilium longiflorum ) flower development at the molecular level, the cDNA library constructed from developing floral buds was screened again and our findings are reported in the chapter three. A clone ( LLSEP3 ) was isolated with high similarity to the SEP3 transcription factor from Arabidopsis . LLSEP3 belongs to the AGL2 subfamily of MADS-box genes and shares its closest relationships with DOMADS1 and OM1 , from the orchid species Dendrobium grex and Arandadeborah, respectively. Expression analysis by Northern hybridisation showed that LLSEP3 was expressed throughout lily flower development and in tepals, stamens and carpel tissues of mature flowers, whereas no expression was detected in leaves. Overexpression of LLSEP3 in Arabidopsis under the CaMV35S promoter induced early flowering but did not induce any floral homeotic changes, which is in accordance with the effect of SEP3 overexpression in this species. Altogether, these data are consistent with the putative role of LLSEP3 as an E functional gene in lily flower development.Drawbacks found during our work on functional characterisation of LLAG1 , by means of complementing the agamous mutant of Arabidopsis thaliana are described and critically discussed in chapter four. Such difficulties are, on the one hand, the nature of the AGAMOUS gene, of which the loss of function induces sterility and, on the other hand, the unavailability of the defective ag-1 allele in another Arabidopsis background than the Landsberg erecta ecotype, which is recognisably difficult to transform by the floral dip method. Even though we did not manage to complement the AG function with LLAG1 in a defective ag genotype so far, we could observe clear floral homeotic changes in those Arabidopsis plants ectopically overexpressing LLAG1 , which together with our data on sequence identities and expression profile described in the previous chapter of this thesis, indicated that LLAG1 is a strong candidate to control the C function in L. longiflorum .This work also contributes towards the improvement of lily transformation procedures. In the chapter five we describe a transformation of bulblet slices by particle bombardment using a vector carrying the ArabidopsisSUPERMAN gene driven by the petunia flower-specific FLORAL BINDING PROTEIN 1 promoter and the bialaphos resistance gene phosphinothricin acetyltransferase under the CaMV35S promoter. Our intentions were improving the transformation parameters for lily transformation in order to reach higher efficiency, and creating novel phenotypes in lily flowers using transcription factors originating from dicot plants. We were capable of obtaining transgenic lines expressing in vitro resistance to bialaphos. The transgenic plants were transferred to the greenhouse, grown and monitored for two flowering seasons. Flowers derived from these plants appeared normal and indistinguishable from wild-type flowers and the possible reasons for this are currently under investigation.Homeoticchanges in floral organs of lily ( Lilium spp.) are described in chapter six. Usually, lily flowers show similar organs in their first and second whorls called tepals. They constitute the appealing and colourful features determining flower appearance. Stamens and the pistil appear as the third and fourth whorls, respectively. A double lily flower shows replacement of stamens by tepals and of its carpel by a new flower in a reiterated manner, similar to what is seen in the agamous mutant of Arabidopsis . A novel floral phenotype of lily, denominated festiva here, has never been reported in other species so far and shows a complete homeotic change of stamens into tepals, but keeps the carpel identity. We tried to explain these phenotypes taking into consideration all the evidence on the genetic mechanisms involved in flower development gathered over the last 15 years. This work launches challenges and encouragement for exploiting the molecular mechanisms involved in flower development of lily.Virus-induced gene silencing (VIGS) system has shown to be of great potential in reverse plant genetics. Advantages of VIGS over other approaches, such as T-DNA or transposon tagging, include the circumvention of plant transformation, methodological simplicity and robustness, and speedy results. These features enable VIGS as an alternative instrument in functional genomics, even in a high throughput fashion. The system is already well established in Nicotiana benthamiana , but efforts are being addressed to improve VIGS in other species, including monocots. Current research is focussed on unravelling the silencing mechanisms of post-transcriptional gene silencing (PTGS) and VIGS, as well as finding novel viral vectors in order to broaden the host species spectrum. In chapter seven, we discuss the advantages of using VIGS to assess gene functions in plants. We address the molecular mechanisms involved in the process, present the available methodological elements such as vectors, inoculation procedures, and we show examples in which the system was applied successfully to characterize gene function in plants. Moreover, we analyse the potential application of VIGS in assessing genetic function of floral transcription factors from monocots.Analyses of gene functions involved in lily flower development and generation of useful information on the molecular breeding potential of this species were the main objectives of the work described in this thesis.The field for studying the molecular aspects of lily flower development is now wide open and the future may uncover very interesting aspects that will produce new tools for ornamental breeders as well as reveal particular features of monocots and the Liliaceae plant family.