BackgroundFor centuries roses have been selected based on a number of traits. Little information exists on the genetic and molecular basis that contributes to these traits, mainly because information on expressed genes for this economically important ornamental plant is scarce.ResultsHere, we used a combination of Illumina and 454 sequencing technologies to generate information on Rosa sp. transcripts using RNA from various tissues and in response to biotic and abiotic stresses. A total of 80714 transcript clusters were identified and 76611 peptides have been predicted among which 20997 have been clustered into 13900 protein families. BLASTp hits in closely related Rosaceae species revealed that about half of the predicted peptides in the strawberry and peach genomes have orthologs in Rosa dataset. Digital expression was obtained using RNA samples from organs at different development stages and under different stress conditions. qPCR validated the digital expression data for a selection of 23 genes with high or low expression levels. Comparative gene expression analyses between the different tissues and organs allowed the identification of clusters that are highly enriched in given tissues or under particular conditions, demonstrating the usefulness of the digital gene expression analysis. A web interface ROSAseq was created that allows data interrogation by BLAST, subsequent analysis of DNA clusters and access to thorough transcript annotation including best BLAST matches on Fragaria vesca, Prunus persica and Arabidopsis. The rose peptides dataset was used to create the ROSAcyc resource pathway database that allows access to the putative genes and enzymatic pathways.ConclusionsThe study provides useful information on Rosa expressed genes, with thorough annotation and an overview of expression patterns for transcripts with good accuracy.

A survey for viruses in rose propagated in Europe resulted in detection of only Prunus necrotic ringspot virus (PNRSV) among seven viruses screened. Four percent of cut-flower roses from different sources were infected with PNRSV. Progression of the disease under greenhouse conditions was very slow, which should make this virus easy to eradicate through sanitary selection. Comparison of the partial coat protein gene sequences for three representative rose isolates indicated that they do not form a distinct phylogenetic group and show close relations to Prunus spp. isolates. However, a comparison of the reactivity of monoclonal antibodies raised against these isolates showed that the most prevalent PNRSV serotype in rose was different from the most prevalent serotype in Prunus spp. All of the 27 rose isolates tested infected P. persica seedlings, whereas three of the four PNRSV isolates tested from Prunus spp. were poorly infectious in Rosa indica plants. These data suggest adaptation of PNRSV isolates from Prunus spp., but not from rose, to their host plants. The test methodologies developed here to evaluate PNRSV pathogenicity in Prunus spp. and rose could also help to screen for resistant genotypes.

Resistant rootstocks offer an alternative to pesticides for the control of soil pests. In Prunus spp., resistance loci to root-knot nematodes (RKN) have been mapped and a transformation method is needed to validate candidate genes. Our efforts have focused on the generation of transformed hairy-roots and composite plants appropriate for nematode infection assays. An efficient and reliable method using the A4R strain of Agrobacterium rhizogenes for the transformation of Prunus roots with an Egfp reporter gene is given. The rooting efficiency, depending on the genotypes, was maximal for the interspecific hybrid 253 (Myrobalan plum × almond-peach), susceptible to RKN, that was retained for subsequent studies. From the agro-inoculated cuttings, 72% produced roots, mainly at the basal section of the stem. Transformed roots were screened by microscope detection of Egfp fluorescence and molecular analyses of the integration of the transgene. The absence of residual agrobacteria in the plants was checked by the non-amplification of the chromosomal gene chvH. Egfp was expressed visually in 76% of the rooted plants. Isolated hairy roots in Petri dishes and composite plants (transformed roots and non-transformed aerial part) in soil containers were inoculated with the RKN Meloidogyne incognita. In both cases, root transformation did not affect the ability of the nematodes to develop in the root tissues. Our results showed that isolated hairy-roots can be used to validate candidate genes and the conditions in which composite plants offer a complementary system for studying the function of root genes in physiological conditions of whole plants are discussed.

No abstract

We developed and evaluated two different methods to improve the detection of the most prevalent virus of rose in Europe, Prunus necrotic ring-spot virus (PNRSV). Immunocapture-reverse transcription-polymerase chain reaction was estimated to be about 100 times more sensitive than double-antibody sandwich-enzyme-linked immunosorbent assay (DAS-ELISA) and showed an equivalent specificity. Based on the observation that PNRSV multiplies actively in young growing tissues (axillary shoots and cuttings), an in vitro culture method allowing rapid (about 15 days) and homogeneous development of dormant axillary buds with high virus titers was standardized. ELISA tests of these young shoots showed, in some cases, a 10(4) to 10(5) increase in sensitivity in comparison to adjacent leaf tissues from the rose mother plants. Between 21 and 98% (depending on the season) more samples were identified as positive by using ELISA on samples from shoot tips grown in vitro rather than on leaves collected directly from the PNRSV-infected mother plants. This simple method of growing shoot tips in vitro improved the confidence in the detection of PNRSV and eliminated problems in sampling appropriate tissues.