Search

Abstract Background Gene expression studies are a prerequisite for understanding the biological function of genes. Because of its high sensitivity and easy use, quantitative PCR (qPCR) has become the gold standard for gene expression quantification. To normalise qPCR measurements between samples, the most prominent technique is the use of stably expressed endogenous control genes, the so called reference genes. However, recent studies show there is no universal reference gene for all biological questions. Roses are important ornamental plants for which there has been no evaluation of useful reference genes for gene expression studies. Results We used three different algorithms (BestKeeper, geNorm and NormFinder) to validate the expression stability of nine candidate reference genes in different rose tissues from three different genotypes of Rosa hybrida and in leaves treated with various stress factors. The candidate genes comprised the classical "housekeeping genes" (Actin, EF-1α, GAPDH, Tubulin and Ubiquitin), and genes showing stable expression in studies in Arabidopsis (PP2A, SAND, TIP and UBC). The programs identified no single gene that showed stable expression under all of the conditions tested, and the individual rankings of the genes differed between the algorithms. Nevertheless the new candidate genes, specifically, PP2A and UBC, were ranked higher as compared to the other traditional reference genes. In general, Tubulin showed the most variable expression and should be avoided as a reference gene. Conclusions Reference genes evaluated as suitable in experiments with Arabidopsis thaliana were stably expressed in roses under various experimental conditions. In most cases, these genes outperformed conventional reference genes, such as EF1-α and Tubulin. We identified PP2A, SAND and UBC as suitable reference genes, which in different combinations may be used for normalisation in expression analyses via qPCR for different rose tissues and stress treatments. However, the vast genetic variation found within the genus Rosa, including differences in ploidy levels, might also influence expression stability of reference genes, so that future research should also consider different genotypes and ploidy levels.

Chrysanthemums are important ornamental plants with abundant phenotypic diversity. Especially in cut-flower breeding, shoot branching is important for the success of new varieties. To assess the genetic regulation of shoot branching and other horticultural important traits, we phenotyped and genotyped two types of chrysanthemum populations: a genotype collection of 86 varieties and a biparental F1-population (MK11/3) of 160 individuals. Using two different statistical approaches, a genome-wide association analysis and a single marker ANOVA, with AFLP marker data and candidate gene markers for shoot branching, we tried to identify markers correlated to the traits of interest. As expected for the outcrossing hexasomic chrysanthemums most of the phenotypic traits showed a continuous variation in both populations. With the candidate gene approach we identified 11 significantly associated marker alleles for all 4 strigolactone pathway genes BRC1, CCD7, CCD8 and MAX2 regulating shoot branching in the genotype collection. In the MK11/3 we detected seven markers for all candidate genes except MAX2 explaining a large proportion of the variation. Using anonymous AFLP markers in the GWA with the 86 genotypes and the single locus analysis with the F1-population we could detect 15 and 17 additional marker-trait associations, respectively. Our analyses indicate a polygenic inheritance of the shoot branching in the chrysanthemum, with a fundamental role of the strigolactone pathway genes BRC1, CCD7, CCD8 and MAX2 and we identified 50 associated markers to all traits under study. These markers could be used in the selection of the parental plants for breeding chrysanthemums to enrich them for positive alleles influencing plant architecture traits.

Polyploidy is a widespread phenomenon among higher plants and a major factor shaping the structure and evolution of plant genomes. The important ornamental chrysanthemum (Chrysanthemum indicum hybrid) possesses a hexaploid genome with 54 chromosomes and was classified based on its evolutionary origin and cytological methods as an allopolyploid. However, it is questionable whether cytological methods are sufficient to determine the type of ploidy, and there are more informative methods available based on molecular marker analyses. Therefore, we collected segregation data for 406 dominant molecular marker alleles [327 amplified fragment length polymorphism (AFLPs), 65 single-strand conformation polymorphism (SSCPs) and 14 microsatellites (EST-SSRs)] in a biparental F1 population of 160 individuals. We analyzed these data for the characteristics that differ between allopolyploids and autopolyploids, including the segregation ratio of each marker, the ratio of single-dose (SD) to multi-dose (MD) markers, the ratio of SD markers in coupling to those in repulsion and the banding patterns of the SSRs. Whereas the analysis of the segregation ratio of each polymorphic marker indicated disomic (13 markers) as well as hexasomic (eight markers) inheritance, the ratio of SD markers in coupling to those in repulsion was 1:0, which is characteristic of autopolyploids. The observed ratio of SD to MD markers was 0.67:0.33 which is significantly different to the expected segregation for auto- and allohexaploids. Furthermore, the three EST-SSR alleles were inherited in all possible combinations and were not independent of each other, as expected for fixed heterozygosity in allopolyploids. Combining our results with published cytological data indicates that cultivated chrysanthemums should be classified as segmental allohexaploids.