This data demonstrates that such low variability genes span a wide range of PD 404,182 expression level, from 7 to almost 16. Setting a threshold of max-min expression of 1.5, we selected an initial set of 49 genes from among the previous 209 candidate genes. Eighteen of them had a mean expression level between 7 and 9, twenty-six between 9 and 11, and five between 11 and 14. Next, we required that the selected genes satisfied the following PalGly criteria: they are associated, as far as possible, with different functional categories to minimize the risk of choosing co-regulated genes; they belong to different transcription units; they show no sign of correlation between the expression profiles and growth conditions. Among the 49 genes, 22 encode metabolic enzymes, 9 genes encode membrane proteins involved in membrane traffic and 6 are predicted to encode transcriptional regulators, the others encoding hypothetical proteins or proteins with unknown function. It is now acknowledged that no gene has a constant expression level regardless of the experimental conditions tested. Every gene is regulated and the expression level may vary slightly. In order to expand the use of the nine reference genes identified here, we determined their expression in new experimental conditions, as follows: exponential and stationary phases in rich medium, as used routinely in the laboratory, and minimal medium supplemented with different carbon sources normally encountered in host plants or in the environment, e.g. xylose, galactose or rhamnose. In addition to these conditions, we also considered the DNA supercoiling state since recent data shows that this parameter modulates virulence gene expression. Transcript accumulation of the retained genes was thus determined in the presence of Coumermycin, a compound which relaxes DNA, in both exponential and stationary phases of growth. We finally extended these studies to the in planta conditions by monitoring the expression levels of the candidate genes during infection of Arabidopsis thaliana by D. dadantii. For this experiment, D. dadantii cells were purified from infected plantlets at two very different time points: 6 hours post-infection, corresponding to the early stage of infection and 24 hours post-infection, corresponding to the maceration stage.