Combining these data with the other experimental conditions described in Brenner et al. (2005), we selected six genes (NDHC, NDHI, RPS2, RPS3, RPS11, RPOC2) that were stable (with exception of NDHI and NDHC in 15 or 120 min BA treatment) under all Antifection chemical the experimental conditions. Stability of reference genes cDNA samples from leaves of transgenic plants with elevated or diminished cytokinin content (Polanská et al. 2007; Synková et al. 1999), as well as from the respective control plants were used to amplify these candidate reference genes. Relative expression data of each cDNA sample were used for geNorm algorithm. The geNorm algoritm calculates a measure M for each reference gene, which reflects

the expression stability of the gene, compared to the other reference genes; a lower M-value Selleck JAK inhibitor means a more stable gene expression. As cytokinins influence

both nuclear- and plastid-encoded genes, it is highly important to know which reference genes (nuclear- and/or plastid-encoded) should be used to normalize our real-time PCR data. Two different geNorm analyses were performed. In a first analysis, when only the nuclear-encoded reference genes were considered, Nt-ACT9, NT-αTUB and Nt-SSU turned out to be the most stable reference genes (Fig. 1a). Analyses of the plastid-encoded reference genes resulted in Nt-RPS3, Nt-NDHC and Nt-IN1 as the best reference genes (Fig. 1b). Fig. 1 Evaluation of reference genes in Nicotiana tabacum (Pssu-ipt/ckx) with the pairwise variation measure. The pairwise variation measure ‘V n/n+1’ measured the effect of adding additional reference genes on the normalisation factor for these treatments. Stepwise exclusion of the reference genes with the highest M value resulted in a ranking of the candidate reference genes when a nuclear-encoded reference genes (18S rRNA (18S), elongationfactor

1α (elongation), actin 9 (actin9), alfa-tubulin (tubulin) and small subunit of RubisCO (rbcS)); or b plastid-encoded reference genes (ribosomal protein S2 (rps2), ribosomal protein S11 Interleukin-2 receptor (rps11), 16S rRNA (16S rRNA), RNA polymerase beta subunit 2 (rpoC2), β subunit of acetyl-CoA carboxylase (accD), NADH dehydrogeanse D3 (ndhC), NADH dehydrogenase subunit (ndhI), initiation factor 1 (ini1) and ribosomal protein S3 (rps3)) were considered The geNorm algorithm also determines the pairwise variation V n/n+1, which indicates how many reference genes should be included, by measuring the effect of adding further reference genes on the normalisation factor. The V-graph of the nuclear-encoded reference genes (Fig. 1a) shows that inclusion of a fourth gene would increase the stability of the normalization, but since this decrease in pairwise variation is not so large, we propose to use only the three most stable nuclear-encoded genes as reference genes. The V-graph of the plastid-encoded reference genes (Fig.