Hybridization, washing and staining of microarrays was performed as recommended in the Affymetrix technical manual. Briefly, the cDNA fragments were hybridized for 16 hours at 45oC to an Affymetrix GeneChip E. coli antisense microarray. Arrays were washed at 25oC with non-stringent wash buffer, followed by a wash at 50oC with stringent buffer. Then, the arrays were stained with phycoerythrin-conjugated streptavidin (Molecular Probes). The arrays were read at 570 nm with a GeneChip Scanner 3000 and the data were analyzed with the Affymetrix Microarray Suite 5.0 software. Sample variations were standardized by scaling the average of all genes to constant target intensity of 500 for all arrays used. The statistical significance of the data was analyzed by using rank-based (non-parametric) methods to generate p-values for each gene. The data were passed through a first screen to discard absent genes, which have p-values > 0.065. Then, from the present (p-value < 0.05) or moderate (p-value between 0.05 and 0.065) genes, only those with a signal ratio (pNM12/pBAD-ryhB) higher than 2 fold, positive or negative, were kept for further studies. Operons were included in the Tables only if at least one gene within the operon had a ratio of over 2. Usually, when the detection level was low (signal less than 200), the genes were discarded, unless otherwise indicated in the text (e.g. bfr). Array experiments were done in duplicate and the reproducibility of the experiments is shown on the scatter plot of the values for all present genes (supplemental data: S2). Genes repressed by RyhB in all three experiments (wild-type, fur, wild-type with added FeSO4) have been placed in Table 1. Table 2 includes genes that are repressed by RyhB only in a fur+ background (unless indicated otherwise in the text). Table 3 includes genes that were increased in expression upon RyhB expression.
