Dear Idrissa,

Have you tried the two serine hydroxymate: (i) MG1655, a wild-type E. coli K-12 strain, and (ii) an isogenic relAΔ251 derivative defective in the stringent response.?

Please note, that induction of amino acid biosynthetic genes is limited to the leader sequences of attenuator-regulated operons. Instead, up-regulated genes with known functions, including both regulators (e.g., rpoE, rpoH, and rpoS) and effectors, are largely involved in stress responses. However, one-half of the up-regulated genes have unknown functions. 

I suggest that you read the paper entitled "Transcription Profiling of the Stringent Response in Escherichia coli▿ " published in J. Bacteriol. February 2008 vol. 190 no. 3 1084-1096.

MATERIALS AND METHODS

Bacterial growth conditions and RNA isolation.Cultures of MG1655 (7) and an isogenic derivative of this strain containing the relAΔ251 mutation (65) were grown aerobically at 37°C in morpholineethanesulfonic acid (MOPS)-based medium supplemented with all 20 amino acids (50 μg/ml each) and 0.1% glucose (66) in shake flasks. The stringent response was induced by addition of 100 μg/ml SHX (Sigma) when cultures reached an optical density at 600 nm of 0.2. At four time points during growth (Fig. 1), samples were processed for transcriptional profiling. Detailed protocols for sample preparation are available on our website (www.genome.wisc.edu). Briefly, 15-ml aliquots were harvested, mixed with 30 ml RNAprotect bacterial reagent (Qiagen), and pelleted. Total nucleic acid was isolated using MasterPure kits according to the manufacturer's specifications (Epicenter Technologies). Nucleic acid pellets were treated with 0.05 U/μl DNase I for 45 min at 37°C and then repurified with MasterPure.

Probe preparation and hybridization.Conversion of RNA to cDNA, labeling, and hybridization were conducted essentially as described previously (www.affymetrix.com). Briefly, 10 μg of purified total RNA was reverse transcribed using Superscript II (Invitrogen), and this was followed by RNase digestion and cDNA purification using Qiaquick PCR purification columns (Qiagen). Note that these columns did not retain DNA smaller than 40 bp and had a lower affinity for molecules smaller than 100 bp, a property which affected quantitative recovery of small RNAs, including tRNAs. Isolated cDNA was fragmented by partial DNase I digestion to obtain an average length of 50 to 100 bp. Fragmented cDNA was 3′ end labeled with biotin-N6-ddATP (Enzo Biochem), mixed with hybridization solution, and loaded onto Affymetrix GeneChip E. coli antisense genome arrays. Following 16 h of hybridization at 45°C, each array was washed and stained with streptavidin-phycoerythrin (Molecular Probes) using an antibody intermediate to enhance the signal. Arrays were scanned at 570 nm with 3-μm resolution using a confocal laser scanner (Hewlett-Packard). Three replicates from independent experiments were analyzed for MG1655, and two replicates were analyzed for therelAΔ251 strain.

Primer extension analyses.MG1655 and the relAΔ251 mutant strain were cultured in the same medium, and RNAs were isolated as described above. The method used for primer extension analysis to detect the transcripts from P1 promoters of four E. coli rRNA operons and from an ompA control promoter and the associated primers have been described previously (103).

Data analysis.Raw data were imported into a relational database and first normalized by scaling the average of the fluorescence intensities for each gene to a constant correcting for variations in sample loading, hybridization, and staining. The average of the normalized intensity values across replicates corresponded to the abundance of each gene. Genes showing a log2 ratio ≥1.5 with an average log2 signal intensity greater than 9 were considered induced. Conversely, genes with a log2 ratio ≤ −1.5 and had average log2 signal intensity greater than 9 in the control were considered repressed. The signal intensity cutoff was used to restrict the analysis to the genes that were unambiguously expressed. The apparent fold changes for some genes should be interpreted with caution when the signal intensity for a gene in either of the profiles being compared is low (log2 value,

Dear Rafik.

Thank you very much for that valuable suggestion !