In all of the following EMSA experiments, 10 fmol of target DNA and 100 μM zinc without addition of EDTA were used in the reaction mixture. Screening

for potential direct Zur targets by computational promoter analysis We further performed computational pattern matching analysis to predict direct Zur targets from the Zur-dependent genes disclosed by microarray. The regulatory consensus elements of Zur were analyzed (Fig. 2), and a position count matrix (Fig. Givinostat cell line 2c) was generated to statistically represent the conserved signals recognized by Zur, and subsequently used to screen for the potential Zur binding sites within the promoter sequences of the Zur-dependent genes uncovered by cDNA microarray. This analysis generated a score value for each promoter sequence, and the larger numbers of these scores would corresponded to the more highly consensus-like sequences in the promoters, i.e., PFT�� cell line the higher probability

of Zur direct binding [20]. Figure 2 The Zur regulatory consensus in γ-Proteobacteria. (a) Original putative Zur binding sites were derived from Panina et al’s study [29]. They were predicted from 13 genes in γ-Proteobacteria including E. coli, Klebsiella pneumoniae, Salmonella typhi, Y. pestis, and Vibrio cholerae, by the comparative genomics analysis [29]. Both coding and non-coding sequences of the above Zur sites were trimmed into 19 bp inverted repeat sequences and then aligned to generate a sequence-logo Suplatast tosilate with a Zur box sequence (AATGTTATAWTATAACATT). (b) A position count matrix was generated as well from the alignment, where each row represented a position and each column a nucleotide. This matrix was subsequently used for the computational pattern matching analysis. Four genes (ykgM, znuC, znuA and

astA) giving the largest score values (Table 1) were picked out for further investigation. The former three genes represent the first genes of three distinct putative operons, namely ykgM-rpmJ2, znuCB and znuA, respectively. ykgM and rpmJ2 encoded ribosomal proteins, while znuA, znuC and znuB encoded the Zn2+ uptake system ZnuABC. The znuCB and znuA operons were transcribed with opposite direction and separated by a short intergenic region (73 bp in length in Y. pestis) [17]. astA is the second gene of the astCADBE operon responsible for the arginine succinyltransferase pathway of arginine catabolism. Zur binds to DNA regions upstream znuA, znuCB and ykgM-rpmJ2 The real-time RT-PCR validated that Zur repressed the first gene of each of the three operons, znuA, znuCB and ykgM-rpmJ2 (Additional file 5). Herein, the DNA regions upstream these first genes (generated as indicated in Fig. 1a) were subjective to EMSA. It was demonstrated that the purified Zur protein bound to each of these potential target promoter regions in a Zur dose-dependent manner in vitro (Fig. 3). Thus, a direct association of Zur with the promoter regions of znuA, znuCB and ykgM-rpmJ2 was detected.