Figure 3. Transcription start sites (TSSs) of the gpmA, eno, pykA, pykF, pdhR and aceE glycolytic genes and operons were determined using a modified 5′RACE methodology, as described in materials and methods.

The cDNAs in the bands were extracted and their nucleotide sequences were determined to their 5′ends to allow TSSs determinations. When multiple PCR bands were present in one gel, these bands were purified and their nucleotide sequences were determined. In some cases, ribosomal or tRNA were detected. In most cases, no specific DNA sequences were obtained indicating nonspecific amplified PCR products (see a precise explanation in the text for each subfigure; the nucleotide sequences of the different cDNA are presented in figure S1 and table 1). The consensus nucleotide sequence of the −10 region for the σ³⁸ sigma subunit (8 nucleotides, framed in the figures) described by Weber et al [5], was utilized for detecting possible σ³⁸ recognition sites and a 10 nucleotides −10 σ⁷⁰ consensus sequence was utilized for detecting possible σ⁷⁰ binding sites. Numbers in parenthesis indicate references in which previously reported TSSs have been described: Vassinova et al [46], Shimada et al [47], Spencer and Guest [50], Quail et al [51], Cunningham et al [52].