The metabolite 5-aminolevulinic acid (ALA) is essential to all cells, as it is the precursor to tetrapyrroles that include vitamin B12, heme and bacteriochlorophyll. Among bacteria, formation of ALA via the C4 or Shemin pathway, in which succinyl-CoA is condensed with glycine in pyridoxal phosphate-dependent reaction catalyzed by ALA synthase, is limited to the ¿ class of proteobacteria, which include the facultative prototroph Rhodobacter sphaeroides. In fact, these bacteria have two ALA synthase isoenzymes, and relatives encode as many as four ALA synthases. Despite decades of studies, only recently has it come to light that the R. sphaeroides enzymes differ with respect to their sensitivity to feedback inhibition by heme (1). However, understanding the full significance of this finding requires knowledge as to the presence of the two enzymes in the cell. It is also necessary to explain how it is that the performance (growth) of wild type strains whose genomes only encode one enzyme appears to equal that of other wild type strains whose genomes encode both enzymes.
To learn more about the distinctive roles of the isoenzymes, the products of the hemA and hemT genes, lacZ transcription reporter plasmids were used to examine their expression in four wild type strains; 2.4.1, 2.4.9 and KD131, all of which encode both hemA and hemT, and 2.4.3 which has only hemA. It was found that, in all four strains, hemA is induced under anaerobic conditions, but that the induction levels differ. The hemT gene is transcriptionally silent in strain 2.4.1 under all growth conditions, including nitrosative stress, while it is actively transcribed in strains 2.4.9 and KD131, and strongly upregulated when cells are respiring on dimethyl sulfoxide (DMSO) compared to aerobic-dark and phototrophic conditions. The picture that emerged from these studies, together with the different susceptibilities of the enzymes to heme-mediated feedback inhibition, is that the bacteria employ different strategies to ensure that adequate amounts of ALA are available for synthesis of whatever kinds and levels of tetrapyrroles are needed according to growth conditions. In some strains hemA transcription is strongly upregulated in order to compensate for inhibition by heme; in other strains in which hemA transcription is weakly upregulated, the less sensitive HemT enzyme is present, when needed, to augment ALA synthase activity.
Further examination of hemT expression in strain 2.4.9 identified cis-acting sequences, as well as two extracytoplasmic function (ECF) sigma factors that are absent from strain 2.4.1, as being important for hemT transcription. Using electrophoretic mobility shift assays with purified sigma factor proteins it was determined that both ECF-type sigma factors directly transcribe hemT. EMSAs also confirmed that a second gene, which was suggested from transcriptomic data, is transcribed by these sigma factors. It encodes a periplasmic protein that binds C4-dicarboxylic acids, which are then transported into the cell by proteins whose genes are co-transcribed with the solute binding protein. For one of the ECF-type sigma factors evidence of the presence of a redox-active disulfide bond within the sigma factor itself was obtained, which explains, in part, upregulation of hemT transcription under reducing conditions. Since genes encoding a transporter of C4-dicarboxylic acids are transcribed by these sigma factors, the influence of one such compound (succinate) on the activities of the sigma factors was evaluated. The evidence suggests that succinate acts as an activating signal for the anti-sigma factor whose partner sigma factor contains the redox-active disulfide. While the signal for the anti-sigma factor of the second sigma factor remains to be determined, it is clear that the transcriptional activity of that sigma factor is also responsive to changes in cellular redox.