We were interested in the role of the SenX3 RegX3 two component regulatory system during growth under different oxygen tensions. Whole genome mRNA profiling and promoter assays were used to profile expression of the operon and to find genes whose expression was differentially controlled in a regX3 deletion strain.
We looked at expression of the operon; PsenX3 activity was dependent on growth conditions, although the fold-changes were small. However, this could still have physiological relevance, since small changes in either SenX3 or RegX3 levels could affect the amount of phosphorylated regulator and subsequently transcriptional activity of the regulon. Promoter activity did correlate with the transcriptome profiles, since the greatest effect on gene expression was seen when the promoter was activated in static cultures.
The over-expression studies were carried out after 7 d in order to provide a direct comparison to the data from our original study in which we looked at global gene expression in aerobic roller culture after 7 days in the RegX3 deletion strain. Further analysis to compare other conditions would be of benefit, for example looking at gene expression in static or anaerobic conditions after over-expression.
The down-regulation of the cydAB
upstream region promoter activity in static culture is in contrast to our previous data showing up-regulation in aerobic roller cultures 
suggesting that the expression of cydAB
is changed under different oxygen tensions. Roller cultures are generally well aerated, whereas the static cultures will have lower oxygen tensions. Recent work has revealed that cydAB
is transiently up-regulated in the mouse lung during the switch from the acute to chronic phase of infection and that a cydC
mutant has a reduced ability to survive this transition 
. We have previously demonstrated that the RegX3 mutant shows an attenuated phenotype in the mouse model, with an early defect in replication (up to day 30) which is recovered at later stages of infection (day 59) 
. If RegX3 is required for the transient induction and controlled expression of cydB
, then this could, at least partly, explain the observed phenotype. It is interesting that cydB
(but not cydA
) is predicted to be essential under normal growth conditions 
, as it should theoretically only be required in low oxygen environments.
was down-regulated in the over-expressor, suggesting that RegX3 is not a simple positive regulator of expression and that the regulatory system is more complex than our original hypothesis and that there may be more than one regulator of cydAB
expression. DNA-binding of unphosphorylated RegX3 has previously been demonstrated 
, but this only occurs with its own promoter and under the conditions used here, the phosphorylated regulator preferentially bound to the target putative promoters for ald
. Further work to characterize the phosphorylation state of RegX3 in various conditions and the phosphorylation/phosphatase activity of SenX3 in response to these conditions would be helpful.
Expression from Pald
was not significantly altered in the mutants in static culture, although changes in mRNA levels were noted and RegX binding to Pald
was demonstrated, suggesting that it is involved in direct regulation of the gene. Ald
is up-regulated in stationary phase (and upon alanine supplementation) 
suggesting that there are other regulatory elements involved; it is possible that the region cloned for the promoter analysis did not contain these elements. Further work to define the promoter and other regulatory elements under multiple conditions would answer these questions.
The SenX3 system is also involved in regulation of phosphate uptake proteins in response to phosphate concentration; under limiting conditions RegX3-dependent up-regulation of the Pst phosphate transporter system is seen 
. However, the stimulus for SenX3 remains unknown, since the sensor for phosphate depletion under these conditions is not SenX3, but a member of the Pst family. It is interesting to note that two of the genes (gltA
) identified as regulon members are up-regulated in low pH 
, suggesting that ionic concentrations may be a stimulus.
Previous work by Rickman et al. 
identified a conserved PAS domain with the SenX3 protein suggesting that it may have role in sensing redox or oxygen status analogous to the ArcAB system of E. coli
. Two of the genes that we show to be differentially regulated in the regX3
deletion strain, namely cydAB
, are controlled by ArcAB in E. coli
. Therefore, we speculate that the role of the SenX3-RegX3 system is similar and acts to integrate the utilisation of carbon sources through the TCA cycle with defence against the potentially damaging reactive oxygen species generated by aerobic metabolism.