Our laboratory investigates the mechanisms by which pathogenic and commensal bacteria modify their gene expression patterns so they can survive and proliferate within host tissues and in abiotic environments. We have focused on the mechanisms utilized by the gastroenteritis- and typhoid fever–causing Salmonella enterica, the bubonic plague agent Yersinia pestis, and the human gut commensal Escherichia coli.
Our research program can be divided into three general areas: (1) the signal transduction pathways by which bacteria detect and integrate multiple signals into a cellular response, (2) the molecular mechanisms by which a regulatory protein or signal elicits distinct responses from coregulated targets, and (3) the genetic basis for the phenotypic differences that distinguish closely related bacterial species.
Signal Transduction by Two-Component Regulatory Systems
Bacteria often rely on two-component systems to modulate their gene expression programs when there is a change in their surroundings. For example, Salmonella experiences a variety of environments during infection of animal hosts, including an acid pH in the stomach, low oxygen and high osmolarity in the small intestine, and for those Salmonella serovars that cause systemic disease, the nutrient-poor and microbicidal-compound-rich milieu of the macrophage phagosome. Because it can replicate both inside and outside host tissues, Salmonella has the means to ascertain whether it is present in an intracellular or extracellular environment and to promote expression of the appropriate set of genes.
Two-component systems typically consist of an integral membrane sensor and a DNA-binding regulator. Sensors respond to the presence of their specific signal(s) by modifying the phosphorylation state of their cognate regulatory proteins, which bind to their DNA targets only when phosphorylated. We have been investigating the PhoP/PhoQ two-component system, a governor of virulence and the adaptation to Mg2+-limiting environments in several bacterial species, and the PmrA/PmrB system, the major controller of modifications of the outermost layer of the bacterial outer membrane. The PhoQ protein is a Mg2+ sensor that promotes the phosphorylated state of the regulator PhoP when experiencing low Mg2+, whereas the sensor PmrB responds to the presence of Fe3+, Al3+, or acid pH by promoting the phosphorylated state of the PmrA protein.
We determined that, surprisingly, the level of phosphorylated regulator changes dynamically even under constant inducing conditions, whereby activation of a two-component system results in a surge of activity before the system reaches steady state. The surge of the PhoP/PhoQ system is critical for Salmonella's ability to cause a disease because a Salmonella strain engineered to reach the same steady state monotonically (i.e., without the surge) was attenuated for virulence. These results demonstrate that organisms care not only about the particular activation level that is achieved in response to a given signal but also about how that activation level is reached.
Differential Gene Control
DNA-binding regulatory proteins often control expression of more than one gene, which may encode products that are required in different amounts. This raises a question: How does a given regulator achieve differential expression of its coregulated targets? To address this question, we developed GPS (gene promoter scan), a computational method that discriminates among coregulated promoters by simultaneously considering multiple cis-acting promoter features. The use of GPS has enabled the discovery of novel targets of regulation of the PhoP protein, as well as regulatory interactions that could not be uncovered using previous approaches.
In addition to controlling genes directly, the PhoP protein utilizes a variety of mechanisms to regulate gene expression indirectly, via other systems and proteins. On the one hand, the PhoP-activated PmrD protein activates the PmrA protein when the signals for the PmrB protein are absent. This allows genes regulated by the PmrA protein to be expressed in low Mg2+, which is the signal activating the PhoQ protein. On the other hand, the PhoP-activated IraP protein promotes an increase in the level of the alternative sigma factor RpoS, thereby promoting expression of RpoS-dependent genes when bacteria experience the low-Mg2+ signal that activates the PhoP/PhoQ system. We refer to PmrD and IraP as connector proteins because they connect the signal sensed by one system with the output mediated by a different system. The kinetic and quantitative properties of the regulatory circuits mediated by connector proteins are different from those in which the PhoP protein functionsas a direct activator.
Certain PhoP-activated genes are also regulated after the initiation of gene transcription, which provides an additional means of differential gene control. For example, we determined that the leader region of the PhoP-activated Mg2+ transporter mgtA gene functions as a Mg2+ sensor that determines whether transcription continues into the mgtA-coding region and the MgtA protein is made. The Mg2+ sensor in the mgtA mRNA enables bacteria to exert separate control over determinants governing cytoplasmic Mg2+ homeostasis (i.e., a Mg2+ transporter) versus extracytoplasmic Mg2+ homeostasis (i.e., enzymes that modify Mg2+-binding sites in the cell envelope) when they are all transcriptionally regulated by the PhoP/PhoQ system.
The Genetic Basis for the Phenotypic Differences between Related Bacterial Species
Closely related bacterial species often display different responses to a given environment. This can be due to the presence/absence of species-specific genes, the differential regulation of the same set of genes, or allelic differences between homologous genes. We have been investigating the genetic basis for the differences that exist between the pathogenic species S. enterica and Y. pestis as well as the gut commensal bacterium E. coli. We are particularly interested in the differences that originate from the differential gene regulation of shared genes, as well as the mechanisms by which newly acquired genes are integrated into ancestral regulatory circuits.
E. coli has PhoP/PhoQ and PmrA/PmrB systems that are closely related to those present in Salmonella and that respond to the same signals. However, the PmrD protein of E. coli is highly divergent and thus unable to mediate the low-Mg2+ activation of PmrA-regulated genes mediated by the Salmonella PmrD protein. Furthermore, E. coli and Salmonella encode highly conserved IraP proteins that can substitute for one another to promote activation of RpoS-regulated genes. The E. coliiraP gene does not, however, activate RpoS in low Mg2+ because its iraP promoter lacks the PhoP-binding site that enables PhoP-dependent iraP expression in Salmonella. Even though it lacks the pmrD gene, Y. pestis can express PmrA-activated genes in low Mg2+ because it harbors binding sites for the transcription factor PhoP in the targets of PmrA control. This indicates that related bacterial species often adopt distinct regulatory circuits to express shared genes. While qualitatively similar, these circuits often display quantitatively different outputs.
