Many bacteria, including V. fischeri, recognize and respond to their environments using two-component regulatory systems (Fig. 2A, reviewed in Stock et al. 2000). These systems are composed of a sensor histidine kinase
Fig. 2. Two-component regulatory systems. A. The phospho-relay in orthodox (top) and hybrid (bottom) two-component systems. Upon detection of signal, phosphate generated from a bound ATP is passed from conserved His to Asp residues until finally being transferred to an Asp in the response regulator, resulting in a response, either altered transcription or protein function. B. RscS is a hybrid sensor kinase. The sensor domain of RscS is composed of two transmembrane helices (TM), a large periplasmic loop and a PAS domain
Fig. 2. Two-component regulatory systems. A. The phospho-relay in orthodox (top) and hybrid (bottom) two-component systems. Upon detection of signal, phosphate generated from a bound ATP is passed from conserved His to Asp residues until finally being transferred to an Asp in the response regulator, resulting in a response, either altered transcription or protein function. B. RscS is a hybrid sensor kinase. The sensor domain of RscS is composed of two transmembrane helices (TM), a large periplasmic loop and a PAS domain protein that recognizes and transmits an environmental signal (through autophosphorylation on a His residue) to a second protein, the response regulator, which (when phosphorylated on a conserved Asp residue) carries out a response. Most frequently, the response consists of a change in gene expression; alternatively, changes in protein activity can result.
Given the changes in environmental conditions that V. fischeri cells experience as they travel from seawater into the LO, it is not surprising that colonization by V. fischeri requires two-component regulators. At least two such regulators are required for efficient initiation of symbiotic colonization: the sensor kinase RscS (Visick and Skoufos 2001) and the response regulator, GacA (Whistler and Ruby 2003). A transcriptional regulator, FlrA, which exhibits limited similarity to response regulators and is required for initiation (Millikan and Ruby 2003) will also be discussed here.
rscS. Mutations in rscS severely reduce the ability of V. fischeri to initiate symbiotic colonization: most animals remain uncolonized following exposure to rscS mutants, although other animals become colonized after a delay of several hours (Visick and Skoufos 2001). These results suggest that mutants are blocked at an early stage of colonization, but that they can occasionally by-pass this block and ultimately achieve what appears to be normal colonization. In culture, rscS mutants do not exhibit defects in growth, motility, or the timing and level of bioluminescence induction, traits known to be important for colonization (Visick and Skoufos 2001). Thus, to date no clues to rscS function have been garnered by phenotypes observed in culture.
The sequence of rscS suggests that it encodes a hybrid sensor kinase similar to ArcB and BvgS (Fig. 2B) (Visick and Skoufos 2001). These proteins contain, in addition to the conserved His residue that serves as the site of autophosphorylation, two additional domains with conserved residues (Asp and His) predicted to be sequentially phosphorylated and that may serve as sites of additional regulation (Fig. 2B). Upon receipt of a colonization signal, RscS is predicted to autophosphorylate and transfer the phosphate to an as-yet-unidentified response regulator, termed RscR, which may regulate genes or activities essential for symbiosis.
What serves as the colonization signal, and how is it detected by RscS? Clearly, many possibilities exist, and include in addition to bacterially produced molecules and seawater signals, components of the LO mucus, cell surface signals and nutrients. Determining the portion of RscS responsible for detecting the colonization signal will advance our understanding of symbiotic signal exchange. In many cases, the amino terminal periplasmic portion of sensor kinase proteins receives the environmental signal (Stock et al. 2000). For example, Salmonella PhoQ detects Mg2+ in the environment through its periplasmic domain; binding of Mg2+ to this domain results in a conformational change and inactivation of the response regulator PhoP (Vescovi et al. 1997). RscS is predicted to possess a periplasmic domain of ~200 residues (Visick and Skoufos 2001); the large size of this region suggests it may play a role in RscS function, possibly signal detection.
In addition to a potential periplasmic signaling domain, RscS contains a second input domain, known as PAS. In other proteins, PAS detects signals such as small ligands, or changes in light levels, oxygen concentration or redox potential (Taylor and Zhulin 1999). Whether the PAS domain contributes to signal detection by RscS during colonization remains unknown. However, the transition from seawater to the nutrient-rich LO could plausibly affect the energy status of the V. fischeri cells thereby altering their redox potential or oxygen concentration, which could be sensed by the PAS domain. Thus, investigations of the PAS and periplasmic domains of RscS will be fruitful for exploring bacteria-host interactions. Perhaps each domain detects a distinct condition, allowing RscS to integrate multiple signals from the squid environment to regulate the initiation of colonization.
What is the identity of the cognate response regulator, RscR, and what genes or proteins are controlled by the RscS/R regulatory system? In many cases, the genes for sensor kinases and their cognate response regulators are linked on the chromosome, and in some cases, genes controlled by the regulators are also nearby. This is not the case for rscS and the gene encoding its response regulator. The advent of the V. fischeri genome sequencing project (http://ergo. integratedgenomics.com/Genomes/VFI), has made it possible to use bioinformatics to look for RscR. Using the sequences of known regulators, we have searched and identified about 40 response regulators (Hussa and Visick, unpubl. data). At least 14 appear unlinked to sensor kinase genes, and thus represent the best candidates for RscR. Current work is aimed at mutagenizing these candidates and asking whether any mutants exhibit rscS-like colonization defects. If rscR encodes a DNA binding protein, then newly available DNA microarrays will be used to explore the regulon controlled by RscS/R. Identification of the targets of RscS/R regulation may also suggest a role for this regulon in symbiosis initiation. Once a target(s) of these regulators is identified, experiments aimed at identifying the colonization signal can be formulated.
gacA. In a number of pathogenic bacteria, the two-component system GacS/A regulates expression of virulence and host association traits, such as production of exoenzymes in Pseudomonas spp. (Heeb and Haas 2001) and motility in Salmonella (Goodier and Ahmer 2001). V. fischeri GacA also plays a role in host association. Mutants defective for gacA exhibit severe defects in initiating colonization: only about 10% of animals become colonized and those animals that become colonized exhibit a nearly 100-fold reduction in the level of colonization (i.e., the number of bacteria residing in the LO) (Whistler and Ruby 2003). The role of GacA is likely to be quite complex. In culture, it is associated with a number of phenotypes known to be important for symbiosis, including motility, nutrient acquisition, siderophore activity and luminescence (Whistler and Ruby 2003). The global control of disparate traits, all of which contribute to host-association, highlights the importance of such regulators in the evolution of symbiotic associations. As with RscS/R, neither the signal nor the gene/protein targets for GacA/S are known. Identification of targets of GacA regulation, possibly through DNA microarray experiments, will help elucidate the role of this regulator in symbiosis and potentially reveal previously unknown traits important for host-microbe interaction.
flrA. FlrA, a transcription regulator with limited sequence similarity to response regulators, functions as a master regulator of flagellar biosynthesis (Millikan and Ruby 2003). Given the absolute requirement for motility in symbiotic initiation, the requirement for FlrA seems straightforward as mutations lead to a lack of flagella. However, complemented flrA mutants showed restored motility but not normal colonization: initiation was delayed and the level of colonization at 48 h post-inoculation was reduced by 10-fold.
One explanation for the above result is that the timing and level of flagellar biosynthesis are critical for optimal initiation and colonization and these characteristics were not properly restored in the complemented strains. In support of this hypothesis, hyper-motile (hyper-flagellated) V. fischeri mutants also exhibit severe delays in initiating colonization and defects in the level of colonization 24 h post-inoculation (Millikan and Ruby 2002). Alternatively, an equally plausible explanation is that FlrA controls genes other than those involved in flagellar biosynthesis (Millikan and Ruby 2003) that are also required for colonization.
Several non-flagellar genes appear to be regulated by FlrA (Millikan and Ruby 2003). One gene that appears to be repressed by FlrA, hvnC, encodes a protein related to HvnA and HvnB, two secreted NAD+ glycohydrolases found in V. fischeri. However, neither hvnA nor hvnB appears necessary for colonization (Stabb et al. 2001); therefore, the relevance of FlrA-mediated regulation of hvnC is unclear. A second putative FlrA-repressed gene is homologous to V. cholerae kefB. In E.coli, KefB is a potassium efflux protein that is important for protecting cells from toxic metabolites during growth on a poor carbon source (Ferguson et al. 2000). Possibly, the V. fischeri KefB homolog provides protection from a LO-specific toxin.
Are FlrA-repressed genes relevant to symbiotic colonization? FlrA-controlled flagella, which are required for initiation, become dispensable to colonized bacteria. Thus, a switch in flagella gene transcription may be coordinated with induction or repression of non-flagellar genes through FlrA. The regulation of FlrA itself may be at the level of transcription, analogous to cAMP-CRP mediated control of the master flagellar regulators flhDC in E. coli (Soutourina et al. 1999). In addition, the limited similarity of FlrA to response regulators suggests its activity could be regulated via phosphorylation by a sensor kinase. Future work will likely focus on determining whether FlrA itself is transcriptionally controlled, whether overexpression of FlrA during colonization affects the level or timing of transcription of putative FlrA-controlled genes and whether such genes themselves promote (or interfere with) colonization.
Was this article helpful?