SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Homologues of the E. coli chemotaxis (Che) signal transduction pathway are present in nearly all motile bacteria. Although E. coli contains only one Che cascade, many other bacteria are known to possess multiple sets of che genes. The role of multiple che-like gene clusters could potentially code for parallel Che-like signal transduction pathways that have distinctly different input and output functions. In this study, we describe a che-like gene cluster in Rhodospirillum centenum that controls a developmental cycle. In-frame deletion mutants of homologues of CheW (ΔcheW3aand ΔcheW3b), CheR (ΔcheR3), CheA (ΔcheA3) and a methyl-accepting chemotaxis protein (Δmcp3) are defective in starvation-induced formation of heat and desiccation resistant cyst cells. In contrast, mutants of homologues of CheY (ΔcheY3), CheB (ΔcheB3), and a second input kinase designated as CheS (ΔcheS3) result in cells that are derepressed in the formation of cysts. A model of signal transduction is presented in which there are three distinct Che-like signal transduction cascades; one that is involved in chemotaxis, one that is involved in flagella biosynthesis and the third that is involved in cyst development.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Rhodospirillum centenum is a bacterium that exhibits a complex life cycle. When grown in liquid medium, this species predominately exists as vibroid-shaped swim cells that contain a single polar flagellum. However, when grown in viscous or agar-solidified media, R. centenum differentiates into rod-shaped swarm cells that contain both a polar flagellum as well as numerous lateral flagella (Ragatz et al., 1995; Jiang et al., 1998). In addition to swarm cell differentiation, these cells are also capable of forming clusters of desiccation resistant cysts when starved for nutrients (Favinger et al., 1989). Cyst cell development is a time- and energy-intensive process that requires at least 48 h for formation (Berleman and Bauer, 2004). During the formation of cysts, R. centenum produces large intracellular storage granules of the polymer poly-hydroxybutyrate (PHB), alters cell shape and loses motility by ejection of flagella. Finally, an exine protective outer coat is produced that typically surrounds four to eight cells that provide extensive resistance to desiccation. Cyst formation is a trait that has been observed in several other bacterial species including Azotobacter and Azospirillum (Stevenson and Socolofsky, 1966; Sadasivan and Neyra, 1987).

The decision to undergo cyst cell development presumably requires complex coordinate monitoring of a variety of cellular and environmental signals. However, in comparison to what is known about the regulation of endospore formation by Bacillus subtilis, and myxospore formation by Myxococcus xanthus, little is known about the molecular mechanism of cyst cell formation in any species. Recently, we have demonstrated that R. centenum mutants can be readily isolated that are defective in the regulation of cyst cell development (Berleman et al., 2004). Surprisingly, among a collection of mutants that constitutively form cysts, five were independent disruptions in a cluster of genes that have significant homology to chemotaxis (che) proteins. Chemotaxis signal transduction pathways normally mediate movement of cells up or down chemical gradients using a two-component system that interacts with multiple chemoreceptors (MCPs) (Falke et al., 1997; Ames et al., 2002). In typical chemotaxis signalling pathways, ligand occupancy of the MCPs is transduced to the histidine kinase CheA that ultimately controls the phosphorylation state of CheY. CheY∼P then interacts with the flagellar motor to alter flagellar rotation (Welch et al., 1993).

Even though Che-type signal transduction pathways normally control motility in response to environmental signals, recent evidence suggests that Che-like signal transduction pathways may also control other cellular processes. For example, the cyanobacterium Synechocystis contains Che-like homologues that regulate synthesis of type IV pili (Bhaya et al., 2001). In M. xanthus, the che3 operon is essential for regulating the timing of fruiting body formation (Kirby and Zusman, 2003). When starved for nutrients, several mutants in the M. xanthus che3 cluster form fruiting bodies more rapidly and at lower cell densities than do wild-type cells. These che3 mutants will also induce fruiting body formation on rich media, but the number of viable myxospores formed under either condition are greatly reduced with respect to wild-type cells. In R. centenum, a che-like signal transduction cascade coded by the che2 operon has previously been shown to control synthesis of polar and lateral flagella (Berleman and Bauer, 2004). Thus, some species appear to utilize the environmental sensing capabilities of Che-like signal transduction pathways to control non-chemotactic processes. In each of these cases, these species have multiple Che-like pathways with at least one cascade dedicated to regulating cell movement.

Previous studies have shown that R. centenum contains a chemotaxis operon, called the che1 operon, which is essential for the control of chemotactic and phototactic responses (Jiang and Bauer, 1997; Jiang et al., 1997). Disruptions in the che1 operon result in phenotypes similar to that observed with E. coli homologues. For example, disruption of the R. centenum cheA1 gene results in a smooth swimming phenotype whereas disruption of cheB1 results in a constitutive tumbly phenotype. We have also shown that R. centenum contains a second che-like operon, che2, which has no involvement in chemotaxis but is essential for regulating flagella expression. Disruption of genes in the che2 operon lead to either aflagellate or hyper-flagellate cells (Berleman and Bauer, 2005). In this study, we demonstrate that R. centenum has a third che-like signal transduction pathway coded by the che3 operon. Mutational analysis indicates that the che3 operon from R. centenum has a role in controlling the regulation of cyst cell development in this species.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Identification of a third che-like operon in R. centenum

A previous screen for mutations that derepress induction of cyst development resulted in the isolation of mutants that contained mini-Tn5 insertions in numerous signal transduction genes, including genes with homology to chemotaxis genes cheA, and cheB, and an insertion in an mcp gene that codes for a chemoreceptor (Berleman et al., 2004). However, that study did not establish if the mcp/che-like genes are part of a complete signal transduction cascade, whether the mini-Tn5 disruptions also affected motility, or if there may be polarity effects caused by linkage of these genes in the chromosome. To establish whether these che-like genes are indeed a part of a complete Che-like signal transduction cascade, we undertook sequence analysis of regions flanking the mini-Tn5 insertions. The result of this analysis (Fig. 1), demonstrates the presence of a third chromosomally encoded che-like gene cluster (che3) that encodes a complete Che-like signal transduction cascade. All of the ‘typical’ chemotaxis sensory transduction components are coded by the che3 gene cluster including an MCP (mcp3) that contains two putative membrane spanning domains, the methylation and demethylation proteins CheR (cheR3) and CheB (cheB3), two CheW linker proteins (cheW3a and cheW3b), a CheA-CheY hybrid protein (cheA3) and a CheY-like output protein (cheY3). The last gene in this cluster also codes for a novel response regulator-sensor kinase hybrid, cheS3. As diagrammed in Fig. 1B, the predicted CheS3 gene product contains two amino-terminal response regulator receiver domains, the first of which contains all the conserved residues essential for phosphorylation, but the second receiver domain lacks the conserved aspartate residue site of phosphorylation. The receiver domains are followed by a low complexity region, a PAS domain, and a kinase domain that is most similar to the recently discovered HWE subfamily of histidine kinases (Karniol and Vierstra, 2004). The genes in the che3 cluster are in a tight linkage order with the majority having no flanking intervening intergenic regions and instead having overlapping stop and start codons, including cheS3 that overlaps with cheB3.

image

Figure 1. A. Genetic organization of the che3 operon. This operon contains several genes homologous to the E. coli che system. B. Of note are the unusual response regulator-sensor kinase hybrid, CheS3, which is not part of the CheA subfamily of histidine kinases. CheS3 has 2 N-terminal receiver domains, one that contains the conserved aspartate residue essential for phosphorylation, the other does not. The C-terminal portion contains a PAS/PAC motif and an HWE histidine kinase motif. C. RT-PCR on deletion strains using a probe to the final gene in the cluster, CheS3. Total RNA was isolated from the given strains. Presence of a band indicates the presence of the cheS3 gene mRNA transcript. Transcript is detectable in all the strains except for ΩcheY3. This strain contains a polar ∧Spectinomycin cassette inserted in the first gene of the cluster, cheY3. The control is a probe to the last gene of the che1 operon.

Download figure to PowerPoint

To determine if the entire che3 gene cluster is expressed in a single transcript, we constructed a polar Ω-spectinomycin resistance insertion in the first gene in the cluster, cheY3. We then used reverse transcription polymerase chain reaction (RT-PCR) to assay for expression of the last gene in the che3 cluster using primers that amplify a portion of the cheS3 mRNA. As a control for mRNA quality and PCR amplification fidelity, we also used a primer set specific to the last gene of the che2 operon. As shown in Fig. 1C, the che2 operon control generated an RT-PCR product in both wild type and ΩcheY3 strains demonstrating fidelity of the mRNA preparations. In contrast, primers that amplify the cheS3 transcript generated an amplification product from mRNA preparations extracted from wild-type cells but not from mRNA from the ΩcheY3 strain. This indicates that the che3 gene cluster is transcribed as an operon that extends from cheY3 through cheS3.

Construction of non-polar disruptions of che3 genes

To genetically characterize the role of individual genes in the che3 operon, we constructed non-polar in-frame deletions of each of the individual che3 genes, as well as a deletion of the entire che3 gene cluster. Deletions were made using a sucrose selection technique that creates an in-frame gene deletion with no antibiotic resistance marker (Masuda and Bauer, 2004). To confirm that the constructed deletions are non-polar, we checked for the presence of cheS3 transcripts using RT-PCR in each of the che3 gene deletions. As shown in Fig. 1C, the cheS3 transcript is present in all of the in-frame deletion mutants (ΔcheY3, ΔcheW3a, ΔcheR3, ΔcheW3b, ΔMCP3, ΔcheA3 and ΔcheB3) with the exception of deletion of cheS3 itself. Thus, none of the individual upstream in-frame deletion constructions are polar on cheS3 transcription.

Previously, we reported that R. centenum vegetative cells form shiny convex colonies when grown on nutrient-rich CENS medium for 48–72 h. This is contrasted by hyper-cyst mutant cell lines that derepress cyst formation, resulting in the formation of large ridged colonies (Berleman et al., 2004). As shown in Fig. 2A, in-frame deletions of genes in the che3 operon lead to striking differences in colony morphology when grown on CENS medium for 48 h. Strains ΔcheY3, ΔcheS3, Δche3 and to a lesser extent, ΔcheB3, exhibit a prominent hyper-cyst colony phenotype with large, ridged colonies typical of strains that contain elevated amounts of cyst cells. However, this phenotype is not observed with all che3 mutants because in-frame deletions of Δmcp3, ΔcheW3a, ΔcheW3b, ΔcheR3 and ΔcheA3, form shiny, convex colonies that are indistinguishable from wild-type colonies grown under similar conditions.

image

Figure 2. Colony morphology of che3 deletions on (A) CENS and (B) CENBA media. Growth of wild-type R. centenum on nutrient-rich CENS medium for 3 days produces shiny convex colonies containing only vegetative cells. In contrast, several che3 strains show the striated features of cyst-forming colonies even on rich medium. When grown on cyst-inducing CENBA medium, wild-type forms the ridged colonies indicative of cyst cell induction. Although most che3 mutants form cyst colonies similar to wild type, several display a phenotype more similar to vegetative cell colonies.

Download figure to PowerPoint

Deletion of cheY3, cheS3, cheB3 and the entire che3 operon, lead to constitutive induction of the cyst developmental cycle

We addressed whether the ridged colonies formed by Δche3, ΔcheY3, ΔcheS3 and ΔcheB3 contain mature cysts by undertaking microscopic and desiccation resistance analysis as described previously (Berleman and Bauer, 2004; Berleman et al., 2004). Microscopic analysis of cells harvested from colonies after 48 h of growth in rich CENS medium shows only the presence of vegetative cells from the wild-type parent strain (Fig. 2A). In contrast, microscopic analysis of strains ΔcheY3, ΔcheS3, Δche3 from the same rich growth medium shows a mixture of vegetative cells, as well as large clusters of cyst cells (Fig. 3C, Q and S respectively). Strain ΔcheB3 also exhibits a mixture of vegetative and cyst cell types with an interesting difference in that the average cyst cell size is at least two to three times that observed with wild-type cells (Fig. 3O). The ΔcheB3 strain also has a smaller number of cyst cells per cluster.

image

Figure 3. Microscopic characterization of che3 deletion mutants. Wet mounts were prepared from colonies after 3 days growth on CENS medium (A, C, E, G, I, K, M, O, Q, S) or 5 days on CENBA medium (B, D, F, H, J, L, N, P, R, T). Phase-contrast analysis of che3 mutants shows that hyper-cyst strains which display a ridged colony phenotype under both conditions also show the constitutive presence of cyst cells. Likewise, strains which appear vegetative on both CENS and CENBA media consist mostly of vegetative cells and show reduced numbers of cyst cells, as well as intermediate stages in cyst cell formation.

Download figure to PowerPoint

To quantify the amount of mature cyst cell formation, wild type and mutant strains were subjected to desiccation resistance analysis. R. centenum vegetative cells are rapidly killed (no detectable survivors) when exposed to 24 h of desiccation. In contrast, mature cyst cells survive up to several months of desiccation (Berleman and Bauer, 2004). Thus, we quantified the percent of cells that formed mature cysts after 48 h of growth on CENS plates by assaying the total viable cell count as well as the number of cells that are resistant to desiccation. The results of this analysis demonstrate that only 0.002% of the cells in the wild-type colony survive 3 days of desiccation (Fig. 4A). In contrast, the Δche3, ΔcheY3, ΔcheS3 and ΔcheB3 strains have greatly increased desiccation resistance at levels that reach as high as ∼20% of the total population for ΔcheY3. Indeed, the amount of cyst production by these hyper-cyst-forming mutants is a remarkable 100 to 10 000-fold higher than wild-type cells when grown for 3 days on nutrient rich CENS medium.

image

Figure 4. Desiccation resistance analysis of che3 deletions on (A) CENS and (B) CENBA medium. Colonies harvested after 3 days growth on CENS plates and 5 days growth on CENBA plates were resuspended in phosphate buffer and sonicated briefly to disperse cells. Resuspensions were pipetted onto 0.45 µm filters, and subjected to desiccation for 4 days at 42°C. Quantification of the level of mature cyst cell formation was determined by comparing the number of colony-forming units for each strain before and after desiccation. Hyper-cyst strains show constitutively high levels of desiccation resistance, whereas cyst-deficient strains show reduced resistance under both conditions tested.

Download figure to PowerPoint

The ability of these hyper-cyst mutant cell lines to produce more cysts then wild-type cells is also observed when cells are grown on minimal medium containing butyrate as a carbon source (CENBA medium). When wild-type cells are grown on nutrient poor butyrate medium, approximately 20% of the cells undergo differentiation onto cysts as measured both microscopically (Fig. 3B) and by resistance to desiccation (Fig. 4B). However, with the hyper-cyst che3 mutants, such as ΔcheY3 (Fig. 3D), the percentage of cyst cells is qualitatively higher than that observed with wild-type cells (Fig. 3B) when grown on butyrate CENBA medium. As shown in Fig. 3, there is also a distinct increase in the size of cyst cell clusters with ΔcheY3, ΔcheS3 and Δche3 mutants forming large multi-cell cyst clusters (> than 16 cyst cell) compared with the more typical four to eight cyst cell clusters observed with the wild-type strain. When measuring desiccation resistance, butyrate grown hyper-cyst cell lines have a 1.5- to fourfold elevation in cyst production relative to that observed with butyrate grown wild-type cells (Fig. 4B). Indeed, the ΔcheY3 mutant colonies consist primarily (∼75%) of desiccation resistant cyst cells. Thus, these mutants not only induce cyst formation under nutrient rich conditions where wild-type cells normally do not undergo cyst formation, but they also produce significantly elevated amounts of cysts relative to wild-type cells when grown under nutrient poor butyrate growth conditions.

Strains ΔcheW3a, ΔcheR3, ΔcheW3b, Δmcp3 and ΔcheA3 are defective in cyst development

As discussed above, individual disruptions of che3 genes, ΔcheW3a, ΔcheR3, ΔcheW3b, Δmcp3 and ΔcheA3 do not exhibit ridged colonies on butyrate minimal medium indicative of cyst formation. Instead, these mutants constitutively display ‘vegetative’-like convex colonies (Fig. 2A and B). To determine if the ΔcheW3a, ΔcheR3, ΔcheW3b, Δmcp3 and ΔcheA3 strains are defective in cyst development, we microscopically analysed cell morphology with wet mounts prepared from these cell lines grown on cyst inducing butyrate medium. In the che3 strains that show a deficiency in the ability to form ridged colonies, there are no observable cyst cells (Fig. 3). This is contrasted by cells from butyrate grown wild-type colonies that exhibited ∼20% cyst cells (Fig. 3B). Furthermore, an intermediate stage of encystment appears to form in these cyst deficient strains, particularly in the Δmcp3 mutant, which consists primarily of oblong cells that contain large phase-bright intracellular PHB granules (Fig. 3L). PHB accumulation  is  the  first  observable  step  in  butyrate-induced cyst-cell formation (Berleman and Bauer, 2004). Later stages of cyst formation, including a spherical shape and production of the refractile outer coat are not observed in this strain during the 5 days period of incubation on butyrate medium that normally leads to formation of cysts in wild-type cells.

Cells from cyst-deficient che3 mutants were also analysed for desiccation resistance to quantify the number of cyst cells that form after 3 days incubation on rich CENS medium or after 5 days incubation on cyst inducing butyrate minimal medium. When cells are harvested from rich CENS medium, strains ΔcheW3b, ΔcheA3 and Δmcp3 produce 0.0005%, 0.0009% and 0.0007% cyst cells respectively (relative to the total cell count). These values are two- to fourfold lower then the 0.002% cyst cells observed in wild-type colonies. Interestingly, strains ΔcheW3a and ΔcheR3 have a modest fivefold increase in cyst production (∼0.01% of total cells are cysts) over the level of cyst production observed with wild-type cells. When these cells were incubated for 5 days on cyst inducing butyrate minimal medium, ∼20% of the cells in the wild-type colony were cysts. In contrast, the ΔcheW3a, ΔcheR3, ΔcheW3b and ΔcheA3 strains formed colonies that contained only 1–3% of the population is mature cysts (∼1/6 that observed from wild-type colonies). The Δmcp3 strain was remarkably low with a scant 0.009% of the total viable cells being cysts. This is a 1000-fold reduction over that observed with wild-type cells.

Prolonged incubation of the cyst defective strains ΔcheW3a, ΔcheR3, ΔcheW3b, Δmcp3 and ΔcheA3 strains for long periods of time (over 2–3 weeks) on butyrate minimal medium does lead to production of microscopically observable cyst cells and there is also a gradual increase in the overall desiccation resistance of the population during extended incubation. Although none of these mutants produce the number of cyst cells that is close to that observed with wild-type cells during extended incubation on butyrate medium (data not shown). Thus, the cyst defective strains are not deficient in essential metabolic or biosynthetic pathways required for constructing a mature cyst. Rather, these mutants are defective in regulating the timing and efficiency of cyst cell development relative to that observed with wild-type cells. Thus, the Che-like signal transduction cascade appears to be essential for the regulation cyst cell induction, with some che3 mutants resulting in increased induction of cyst development and other che3 mutants resulting in reduced induction of cyst development relative to that of wild-type cells.

Motility and chemotactic proficiencies of che3 mutants

To determine if che3 mutants also affect the ability of R. centenum to respond to chemical gradients, we analysed these strains with soft agar chemotaxis assays. Although many of these strains have a large portion of cells in a non-motile cyst cell state, phase-contrast microscopic analysis shows that the vegetative cell population is still actively motile in the majority of these strains. To enrich for vegetative cells, che3 mutant strains were grown photosynthetically in CENS liquid medium, which suppresses encystment even in hyper-cyst strains, to levels below 5% of the total population (data not shown). As shown in Fig. 5, with the exception of ΔcheB3, ΔcheR3 and ΔcheA3, all of the che3 mutant strains form chemotactic rings similar to wild type, indicating that the majority of these deletions have no direct effect on chemotaxis. Additional analysis of tethered cells demonstrates that they retain the ability to change direction of rotation in response to step down in light intensity indicating that phototaxis also remains intact in these mutant cell lines (data not shown). Microscopic examination of ΔcheB3, ΔcheR3 and ΔcheA3 indicates that greater than 99% of these cells are non-motile which would inhibit their ability to form a chemotactic ring.

image

Figure 5. Soft-agar chemotaxis assays on che3 deletion mutants. Overnight R. centenum cultures were washed and pipetted in 3 µl aliquots onto 0.25% agarose minimal medium with 20 mM pyruvate supplied as a chemoattractant. Analysis of (A) wild type, ΔcheY3, ΔcheW3a, ΔcheR3, ΔcheW3b, Δmcp3 and (B) wild type, ΔcheA3, ΔcheB3, ΔcheS3, Δche3 and cheA1- strains indicates that most of these strains are chemotactic similar to wild type. The only exceptions are the smooth swimming cheA1- mutant and also ΔcheR3, ΔcheB3 and ΔcheA3.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Che proteins constitute a well-characterized signal transduction cascade that regulates bacterial movement up and down chemical gradients. Most motile bacteria contain at least one Che-like signal transduction cascade that is dedicated to the control of motility, and there are numerous examples where organisms contain multiple (up to eight) Che-like signal transduction cascades (Szurmant and Ordal, 2004). In most species where there is more then one set of che-like genes, there is little known about what role these additional che genes perform. However, there are a few cases in which multiple Che-like signal transduction cascades appear to have very different roles in the same species. For example, in Pseudomonas aeruginosa one Che-like signal transduction cascade controls the direction of flagella rotation and the other controls pili-mediated twitching motility (Kato et al., 1999; Whitchurch et al., 2004). In M. xanthus, there are eight clusters that code for che-like genes (Kirby and Zusman, 2003). The frz che-like gene cluster controls the frequency of gliding motility reversal (McBride et al., 1989), the dif che-like cluster controls EPS synthesis that results in hyperpiliated cells (Yang et al., 2000) and a third che-like cluster (che3) appears to control entry in to the myxospore developmental pathway and development specific gene expression (Kirby and Zusman, 2003).

In R. centenum, we have genetically identified three che-like operons that regulate distinct processes (Fig. 6) (Jiang et al., 1997; Berleman et al., 2004). The che1 operon codes for components of a typical Che signal transduction cascade that is dedicated to the control of chemotaxis and phototaxis with che1 mutants chemotactically and phototactically deficient (Jiang and Bauer, 1997; Jiang et al., 1997). We have recently reported the presence of a second complete Che-like signal transduction cascade (Che2) that is involved in controlling lateral and polar flagella biosynthesis in R. centenum (Berleman and Bauer, 2005). In this study, we have characterized a third Che-like signal transduction cascade from R. centenum that contains a complete set of signal transduction components comprised of a membrane bound MCP receptor, CheR-like and CheB-like methyltransferase and methylesterases, a CheW-like linker protein, a CheA kinase, and a CheY-like output protein (Fig. 6). There is also a unique response regulator-sensor kinase hybrid, CheS3, that is also involved in this developmental signal transduction cascade.

image

Figure 6. Model for the role of the R. centenum che3 signal transduction pathway.

Download figure to PowerPoint

Interestingly, mutations in all of the three known R. centenum che-like gene clusters yield classic opposing phenotypes that are typically observed with mutations in a chemotactic signal transduction pathway. For example, mutations in cheA1 are smooth swimming while mutations in cheY1 are tumbly, mutations in cheA2 are hyper-flagellated while mutations in cheY2 are non-flagellated, and finally, mutations in cheA3 are cyst-deficient while mutations in cheY3 are hyper-cyst forming. Thus, all indications suggest that the che3 signal transduction cascade functions analogous to that of a classic motility-based chemotaxis cascade complete with MCP receptors that are apparently methylated and demethylated by analogous CheR and CheB-like adaptation proteins. What is clearly different in the three distinct R. centenum Che-like cascades are the type of cellular responses that they control.

A major benefit for utilizing Che-like signal transduction cascades is the presence of an MCP as an input. Methylation and demethylation of MCPs by CheR-like and CheB-like methyltransferase and methylesterases allow the signal transduction pathway to be tuned to differing concentrations of input ligand (Springer and Koshland, 1977; Stock and Koshland, 1978). It also provides the cell with a molecular memory such that the current environmental status can be compared with the recent past to determine if conditions are improving or deteriorating for the cell. Recent evidence of bacteria utilizing Che-like signal transduction pathways to regulate processes other than chemotaxis indicates that these processes may also function best, when regulated temporally to respond to a changing environment. The fact that some of the che3 mutants are somewhat blocked at an early stage of cyst cell formation, such as Δmcp3, indicates that the che3 pathway may be used as a checkpoint after PHB accumulation to ensure that environmental conditions still warrant formation of cysts.

One interesting difference of the Che3 signal transduction cascade, from that of a classic motility-based chemotactic cascade, is the presence of the additional input sensor kinase CheS3. This kinase has a typical response regulator domain at the amino terminus, followed by a partial response regulator domain, a PAS domain and finally a histidine kinase domain. Interestingly, the cheS3 disruption shows a hyper-cyst phenotype, opposite that of the cyst-deficient phenotype observed upon disruption of cheA3. Biochemical analysis of purified CheS3, and of other components of the Che3 cascade is ongoing, with evidence to date indicating that the CheS3 kinase is able to autophosphorylate in the presence of ATP and that dephosphorylated CheY3 is capable of inhibiting the accumulation of phosphate on CheS3 (J. Marden and C. E. Bauer, unpubl.). These results suggest that there is an interesting flow of phosphates between CheS3 and other components of the Che3 cascade that will require extensive genetic and biochemical analysis to unravel. It also suggests that here are multiple input signals that are integrated to control the ultimate decision of whether to undergo encystment. This is not unlike sporulation in Bacillus that also utilizes numerous input signals to control the decision to undergo spore formation (Stephenson and Hoch, 2002).

We also observed that ΔcheB3, ΔcheR3 and ΔcheA3 fail to form a chemotactic ring because of lack of motility. TEM and phase-contrast microscopic analysis of these strains demonstrates that these mutants lack attached flagella, which is very similar to the phenotype of several mutants in the che2 operon (data not shown). This suggests that there may be some interactions between the Che2 and Che3 signal transduction cascades. Ejection of flagella occurs early in the cyst developmental pathway (Berleman and Bauer, 2004), suggesting that the Che2 and Che3 pathways may coordinate the loss of flagella during encystment. Interestingly flagella defects of che3 deletions are limited to proteins involved in adaptation such as CheR3, CheB3 and CheA3 that presumably phosphorylates CheB3. We thus propose a model by which che3 signalling is a bifurcated pathway with two output responses with the first output dependent on all che3 proteins that modulates the timing and efficacy of cyst cell induction through a phosphorylation cascade from CheA3 to CheS3 or CheY3. The mechanism for how CheY3∼P or CheS3 can effect gene expression is unclear, because they contain no readily identifiable DNA-binding domains. This suggests that other downstream components involved in this signalling pathway are yet to be identified. The second output of the che3 pathway is adaptation that affects the signalling state of the Che2 pathway that controls flagella stability as well as the flow of phosphates in the Che3 encystment pathway. Clearly, this species has an interesting set of Che-like signalling pathways that controls very different processes. Additional detailed genetic and biochemical analysis of these signalling pathways will be needed to unravel the complex level of control that governs bacterial encystment.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Bacterial strains and growth conditions

The parental strain used in this study was wild-type R. centenum strain ATCC51521. All R. centenum strains were cultured aerobically either in liquid CENS media (Stadtwald-Demchick et al., 1990) at 37° or at 42°C on agar solidified CENS media or agar solidified CENBA media. Escherichia coli strains S17-1 (λ pir) and JM109 λ pir were used for conjugation and cloning respectively (Simon et al., 1983; Miller and Mekalanos, 1988). Plasmid pUC18 (Norrander et al., 1983) was used for routine cloning and plasmid pZJD29a was used for constructing in frame deletions using sucrose selection (Masuda and Bauer, 2004). E. coli strains were cultured in Luria–Bertani (LB) media at 37°C, with ampicillin and gentamicin (Gm), kanamycin (Km) used when appropriate at 150 µg ml−1, 50 µg ml−1 and 50 µg ml−1 respectively. For R. centenum, Gm and Km were used at 10 µg ml−1 and 40 µg ml−1 respectively.

Deletion construction

In-frame deletion mutants were constructed using a modified sucrose (Suc) selection technique as previously described (Blomfield et al., 1991; Berleman and Bauer, 2005; Masuda and Bauer, 2004). The primers used for amplification of gene-flanking recombination sites are described in Supplementary materials (Table S1). PCR-amplified fragments were initially cloned into pUC18 and then subcloned into the suicide vector, pZJD29a. The resulting plasmids were transferred into wild-type R. centenum via conjugation with E. coli S17-1 λ pir. Correct deletions were determined through phenotypic (GmS/SucR colonies) and colony PCR analysis.

Reverse transcriptase-PCR assays

Total RNA was isolated from cells using a TRI method from Molecular Research Center. RNA preps were normalized by reading the absorbance at 260 nm, and adjusting the samples to a final concentration of 100 ng µl−1 RNA. Trace DNA contamination was removed using a DNA-free protocol from Ambion. RT-PCR reactions were performed using Superscript III First Strand Synthesis System (Invitrogen) using standard PCR conditions as described by Berleman and Bauer (Berleman and Bauer, 2005). For each strain to be analysed, PCR reactions on the RNA template were performed with and without reverse transcriptase using gene specific primers. For PCR amplification of the cDNA template, primers designed to amplify a portion of the cheS3 gene were cheS3RTset3F, 5′-AGC GTG ACC TTC CTC AAC-3′ and cheS3RTset3R, 5′-GTG TTC TTC ACC CGA TGG-3′. For the control reaction, primers specific for the last gene of the che2 operon, ORF74 were ORF74for1, 5′-AAG GAG ATC GTG CGA TGA TC-3′ and ORF74rev1, 5′-GAC GAG CTG CAG CAC GGC-3′.

Sequence analysis

The entire che3 gene cluster was sequenced by primer walking from mini-Tn5 insertions in the mcp3, cheB3 and cheS3 genes described in a previous study (Berleman et al., 2004). Sequences are deposited in GenBank under Accession number AY260903. Analysis of translated DNA was performed using commercial GCG software.

Colony morphology and phase contrast microscopy

For analysis of colony morphology, overnight liquid cultures of R. centenum were harvested, washed three times in phosphate buffer (40 mM KH2PO4/K2HPO4; pH 7.0) resuspended in 1/10 volume and pippetted as five µl aliquots onto agar-solidified CENS or CENBA plates. Colonies were photographed using a Sony DSC-F707 digital camera, and microscopically analysed, after 3 days growth at 42°C on CENS or 5 days growth on CENBA medium.

Rhodospirillum centenum wet mounts were prepared on cells harvested from plates or from liquid cultures and viewed with a Nikon E800 light microscope equipped with a 100× Plan Apo oil objective. Image capture was carried out with a Princeton Instruments cooled charge-coupled device (CCD) camera and Metamorph imaging software, v.4.5.

Desiccation resistance assays

Quantitative analysis of the maturation of cysts in R. centenum has previously been described in detail (Berleman and Bauer, 2004; Berleman et al., 2004). Briefly, R. centenum overnight cultures were harvested and washed three times in phosphate buffer with 5 µl aliquots pipetted onto either CENS or CENBA plates. Cells were harvested after 3 days on CENS and 5 days on CENBA, resuspended in 1 ml of phosphate buffer and sonicated for 5 s at low power to disperse cell aggregates. To quantify the total number of viable cells (vegetative cells plus cyst cells), the resuspended cells were serially diluted onto CENS plates and incubated at 42°C for 3 days. To quantify the number of cyst cells, replicates of the total viable cell diluents were pipetted onto 0.45 µm filters, dried for 20 min at 22°C, and then desiccated at 42°C for 3 days. Desiccated filters were then placed onto CENS plates for 2 days at 42°C to allow outgrowth of surviving cells. Total colonies before and after desiccation were counted with assays repeated in triplicate.

Chemotaxis assays

To assay chemotaxis of swim cells, overnight cultures of R. centenum strains were washed in phosphate buffer and pipetted in 3 µl aliquots onto 0.25% agarose CENMED minimal medium with 20 mM pyruvate supplied as a chemoattractant. Plates were incubated for 48 h at the temperatures stated.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

J.B. was supported by a National Institutes of Health Training Grant GM007757.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information
  • Ames, P., Studdert, C.A., Reiser, R.H., and Parkinson, J.S. (2002) Collaborative signaling by mixed chemoreceptor teams in Escherichia coli. Proc Natl Acad Sci USA 99: 70607065.
  • Berleman, J.E., and Bauer, C.E. (2004) Characterization of cyst cell formation in the purple photosynthetic bacterium Rhodospirillum centenum. Microbiology 150: 383390.
  • Berleman, J.E., and Bauer, C.E. (2005) A che-like signal transduction cascade involved in controlling flagella biosynthesis in Rhodospirillum centenum. Mol Microbiol 55: 13901402.
  • Berleman, J.E., Hasselbring, B.M., and Bauer, C.E. (2004) Hypercyst mutants in Rhodospirillum centenum identify regulatory loci involved in cyst cell differentiation. J Bacteriol 186: 58345841.
  • Bhaya, D., Takahashi, A., and Grossman, A.R. (2001) Light regulation of type IV pilus-dependent motility by chemosensor-like elements in Synechocystis PCC6803. Proc Natl Acad Sci USA 98: 75407545.
  • Blomfield, I.C., Vaughn, V., Rest, R.F., and Eisenstein, B.I. (1991) Allelic exchange in Escherichia coli using the Bacillus subtilis sacB gene and a temperature-sensitive pSC101 replicon. Mol Microbiol 5: 14471457.
  • Falke, J.J., Bass, R.B., Butler, S.L., Chervitz, S.A., and Danielson, M.A. (1997) The two-component signaling pathway of bacterial chemotaxis: a molecular view of signal transduction by receptors, kinases, and adaptation enzymes. Annu Rev Cell Dev Biol 13: 457512.
  • Favinger, J., Stadtwald, R., and Gest, H. (1989) Rhodospirillum centenum, sp. nov., a thermotolerant cyst-forming anoxygenic photosynthetic bacterium. Antonie Van Leeuwenhoek 55: 291296.
  • Jiang, Z.Y., and Bauer, C.E. (1997) Analysis of a chemotaxis operon from Rhodospirillum centenum. J Bacteriol 179: 57125719.
  • Jiang, Z.Y., Gest, H., and Bauer, C.E. (1997) Chemosensory and photosensory perception in purple photosynthetic bacteria utilize common signal transduction components. J Bacteriol 179: 57205727.
  • Jiang, Z.Y., Rushing, B.G., Bai, Y., Gest, H., and Bauer, C.E. (1998) Isolation of Rhodospirillum centenum mutants defective in phototactic colony motility by transposon mutagenesis. J Bacteriol 180: 12481255.
  • Karniol, B., and Vierstra, R.D. (2004) The HWE histidine kinases, a new family of bacterial two-component sensor kinases with potentially diverse roles in environmental signaling. J Bacteriol 186: 445453.
  • Kato, J., Nakamura, T., Kuroda, A., and Ohtake, H. (1999) Cloning and characterization of chemotaxis genes in Pseudomonas aeruginosa. Biosci Biotechnol Biochem 63: 155161.
  • Kirby, J.R., and Zusman, D.R. (2003) Chemosensory regulation of developmental gene expression in Myxococcus xanthus. Proc Natl Acad Sci USA 100: 20082013.
  • McBride, M.J., Weinberg, R.A., and Zusman, D.R. (1989) ‘Frizzy’ aggregation genes of the gliding bacterium Myxococcus xanthus show sequence similarities to the chemotaxis genes of enteric bacteria. Proc Natl Acad Sci USA 86: 424428.
  • Masuda, S., and Bauer, C.E. (2004) Null mutation of HvrA compensates for loss of an essential relA/spoT-like gene in Rhodobacter capsulatus. J Bacteriol 186: 235239.
  • Miller, V.L., and Mekalanos, J.J. (1988) A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR. J Bacteriol 170: 25752583.
  • Norrander, J., Kempe, T., and Messing, J. (1983) Construction of improved M13 vectors using oligodeoxynucleotide-directed mutagenesis. Gene 26: 101106.
  • Ragatz, L., Jiang, Z.Y., Bauer, C.E., and Gest, H. (1995) Macroscopic phototactic behavior of the purple photosynthetic bacterium Rhodospirillum centenum. Arch Microbiol 163: 16.
  • Sadasivan, L., and Neyra, C.A. (1987) Cyst production and brown pigment formation in aging cultures of Azospirillum brasilense ATCC 29145. J Bacteriol 169: 16701677.
  • Simon, R., Priefer, U., and Puhler, A. (1983) A broad host range mobilization system for in vitro genetic engineering: transposition mutagenesis in Gram-negative bacteria. Biotechnology 1: 784791.
  • Springer, W.R., and Koshland, D.E., Jr (1977) Identification of a protein methyltransferase as the cheR gene product in the bacterial sensing system. Proc Natl Acad Sci USA 74: 533537.
  • Stadtwald-Demchick, R., Turner, F.R., and Gest, H. (1990) Physiological properties of the thermotolerant photosynthetic bacterium, Rhodospirillum centenum. FEMS Microbiol Lett 67: 139144.
  • Stephenson, K., and Hoch, J.A. (2002) Evolution of signalling in the sporulation phosphorelay. Mol Microbiol 46: 297304.
  • Stevenson, L.H., and Socolofsky, M.D. (1966) Cyst formation and poly-beta-hydroxybutyric acid accumulation in Azotobacter. J Bacteriol 91: 304310.
  • Stock, J.B., and Koshland, D.E., Jr (1978) A protein methylesterase involved in bacterial sensing. Proc Natl Acad Sci USA 75: 36593663.
  • Szurmant, H., and Ordal, G.W. (2004) Diversity in chemotaxis mechanisms among the bacteria and archaea. Microbiol Mol Biol Rev 68: 301319.
  • Welch, M., Oosawa, K., Aizawa, S., and Eisenbach, M. (1993) Phosphorylation-dependent binding of a signal molecule to the flagellar switch of bacteria. Proc Natl Acad Sci USA 90: 87878791.
  • Whitchurch, C.B., Leech, A.J., Young, M.D., Kennedy, D., Sargent, J.L., Bertrand, J.J., et al. (2004) Characterization of a complex chemosensory signal transduction system which controls twitching motility in Pseudomonas aeruginosa. Mol Microbiol 52: 873893.
  • Yang, Z., Ma, X., Tong, L., Kaplan, H.B., Shimkets, L.J., and Shi, W. (2000) Myxococcus xanthus dif genes are required for biogenesis of cell surface fibrils essential for social gliding motility. J Bacteriol 182: 57935798.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

The following supplementary material is available for this article: Appendix S1. Derivation of an expression for the dependency of initial rate of Trg modification on the ratio of assisting Tar to assisted Trg. See text for assumption and logic.

FilenameFormatSizeDescription
MMI_4646_sm_TableS1.doc118KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.