The photosensor protein Ppr of Rhodocista centenaria is linked to the chemotaxis signalling pathway
© Kreutel et al. 2010
Received: 28 March 2010
Accepted: 9 November 2010
Published: 9 November 2010
Skip to main content
© Kreutel et al. 2010
Received: 28 March 2010
Accepted: 9 November 2010
Published: 9 November 2010
Rhodocista centenaria is a phototrophic α-proteobacterium exhibiting a phototactic behaviour visible as colony movement on agar plates directed to red light. As many phototrophic purple bacteria R. centenaria possesses a soluble photoactive yellow protein (Pyp). It exists as a long fusion protein, designated Ppr, consisting of three domains, the Pyp domain, a putative bilin binding domain (Bbd) and a histidine kinase domain (Pph). The Ppr protein is involved in the regulation of polyketide synthesis but it is still unclear, how this is connected to phototaxis and chemotaxis.
To elucidate the possible role of Ppr and Pph in the chemotactic network we studied the interaction with chemotactic proteins in vitro as well as in vivo. Matrix-assisted coelution experiments were performed to study the possible communication of the different putative binding partners. The kinase domain of the Ppr protein was found to interact with the chemotactic linker protein CheW. The formation of this complex was clearly ATP-dependent. Further results indicated that the Pph histidine kinase domain and CheW may form a complex with the chemotactic kinase CheAY suggesting a role of Ppr in the chemotaxis signalling pathway. In addition, when Ppr or Pph were expressed in Escherichia coli, the chemotactic response of the cells was dramatically affected.
The Ppr protein of Rhodocista centenaria directly interacts with the chemotactic protein CheW. This suggests a role of the Ppr protein in the regulation of the chemotactic response in addition to its role in chalcone synthesis.
Rhodocista centenaria, first described as Rhodospirillum centenum  is a thermotolerant phototrophic purple bacterium of the α-proteobacteria group isolated from hot springs in Wyoming 1985. The slightly spiroid or vibrioid shaped cells are motile by means of a single long flagellum, their intracellular photosynthetic membranes are lamellar and their in vivo absorption spectra show features almost indistinguishable from those of Rhodospirillum rubrum . However, 16S rRNA analysis elucidated considerable differences between the species, hence Rhodocista was separated into a new genus , now consisting of three species [4, 5]. R. centenaria is closely related to the plant-associated genus Azospirillum . As virtually all phototrophic organisms, R. centenaria exhibits a sensory response to light originally described as "Schreckbewegung" . Engelmann and also Manten  found that R. rubrum cells accumulated in the most intense area of light gradients between wavelengths 800 and 900 nm. R. centenaria shows a particularly unique form of macroscopic phototactic behaviour, first described in 1994 by Gest and coworkers . On solid media, the phototactic colonies move towards longwave light and away from light with wavelengths less than 650 nm . R. centenaria develops lateral flagella in viscous media or on solidified surfaces. These flagella consist of a distinct flagellin whose expression is controlled by specific mot and fli genes .
For R. centenaria, a close relationship between chemotaxis and the phototactic response has been found . As seen with many other photosynthetic bacteria, R. centenaria has multiple chemotaxis operons with distinct functions [13–15]. The chemotaxis gene cluster has been well characterized and most of the genes are similar to those of other Gram negative bacteria like Escherichia coli. In brief, the histidine kinase CheA is linked to the chemotactic receptors (MCPs) by the CheW protein . This trimeric receptor complex controls the phosphorylation level of the response regulator CheY. Activated CheY (CheY-P) interacts directly with the motor of the flagella to control swimming direction. The dephosphorylation of CheY-P occurs spontaneously, only in enterobacteria this reaction is accelerated by the phosphatase CheZ. For adaptation, CheB and its antagonist CheR remove or add methyl groups to the receptors, respectively.
In R. centenaria, the two central components of the chemotactic signal transduction cascade, namely CheA and CheY, are present as the fusion protein Rc-CheAY located in the first chemotactic operon , a situation that is also observed in Helicobacter . Whereas the role of the CheY-domain of the CheAY protein in H. pylori seems to be a phosphate sink, in R. centenaria, the function of Rc-CheAY remains still unclear. While Che proteins are generally involved in chemotactic responses, they were also shown to affect the phototactic response in R. centenaria as demonstrated by the analysis of many che mutants .
In the last decade, bacterial photoreactive proteins like phytochromes, previously thought to be a unique feature in plants, have been identified as photoactive yellow proteins (Pyp) and have now been extensively studied in a variety of eubacterial species (for review see [20, 21]). For R. centenaria, a Pyp-like protein, Ppr, was described in 1999 by Bauer and colleagues . The large fusion protein Ppr consists of three functional domains, an N-terminal Pyp domain with the cinnamic acid chromophore, the central phytochrome-like bilin attachment domain Bbd and the C-terminal histidine kinase domain Pph which autophosphorylates an essential histidine residue . Although some Pyp proteins have been crystallized and biophysically characterized in great detail (reviewed by ), no distinct physiological role could be attested to these unique proteins. A Ppr-deletion mutant lacking amino acid residues 114-750 did not show any alterations in phototactic behaviour, instead exhibited a strongly deregulated expression of the chalcone synthase gene suggesting a regulatory function in the polyketide synthesis . Although there is no obvious direct involvement of Ppr in the phototactic or scotophobic reaction, an interaction with the chemotactic signal transduction components is plausible to regulate general phosphorylation levels or transduce phosphoryl groups to a yet unknown light-dependent signal transducing protein. We therefore analysed whether the Ppr protein and in particular its phosphorylating kinase domain Pph interacts with the Rc-Che proteins.
Taken together, the results give preliminary evidence that the C-terminal histidine kinase domain Pph of the photosensor protein Ppr assembles in vitro into a trimeric complex of Pph, Rc-CheW and Rc-CheAY.
Since photosynthetic bacteria have to locate their habitat with optimal light conditions, specialized sensor systems and signal transduction cascades involving different chromophores arose during evolution (for review see ). The blue light sensitive Ppr protein of R. centenaria consists of three distinct domains, the Pyp domain containing a cinnamic acid chromophore, the phytochrome-like bilin binding domain and the histidine kinase domain Pph (Figure 1; ). The structural organization suggests that the protein is involved in a light-dependent signaling pathway similar to chemotaxis. Since R. centenaria exhibits a strikingly obvious phototactic behavior it is compelling to assume that the Ppr protein is involved in this reaction. Light with a wavelength of above 650 nm is attractive, whereas light with less than 650 nm acts as a repellent . The absorption maximum of a prototypical cinnamic acid chromophore in a Pyp light sensor is at about 450 nm , whereas the phytochrome-linked biliverdin absorbs red light, suggesting that the latter could function as an attractant sensor. Recently, Cusanovich and co-workers showed that the holo-Ppr of R. centenaria has absorption maxima at 425 nm (Pyp), 400, 642 and 701 nm (phytochrome)  corresponding to the typical absorption spectrum of Pyp  and phytochromes . The phytochromes TaxD1, Cph2 and PlpA were found to be involved in the phototactic reaction of Synechocystis sp. PCC 6803, a finding that supports the idea of a participation of the Ppr sensor in the phototactic response of R. centenaria [42, 43].
The data presented here show that the histidine kinase Pph domain of the Ppr receptor is found in a complex with Rc-CheW when isolated from R. centenaria (Figure 8). The interaction of Pph and Rc-CheW was also observed in vitro with purified components showing a strong affinity with a KD of about 130 nM (Figure 5). In contrast to the trimeric Tsr-CheA-CheW complex that is formed in E. coli with an affinity of about 3 μM  we observed that the complex formation of Pph and Rc-CheW is clearly ATP-dependent (Figure 4B). It is likely that the Pph-CheW complex is capable to bind Rc-CheAY (Figure 6) consistent with the idea that the chemotactic network is functioning in the presence of Pph. However, the function of the Rc-CheAY fusion protein in this signaling cascade remains unclear. Preliminary transphosphorylation experiments that we perfomed indicate that the CheY domain of the Rc-CheAY protein acts as a phosphate receiver domain and that the CheY domain acts as a phosphate sink similar as it has been described for the chemotactic system in Rhizobium meliloti and Helicobacter pylori [44, 45]. The involvement of Ppr in chemotaxis is also supported from the experiments we performed with E. coli. The heterologous expression of Pph has a strong inhibitory effect on chemotaxis as demonstrated by the swarm assay (Figure 2) and the capillary assay (Figure 3). Both assays showed that upon expression of Ppr or Pph the chemotaxis of E. coli is turned off whereas expression of the R. centenaria histidine kinase KdpE had no effect. This suggests that the Ppr protein interacts with Ec-CheW although the CheW proteins of E. coli and R. centenaria show a homology of only about 59% and an identity of 28% . However, the structural analysis suggests that all CheW proteins of different species share common features [46, 47]. We propose that the binding of the Ppr protein results in a non-functional Ec-CheW-Ppr complex that is inhibitory for chemotaxis (Figures 2 and 3) due to the inactivation of Ec-CheW. Remarkedly, a mutant of the predicted phosphorylation site of Pph with the histidine at position 670 being changed to an alanine residue had a less inhibitory effect on chemotaxis, suggesting that the kinase activity of Pph has a functional role in CheW binding. Similar inhibitory effects on chemotaxis have been observed for E. coli when Ec-CheW, Ec-CheA or the MCP-receptors were overproduced [23, 25, 27]. In addition, such an inhibitory effect was also observed when chemotactic proteins from other organisms like Rhodobacter capsulatus  or Leptospira interrogans  were heterologously expressed in E. coli.
We found that the histidine kinase domain Pph was mainly present as a monomer when expressed in E. coli (Figure 7) and only a minor fraction was found as dimers. Most other bacterial histidine kinases that have been investigated so far were found to be homodimers . Accordingly, when the plasmid encoded Pph protein was isolated from R. centenaria it appeared in a complex consisting of CheW and most likely a dimer of Pph (Figure 8).
Working in a network together with the perception of chemical substrates by the chemotactic signal transduction cascade, Ppr could very well be a part of a complex sensory machinery to tune metabolic and phototrophic processes in phototrophic bacteria. The perception of light may only be an oblique indicator for the metabolic state of a R. centenaria cell as is suggested by its influence on cyst formation [13, 22]. Therefore, Ppr could work in parallel with the photosynthetic electron transport sensor Ptr of R. centenaria  to specifically regulate cellular motility and sense the metabolic state of the cell.
All genetic manipulations were performed according to standard methods in E. coli XL1-Blue (recA1 thi supE44 endA1 hsdR17 gyrA96 relA1 lac F′ (proAB+ lacI q lacZΔM15 Tn10) as described . For expression of Rc-CheW and Pph, E. coli C41  was used. For genetic transfer into R. centenaria, E. coli RR28  and in the swarm assays, E. coli MM500  was used. For E. coli, antibiotics were added at final concentrations of 200 μg/ml ampicillin, 10-50 μg/ml kanamycin and 5 μg/ml gentamycin and for R. centenaria 5 μg/ml gentamycin, 10 μg/ml kanamycin. All E. coli strains were cultured in LB medium at 37°C if not indicated otherwise. R. centenaria (ATCC 43720) was obtained from the culture collection. (For anaerobic photosynthetic growth R. centenaria was cultured in screw cap bottles filled to the top with PYVS medium  and illuminated by an 80 W tungsten bulb (Concentra, Osram, Germany) at 42°C.
Constructs and plasmids used in this study
Ampr, pET16b, His10-Ppr
Ampr, araC, PBAD, pBR322 ori, expression vector
Ampr, pBAD18, His10-PprΔ1-601-StrepTag II
Ampr, pBAD18, His10-PprΔ1-601; H670A-StrepTag II
Ampr, pBAD18, His10-Ppr-StrepTag II
Ampr, pBAD24, Rc-KdpE
Ampr, T7 promoter, N-terminal deca histidyl tag
Ampr, pET16b, His10-PprΔ1-601-StrepTag II
Ampr, pET16b, His10-PprΔ1-601; H670A-StrepTag II
Ampr, pET16b, His10-Ppr-StrepTag II
Kanr, pET28a(+), His6-CheAY
Kanr, Tra, helper plasmid
Ampr, pT7-7, PprΔ1-601-StrepTag
Ampr, T7 promoter, expression vector
Ampr, pT7-7, His6-Rc-CheW
Ampr, Kanr, lac promoter, lacZ
Genr, pRK2 derived plasmid, lacZ
A ppr-strep tag II fusion gene was constructed as follows. pET16b containing the entire ppr gene (pNB10), as well as pET16b-Pph were cut by NcoI and the resulting fragments (~6.0 kb and ~2.5 kb) were ligated. The orientation of the ppr-insert was checked by DNA-sequencing and the resulting plasmid was named pET16b-Ppr. To construct an arabinose inducible full length ppr, the gene was excised by XbaI and HindIII from pET16b-Ppr and ligated into the pBAD18 vector.
The putative phosphorylation site (the histidine at position 670 in the Ppr protein) was changed to an alanine (CAC→GCG) using site directed mutagenesis with the primers (5'-CTGGCGAACATGAGCGCGGAGCTGCGGACTCCG-3') and (5'-CGGAGTCCGCAGCTCCGCGCTCATGTTCGCCAG-3') and pSK4 as a template. The resulting mutant was digested by NdeI and BamHI and subcloned into the pET16b vector generating pET16b-PphH670A. Then the pphH670A mutant was excised by XbaI and HindIII and the fragment was inserted into the pBAD18 vector to create pBAD-PphH670A.
To express the histidine kinase domain Pph with an N-terminal his10-tag and a C-terminal strep-tag II in R. centenaria, the plasmid pZJD11 (kindly provided by C. Bauer) was used . We used the oxygen regulated puc promoter and the puhA Shine Dalgarno sequence from Rhodobacter capsulatus to initiate translation. Therefore, a PCR reaction with the primers (5'-TACGTAGGGCCCTAAGCTAAAGGAGGACTAACATGGGCCATCATCAT-3') and (5'-TACGTAGGCGCGAATTCGGCTTGATCAGGC-3') and pET16b-Pph as a template was conducted. Simultaneously, a SnaBI restriction site was introduced at the 3' end of the gene. The resulting fragment was subcloned into pGEM T-easy vector (Promega) and verified by DNA sequencing. This plasmid was used as a template to insert the puc promoter via a second PCR. The primers (5'-GGTAACCTTGATCGCCGACACTTGGGCTCCCA TAGTGGAGCTCGGGCCCTAAG-3') and (5'-TACGTAGGCGCGAATTCGGCTTGATCA GGC-3') were used to introduce a BstEII site at the 5' end. The resulting fragment was inserted into pGEM T-easy vector. After sequencing, the pph construct was excised by BstEII and SnaBI and ligated into the corresponding sites of pZJD11 to generate pSK10.
To express the Rc-CheW protein in E. coli, the cheW gene was amplified by PCR from the R. centenaria genome using the primers (5'CATATGCATGCCCGCCTGCCCGTTCCC-3') and (5'GGGAATCGTTCATTGCGATCAGTTTCCGG-3'), respectively. The resulting fragment was first cloned into pT-Adv. Then the cheW gene was excised by NdeI and EcoRI and ligated into the corresponding restriction sites of a pT7-7 derivative containing a decahistidine sequence to create the IPTG inducible expression vector pT7-7-CheW.
TB swarm agar plates (1% bacto-tryptone, 0.8% NaCl; 0.35% bacto-agar) containing 0.2% arabinose or 0.2% fructose, respectively, were inoculated with a single colony of E. coli MM500 or MM500 harbouring one of the plasmids pBAD-Ppr, pBAD-Pph, pBAD-PphH670A, pBADKdpE and pBAD, respectively. The plates were incubated for 6 hours at 37°C.
2 ml minimal medium A (MMA)  containing an amino acid mixture (threonine, leucine, histidine, methionine), vitamin B1 (final concentration 10 μg/ml each), 200 μg/ml ampicillin and 0.2% fructose were inoculated with an overnight culture of E. coli MM500 or cells harbouring pBAD-Pph, pBAD-PphH670A, pBAD-KdpE or pBAD18, respectively. When the cultures reached an OD600 = 0.6 the cells were washed twice with MMA without sugar and finally either 0.2% arabinose to induce protein expression or 0.2% fructose (as a control) were added. The cultures were incubated for 60 min at 37°C. For the kinetic analysis the incubation times are indicated in Figure 3B. Again, the cells were washed twice with MMA without carbon source and were back diluted to an OD600 = 0.6. The chemotactic assays were performed as follows. 300 μl of the cell suspension were filled in each drilling of the chamber and a capillary containing either 2 μl 1 mM aspartate or 2 μl H2O as a control was placed into the channel between the two cylindrical compartments. The chamber was incubated at 37°C for 30 minutes. The outside of the capillary was washed extensively with sterile water and the content of the capillary was blown out and a dilution series was streaked on agar plates. After overnight incubation at 37°C the colonies were counted and the chemotactic inhibition (CI) was calculated as the ratio of colonies of the water containing capillary to the colonies from the aspartate containing capillary. Therefore, a low CI indicates an undisturbed chemotactic response whereas a high CI reflects an inhibition of the E. coli chemotactic system.
E. coli strain C41  harbouring the plasmid pET16b-Pph were grown at 37°C in 1 l LB medium containing 200 μg/ml ampicillin. When cells reached the midlogarithmic phase, IPTG was added at a final concentration of 1 mM and the cells were grown for an additional 4 hours at 37°C. Then the cells were harvested by centrifugation. The resulting pellets were resuspended in 100 mM Tris-HCl pH 8.0, 150 mM NaCl (buffer W) and lysed by a French Press. Inclusion bodies were precipitated by centrifugation and resuspended in buffer W containing 0.5% N-lauroylsarcosine. The inclusion bodies were solubilized overnight at 4°C with gentle shaking. To the filtered extract 10 mM imidazole was added and applied to a Sepharose 6b (GE Healthcare) column precharged with Cu(II) ions. Unbound proteins were removed by washing the column with 15 column volumes of buffer W containing 0.5% N-lauroylsarcosine and 10 mM imidazole. The bound protein was eluted by a linear gradient up to 500 mM imidazole in buffer W + 0.5% N-lauroylsarcosine. The Pph protein containing fractions were pooled, diluted 1:40 with buffer W (final detergent concentration = 0.01%) and applied to a streptactin-sepharose column (IBA, Göttingen, Germany) to remove contaminating proteins. After washing the column with five column volumes buffer W + 0.01% N-lauroylsarcosine, the protein was eluted with buffer W + 0.01% N-lauroylsarcosine containing 2.5 mM desthiobiotin. The protein was dialyzed against buffer W + 0.01% N-lauroylsarcosine and the purity was checked by SDS-PAGE analysis as described . Protein marker SM0431 and SM0441(Fermentas) were used.
1 Liter of LB medium containing 200 μg/ml ampicillin was inoculated with a freshly transformed single colony of E. coli C41 harbouring the plasmid pT7-7-CheW. The cells were grown to a cell density of 2 × 108 cells per ml at 37°C, then IPTG was added to a final concentration of 1 mM. The cells were incubated for an additional 4 hours and harvested by centrifugation. The pellet was resuspended in TBS (50 mM Tris-HCl pH 7.4, 150 mM NaCl) and lysed by a French Press. Cell debris was removed by centrifugation and a final concentration of 10 mM imidazole was added. This crude extract was applied to a Cu(II)-charged Sepharose 6b column and unbound proteins were washed out with 10 column volumes of TBS + 10 mM imidazole. The protein was eluted with a linear gradient from 10 to 500 mM imidazole and fractions containing Rc-CheW were dialyzed against TBS-buffer. The homogeneity of the protein was monitored by SDS-PAGE.
The plasmid pSK10 was transferred to wild type R. centenaria by triparental conjugation using E. coli RR28 , the helper plasmid pRK2013  and the filter-mating technique as described previously . After conjugation, about 109 T7 phages were added, and the mixture was incubated for 30 minutes at 37°C to eliminate remaining E. coli cells. Finally, conjugants were selected on the basis of gentamycin resistance on PYVS plates containing 5 μg/ml gentamycin under anaerobic conditions. 2L PYVS media containing 5 μg/ml gentamycin and 10 μg/ml kanamycin (R. centenaria is naturally resistant to kanamycin ) was inoculated with a culture of pSK10 containing R. centenaria cells. The cells were grown under anaerobic and illuminated conditions for 96 h and harvested by centrifugation, resuspended in 100 mM Tris pH 8.0, 150 mM NaCl (buffer W) and lysed by a French Press. The cell debris and the photosynthetic membranes were removed by centrifugation. The cleared extract was applied to a streptactin-sepharose column (IBA). The unbound proteins were removed by extensively washing the column with buffer W and the bound proteins were eluted with buffer W containing 2.5 mM desthiobiotin. The success of the purification was verified by SDS-PAGE, silver staining and Western blot analysis with the antibodies raised against the his-tag or the strep-tagII, respectively.
The Pph protein was purified from inclusion bodies as described above and the aminosilane cuvette was activated as described by the manufacturer (Iasys, Biosensors). 200 μl of the purified Pph protein (50 μg/ml) was added to the activated cuvette and the immobilization was recorded for 30 minutes. The unbound protein was removed by extensive washing and increasing amounts of purified Rc-CheW (see above) were added. After 30 minutes of incubation the free Rc-CheW was washed out and the amount of bound Rc-CheW was determined for each experiment. The fractional saturation was calculated and depicted against the amount of the added Rc-CheW concentration. The resulting Scatchard Plot is illustrated as the inlet of Figure 5.
The histidine kinase domain Pph as well as Rc-CheAY were transcribed in vitro from the plasmids pSK4 and pET28-CheAY, respectively, using a T7 transcription kit (Fermentas) according to the manufacturers manual. The translation reaction was performed as described previously  by using an E. coli based cell free expression system The proteins were labeled with 10 μCi of [35S]methionine (ICN) in each experiment. The high speed supernatant (S-135) was prepared as described from E. coli MRE600 .
50 μg of the purified his6-Rc-CheW protein was mixed with 25 μl of the in vitro translated Pph protein and 25 μl Rc-CheAY when indicated. The protein mixture was incubated overnight at 37°C. Then, the his6-Rc-CheW protein was bound to a column containing 50 μl Sepharose 6b (GE Healthcare) charged with Cu(II) ions and pre-equilibrated with buffer I (20 mM sodium phosphate pH 7.7, 200 mM NaCl, 50 mM imidazole pH 8.0). After 30 minutes at room temperature, the unbound proteins were removed by washing the column five times with 500 μl buffer I followed by an elution with 1.5 ml buffer II (20 mM sodium phosphate pH 7.7, 200 mM NaCl, 500 mM imidazole pH 8.0). All fractions were TCA precipitated and analyzed by SDS-PAGE. The gels were stained with coomassie brilliant blue and the radiolabeled bands were quantified using a Fuji BAS 1500 phosphorimager.
1L terrific broth  in a Fernbach flask was inoculated with an overnight culture of E. coli C41 (DE3) harbouring pET16b-Pph. The cells were incubated at 18°C with gentle shaking for 48 hours. This procedure prevents the formation of inclusion bodies . Then the cells were harvested by centrifugation and resuspended in 20 mM Tris pH 7.4, 40 mM NaCl, 20% glycerol. The cells were lysed by 3 passages through a French Press and the cell debris was removed by centrifugation. The filtered crude protein extract was applied on a Sephadex G-200 gelfiltration column (GE Healthcare) and separated according to the manufacturer's manual. The resulting fractions were analyzed by SDS-PAGE and Western blotting with an antibody to strep tag II (IBA, Göttingen, Germany).
We thank Dr. Robin Ghosh for his generous support and scientific input and Dr. Birgit Scharf for critical reading of the manuscript.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.