Background

In their natural environment bacteria are exposed to changing conditions, which often cause stress. To survive these conditions they have evolved defense systems that allow adaptation to the changing environment. Many previous studies revealed a big overlap of the responses to a variety of different stress factors, considered as general stress response.

The PhyR-NepR-σ (EcfG) cascade was recognized as a core pathway regulating the general stress response in Alphaproteobacteria [1]. PhyR (phyllosphere regulator) was identified as a response regulator essential for plant colonization by Methylobacterium extorquens and promotes resistance to various stresses by controlling stress-related genes [2, 3]. In several Alphaproteobacteria, the phyR gene is in close proximity to the nepR and ecfG genes [1]. The σ(EcfG)-orthologous sigma factors RpoE2 (Sinorhizobium meliloti) and SigT (Caulobacter crescentus) induce large regulons in response to various stresses [4, 5]. In M. extorquens, NepR (negative regulator of the PhyR response) and PhyR control the activity of the σ(EcfG) sigma factor by acting as anti-sigma factor and as anti-sigma factor antagonist, respectively [6]. In the current model [1] NepR sequesters σ(EcfG) under non-stress conditions. Under stress conditions, the response regulator PhyR becomes phosphorylated, interacts with NepR and consequently releases σ(EcfG), which in turn associates with RNA polymerase.

Rhodobacter sphaeroides is a facultative phototrophic Alphaproteobacterium, which has been intensively studied in regard to its oxidative stress response, including singlet oxygen stress [7,8,9]. In photosynthetic bacteria (bacterio-) chlorophylls act as photosensitizers, which generate this harmful oxygen species when light and oxygen are present. The ECF sigma factor RpoE (RSP_1092) plays an important role in the singlet oxygen response of R. sphaeroides. Many genes are activated by RpoE [10, 11], however only a small subset directly [11]. Among the genes activated by RpoE is the alternative sigma factor RpoHII, which directly activates a much larger set of genes [12]. The RpoHII regulon has an overlap with the RpoHI regulon [13, 14], both sigma factors also have an important role in the general stress response [15]. Under non-stress conditions, RpoE is sequestered by the anti-sigma factor ChrR [16, 17], which is degraded upon singlet oxygen stress by the proteases DegS and RseP [18]. The RSP_2681 protein was also annotated as RpoE, but shares only 37% similarity with RSP_1092. No role of RSP_2681 in stress responses was reported to date. ChrR homologs, are also present in some Alphaproteobacteria besides Rhodobacter but lacking in others. While many Alphaproteobacteria harbour the PhyR-NepR-σ (EcfG) cascade as well as an RpoE/ChrR system, others like Methylobacterium extorquens or Brucella melitensis harbour the PhyR-NepR-σ (EcfG) cascade but lack a ChrR homolog.

A single PhyR homolog, RSP_1274, is also encoded in the R. sphaeroides chromosome and considering its function in other Alphaproteobacteria we hypothesized a function in the oxidative stress response and / or heat shock response in this organism. Our results do not support an important function of PhyR in theses stress responses for this phototrophic bacterium, however we observed an effect on membrane stress.

Methods

Bacterial strains and growth conditions

All strains, plasmids, and oligonucleotides used in this study are listed in Additional file 1. Rhodobacter sphaeroides was cultivated at 32 °C in minimal malate salt medium [19] in 50 ml Erlenmeyer flasks or flat bottles. Aerobic growth conditions with 160 to 180 μM dissolved oxygen were established by continuous shaking of 50 ml Erlenmeyer baffled flasks with 20 ml medium at 140 rpm or by gassing air into cultures in flat bottles. For microaerobic growth conditions, Erlenmeyer flasks with a culture volume of 80% were agitated at 140 rpm, resulting in a constant dissolved oxygen concentration of 25 to 30 μM during the mid exponential growth phase. Escherichia coli strains were grown in LB medium at 37 °C with shaking at 180 rpm or on solid growth medium, which contained 1.6% (w/v) agar. As necessary, antibiotics were added into liquid or solid medium at the following concentration: kanamycin (25 μg ml− 1); spectinomycin (10 μg ml− 1); tetracycline (1.5 μg ml− 1) (for R. sphaeroides); trimethoprim (50 μg ml− 1). Antibiotics were omitted from cultures and agar plates used for R. sphaeroides during stress experiments and zone inhibition assay.

Construction of a R. sphaeroides PhyR (RSP_1274) deletion mutant

Rhodobacter sphaeroides strain ΔPhyR was generated by transferring the suicide plasmid pPHUΔRSP_1274:Sp into R. sphaeroides 2.4.1, and screening for insertion of the spectinomycin resistance cassette into the chromosome by homologous recombination. The suicide plasmid was also transferred to the strains ΔChrR and TF18 (lacks RpoE and ChrR) to create the respective double and triple mutant. Parts of the RSP_1274 gene of R. sphaeroides 2.4.1, together with upstream and downstream sequences were amplified by polymerase chain reaction (PCR) using oligonucleotides 1274_for1_KpnI/1274_rev2_EcoRI and 1274_for3_EcoRI/1274_rev4_XbaI (Additional file 1: Table S2). The amplified PCR fragments were cloned into the XbaI-EcoRI and EcoRI-KpnI sites of suicide plasmid pPHU281 [20], generating plasmid pPHUΔRSP_1274. A 2.2 kb fragment containing the spectinomycin cassette from pHP45Ω [21] was inserted into the EcoRI site of pPHUΔRSP_1274 to generate pPHUΔRSP_1274:Sp. This plasmid was transferred into E. coli strain S17-I and biparentally conjugated into R. sphaeroides 2.4.1 wild-type strain. Conjugants were selected on malate minimal medium agar plates containing spectinomycin (10 μg ml− 1) and subsequently tested for tetracycline sensitivity.

RNA extraction, northern blot analysis, and RNAseq

For Northern blot analysis, samples from stress experiment were collected before (0 min) and 7 min after organic peroxide (360 μM of t-Butyl hydroperoxide (t-BOOH)) stress condition. Total RNA was isolated by the hot phenol method [22]. A total of 10 μg RNA was loaded per lane and separated on 10% polyacrylamide gels containing 7 M urea. RNA was transferred onto Biodyne B 0.45-μm membranes (Pall) by semidry electroblotting. For detection of Pos19 (photo-oxidative stress induced sRNA 19), the end-labeled oligonucleotide p-0019 was used. Membranes were exposed on phospho imaging screens (Bio-Rad) and analyzed with the Quantity One software (Bio-Rad).

RNAseq and dRNAseq are described in Remes et al. [23]. The RNA-seq data are available at the NCBI Gene Expression Omnibus database under accession number GSE71844.

Quantitative real-time RT-PCR

The one step Brilliant III Ultra-Fast SYBR® QRT-PCR Master Mix Kit (Agilent) was used for reverse transcription followed by PCR as described in the manufacturer’s manual. For real-time RT-PCR, a final concentration of 4 ng μl− 1 of total RNA was run in a C1000 Thermal cycler (Bio-Rad) for relative quantification of mRNAs in each of the three independent experiments. Relative expression of target genes was calculated relative to expression of untreated samples and normalized to the housekeeping gene rpoZ according to Pfaffl [24].

β-galactosidase assay

β-galactosidase assays were performed according to the method of Miller [25, 26]. At least three biological repeats were measured. In brief, cultures were inoculated from a single colony into 40 ml minimal salt medium and grown under microaerobic growth condition. Cultures were diluted to an OD660 of 0.2 in a flat bottle and allowed to double once under aerobic growth conditions in darkness. β-galactosidase activity was measured before (0 h), 1 and 3 h under singlet oxygen (high light 880 W m− 2 and 50 nM methylene blue) and organic peroxide (360 μM of t-Butyl hydroperoxide (t-BOOH)) stress.

Zone of inhibition assay

All strains were grown in minimal salt medium to OD660 of 0.5. 500 μl of culture were mixed with 5 ml pre-warmed soft agar and poured on solid minimal salt medium. Disks were placed at the center of plates, 5 μl of 10 mM methylene blue, 750 mM t-BOOH, 700 mM diamide or 700 mM paraquat were added on the filter disks. The plates of methylene blue were incubated in the light (60 W lamp), other plates were incubated in the dark. The diameters of growth inhibition areas were measured after incubation at 32 °C for 3 days [18].

Ultraviolet assay

All strains were cultured to an OD660 of 0.7–0.8 and diluted to a final dilution of 0.5 × 10− 6. 50 μl of the dilution was distributed on four plates. Plates were exposed to Ultraviolet (UV) light of 100 J m− 2 (254 nm) and incubated under the indicated temperature in the light or in the dark. The plates incubated in the light overnight were then transferred to the dark. Survival rates of UV exposed cells compared with non-exposed cells were calculated after incubating 3 days in the dark.

Survival assays

All strains were cultured to an OD660 of 0.5 in microaerobic condition. For spot survival assays, ethanol (12%), SDS (0.015%) and EDTA (30 mM), polymyxin B (2.5 μg/ml) or CCCP (25 μM) were added. Growth of viable cells was monitored by spotting 5 μl from consecutive 10-fold dilutions onto agar plates after 30 and 60 min of ethanol, 15 and 30 min of SDS and EDTA. For counting survival cell numbers, cultures treated in the same way or treated for 60 min with polymyxin B (2.5 μg/ml), or 60 min of CCCP(25 μM) were further dilutes 10− 1 to 10− 6.(dependent on the strain) and 50 μl were spread on the plate. Colonies were counted after three days incubation at 32 °C in the dark.

Results

Genomic context of the phyR gene RSP_1274

PhyR consisting of an amino terminal effector domain and a carboxy-terminal domain, was described first in Methylobacterium extorquens AM1 as a protein that regulates genes expression involved in general stress response. The amino terminal effector domain is a conserved general stress regulator in Alphaproteobacteria orthologous to sigma factors SigT of Caulobacter crescentus [4] and RpoE2 of Sinorhizobium meliloti [27]. The carboxy-terminal domain is also conserved. Phosphorylation of an aspartate in this domain leads to a conformational change which subsequently activates the effector domain [28]. Gene RSP_1274 of R. sphaeroides encodes a protein that shares 50%, 49%, 49% and 48% identity with the PhyR proteins from Sinorhizobium meliloti, Methylobacterium extorquens AM1, Bradyrhizobium japonicum and Caulobacter crescentus, respectively. As found for several other Alphaproteobacteria [1], an RNA polymerase sigma factor (RSP_1272) and a sensor histidine kinase (RSP_1271) are encoded upstream of the phyR gene (RSP_1274) on the opposite strand (Fig. 1). The RSP_1272 gene product shares 45% identity with RpoE2 from S. meliloti and 42% identity with SigT from C. crescentus. Other bacteria harbour the gene for the 61 aa NepR protein between the ecfG gene and the phyR gene. In this position R. sphaeroides encodes a 68 aa protein with 25% identity to NepR. Downstream of phyR a protein of the Crp-Fnr family is encoded. The gene arrangement for R. sphaeroides and selected Alphaproteobacteria is shown in Fig. 1.

Fig. 1
figure 1

Genetic organization of the phyR locus in different Alphaproteobacteria

Full size image

Differential RNA sequencing (dRNAseq) revealed the transcriptional organization for these genes. This method compares RNA samples, which were treated with terminal exonuclease with untreated samples and thus discriminates between RNA 5′ ends with triphosphate (TSS, transcriptional start sites) and RNA with monophosphate at the 5′ end (processing sites) [29]. The data strongly suggest that RSP_1273 and RSP_1272 are transcribed together from a promoter which is located upstream of RSP_1273, opposite to the coding region of phyR (Additional file 2). Another TSS seems to be present for RSP_1271, however slightly downstream of the ATG start codon of the annotated protein. This ATG overlaps with the TAG terminator codon of RSP_1272 suggesting translational coupling.

PhyR affects the resistance of R. sphaeroides to membrane stress but not to oxidative stress

In order to test the function of the PhyR homolog in R. sphaeroides stress response, we constructed a deletion strain which lacks the phyR gene and has a spectinomycin cassette inserted instead the doubling time of strain. ΔPhyR did not differ from the doubling time of the parental wild type 2.4.1 in aerobic conditions or during exposure to singlet oxygen (data not shown). There was also no difference in doubling time during osmotic stress (200 mM NaCl) in aerobic growth condition. Wild type and mutant strain also showed identical absorption spectra excluding an effect of PhyR on photosynthesis gene expression (data not shown).

Zone inhibition assays revealed no difference in resistance to singlet oxygen, t-BOOH, superoxide, or diamide between wild type and mutant (Fig. 2). Reaction of singlet oxygen with proteins, lipids and photopigments results in both direct damage and the formation of long-lived reactive organic peroxides. t-BOOH is used as model organic peroxide in our assays. In the same experiments we observed significantly increased resistance of a mutant lacking ChrR (RSP_1093) and decreased resistance in a mutant lacking RpoE (RSP_1092) and ChrR (strain TF18) against these stress factors (Fig. 2). Strains lacking ChrR and PhyR together or lacking RpoE/ChrR/PhyR showed the same stress resistance as the parental strains harbouring PhyR (Fig. 2). Thus an additional lack of PhyR did not alter the effects of higher RpoE activity (strain ΔChrR) or a lack of RpoE (strain TF18).

Fig. 2
figure 2

Inhibition of growth of the R. sphaeroides wild type strain 2.4.1 and of mutants lacking the phyR or chrR gene or chrR and rpoE (strain TF18) or phyR together with chrR or rpoE/chrR. Following agents were used in zone inhibition assays: t-BOOH (white bars), methylene blue (black bars), diamide (dark grey bars), or paraquat (light grey bars). Error bars indicate the standard deviation of zones of inhibition from three biological replicates. According to Students t-test none of the changes between PhyR deletion strains and the parental strains is significant (P ≤ 0.05)

Full size image

Since PhyR-dependent signaling also contributes to the heat stress response in other bacteria [6], we tested the effect of the deletion on growth at elevated temperature. Plates were incubated at 42 °C for 24 h after streaking and further incubation was at 32 °C for 72 h. These plates were compared to plates permanently grown at 32 °C. As previously described, we observed a clear growth defect of strains lacking RpoHI or RpoHI and RpoHII after heat stress [13]. Both sigma factors are known to be involved in the heat shock response of R. sphaeroides [12, 14]. There was no significant effect of elevated temperature on growth of the strain lacking PhyR (Additional file 3).

When R. sphaeroides cells were exposed to UV light and kept in the dark, the survival rate of the mutant strain was significantly lower than that of the wild type. When the cells were incubated in the light allowing photoreactivation, no difference in survival rates was observed for the two strains (Fig. 3). This suggests that PhyR does not affect the photolyase activity of R. sphaeroides [30, 31], but rather processes for UV repair that are light-independent.

Fig. 3
figure 3

Survival of R. sphaeroides after UV light exposure. Cells of the indicated strains were exposed to UV light of 100 J m− 2 (254 nm), spread on agar plates and kept in the light (60 W lamp) or in the dark. The survival rates are given as the mean of three experiments with standard deviation. ** indicates a highly significant change (P ≤ 0.01) according to Students t-test

Full size image

Additional experiments addressed the resistance of R. sphaeroides wild type and mutant to membrane stress. Survival was monitored by spot plating assays following treatment with SDS and EDTA or treatment with ethanol. After 15 min of treatment with 0.015% SDS and 30 mM EDTA survival of the wild type was much better than that of the mutant (Additional file 4A) indicating a role of the PhyR homolog in defense of membrane stress. 30 min or 60 min after treatment with ethanol the wild type showed slightly better survival than the ΔPhyR mutant (Additional file 4B). When the PhyR mutant was complemented by a plasmid-encoded phyR gene (strain 2.4.1ΔPhyR (pBE::phyR eCFP)), survival was identical as for the wild type for both stresses (Additional file 4A and B).

We complemented these experiments by spread plating assays which confirmed a significant effect of PhyR on the survival in presence of SDS/EDTA or ethanol (Fig. 4c). We also included treatment with polymyxin that alters membrane permeability and the uncoupling reagent CCCP (carbonyl cyanide m-chlorophenyl hydrazone). The effect of the phyR mutation for these reagents was less pronounced than for SDS/EDTA or ethanol, nevertheless the survival rate in the mutant was significantly lower than in the wild type strain (Fig. 4).

Fig. 4
figure 4

Survival rates of R. sphaeroides strains as determined by spread plating assays and colony counting before and after 15 min of addition of SDS (0,015%) and EDTA (30 mM), 30 min of ethanol (12%), 60 min of polymyxin (2.5 mg/ml), or 60 min of CCCP (25 μM). Error bars indicate the standard deviation from three biological replicates. ** indicates a highly significant change (P ≤ 0.01) according to Students t-test

Full size image

To test whether the effect of PhyR on membrane stress is mediated via the RpoE/ChrR system we also constructed mutants lacking ChrR or ChrR/RpoE together with PhyR. The results clearly demonstrate that the effect on membrane stress is solely due to PhyR. Strains ΔChrR and TF18 showed the same resistance to membrane stress as the wild type (Fig. 4).

Effect of R. sphaeroides PhyR on RpoE-dependent gene activation

RpoE (RSP_1092) of R. sphaeroides has an important function in the singlet oxygen stress response and also activates RpoHII in the general stress response. To elucidate a possible effect of PhyR on the activity of this sigma factor, we determined the β-galactosidase activity of a phrA-lacZ reporter plasmid. The phrA gene encodes a photolyase and is under direct control of RpoE. In strain TF18, which lacks RpoE and ChrR the β-galactosidase levels are very low in absence or presence of singlet oxygen (generated by addition of methylene blue and illumination) [30]. The ΔPhyR deletion strain showed an increase in β-galactosidase activity after 1 h and 3 h of singlet oxygen stress, which was however less than observed for the wild type (Fig. 5a). As a consequence the wild type showed significantly higher activity after 1 h and 3 h of stress (Fig. 5a). phrA-lacZ activity was also monitored after treatment with t-BOOH (Fig. 5b). The β-galactosidase activity was significantly lower in the mutant after 1 h and 3 h of stress compared to the wild type. When the phyR gene was introduced into the strain that has phyR deleted from the chromosome, we observed ß-galactosidase activity similar or higher to that of the wild type, proving that altered RpoE activity in the mutant strain is indeed a consequence of the lack of PhyR.

Fig. 5
figure 5

β-galactosidase actvity of R. sphaeroides strains harboring the reporter plasmid pPHUphrAlacZ. Cells were grown in aerobic conditions in the dark and were exposed to high light intensity (880 W m− 2) and 50 nM methylene blue (a) or to 360 μM of t-BOOH (b) for the indicated time periods. The mean of three experiments and standard deviations are shown. * indicates a significant change (P ≤ 0.05), ** indicates a highly significant change (P ≤ 0.01) according to Students t-test

Full size image

Another gene, which is under direct control of RpoE encodes the small RNA Pos19 [32, 33]. Northern Blot analysis revealed a stronger induction of Pos19 in the wild type compared to the phyR mutant upon treatment with organic peroxide (Additional file 5).

Effect of R. sphaeroides PhyR on stress-dependent mRNA levels

Our data strongly indicate reduced RpoE-dependent gene activation in the strain lacking PhyR. Real time RT-PCR analyses revealed that the rpoE mRNA level after treatment with t-BOOH increases significantly stronger in the wild type and in the complemented mutant than in ΔPhyR (Fig. 6a). We also tested the effect of PhyR on some other genes with a role in the oxidative stress response in R. sphaeroides [34, 35]. After 7 min of hydrogen peroxide the catA (RSP_2779) mRNA level for catalase increased significantly more in the PhyR mutant than in the wild type (Fig. 6c). For the mRNAs gloA (RSP_0392) and gloB (RSP_2294) for putative glyoxalases, higher induction in response to singlet oxygen was however observed for the wild type and the complemented PhyR mutant compared to the PhyR mutant(Fig. 6b). The gloA and gloB genes are preceded by RpoHII-dependent promoters (14), while catA is not preceded by a promoter sequence for an alternative sigma factor. Since the rpoHII gene is under control of RpoE, changed expression levels of gloA and gloB mRNAs are most likely a consequence of the PhyR effect on RpoE activity.

Fig. 6
figure 6

a Levels of rpoE mRNA as determined by real time RT PCR in wild type 2.4.1 (white bars), phyR mutant (black bars) and complemented strain (grey bars). The fold change of 7 min versus 0 min under t-BOOH after normalization to rpoZ mRNA level is shown. b Levels of relative expression are shown for catA in response to hydrogen peroxide and gloA, gloB in response to 1O2 exposure in wild type 2.4.1 (white bars), phyR mutant (black bars) and complemented strain (grey bars). Exposure to hydrogen peroxide or 1O2 was performed for 7 min. Values for relative expression levels represent the increase in gene expression compared to that of the control at time point 0 min and were normalized to mRNA levels determined for rpoZ. The mean of three experiments with standard deviation is shown. c Relative rpoE mRNA levels under different stress conditions as determined by real time RT PCR in the wild type and the mutant lacking PhyR. * indicate a significant change (P ≤ 0.05), ** indicate a highly significant change (P ≤ 0.01) according to Students t-test

Full size image

We also followed rpoE mRNA levels in response to other stress conditions in the wild type and the ΔPhyR mutant strain (Fig. 6c). A significant difference in rpoE mRNA level between the two strains was observed for singlet oxygen stress (methylene blue in the light). Diamide, heat, SDS/EDTA or ethanol had only minor effects on rpoE mRNA levels, while superoxide (paraquat treatment) and hydrogen peroxide resulted in increased rpoE mRNA levels, which were slightly lower in the mutant, but these differences were statistically not significant (Fig. 6b). We also tested how diverse stresses affect phyR and RSP_1272 (−σ (EcfG)) mRNA levels by real time RT PCR (Fig. 7). While singlet oxygen (methylene blue), superoxide (paraquat) and diamide resulted in 2–3 fold increase of the phyR mRNA level, none of the other stresses including membrane stress led to marked changes in phyR mRNA levels. Membrane stress was however the only stress factor that resulted in increased RSP_1272 mRNA levels and this increase is clearly dependent on PhyR.

Fig. 7
figure 7

Levels of phyR (a) and RSP_1272 (b) mRNAs under various stresses as determined by real time RT PCR. The following reagents were added to aerobic cultures at OD660 of 0.4 and samples were collected immediately before (0 min) and 7 min after addition: 0.2 μM methylene blue and high light (880 W m− 2), 360 μM t-BOOH, 1 mM H2O2, 250 μM paraquat, 500 μM diamide, 500 mM NaCl, or 10 μM CdCl2. For heat shock, microaerobic cultures were shifted to 42 °C at time 0 min or the following reagents were added: 0.005% SDS and 1 mM EDTA, 2.5% ethanol, 1 μg/ml polymyxin B or 10 μM CCCP, and samples were collected at 0 min and 7 min. The mean of three experiments with standard deviation is shown. * indicate a significant change (P ≤ 0.05), ** indicate a highly significant change (P ≤ 0.01) according to Students t-test

Full size image

Discussion

An important role for the PhyR-NepR-σ (EcfG) cascade in regulating the general stress response was observed in several members of the Alphaproteobacteria [1]. Genes encoding proteins with good homology to the proteins of the PhyR-NepR-σ (EcfG) cascade are also present in R. sphaeroides and are found in a similar chromosomal arrangement (Fig. 1). A similar role of PhyR in stress responses of R. sphaeroides as in other Alphaproteobacteria was conceivable. A R. sphaeroides strain lacking PhyR showed similar response as the parental wild type to singlet oxygen, hydrogen peroxide, superoxide, organic peroxides, diamide, heat or salt stress. We conclude that other proteins like the alternative sigma factors RpoE, RpoHI, and RpoHII are indeed the main regulators of the general stress response in R. sphaeroides.

Our data demonstrate however that expression of the phyR gene is modulated in response to some stresses and we found significant differences between wild type and mutant in response to membrane and UV stress. This suggests that PhyR has a more specialized role in R. sphaeroides. Up to now, the role of PhyR in stress responses was not analyzed for any member of the Rhodobacterales. Therefore it is possible that a more specialized function of PhyR is not limited to R. sphaeroides but may apply to a certain sub-branch of the Alphaproteobacteria.

While the phyR mRNA levels did not respond to membrane stress, the levels of the RSP_1272 mRNA encoding the σ(EcfG) protein of the PhyR-NepR-σ (EcfG) locus, was clearly increased in response to membrane stress, but not by other stress factors. This supports the view that the PhyR-NepR-σ (EcfG) cascade in R. sphaeroides has a main role in the defense of membrane stress. Increased RSP_1272 mRNA levels were dependent on PhyR in agreement with signal transfer within the PhyR-NepR-σ (EcfG) cascade.

Furthermore our data provide clear evidence for an effect of PhyR on rpoE mRNA level and RpoE-dependent gene activation (Figs. 5, 6 and Additional file 5). Why is the influence of PhyR on RpoE activity not manifested as altered stress resistance in the mutant? In mutants lacking ChrR or RpoE and ChrR together (strain TF18) the change in rpoE mRNA level is of course much more pronounced than in the PhyR mutant under the tested conditions. ΔChrR or TF18 mutants are clearly affected in stress resistance, but additional mutation of phyR did not increase this effect implying that PhyR can not even partially compensate the loss of these main regulators. Many investigations in the past have demonstrated that complex regulatory networks including protein and RNA regulators are involved in controlling and balancing stress responses [36,37,38]. The moderate changes of RpoE activity caused by the lack of PhyR may not be sufficient to cause a clear phenotype since compensation by other players in the regulatory network may take place. In R. sphaeroides the PhyR-NepR-σ(EcfG) cascade may also have a function in balancing some stress responses rather than triggering such responses. A cross talk of the PhyR-NepR-σ(EcfG) cascade to the RpoE/ChrR system has not been reported for other Alphaproteobacteria to date.

Conclusions

Our results demonstrate that PhyR has no major role in the general stress response of the Alphaproteobacterium R. sphaeroides as reported before for other bacterial species [1, 39, 40]. We could attribute a role of PhyR in defense of membrane stress and survival of UV light in the dark in R. sphaeroides, supporting a more specialized function in this bacterium. PhyR has no major contribution to the complex regulatory network (e.g.: 8, 9, 15, 29, 30) of protein and sRNA regulators that control the oxidative stress response in R. sphaeroides.