Skip to main content
NCBI home page
Microbiology logoLink to Microbiology
. 2015 Nov;161(Pt 11):2256–2264. doi: 10.1099/mic.0.000172

DNA-binding properties of a cGMP-binding CRP homologue that controls development of metabolically dormant cysts of Rhodospirillum centenum

Sugata Roychowdhury 1,†, Qian Dong 1,†, Carl E Bauer 1,✉
PMCID: PMC4806592  PMID: 26362215

Abstract

Rhodospirillum centenum utilizes 3′,5′-cyclic guanosine monophosphate (cGMP) as a messenger to regulate development of desiccation-resistant cysts. In this study, we demonstrated that gcyA, gcyB and gcyC, coding for putative subunits of a guanylyl cyclase, increase expression from 8- to 500-fold when cells transition from vegetative to cyst phases of growth. This induction did not occur in a strain that is defective in cGMP synthesis or in a strain that contains a deletion of cgrA that codes for a cGMP-binding homologue of Escherichia coli catabolite repressor protein (CRP). We also demonstrated that cgrA auto-induces its own expression in the presence of cGMP, indicating that a feed-forward loop is used to ramp up cGMP production as cells undergo encystment. Inspection of an intragenic region upstream of gcyB revealed a sequence that is identical to the CRP consensus sequence from E. coli. DNase I and fluorescence anisotropy analyses demonstrated that CgrA bound to this target sequence at a protein : cGMP ratio of 1 : 2 with Kd ∼61 nM. This was in contrast to CgrA in the presence of cAMP, which exhibited Kd ∼1795 nM. CgrA thus constitutes a novel variant of CRP that utilizes cGMP to regulate production of cGMP synthase for the control of cyst development.

Introduction

We recently established that several members of the Azospirillum clade, specifically Azospirillum brasilense, and the photosynthetic member Rhodospirillum centenum (Fani et al., 1995), utilize cGMP as a signalling molecule to control development of desiccation-resistant cysts (Marden et al., 2011; Gomelsky, 2011). Azospirillum spp. are known members of the plant root rhizosphere, where they associate with plant roots and provide plants with bacterially fixed nitrogen and several bacterially synthesized plant hormones (Bashan et al., 2004). Cyst production by members of this genus, and by other Gram-negative soil micro-organisms such as Azotobacter spp., allows survival during periods of desiccation and nutrient limitation (Berleman & Bauer, 2004).

The use of cGMP as a signalling molecule by Azospirillum spp. was surprising as cGMP was not thought to have a role in controlling bacterial physiology (Gomelsky, 2011). However, this assumption was not correct as the plant pathogen Xanthomonas campestris has also recently been shown to utilize cGMP as a signalling molecule to regulate bacterial phytopathogenesis (An et al., 2013; Gomelsky & Galperin, 2013). There is also evidence for the production of cGMP by Synechocystis, where it is thought to be used for UV-B-induced photoacclimation (Cadoret et al., 2005; Rauch et al., 2008). Ryu et al. (2015) also recently described the construction of an Escherichia coli reporter strain that allows detection of bacterial guanylyl cyclases and have confirmed that Azs. brasilense and Rhs. centenum do indeed express a guanylyl cyclase.

Among sequenced members of the Azospirillum clade, there is a conserved cluster of three guanylyl cyclase genes, gcyA, gcyB and gcyC, that are required for cGMP synthesis (Marden et al., 2011). The GcyA peptide is clearly the catalytic subunit as it is capable of synthesizing cGMP from GTP alone in vitro. Synthesis of GcyB and GcyC is required for in vivo production of cGMP; however, their roles are not yet established (Marden et al., 2011). Linked to the guanylyl cyclase genes is cgrA, which codes for a CRP homologue that binds cGMP preferentially over cAMP (Marden et al., 2011). A chromosomal deletion of cgrA leads to loss of cGMP production, suggesting that CgrA is required for expression of the gcy gene cluster (Marden et al., 2011).

Members of the CRP family of transcription factors are known to respond to a variety of small signalling molecules such as cyclic nucleotides cAMP, cGMP and cyclic-di-GMP (c-di-GMP) as well as to the gases O2, NO and CO (Marden et al., 2011; Crack et al., 2004; Spiro & Guest, 1990; Zumft, 2002; Shelver et al., 1995, 1997; Körner et al., 2003). In several species tested, CRP homologues act as global regulators to control a wide range of physiological functions, from carbon metabolism, aerobic/anaerobic physiology, flagella biosynthesis, pathogenesis, and biofilm formation to bacterial development (Marden et al., 2011; Crack et al., 2004; Spiro & Guest, 1990; Shimada et al., 2011; Bian et al., 2013; Österberg et al., 2013; An et al., 2013). CRP from E. coli is best studied where it regulates genes involved in carbon metabolism in response to cAMP synthesis in a process known as catabolite repression (Shimada et al., 2011; Aiba 1986). In this process, a sugar such as glucose is transported into the cell by the sugar phosphotransferase system, which also phosphorylates the sugar during transport. The phosphotransferase system transport complex can sense the efficiency of glycolysis by monitoring the phosphoenolpyruvate : pyruvate ratio and in response can regulate the activity of adenylyl cyclase (Postma et al., 1993; Kolb et al., 1993). CRP in the absence of cAMP (apo-CRP) cannot effectively bind DNA because recognition helix F in the DNA-binding domain is not oriented properly to interact with the major groove of DNA (Görke & Stülke, 2008; Popovych et al., 2009). However, when CRP binds cAMP there is a coil-to-helix conformation change that extends a coiled-coil dimerization interface which promotes a 60° rotation of the F-helix. This conformational change allows the DNA-binding domain to properly interact with a conserved 22 bp palindrome sequence (Kim et al., 1992; Lawson et al., 2004; Harman, 2001; Passner et al., 2000; Sharma et al., 2009).

An interesting variation of the E. coli CRP paradigm exists in Xanthomonas, where the CRP homologue Clp exhibits specific DNA binding in the absence of its effector molecule, c-di-GMP (Tao et al., 2010). The binding of c-di-GMP to Clp results in a conformational change that releases it from DNA (Tao et al., 2010; Chin et al., 2010; Leduc & Roberts, 2009). Thus, c-di-GMP functions as an inhibitory ligand for Clp, which is converse to the activating role of cAMP for CRP. The opposing effects of cAMP and c-di-GMP on the DNA-binding activities of CRP and Clp highlight that CRP homologues can exhibit very different output responses when interacting with different cyclic nucleotides.

In this study, we demonstrate that expression of the three genes needed for cGMP synthesis, gcyA, gcyB and gcyC, continually ramps up as cells transition from vegetative to cyst form. Expression of gcyA, gcyB and gcyC is also shown to be dependent on the presence of both cGMP and CgrA. Given that cGMP continuously rises during cyst development (Marden et al., 2011), these results indicate that a cGMP-dependent feed-forward loop drives cGMP synthesis. We have also utilized DNase I footprint analyses to identify a DNA-binding site for CgrA upstream of the gcyB-gcyC-cgrA gene cluster. DNA-binding studies using fluorescence anisotropy demonstrate that cGMP stimulates the binding of CgrA to its target DNA sequence 25-fold. Collectively, these results indicate that CgrA exhibits similar properties to those described for E. coli CRP, with the caveat that CgrA responds to cGMP instead of cAMP to regulate its DNA-binding activity.

Methods

Bacterial strains, media, and growth conditions

WT, and ΔgcyA and ΔcgrA in-frame deletion strains of Rhs. centenum utilized in this study have been described previously (Marden et al., 2011). All Rhs. centenum strains were cultured aerobically in vegetative CENS medium or in cyst-inducing CENBA (Stadtwald-Demchick et al., 1990) or CENS-8xN (Marden et al., 2011) medium at 37 °C when grown in liquid medium or at 42 °C when grown in agar-solidified medium. E. coli strain DH5α was used for cloning, BL21 Rosetta II (Invitrogen) for overexpression of CgrA, and S17-1 λ pir for plasmid conjugation. E. coli strains were grown in Luria–Bertani medium at 37 or 16 °C for protein overexpression, with appropriate antibiotics for plasmid selection.

RNA isolation, quantitative reverse transcription PCR (qRT-PCR)

Strains were grown overnight in non-cyst-inducing CENS medium, subcultured into CENS or cyst-inducing minimal CENBA medium at a 1 : 50 ratio, and then grown until OD600 ∼0.4–0.6 at 37 °C. RNA was extracted using FastRNA Pro Blue kit from MP Biomedicals, followed by further purification using a commercial RNA clean-up protocol (RNeasy Mini kit; Qiagen). RNA samples were then treated with TURBO DNase (Ambion) to remove contaminating genomic DNA, with final RNA concentrations measured using NanoDrop (Thermo Scientific). Samples with an OD260 to OD280 ratio >1.8 were considered for further analysis. qRT-PCR analysis of gene expression was performed with 8 ng RNA sample in a 20 μl reaction using the SensiFAST SYBR Hi-ROX kit (Bioline). Synthesis of complementary DNA was carried out at 45 °C for 10 min, followed by qRT-PCR. A two-step PCR was used, consisting of DNA strand separation at 95 °C for 5 s, followed by primer annealing and DNA polymerase elongation at 60 °C for 20 s. All PCRs contained the housekeeping gene rpoZ as an internal control, which was chosen based on BestKeeper analysis (Pfaffl et al., 2004), which indicated minimal variability in expression across different growth conditions. The list of primers used in this study is provided in Table S1, available in the online Supplementary Material.

Overexpression and purification of CgrA

CgrA purification was carried out as described by Marden et al. (2011), with the exception that CgrA was eluted from the size-exclusion column in 20 mM Tris, pH 7.5, 100 mM NaCl, 5 mM DTT and 15 % glycerol v/v. Protein concentration was assayed with a Bradford assay kit (Bio-Rad) and stored at 4 °C for all subsequent analyses.

DNase I footprinting

A 215 bp double-stranded DNA was used for this study that represented the intergenic site between RC1_3784 and RC1_3785. DNA was amplified by PCR using the 5′ FAM-labelled forward primer SiteVI FwdFP and 5′ HEX-labelled reverse primer SiteVI RevFP (Table S1). PCR products were cleaned using the QIAquick PCR purification kit (Qiagen) and eluted in EB buffer (supplied with the kit). Cyclic nucleotide (Sigma) stock solutions were made in water. Footprint analyses were carried out as described previously (Willett et al., 2007). Briefly, 10 nM DNA probe was digested with DNase I (2 U μl− 1; NEB) in the presence of varying concentrations of protein–cGMP (1 : 2) in footprint buffer consisting of 20 mM Tris/HCl, pH 7.5, 3 mM MgCl2, 5 mM CaCl2, 100 mM NaCl and 50 μg ml− 1 BSA at 25 °C. Simultaneously, DNase I digestion was carried out with DNA only and in the presence of protein only. DNase I activity was quenched by the addition of 0.24 M EDTA, pH 8.0. The reaction products were purified using the MinElute PCR Purification kit (Qiagen) and eluted in EB buffer (supplied with the kit). One microlitre of GeneScan-500 Liz Size Standard (Applied Biosystems) was added to all the tubes. Samples were loaded in a natural-coloured 96-well plate (Thermo Fisher Scientific) and covered with a septum, heated on a thermocycler at 95 °C for 5 min and cooled immediately by placing on ice for at least 5 min. Sequencing reactions were run on an ABI Prism Gene Scanner 3730 and data were analysed using Peak Scanner v1.0 software.

Fluorescence anisotropy-based DNA-binding assay

A 35-mer 5′-fluorescein-tagged DNA probe containing the CRP consensus site (PcgrA) located at the intergenic region between RC1_3784 and RC1_3785 was used for determining the CgrA-binding constant to the target DNA. To test specific DNA-binding activity by CgrA, a non-specific probe (PbchC) from Rhodobacter capsulatus was also prepared. Sequence details of DNA probes used for fluorescent anisotropy experiments are listed in Table S1.

The double-stranded DNA probe was prepared by heating equimolar amounts of both DNA strands at 95 °C for 10 min, followed by slowly cooling to room temperature in the dark for 2 h. All reactions were performed in TE (10 mM Tris/HCl pH 8.0, 1.0 mM EDTA) buffer with 100 mM NaCl.

Fluorescence anisotropy experiments were carried out with a Synergy H1 Hybrid Multi-Mode Microplate Reader at 25 °C. DNA was incubated with increasing concentrations of CgrA in the presence of the same buffer as used in the DNase I footprint experiment, before loading onto a black 96-well flat-bottom plate (BD Falcon). The binding of the CgrA dimer (CgrA chromatographs as a dimer – data not shown) to DNA was performed by monitoring the change in the apparent fluorescence anisotropy of the 5′-fluorescein-tagged DNA. The excitation and emission wavelengths were measured at 485 and 528 nm, respectively, using a filter cube with 510 nm as the cut-off value. To test the effect of ligand on CgrA–DNA-binding activity, the assay was carried out in the presence of either cGMP or cAMP at a fixed 1 : 2 protein : ligand ratio. Background signal contributed by the buffer alone was subtracted from all readings. The anisotropy is calculated using the equation: r = (Iparallel–Iperpendicular)/(Iparallel+2Iperpendicular) (G = 1).

The sample volume per well was kept constant at 100 μl and measurements were repeated three times in duplicate reactions and from at least three different protein batches.

Results

Genes needed for cGMP synthesis increased expression during cyst formation

In a previous study, we demonstrated a continuous rise of cGMP levels in the culture supernatant as Rhs. centenum cells transition from vegetative to cyst form (Marden et al., 2011). Null mutations in any of three genes gcyA, gcyB or gcyC (Fig. 1) inhibited cGMP production and subsequent cyst formation. Inhibition of cGMP production, and cyst formation were also observed upon disruption of the linked gene cgrA, which codes for a cGMP-binding CRP homologue (Marden et al., 2011). This latter result suggested that CgrA regulates expression of guanylyl cyclase genes in response to cGMP levels in a positive manner.

Fig. 1.

Fig. 1.

Open in a new tab

Genetic organization of the gcyA–cgrA gene cluster regulating cGMP production and encystment in Rhs. centenum. White arrows are ORFs with unknown functions, black arrows indicate genes encoding putative subunits of the guanylyl cyclase, and the grey arrow denotes a gene that codes for a CRP homologue that binds cGMP.

To test whether guanylyl cyclase gene expression is indeed rising during the induction of cyst development, we assayed expression of gcyA, gcyB and gcyC using qRT-PCR. The results in Fig. 2 show that shifting WT cells from vegetative CENS medium into cyst-inducing minimal CENBA medium results in a rapid rise in gcyA, gcyB and gcyC expression over a 4 day period (cysts become visible by day 4). Interestingly, gcyA increases expression five- to eightfold, while gcyB and gcyC increase expression ∼500-fold. (The gene gcyA codes for the catalytic subunit of guanylyl cyclase, while gcyB and gcyC code for subunits with unknown functions that are required for synthesis of cGMP in vivo.) Expression of cgrA also rapidly increased approximately eightfold during the first few hours after the shift to cyst-inducing CENBA medium, but unlike continuous increases in gcy expression, cgrA expression did not increase much beyond the first 1 h time point (Fig. 2). This suggests that there is an undefined mechanism to limit the increase in cgrA expression.

Fig. 2.

Fig. 2.

Open in a new tab

Expression of gcyA, gcyB, gcyC and cgrA after a CENS to CENBA shift in the WT strain at periods shown during 4 days of growth. The values represent fold changes in mRNA levels in cells that underwent a nutrient downshift versus mRNA levels present in the cells just prior to the nutrient downshift. Expression levels were measured using qRT-PCR. These data are mean values from three independent experiments, with error bars representing se (n = 3). P-values were determined using t-tests, with *P < 0.05, **P < 0.01 and ***P < 0.001. The calculated P-values compare means between fold change of samples shifted to nutrient-poor CENBA (three replicates) relative to the fold changes of the vegetative control group in CENS (which is set to 1) with mRNA harvested just prior to the nutrient downshift.

cGMP stimulates guanylyl cyclase expression via CgrA

We tested if the observed increase in gcyB, gcyC and cgrA expression is dependent on cGMP by assaying expression in a ΔgcyA deletion strain that is defective in cGMP production. We also addressed whether expression was dependent on CgrA by assaying expression in a ΔcgrA deletion strain. The results in Fig. 3 show that an increase in gcyB and gcyC expression was observed in WT cells after a 1 h growth shift from vegetative CENS medium to CENBA encystment medium, but that this increase did not occur in the ΔgcyA mutant strain or in the ΔcgrA strain. In contrast, gcyA expression was not significantly affected by a deletion of CgrA, indicating that its increase in expression is dependent on other undefined encystment regulatory events. In assaying cgrA expression, we observed that a portion of the increase was dependent on cGMP production, as the WT strain had significantly higher levels of cgrA transcription than the ΔgcyA strain (Fig. 3). However, there was still a basal level of cgrA expression even in the absence of cGMP (Fig. 3).

Fig. 3.

Fig. 3.

Open in a new tab

Expression of gcyA, gcyB, gcyC and cgrA genes 1 h after a CENS to CENBA nutrient downshift in WT cells, in a ΔgcyA strain and in a ΔcgrA strain. The values represent fold changes in expression in cells that underwent a nutrient downshift versus cells that remained in rich CENS medium. Expression levels were measured using qRT-PCR. These data are mean values from three independent experiments, with error bars representing se (n = 3). P-values were determined using t-tests, with *P < 0.05, **P < 0.01 and ***P < 0.001.

Identification of a CgrA DNA recognition sequence

We identified a potential CgrA-binding site upstream of the Rc1_3785-gcyB-gcyC-cgrA gene cluster by visually scanning for a CRP-like palindrome sequence in the gcyA–cgrA gene cluster. As shown in Fig. 4, there are six sequences in this region (sites 1–6) that exhibit similarity to the E. coli CRP consensus binding sequence TGTGA-N6-TCACA (Shimada et al., 2011). In vitro binding of CgrA to DNA segments containing these palindromic sequences was assayed using gel mobility shift and fluorescence anisotropy analyses, with only site 2 exhibiting binding. This sequence has complete conservation of the CRP consensus sequence and is located in the intergenic region between divergent ORFs RC_13784 and RC1_3785. Furthermore, sequence analysis of a mini-Tn5 transposition mutation that renders cells deficient in cGMP production (strain HSM057; Marden et al., 2011) showed that the mini-Tn5 insertion occurs adjacent to this palindrome (strain HSM057; Marden et al., 2011). We therefore performed DNase I footprint analysis of CgrA binding to this intergenic region by PCR amplifying a 200 bp fluorescently labelled DNA segment. Footprint results for top- and bottom-strand DNA segments (Fig. 5a, b) showed that CgrA+cGMP protected numerous DNase I-generated peaks in the region that contains the TGTGA-N6-TCACA palindrome. The protection was not observed with CgrA alone or with CgrA incubated with the DNA probe in the presence of cAMP (Fig. 5).

Fig. 4.

Fig. 4.

Open in a new tab

Regions in the gcyA–cgrA gene cluster that contain sequences with similarity to the E. coli CRP consensus sequence. Capital bold bases denote conserved sequences. Carets denote location of a mini-Tn5 insertion that disrupts cGMP production (Marden et al., 2011).

Fig. 5.

Fig. 5.

Open in a new tab

DNase I footprint assay of the CRP-like palindrome present in the intergenic region between RC1_3784 and RC1_3785 (site 2 in Fig. 4). Fluorograms indicate the fluorescence intensities after DNase I digestion of the fragments containing the 5′-FAM-labelled coding strand (a) and 3′-HEX-labelled non-coding strand (b) in the presence of CgrA with or without cyclic nucleotides. The region containing arrows encompasses the sites protected from, as well as those hypersensitive to, DNase I digestion upon binding by CgrA.

DNA-binding affinity of CgrA is stimulated by cGMP in vitro

We determined the DNA-binding affinity of CgrA to the CgrA-protected palindrome using fluorescence anisotropy. This analysis was carried out with a 5′-fluorescein-labelled 35 bp DNA probe that contains the 5′-TGTGA-N6-TCACA-3′ palindrome (probe PcgrA) centred on the primer. We also tested the binding of CgrA to a 35 bp non-specific DNA control segment (probe PbchC) designed from the bchC promoter region from Rhb. capsulatus, which has a similar GC content to Rhs. centenum (Fig. 6). The results of this analysis showed that there was no significant binding of CgrA to the non-specific DNA control as well as no significant binding of CgrA in the absence of cGMP to the probe that contains the TGTGA-N6-TCACA palindrome (Fig. 6). In contrast, CgrA in the presence of cGMP at a 1 : 2 protein : ligand ratio exhibited a very good DNA-binding isotherm with a dissociation constant (Kd) of 61.3 ± 16.97 nM (Fig. 6). CgrA also bound to its target DNA in the presence of cAMP (also at a 1 : 2 ratio), but with a much weaker binding affinity, Kd = 1794.64 ± 179.4 nM, than observed with cGMP (Table 1).

Fig. 6.

Fig. 6.

Open in a new tab

Cyclic-nucleotide-dependent DNA binding by CgrA. Representative fluorescence anisotropy-based isotherms of CgrA binding to double-stranded DNA containing the palindromic region in site 2 (DNA probe PcgrA). CgrA–DNA binding was carried out at a 1 : 2 protein : ligand ratio with no nucleotide (squares), cAMP (filled circles) or cGMP (triangles) added to the reaction mixture and incubation with DNA for 30 min before taking the reading. As a control, a binding isotherm of a fluorescein-labelled non-specific DNA probe (open circles) was obtained as mentioned previously. The normalized anisotropy change calculated for either cGMP or cAMP at a 1 : 2 ratio and for apo-protein and non-specific DNA (DNA probe PbchC) was calculated relative to the total change observed in the presence of cGMP in the respective conditions. The black continuous line through each binding isotherm represents a nonlinear fit to the Hill equation. Kd values shown in Table 1 were generated as the mean of at least three binding experiments performed with three independent protein preparations.

Table 1. Summary of fluorescein anisotropy-based DNA-binding affinities of WT CgrA with or without cyclic nucleotides.

A final concentration of 10 nM DNA was used in all assays, with a fixed twofold molar excess of cGMP or cAMP present relative to the molar concentration of CgrA at each data point. nb, No binding detected.

DNA-binding complex Protein : cyclic nucleotide ratio Kd (nM)
apo-CgrA None added nb
cAMP–CgrA 1 : 2 1794.64 ± 179.4
cGMP–CgrA 1 : 2 61.3 ± 16.97
Open in a new tab

Discussion

In this study, we demonstrated that a nutrient downshift that promoted cyst formation also increased expression of gcyA, gcyB and gcyC, which are involved in cGMP production, and cgrA, which codes for a cGMP-binding homologue of CRP. These results provide a framework for a model to explain the role of cGMP in inducing encystment. As shown in Fig. 7, under nutrient-rich conditions there are low amounts of guanylyl cyclase and cGMP owing to low levels of gcyA, gcyB and gcyC expression. Under these conditions, CgrA+cGMP-dependent cyst developmental gene expression is not induced. However upon a nutrient downshift, there is a rapid increase in expression of gcyA, gcgB, gcyC and cgrA, leading to a rapid increase in further expression of these genes as well as increasing production of cGMP. Interestingly, a large amount of cGMP is actually excreted from the cell (Marden et al., 2011), with presumably enough cGMP retained inside the cell to allow the formation of a CgrA–cGMP complex that binds to target promoters.

Fig. 7.

Fig. 7.

Open in a new tab

Proposed model explaining regulation of cgrA and cyclase subunits. In the presence of high-nutrient-level non-cyst-inducing conditions, the expression of gcyA (which codes for the catalytic subunit of guanylyl cyclase), gcyB and gcyC, and cgrA are low, leading to low amounts of cGMP and low amounts of apo-CgrA. Under nutrient-starving cyst-inducing conditions, expression of gcyA and cgrA increases four- to fivefold while expression of gcyB and gcyC increases 100- to 500-fold. Under these conditions, GcyA presumably forms an active complex with GcyB and GcyC to synthesize large amounts of cGMP (light orange pentagons), a significant portion of which is transported out of the cell (Marden et al., 2011). CgrA forms an active complex with cGMP that triggers feed-forward further expression of the GcyB and GcyC cyclase components, CgrA itself and other genes necessary for encystment.

There are several outstanding questions regarding the regulation, biochemical composition and activity of this guanylyl cyclase gene cluster. For example, the expression of gcyA, which codes for the catalytic subunit (Marden et al., 2011), only increased approximately eightfold over the 4 day cyst induction period, while expression of gcyB and gcyC, which presumably code for non-catalytic subunits of the cyclase, increased ∼200- and ∼600-fold, respectively (Fig. 2). It is possible that the observed CgrA-dependent increase in gcyA expression was a result of indirect activation by CgrA as there was no observable CgrA-mediated footprint protection pattern upstream of gcyA (data not shown). This is in contrast to observable DNase I protection upstream of the Rc1_3785-gcyB-gcyC-cgrA gene cluster with CgrA+cGMP. There is also no information on the function of GcyB or GcyC other than their need for cGMP production in vivo (Marden et al., 2011) and no information on individual subunit stoichiometry, subunit stability or translation efficiency. Indeed it is possible that GcyB/GcyC transiently regulates guanylyl cyclase activity of GcyA, not unlike phosphorylation-dependent binding of EIIAGlc to adenylyl cyclase in E. coli (Feucht & Saier, 1980; Harwood et al., 1976). Clearly, future studies of the guanylyl cyclase enzyme complex are needed to understand why gcyB and gcyC expression is ramped up so much higher than gcyA expression during cyst cell development.

While guanylyl cyclase expression underwent a continuous rise over the 4 day period, the expression of CgrA rapidly (within 4 h) increased expression three- to fourfold and then remained at this level even while expression of the guanylyl cyclase subunits continued to rise (Fig. 2). This suggests the presence of a mechanism that limits cgrA expression at a three- to fourfold level during cyst induction. Such an event could involve feed-forward activation by CgrA+cGMP balanced with a feedback mechanism that is yet to be defined. Presumably, limiting the amount of cgrA expression ensures that the gene targeted by CgrA remains dependent on cGMP as a rise in CgrA to levels that are too high could potentially promote DNA binding in the absence of the cGMP.

The continuous increase in gcy expression that occurs as cyst cells develop suggests that a feed-forward increase in this signalling pathway may be used by these cells to lock-in this developmental pathway (Fig. 6). The feed-forward nature of gcy and cgrA expression differs from crp expression in E. coli, where adenylyl cyclase expression is repressed by CRP–cAMP to limit cAMP production (Aiba, 1985; Mori & Aiba, 1985; Ishizuka et al., 1994). In E. coli there is a basal amount of CRP expression that occurs irrespective of cAMP (Cossart & Gicquel-Sanzey, 1985); however, under conditions where cAMP is increased, then the CRP–cAMP complex binds and represses further CRP synthesis (Ishizuka et al., 1994; Aiba, 1983; Hanamura & Aiba, 1992). Glucose also exerts crp repression in a cAMP-independent way, suggesting that there might be more than one regulatory mechanism other than the direct involvement of the CRP–cAMP complex (Cossart & Gicquel-Sanzey, 1985). It should be noted that repeated gel shift and DNase I footprint analyses (not shown) have not identified any CgrA-binding sites located immediately upstream of cgrA or within its coding region. Thus, even though we cannot definitively rule out that CgrA does not bind immediately upstream of its coding sequence, it seems likely that its expression occurs as part of an operon that also includes gcyB and gcyC.

In the future, we need to address whether CgrA activated by cGMP binds upstream of genes involved in encystment. Recently we utilized RNA-Seq analysis to demonstrate that 812 genes (∼29 % of annotated genes) undergo an alteration in expression as Rhs. centenum transitions from vegetative to cyst form (Dong and Bauer, 2015). It will be of interest to utilize deep-sequencing approaches such as RNA-Seq and chromosome immunoprecipitation sequencing (ChIP-Seq) to determine how many of these genes are directly or indirectly regulated by CgrA. We also need to ascertain how the CgrA+cGMP regulatory circuit fits in with other known encystment regulators such as the novel Che-like signalling pathway (He & Bauer, 2014; He et al., 2013), histidine kinases (Din et al., 2011) and other cell cycle regulators that are known to have a role in controlling this developmental process.

Acknowledgements

We thank Daniel Kearns for helpful advice. This study was funded by a National Institutes of Health grant GM099703 awarded to C. E. B.

Supplementary Data

Supplementary Data

Click here for additional data file. (65.1KB, pdf)

Abbreviations:

cGMP

3′5′-cyclic guanosine monophosphate

CRP

catabolite repressor protein

Kd

dissociation constant

qRT-PCR

quantitative reverse transcription PCR

References

  1. Aiba H. (1983). Autoregulation of the Escherichia coli crp gene: CRP is a transcriptional repressor for its own gene Cell 32 141–149 10.1016/0092-8674(83)90504-4 . [DOI] [PubMed] [Google Scholar]
  2. Aiba H. (1985). Transcription of the Escherichia coli adenylate cyclase gene is negatively regulated by cAMP-cAMP receptor protein J Biol Chem 260 3063–3070 . [PubMed] [Google Scholar]
  3. Aiba H. (1986). Negative control of transcription by cAMP and cAMP receptor protein in Escherichia coli Adv Biophys 21 193–204 10.1016/0065-227X(86)90023-7 . [DOI] [PubMed] [Google Scholar]
  4. An S.Q., Chin K.H., Febrer M., McCarthy Y., Yang J.G., Liu C.L., Swarbreck D., Rogers J., Maxwell Dow J., other authors (2013). A cyclic GMP-dependent signalling pathway regulates bacterial phytopathogenesis EMBO J 32 2430–2438 10.1038/emboj.2013.165 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bashan Y., Holguin G., de-Bashan L.E. (2004). Azospirillum-plant relationships: physiological, molecular, agricultural, and environmental advances (1997-2003) Can J Microbiol 50 521–577 10.1139/w04-035 . [DOI] [PubMed] [Google Scholar]
  6. Berleman J.E., Bauer C.E. (2004). Characterization of cyst cell formation in the purple photosynthetic bacterium Rhodospirillum centenum Microbiology 150 383–390 10.1099/mic.0.26846-0 . [DOI] [PubMed] [Google Scholar]
  7. Bian J., Liu X., Cheng Y.-Q., Li C. (2013). Inactivation of cyclic di-GMP binding protein TDE0214 affects the motility, biofilm formation, and virulence of Treponema denticola J Bacteriol 195 3897–3905 10.1128/JB.00610-13 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cadoret J.C., Rousseau B., Perewoska I., Sicora C., Cheregi O., Vass I., Houmard J. (2005). Cyclic nucleotides, the photosynthetic apparatus and response to a UV-B stress in the cyanobacterium Synechocystis sp. PCC 6803 J Biol Chem 280 33935–33944 10.1074/jbc.M503153200 . [DOI] [PubMed] [Google Scholar]
  9. Chin K.H., Lee Y.C., Tu Z.L., Chen C.H., Tseng Y.H., Yang J.M., Ryan R.P., McCarthy Y., Dow J.M., other authors (2010). The cAMP receptor-like protein CLP is a novel c-di-GMP receptor linking cell-cell signaling to virulence gene expression in Xanthomonas campestris J Mol Biol 396 646–662 10.1016/j.jmb.2009.11.076 . [DOI] [PubMed] [Google Scholar]
  10. Cossart P., Gicquel-Sanzey B. (1985). Regulation of expression of the crp gene of Escherichia coli K-12: in vivo study J Bacteriol 161 454–457 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Crack J., Green J., Thomson A.J. (2004). Mechanism of oxygen sensing by the bacterial transcription factor fumarate-nitrate reduction (FNR) J Biol Chem 279 9278–9286 10.1074/jbc.M309878200 . [DOI] [PubMed] [Google Scholar]
  12. Din N., Shoemaker C.J., Akin K.L., Frederick C., Bird T.H. (2011). Two putative histidine kinases are required for cyst formation in Rhodospirillum centenum Arch Microbiol 193 209–222 10.1007/s00203-010-0664-7 . [DOI] [PubMed] [Google Scholar]
  13. Dong Q., Bauer C.E. (2015). Transcriptome analysis of cyst formation in Rhodospirillum centenum reveals large global changes in expression during cyst development BMC Genomics 16 68 10.1186/s12864-015-1250-9 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fani R., Bandi C., Bazzicalupo M., Ceccherini M.T., Fancelli S., Gallori E., Gerace L., Grifoni A., Miclaus N., Damiani G. (1995). Phylogeny of the genus Azospirillum based on 16S rDNA sequence FEMS Microbiol Lett 129 195–200 . [DOI] [PubMed] [Google Scholar]
  15. Feucht B.U., Saier M.H., Jr (1980). Fine control of adenylate cyclase by the phosphoenolpyruvate:sugar phosphotransferase systems in Escherichia coli and Salmonella typhimurium J Bacteriol 141 603–610 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gomelsky M. (2011). cAMP, c-di-GMP, c-di-AMP and now cGMP: bacteria use them all! Mol Microbiol 79 562–565 10.1111/j.1365-2958.2010.07514.x . [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gomelsky M., Galperin M.Y. (2013). Bacterial second messengers, cGMP and c-di-GMP, in a quest for regulatory dominance EMBO J 32 2421–2423 10.1038/emboj.2013.193 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Görke B., Stülke J. (2008). Carbon catabolite repression in bacteria: many ways to make the most out of nutrients Nat Rev Microbiol 6 613–624 10.1038/nrmicro1932 . [DOI] [PubMed] [Google Scholar]
  19. Hanamura A., Aiba H. (1992). A new aspect of transcriptional control of the Escherichia coli crp gene: positive autoregulation Mol Microbiol 6 2489–2497 10.1111/j.1365-2958.1992.tb01425.x . [DOI] [PubMed] [Google Scholar]
  20. Harman J.G. (2001). Allosteric regulation of the cAMP receptor protein Biochim Biophys Acta 1547 1–17 10.1016/S0167-4838(01)00187-X . [DOI] [PubMed] [Google Scholar]
  21. Harwood J.P., Gazdar C., Prasad C., Peterkofsky A., Curtis S.J., Epstein W. (1976). Involvement of the glucose enzymes II of the sugar phosphotransferase system in the regulation of adenylate cyclase by glucose in Escherichia coli J Biol Chem 251 2462–2468 . [PubMed] [Google Scholar]
  22. He K., Bauer C.E. (2014). Chemosensory signaling systems that control bacterial survival Trends Microbiol 22 389–398 10.1016/j.tim.2014.04.004 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. He K., Marden J.N., Quardokus E.M., Bauer C.E. (2013). Phosphate flow between hybrid histidine kinases CheA3 and CheS3 controls Rhodospirillum centenum cyst formation PLoS Genet 9 e1004002 10.1371/journal.pgen.1004002 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ishizuka H., Hanamura A., Inada T., Aiba H. (1994). Mechanism of the down-regulation of cAMP receptor protein by glucose in Escherichia coli: role of autoregulation of the crp gene EMBO J 13 3077–3082 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kim J., Adhya S., Garges S. (1992). Allosteric changes in the cAMP receptor protein of Escherichia coli: hinge reorientation Proc Natl Acad Sci U S A 89 9700–9704 10.1073/pnas.89.20.9700 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kolb A., Busby S., Buc H., Garges S., Adhya S. (1993). Transcriptional regulation by cAMP and its receptor protein Annu Rev Biochem 62 749–797 10.1146/annurev.bi.62.070193.003533 . [DOI] [PubMed] [Google Scholar]
  27. Körner H., Sofia H.J., Zumft W.G. (2003). Phylogeny of the bacterial superfamily of Crp-Fnr transcription regulators: exploiting the metabolic spectrum by controlling alternative gene programs FEMS Microbiol Rev 27 559–592 10.1016/S0168-6445(03)00066-4 . [DOI] [PubMed] [Google Scholar]
  28. Lawson C.L., Swigon D., Murakami K.S., Darst S.A., Berman H.M., Ebright R.H. (2004). Catabolite activator protein: DNA binding and transcription activation Curr Opin Struct Biol 14 10–20 10.1016/j.sbi.2004.01.012 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Leduc J.L., Roberts G.P. (2009). Cyclic di-GMP allosterically inhibits the CRP-like protein (Clp) of Xanthomonas axonopodis pv. citri J Bacteriol 191 7121–7122 10.1128/JB.00845-09 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Marden J.N., Dong Q., Roychowdhury S., Berleman J.E., Bauer C.E. (2011). Cyclic GMP controls Rhodospirillum centenum cyst development Mol Microbiol 79 600–615 10.1111/j.1365-2958.2010.07513.x . [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Mori K., Aiba H. (1985). Evidence for negative control of cya transcription by cAMP and cAMP receptor protein in intact Escherichia coli cells J Biol Chem 260 14838–14843 . [PubMed] [Google Scholar]
  32. Österberg S., Åberg A., Herrera Seitz M.K., Wolf-Watz M., Shingler V. (2013). Genetic dissection of a motility-associated c-di-GMP signalling protein of Pseudomonas putida Environ Microbiol Rep 5 556–565 10.1111/1758-2229.12045 . [DOI] [PubMed] [Google Scholar]
  33. Passner J.M., Schultz S.C., Steitz T.A. (2000). Modeling the cAMP-induced allosteric transition using the crystal structure of CAP-cAMP at 2.1 Å resolution J Mol Biol 304 847–859 10.1006/jmbi.2000.4231 . [DOI] [PubMed] [Google Scholar]
  34. Pfaffl M.W., Tichopád A., Prgomet C., Neuvians T.P. (2004). Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper - Excel-based tool using pair-wise correlations Biotechnol Lett 26 509–515 10.1023/B:BILE.0000019559.84305.47 . [DOI] [PubMed] [Google Scholar]
  35. Popovych N., Tzeng S.R., Tonelli M., Ebright R.H., Kalodimos C.G. (2009). Structural basis for cAMP-mediated allosteric control of the catabolite activator protein Proc Natl Acad Sci U S A 106 6927–6932 10.1073/pnas.0900595106 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Postma P.W., Lengeler J.W., Jacobson G.R. (1993). Phosphoenolpyruvate : carbohydrate phosphotransferase systems of bacteria Microbiol Rev 57 543–594 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Rauch A., Leipelt M., Russwurm M., Steegborn C. (2008). Crystal structure of the guanylyl cyclase Cya2 Proc Natl Acad Sci U S A 105 15720–15725 10.1073/pnas.0808473105 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Ryu M.-H., Youn H., Kang I.H., Gomelsky M. (2015). Identification of bacterial guanylate cyclases Proteins 83 799–804 10.1002/prot.24769 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sharma H., Yu S., Kong J., Wang J., Steitz T.A. (2009). Structure of apo-CAP reveals that large conformational changes are necessary for DNA binding Proc Natl Acad Sci U S A 106 16604–16609 10.1073/pnas.0908380106 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Shelver D., Kerby R.L., He Y., Roberts G.P. (1995). Carbon monoxide-induced activation of gene expression in Rhodospirillum rubrum requires the product of cooA, a member of the cyclic AMP receptor protein family of transcriptional regulators J Bacteriol 177 2157–2163 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Shelver D., Kerby R.L., He Y., Roberts G.P. (1997). CooA, a CO-sensing transcription factor from Rhodospirillum rubrum, is a CO-binding heme protein Proc Natl Acad Sci U S A 94 11216–11220 10.1073/pnas.94.21.11216 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Shimada T., Fujita N., Yamamoto K., Ishihama A. (2011). Novel roles of cAMP receptor protein (CRP) in regulation of transport and metabolism of carbon sources PLoS One 6 e20081 10.1371/journal.pone.0020081 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Spiro S., Guest J.R. (1990). FNR and its role in oxygen-regulated gene expression in Escherichia coli FEMS Microbiol Rev 6 399–428 . [DOI] [PubMed] [Google Scholar]
  44. Stadtwald-Demchick R., Turner F.R., Gest H. (1990). Physiological properties of the thermotolerant photosynthetic bacterium, Rhodospirillum centenum FEMS Microbiol Lett 67 139–143 10.1111/j.1574-6968.1990.tb13851.x. [DOI] [Google Scholar]
  45. Tao F., He Y.W., Wu D.H., Swarup S., Zhang L.H. (2010). The cyclic nucleotide monophosphate domain of Xanthomonas campestris global regulator Clp defines a new class of cyclic di-GMP effectors J Bacteriol 192 1020–1029 10.1128/JB.01253-09 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Willett J., Smart J.L., Bauer C.E. (2007). RegA control of bacteriochlorophyll and carotenoid synthesis in Rhodobacter capsulatus J Bacteriol 189 7765–7773 10.1128/JB.00853-07 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zumft W.G. (2002). Nitric oxide signaling and NO dependent transcriptional control in bacterial denitrification by members of the FNR-CRP regulator family J Mol Microbiol Biotechnol 4 277–286 . [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data

Click here for additional data file. (65.1KB, pdf)

Articles from Microbiology are provided here courtesy of Microbiology Society

RESOURCES