INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The -proteobacterium Sinorhizobium meliloti forms nitrogen-fixing nodules on the roots of certain legumes of the genera Medicago, Melilotus, and Trigonella. The bacterium and plant exchange chemical signals during initiation of this process (7, 24). Bacteria attach to root hairs and invade plant cells via an infection thread which encases the bacterial colony as the infection thread grows into the root; bacteria are released from the infection thread into the cytoplasm of plant cells (reviewed in references 4 and 15). After release, one or more rounds of bacterial cell division occur before the bacteria differentiate into nitrogen-fixing bacteroids. These bacteroids are four to seven times longer than vegetative bacteria and are often Y shaped. Bacteroids remain in the host cytoplasm, each one enclosed in a peribacteroid membrane, a structure formed from both bacterial and plant components (4). They continue to fix nitrogen until senescence of the plant cell. Nitrogen-fixing bacteroids are thought to represent a terminally differentiated state; however, it is possible that nondifferentiated bacteria are present in the plant cell and can multiply upon release (40, 44).

While much is known about the genetics and biochemistry of nitrogen fixation, little is known about bacterial release and differentiation. Specific regulatory circuits must exist to coordinate the cessation of DNA replication and cell division with the other processes of bacteroid differentiation (reviewed in reference 30). Some S. meliloti genes involved in cell division have been characterized, but neither the mechanisims of control of these genes nor their role in the differentiation process is known (25-27).

In contrast, control of the cell cycle in Caulobacter crescentus, a closely related member of the -proteobacteria, is better understood. Several regulatory proteins that play essential roles in cell cycle progression have been identified in this organism (17). These include members of the two-component family of signal transduction proteins, the CtrA response regulator (32), the CckA histidine kinase that phosphorylates CtrA (18), and the DivK histidine kinase (43), as well as a DNA adenine methyltransferase, CcrM, that methylates newly replicated DNA at the end of S phase (39). CtrA is a global regulator that plays a pivotal role in orchestrating the cell cycle. It is involved in the control of a quarter of the cell cycle-regulated genes (23) including genes required for DNA replication (32, 33), DNA methylation (32, 34), cell division (20), and biogenesis of flagella and pili (32, 38). CtrA binds to the chromosomal origin of replication and inhibits DNA replication initiation; it must be cleared from the cell before it enters S phase (9, 33). Transcription of ctrA is activated after initiation of DNA replication and reaches threshold levels by an autoregulated positive feedback loop. ctrA gene is transcribed from two promoters: P1, a promoter that is active in early S phase and is negatively controlled by CtrA; and a stronger promoter, P2, which is active later in S phase and is positively controlled by CtrA (10). The level of intracellular CtrA is controlled not only by autoregulated transcription but also by specific degradation of this protein when the cells enter S phase and by cell cycle-controlled phosphorylation which activates the protein (9, 18).

Here we report the isolation and characterization of ctrA from S. meliloti. As a first step in dissecting the regulatory circuit controlling the cell cycle and bacteroid differentiation in S. meliloti, we show that ctrA is essential for viability of S. meliloti. The promoter region was identified and shown to contain multiple CtrA binding sites, suggesting that the S. meliloti gene may be autoregulated, as is the case in C. crescentus.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial genetic techniques. Strains and plasmids used in this study are listed in Table 1. Escherichia coli strains were grown in either Luria broth (LB) or 2XYT medium at 37°C. S. meliloti strains were grown on LB agar plates or in liquid TY medium at 30°C. Antibiotic concentrations were for S. meliloti, neomycin and spectinomycin, 200 µg/ml; gentamicin, 25 µg/ml; tetracycline, 10 µg/ml; streptomycin, 500 µg/ml; and trimethoprim, 600 µg/ml; for E. coli, gentamicin, 5 µg/ml; kanamycin, 25 µg/ml; tetracycline, 10 µg/ml; ampicillin, 50 or 100 µg/ml; chloramphenicol, 50 µg/ml; and spectinomycin, 40 µg/ml. Triparental conjugations were performed using MT616 as the helper plasmid strain (12). C. crescentus was grown in PYE broth as previously described (11). Kanamycin (20 µg/ml) was used for C. crescentus. Agrobacterium tumefaciens was grown in LB medium.

Southern hybridization. DNA from S. meliloti, C. crescentus, and A. tumefaciens was purified using the PureGene kit (Gentra systems), and Brucella abortus strain 2308 DNA was kindly provided by R. M. Roop, Louisiana State University. Genomic DNA was digested with various restriction enzymes and electrophoresed on a 1% agarose gel. The DNA was transferred to Hybond N+ (Amersham) as described previously (37). Hybridization was performed by the Church and Gilbert method (5); a random-primed EcoRI-SalI restriction fragment from pSALF1, containing C. crescentus ctrA, was the probe (32). The blot was hybridized for 4 h at 65°C and then washed once in 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) for 1 h, twice in 1× SSC (20 min each time), and finally once in 0.1× SSC for 20 min.

Isolation of the S. meliloti ctrA clone. An S. meliloti partial Sau3A genomic library, constructed in a  FixII vector (1), was screened using Church and Gilbert hybridization (5). The probe was a random-primed 695-bp PCR product containing C. crescentus ctrA; the primers used to amplify this DNA are shown in Table 1. Following hybridization, filters were washed three times in 2X SSC-0.1% sodium dodecyl sulfate and then once in 0.5X SSC-0.1% sodium dodecyl sulfate. Positive plaques were purified and rescreened. DNA was isolated from putative ctrA-containing phage, mapped, subcloned, and sequenced. The University of Wisconsin Genetics Computer Group (8) and Sequencher (Gene Codes Corp.) software were used for sequence assembly and analysis.

Construction and complementation of S. meliltoti ctrA mutant. To introduce an insertion mutation into the S. meliloti ctrA coding sequence we used pJQ200, which contains the levan sucrase (sacB) gene, which confers lethality when strains are grown on sucrose (31). This vector has been previously used to create null mutations in S. meliloti ccrM (42). A derivative of pJQ200 lacking polylinker sites from Ecl136 to SmaI was created. A 3-kb PstI-SalI fragment from pMB453 was cloned into this derivative to form pMB491. A blunted, BamHI-HindIII fragment containing a neomycin resistance (Nm^r) cassette was ligated into pMB491 that had been digested with BamHI, thereby removing the 5' half of ctrA, and blunted with Klenow to form pMB492. In pMB492, the Nm^r cassette is oriented such that the neomycin promoter reads in the opposite direction from the ctrA promoter. 1.35 and 0.95 kb of S. meliloti DNA are present upstream and downstream, respectively, of the Nm^r cassette. After conjugation into S. meliloti strain Rm1021, colonies were selected on LB agar plates containing gentamicin, neomycin, and streptomycin; these are the putative single recombinants containing both an intact and a disrupted copy of ctrA. Either a ctrA plasmid (pMB467 or pSAL290) or the vector only (pMB393 or pRK290) was introduced into this strain by conjugation. The resulting strains were selected on plates lacking gentamicin and containing 5% sucrose. Colonies viable on these plates were screened for gentamicin sensitivity (Gm^s) in the presence of sucrose (Gm^r sucrose-resistant mutants likely contain mutations in the sacB gene or in host genes that suppress the sacB phenotype). Double recombinants, containing a single mutated ctrA, are expected to be Gm^s, Nm^r, and sucrose resistant. The frequency of apparent double recombinants was calculated for a representative experiment by dividing the number of Gm^s Nm^r sucrose-resistant colonies by the total number of Nm^r sucrose-resistant CFU.

Complementation tests of C. crescentus ctrA mutants with S. meliloti ctrA. To determine if the S. meliloti ctrA gene could replace the C. crescentus ctrA, the S. meliloti ctrA promoter and gene were ligated into the low-copy-number vector pMR10, generating pDYH218. Both pDYH218 and pMR10 were mated into strains NA1000 (wild-type C. crescentus) and LS2195 (NA1000 ctrA401, a temperature-sensitive allele of ctrA). Logarithmic-phase cultures were diluted serially, plated in duplicate on PYE agar containing kanamycin (20 µg/ml), and incubated at either 30 or 37°C. Viable cell number at the permissive (30°C) and restrictive (37°C) temperatures was measured as CFU per milliliter.

Primer extension analysis. S. meliloti total RNA was purified from LB-grown cells, using Trizol reagent (BRL Corp.), as previously described (3). Primer extension reactions were performed as before (3), using an initial annealing temperature of 65°C for 1 h and then cooling to 40°C over 1.5 h. Extensions were performed at 49°C. The nucleotide sequences of PE-1, PE-2, and PE-3 used for primer extension are shown in Table 1. Extension products were analyzed on a urea-polyacrylamide sequencing gel as described elsewhere (13). Sequencing ladders using each of the three primers and pMB464 as template were used to determine the transcription start site.

DNase I protection experiments. DNase I footprinting experiments were performed with purified C. crescentus CtrA that was phosphorylated using a maltose-binding protein (MBP)-EnvZ fusion protein as previously described (34). Native C. crescentus CtrA was overexpressed in E. coli using the pET expression system (Novagen), solubilized from inclusion bodies with 4 M guanidine-HCl, and purified by Q-Sepharose chromatography (K. Ryan and L. Shapiro, unpublished data). Template DNA for footprinting the P1 and P2 promoters was generated by PCR using the oligonucleotides in Table 1. Both reverse primers (P1-rev and P2-rev) were end labeled with [-³²P]ATP and T4 DNA polynucleotide kinase prior to PCR amplification of the template.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

CtrA is conserved among -proteobacteria. To determine the extent of conservation of the gene encoding the CtrA response regulator among the -proteobacterial family, we generated a probe containing the entire C. crescentus ctrA coding region and part of the 5' and 3' flanking sequences (32). We hybridized this SalI-EcoRI DNA fragment to digested genomic DNA from S. meliloti, B. abortus, and A. tumefaciens (Fig. 1) using fairly stringent conditions. The C. crescentus ctrA probe hybridized to at least one band in genomic digests of all these organisms.

View larger version (80K): [in this window] [in a new window]   FIG. 1.   ctrA is present in several members of the -proteobacteria. The Southern blot shows hybridization of the C. crescentus ctrA probe to chromosomal DNA from C. crescentus (Cc), S. meliloti (Sm), B. abortus (Ba), and A. tumifaciens (At). Restriction enzymes used are BamHI (B), EcoRI (E), HindIII (H), NotI (N), PstI (P), SalI (S), and XhoI (X).

View larger version (46K): [in this window] [in a new window]   FIG. 2.   Sequence alignment of S. meliloti CtrA with CtrA family proteins using Pileup (8). Residues that differ from S. meliloti CtrA are in red; the conserved aspartate phosphorylation site is boxed in blue; the acidic pocket conserved residues are boxed in yellow. Secondary structures ( helices and  strands) corresponding to those in the OmpR C-terminal domain are indicated (29). Wing1 is is positioned between the 1 helix and 4 (29). Wing2 connects 6 and 7 of the C-terminal hairpin (29). We aligned E. coli OmpR (accession no. P03025) with the CtrA family, ignoring the gap caused by differing lengths in the N terminus-C terminus linker region. Residues conserved in S. meliloti CtrA and OmpR are indicated by asterisks below the alignment. Conserved groupings were Val, Ile; Leu, Met; Asp, Glu; and Arg, Lys. Proteins included in this lineup are S. meliloti (Sm) CtrA (accession no. AF288464), B. abortus (Ba) CtrA (accession no. AAC69920), C. crescentus (Cc) CtrA (accession no. AAF1377), R. capsulatus (Rc) CtrA (accession no. AAF1377), and Rickettsia prowazekii (Rp) CzcR (accession no. CAA14542).

S. meliloti ctrA encodes an essential function. To determine if ctrA is essential for viability in S. meliloti, as is the case in C. crescentus (32), we disrupted the ctrA gene by deleting a 390-bp BamHI fragment by single-crossover, Campbell recombination. This deletion removes the first 112 amino acids of CtrA. Counterselection with the single-crossover strain MB492X was performed as described in Materials and Methods, with either S. meliloti ctrA present on a plasmid (pMB467) or the vector only (pMB393) (Table 2). Double recombinants (Nm^r Gm^s sucrose resistant) were obtained only when S. meliloti ctrA was present on a complementing plasmid, indicating that ctrA is necessary for viability in S. meliloti.

Exchange of S. meliloti and C. crescentus ctrA. To test if C. crescentus ctrA functions in S. meliloti, sucrose counterselection with MB492X was performed as described above, in the presence of either C. crescentus ctrA on a plasmid or the vector alone. Table 2 shows that Nm^r Gm^s sucrose-resistant colonies were obtained only when ctrA was present on a plasmid (pSAL290); therefore, C. crescentus ctrA can substitute for S. meliloti ctrA for viability. To test if S. meliloti ctrA could complement a C. crescentus ctrA temperature-sensitive mutation, we mated a plasmid containing S. meliloti ctrA expressed from its own promoter (pDYH218) into the temperature-sensitive strain at the permissive temperature (Materials and Methods). This plasmid failed to complement the C. crescentus ctrA mutant at the restrictive temperature (data not shown). Since correct function of ctrA in C. crescentus requires coordinated expression of the gene, phosphorylation, and proteolysis of the gene product, the heterologous product may have failed at any one of these levels.

The CtrA response regulator binds the S. meliloti ctrA promoter. To define the S. meliloti ctrA promoter region, we determined the transcriptional start site of the ctrA gene by primer extension assays using RNA isolated from wild-type cells. We identified two transcripts initiating 69 and 291-bp upstream of the ctrA translational start site (Fig. 3B).

View larger version (43K): [in this window] [in a new window]   FIG. 3.   Identification of the ctrA transcription start sites. (A) Sequence upstream of ctrA. Locations of the P1 and P2 start sites are shown by bent arrows. The 10 and 35 regions of the two promoters are marked below the sequence. The sequence protected from DNase I digestion by CtrA~P is underlined (see Fig. 4), and the CtrA recognition motifs are in bold. The final ATG is the presumed translational start codon. (B) Autoradiograms of primer extension products (P) and sequencing reactions (labeled A, C, G, and T) for the P1 and P2 start sites of the S. meliloti ctrA promoter. The P1 start site was mapped using primer PE-3 (left); the P2 promoter was mapped using primers PE-1 and PE-2. Results obtained with PE-2 are shown (right). (C) Alignment of the five S. meliloti CtrA binding sites in the S. meliloti ctrA promoter with the S. meliloti and C. crescentus CtrA consensus sequences (32).

View larger version (55K): [in this window] [in a new window]   FIG. 4.   P CtrA~P binds to five sites in the ctrA promoter. (A) Diagram of the ctrA promoter regions in C. crescentus (top) and S. meliloti. Locations of the ctrA P1 and P2 transcription start sites are marked by arrows, and the regions protected from DNase I digestion by CtrA~P are shown as gray boxes (labeled a to e to match the footprinted regions in panel B). (B) DNase I protection of the ctrA P1 and P2 promoters. CtrA purified from C. crescentus was phosphorylated with MBP-EnvZ and used in the footprinting reactions (0.5 to 1.7 µM). CtrA-P was omitted from the reactions in the lanes marked with a minus sign. The vertical lines labeled a to e mark the protected regions.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The CcrM methyltransferase is conserved in -proteobacteria (42). Here we report the conservation of a second regulatory protein, the CtrA response regulator, in this group of bacteria. We isolated and characterized CtrA, a global regulator of cell cycle progression in C. crescentus (32), from S. meliloti. A ctrA homologue appears to be present in another member of the Rhizobiaceae, A. tumefaciens. Our data show that by sequence similarity CtrA belongs to a unique subfamily of the OmpR family of helix-turn-helix transcriptional regulators. Thus far, all members of this subfamily belong to the -proteobacteria, which are characterized by specialized mechanisms to adapt to unique environments, including mechanisms for pathogenicity and symbiosis (2). Some, particularly the Rhizobiaceae, possess large, multipartite genomes (2, 19).

The primary sequence of S. meliloti CtrA is very similar to C. crescentus CtrA. In addition, antibodies to C. crescentus CtrA cross-react with S. meliloti CtrA (A. Reisenauer, unpublished data). The sequence conservation among CtrA-like proteins is especially strong for the C-terminal DNA binding domain, suggesting that the DNA target site recognized by these proteins is also conserved. In fact, the promoter region for S. meliloti ctrA has five CtrA binding motifs that are nearly identical to the C. crescentus consensus motif.

The S. meliloti and C. crescentus ctrA promoters are strikingly similar. In both species there are two transcriptional start sites, and each promoter (called P1 and P2) contains at least one CtrA binding motif. In addition, purified and phosphorylated C. crescentus CtrA binds the consensus CtrA motifs in vitro in each of the ctrA promoters, suggesting that ctrA transcription is autoregulated in both species. In fact, CtrA has been shown to repress transcription from P1 and activate transcription from P2 in vivo in C. crescentus (10). However, the spacing between P1 and P2 and the number of CtrA binding sites in the S. meliloti and C. crescentus ctrA promoters are distinct (Fig. 4A). P1 and P2 are separated by 222 bp in S. meliloti but only 57 bp in C. crescentus. The most distal promoter in S. meliloti (P1) is far upstream from the ATG translational start site; however, it is not unusual for S. meliloti genes to have long mRNA leader sequences (3, 13).

In the S. meliloti ctrA promoter, CtrA~P protects five distinct sites from DNase I digestion; two overlap the 35 regions of P1 and P2, respectively, and there are three additional sites between P1 and P2. Moreover, the P1 transcriptional start site itself is protected from DNase I digestion by CtrA~P. In contrast, CtrA protects only two regions of the C. crescentus ctrA promoter, one overlapping the 10 region of P1 and the second overlapping the 35 region of P2 (10). It is tempting to speculate that CtrA molecules bound to the different recognition motifs in each of these promoters act cooperatively to regulate ctrA transcription. The CtrA binding sites in the S. meliloti ctrA promoter are very similar to the C. crescentus CtrA consensus motif. Given the similarity of the C-terminal DNA binding domains and the fact that C. crescentus CtrA can substitute for S. meliloti CtrA in vivo, it is likely that CtrA recognizes these same binding sites in both species.

As in C. crescentus, S. meliloti ctrA is essential for viability. In contrast, ctrA is not essential in R. capsulatus where insertion mutations in ctrA were obtained (22). CtrA in R. capsulatus is responsible for expression of the gene transfer agent structural genes. The gene transfer agent is a small phage-like particle that can transfer genes between R. capsulatus cells. Possibly, ctrA is present in multiple copies in R. capsulatus or is not involved in critical cell cycle functions. It is not known if ctrA is an essential gene in B. abortus or R. prowazekii.

Based on sequence similarity, we thought it likely that C. crescentus ctrA could substitute for S. meliloti ctrA, as is the case for the conserved ccrM in these organisms (42). C. crescentus ctrA, expressed from its own promoter, complements a S. meliloti null mutant for viability. Thus, S. meliloti contains the cellular machinary necessary for transcription and phosphorylation of C. crescentus CtrA, at least to an extent that supports viability.

Little is known of cell division control in S. meliloti, but examination of symbiotic behavior suggests the likelihood of specific controls related to invasion and differentiation. We and others have suggested that bacterial cell division is coordinately controlled to match the expansion rate of the infection thread, in order to keep the infection thread filled but not overrun with bacteria (14, 15). After invasion is complete and bacteria are released into the plant cell cytoplasm, it appears that one or several rounds of bacterial cell division likely occur. Circuitry must exist to halt DNA replication and cell division after this step, as the bacteria differentiate. The essential functions of S. meliloti CtrA are not known, but it is an attractive candidate to study in regard to these inferred regulatory steps. In C. crescentus, CtrA is known to directly prevent DNA synthesis in the swarmer cell by binding to the origin of replication (33). It will be interesting to determine if perturbation of ctrA circuiting in free-living S. meliloti cells results in disruption of the invasion process.