Skip to main content
NCBI home page
Journal of Bacteriology logoLink to Journal of Bacteriology
. 1999 Aug;181(15):4576–4583. doi: 10.1128/jb.181.15.4576-4583.1999

Cloning, Sequencing, and Characterization of the cgmB Gene of Sinorhizobium meliloti Involved in Cyclic β-Glucan Biosynthesis

Ping Wang 1,†, Cheryl Ingram-Smith 1,‡, Jill A Hadley 1,§, Karen J Miller 1,2,3,*
PMCID: PMC103589  PMID: 10419956

Abstract

Periplasmic cyclic β-glucans of Rhizobium species provide important functions during plant infection and hypo-osmotic adaptation. In Sinorhizobium meliloti (also known as Rhizobium meliloti), these molecules are highly modified with phosphoglycerol and succinyl substituents. We have previously identified an S. meliloti Tn5 insertion mutant, S9, which is specifically impaired in its ability to transfer phosphoglycerol substituents to the cyclic β-glucan backbone (M. W. Breedveld, J. A. Hadley, and K. J. Miller, J. Bacteriol. 177:6346–6351, 1995). In the present study, we have cloned, sequenced, and characterized this mutation at the molecular level. By using the Tn5 flanking sequences (amplified by inverse PCR) as a probe, an S. meliloti genomic library was screened, and two overlapping cosmid clones which functionally complement S9 were isolated. A 3.1-kb HindIII-EcoRI fragment found in both cosmids was shown to fully complement mutant S9. Furthermore, when a plasmid containing this 3.1-kb fragment was used to transform Rhizobium leguminosarum bv. trifolii TA-1JH, a strain which normally synthesizes only neutral cyclic β-glucans, anionic glucans containing phosphoglycerol substituents were produced, consistent with the functional expression of an S. meliloti phosphoglycerol transferase gene. Sequence analysis revealed the presence of two major, overlapping open reading frames within the 3.1-kb fragment. Primer extension analysis revealed that one of these open reading frames, ORF1, was transcribed and its transcription was osmotically regulated. This novel locus of S. meliloti is designated the cgm (cyclic glucan modification) locus, and the product encoded by ORF1 is referred to as CgmB.

The cell surface carbohydrates of bacteria within the Rhizobiaceae family provide important functions during plant infection (9, 19, 22, 35, 42). One class of cell surface carbohydrate, the periplasmic cyclic β-glucans, has additionally been shown to provide important functions for the free-living forms of these bacteria during hypo-osmotic adaptation (9).

In species of Rhizobium, Sinorhizobium, and Agrobacterium, cyclic β-glucans contain 17 to 25 glucose residues linked solely by β-(1,2) glycosidic bonds (9). These molecules may become highly modified with anionic substituents which include sn-1 phosphoglycerol and succinyl moieties. The cyclic β-glucans of Bradyrhizobium species are smaller (i.e., 10 to 13 glucose residues) and are linked by both β-(1,6) and β-(1,3) glycosidic bonds (9). These glucans may become modified with the zwitterionic substituent phosphorylcholine.

Until recently, only two classes of mutants defective for cyclic β-(1,2)-glucan biosynthesis had been described (9). These correspond to the ndvA and ndvB mutants of Rhizobium and Sinorhizobium species and the chvA and chvB mutants of Agrobacterium species (the chvA and chvB genes are functionally and structurally homologous with ndvA and ndvB, respectively). The ndvB (chvB) gene encodes a high-molecular-weight (319-kDa) membrane protein that is involved in the biosynthesis of the cyclic β-(1,2)-glucan backbone from UDP-glucose (15, 25, 50, 51). The ndvA (chvA) gene encodes a protein involved in the transport of the cyclic β-glucans to the periplasm and extracellular medium (14, 26, 43). Mutants at the ndvA (chvA) and ndvB (chvB) loci are impaired for host plant infection and for growth in hypo-osmotic media. Thus, studies with these mutants have provided important insight concerning the functions of the cyclic β-(1,2)-glucans.

During the past few years, additional loci linked to cyclic β-glucan biosynthesis have been identified in Bradyrhizobium japonicum and Sinorhizobium meliloti. These include the identification of a ndvB-like locus and the ndvC locus in B. japonicum by Bhagwat et al. (5–7). The ndvC locus appears to be involved in the biosynthesis of β-(1,6) linkages within the B. japonicum cyclic glucan backbone (7). Recently, we have identified a novel cyclic β-(1,2)-glucan mutant of S. meliloti (also known as Rhizobium meliloti) which we refer to as mutant S9 (8). Mutant S9, created by Tn5 insertional mutagenesis, is specifically impaired in its ability to transfer sn-1 phosphoglycerol substituents to the cyclic β-(1,2)-glucan backbone. Although the cyclic β-(1,2)-glucans of mutant S9 lack phosphoglycerol substituents, high levels of succinyl substituents are present on these molecules. Indeed, the overall anionic charge on the cyclic β-(1,2)-glucans of this mutant is similar to that found in wild-type cells. Interestingly, this mutant is able to effectively nodulate alfalfa and can grow as well as wild-type cells in hypo-osmotic media. These results reveal that the phosphoglycerol substituent is not required for either process and suggest that it is the overall anionic charge on the cyclic β-glucans that may be important for nodulation and/or hypo-osmotic adaptation. In the present study, we have characterized the mutation within mutant S9 at the molecular level and have identified a novel locus in S. meliloti which we refer to as the cgm locus (for cyclic glucan modification).

MATERIALS AND METHODS

Bacterial strains, cosmids, and plasmids.

The strains, plasmids, and cosmids used in this study are described in Table 1. S. meliloti strains were grown in glutamate mannitol salts (GMS) medium (11) or Luria-Bertani (LB) medium (36) at 30°C. Rhizobium leguminosarum bv. trifolii TA-1JH was grown at 30°C in GMS medium containing 400 μg of streptomycin per ml. Escherichia coli strains were grown in LB medium at 37°C. LB-MC medium (LB medium containing 2.5 mM MgSO4 and 2.5 mM CaCl2) was used in triparental mating experiments. An S. meliloti 1021 genomic library, constructed within cosmid pLAFRI, was kindly provided by B. Tracy Nixon (Department of Biochemistry and Molecular Biology, Pennsylvania State University, University Park, Pa.). The library was prepared by using S. meliloti 1021 genomic DNA partially digested with EcoRI and consists of a total of 1,920 clones containing an average genomic insert size of approximately 23 kbp (20).

TABLE 1.

Bacterial strains, cosmids, and plasmids

Strain, plasmid, or cosmid Descriptiona Reference or source
Bacteria
 E. coli MM294 endA thi-1 hsdR17 supE44 pro F. Ausubel
 E. coli DH5α recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 deoR lacU169 (φ80dlacZΔM15) 23
 S. meliloti 1021 Smr, derivative of SU47, parent of S9 37
 S. meliloti S9 Smr, Nmr, S. meliloti 1021 cgmB::Tn5, cyclic β-glucans lack phosphoglycerol substituents 8
 R. leguminosarum bv. trifolii TA-1 Synthesizes only neutral, unsubstituted cyclic β-(1,2)-glucans 12
 R. leguminosarum bv. trifolii TA-1JH Spontaneous streptomycin-resistant mutant of strain TA-1 This study
Cosmids
 pLAFRI 21.6 kb; Tcr, S. meliloti 1021 genomic DNA was cloned into this vector to generate a cosmid library 20
 12A7 25-kb S. meliloti 1021 genomic insert within pLAFRI; Tcr, complements S9 This study
 13H9 27.5-kb S. meliloti 1021 genomic insert within pLAFRI; Tcr, complements S9 This study
Plasmids
 pRK2013 48 kb; Nmr, ColE1 replicon with RK2 tra genes 17
 pRK602 56 kb; Cmr, Nmr, pRK2013 nm::Tn9 ΩTn5 34
 pUC19 2.7 kb; Apr, ColE1ori 49
 pTR100 7.3 kb; Tcr, Apr, RK2ori 47
 pFL6 7.6 kb; Apr, pUC19 with 4.9-kb partially digested EcoRI fragment from cosmid 12A7 inserted in the orientation such that HindIII digestion will generate a 3.1-kb fragment containing 50 bp of polylinker region from pUC19 This study
 pEW1 10.2 kb; Tcr, Apr, pTR100 with 2.9-kb BamHI fragment from 12A7 This study
 pEW2 10.4 kb; Tcr, Apr, pTR100 with 3.1 kb HindIII fragment from pFL6, complements S9 This study
Open in a new tab
a

Ap, ampicillin; Cm, chloramphenicol; Nm, neomycin; Sm, streptomycin; Tc, tetracycline. 

DNA manipulations.

Genomic DNA was purified from S. meliloti mutant S9 by the method described by Streit et al. (44) with minor modifications. Specifically, the DNA was precipitated with 0.5 volume of 7 M ammonium acetate and 2 volumes of ethanol at −20°C overnight. Cosmid DNA was purified from E. coli with the Qiagen plasmid kit (Qiagen Inc., Chatworth, Calif.). Plasmid DNA was purified by using either the Qiagen plasmid kit or the Wizard Plus Miniprep DNA purification system (Promega Corp., Madison, Wis.). Standard methods were used for restriction digestions, agarose gel electrophoresis, and ligations (40). E. coli strains were transformed by electroporation by using the E. coli Pulser transformation apparatus (Bio-Rad Laboratories, Richmond, Calif.). Experimental conditions were those recommended by the manufacturer.

Preparation of biotinylated DNA probes.

Biotinylated DNA probes were synthesized by PCR. Biotin-16-dUTP, the GeneAmp PCR reagent kit with AmpliTaq DNA polymerase (Perkin-Elmer Cetus, Norwalk, Conn.), and the GeneAmp PCR System 9600 (Perkin-Elmer Cetus) were used for these experiments. A biotinylated probe corresponding to a 1,063-bp fragment internal to Tn5 (beginning at nucleotide 1708 and ending at nucleotide 2770; GenBank accession no. U00004) was synthesized by using oligonucleotide primers 5′-TGTCCGGTGCCCTGAATGAA-3′ and 5′-CAGGCGGAAAACGGGAAGAC-3′. A second biotinylated probe (of approximately 3.3 kb) for DNA sequences flanking the Tn5 insertion within S. meliloti S9 genomic DNA was synthesized by inverse PCR (39). This probe contains 3.2 kb of flanking DNA sequence as well as 74 bp of Tn5 sequence (derived from both ends of Tn5). For these experiments, genomic DNA from mutant S9 was first digested with EcoRI, and the fragments were then self ligated. A single primer (5′-GGTTCCGTTCAGGACGCTAC-3′), complementary to bases 18 to 37 of the top strand and 5782 to 5801 of the bottom strand of the Tn5 sequence, was then used for PCR amplification.

Southern hybridization.

Genomic DNA preparations (from S. meliloti 1021 and mutant S9) and cosmid DNA preparations (cosmids 12A7 and 13H9) were digested with BamHI, EcoRI, SalI, ApaI, BglII, or combinations of these enzymes. In some experiments, restriction fragments were purified from the gel and then digested with a second (or third) restriction enzyme. The digested fragments were then examined by agarose gel electrophoresis. Gels were subsequently blotted onto nitrocellulose membranes (supported nitrocellulose-1; Life Technologies, Inc., Gaithersburg, Md.) following the technique described by Sambrook et al. (40). Two biotinylated probes were used in these experiments: the 1,063-bp fragment internal to Tn5 and the 3.3-kb fragment (generated by inverse PCR) representing DNA sequences flanking the Tn5 insert within mutant S9. Detection of probe-target DNA complexes was performed with the BluGene nonradioactive detection system (Gibco BRL, Gaithersburg, Md.). Hybridization, washing, and color development procedures were those recommended by the manufacturer.

Screening of the S. meliloti cosmid library.

The DNA probe (generated by inverse PCR) for sequences flanking the Tn5 insertion mutation within S. meliloti mutant S9 was used to probe a cosmid library prepared from genomic DNA of S. meliloti 1021 (the wild-type parent strain). The first round of screening this library was performed after pooling the 1,920 clones into 384 wells within four microtiter plates (five clones per well). After overnight growth of cells in LB medium containing 12.5 μg of tetracycline per ml, lysozyme (final concentration, between 0.5 and 0.8 mg/ml) was added to each well to lyse the cells. Lysates were then transferred to nitrocellulose membranes (supported nitrocellulose-1; Gibco BRL Life Technologies, Inc., Grand Island, N.Y.) by using the Milli Blot-D system (Millipore Corp., Bedford, Mass.). Prior to hybridization, the membranes were treated with NaOH, neutralized, and heated as described (3). The membranes were then treated with proteinase K and hybridized overnight at 42°C with the biotinylated DNA probe generated by inverse PCR. Hybridization, washing, and development procedures were those recommended by the manufacturer of the BluGene nonradioactive detection system (Gibco BRL). Because each positive well corresponded to five clones from the original library, a second round of screening was performed with all potentially positive clones from the original library. The second round of screening was performed with the procedures described above.

Complementation studies.

Cosmids which hybridized with the inverse PCR probe were examined for their ability to complement the mutation within S. meliloti mutant S9. For these experiments, each cosmid was transferred into mutant S9 through a triparental mating. The donor strain, E. coli MM294 containing cosmid, was grown overnight at 37°C in LB medium containing 12.5 μg of tetracycline per ml. The helper strain, E. coli MM294 containing plasmid pRK2013 (with transfer functions), was grown overnight at 37°C in LB medium containing 50 μg of kanamycin per ml. The recipient strain, S. meliloti mutant S9, was grown in LB medium containing 150 μg of streptomycin and 10 μg of neomycin per ml for 2 days at 30°C. All three cultures were harvested, washed, and resuspended in LB-MC medium. Donor, helper, and recipient cells were mixed in a ratio of approximately 1:1:1. The mixtures were incubated at 30°C overnight without shaking. The mixtures were then centrifuged, washed with 0.9% (wt/vol) NaCl, and plated onto GMS medium containing 10 μg of tetracycline, 150 μg of streptomycin, and 10 μg of neomycin per ml. Colonies isolated on this selective medium were inoculated into liquid GMS medium containing tetracycline (5 μg/ml) and neomycin (5 μg/ml). After growth to an optical density at 650 nm of approximately 1.0 to 1.5, periplasmic cyclic β-(1,2)-glucans were extracted from cells by using 70% ethanol and analyzed by thin-layer chromatography as previously described (8).

Additional complementation experiments were performed with subcloned fragments derived from the original cosmids. The plasmid vector pTR100 (47) used in these experiments was kindly provided by D. R. Helinski (Department of Biology and Center for Molecular Genetics, University of California at San Diego, La Jolla, Calif.). These plasmids were transferred to mutant S9 by triparental mating (as described above) or by electroporation.

A parallel series of experiments was performed with R. leguminosarum bv. trifolii TA-1JH, which is a spontaneous streptomycin-resistant mutant of R. leguminosarum bv. trifolii TA-1 (Table 1). R. leguminosarum bv. trifolii TA-1 has previously been shown to synthesize only neutral, unsubstituted cyclic β-(1,2)-glucans (12). Triparental matings were performed as described above, except that R. leguminosarum bv. trifolii TA-1JH was used as the recipient strain; all three cultures were resuspended and mated in GMS medium containing 2.5 mM CaCl2, 2.5 mM MgSO4, and 0.1 g of yeast extract and 0.2 g of bactotryptone per liter; and the final selective medium was GMS containing 10 μg of tetracycline and 100 μg of streptomycin per ml.

Nucleotide and protein sequence analysis.

DNA restriction fragments (obtained from cosmid clones) which hybridized with the inverse PCR probe were subcloned into plasmid pUC19 for sequencing studies. The inverse PCR product itself (not labeled with biotin-16-dUTP) was sequenced directly to locate the Tn5 insertion site within mutant S9. DNA sequencing was performed by the dideoxynucleotide chain termination method (41) with Sequenase 2.0 (U.S. Biochemical Corp., Cleveland, Ohio) and α-35S-labeled dATP as the radioactive label. In some experiments, automated DNA sequencing was performed with fluorescent dye terminator labeling at the nucleic acid facility at the Pennsylvania State University. The ABI 377 Prism sequencer (Perkin-Elmer) was used in these experiments. Sequences were determined for the entire length of both strands.

BLASTp, BLASTn, tBLASTn, and PSI-BLAST searches of the nonredundant sequence database and the database of unfinished microbial genomes at the National Center for Biotechnology Information (NCBI) were performed with nucleotide and deduced amino acid sequences as query sequences (1, 2). Hydropathy profiles and transmembrane-spanning residue predictions were performed by the method of Kyte and Doolittle (32) and with the TMpredict program (24).

RNA isolation and primer extension reactions.

RNA was isolated from S. meliloti 1021, R. leguminosarum bv. trifolii TA-1JH containing pTR100, and R. leguminosarum bv. trifolii TA-1JH containing pEW2 by the method of Moran (38). Primer extension reactions were performed as described (38) with [γ-32P]ATP-end-labeled oligonucleotide primers and annealing and extension temperatures of 60 and 45°C, respectively. Primer extension products were electrophoresed adjacent to dideoxy sequencing reaction mixtures prepared with the same primer. Sequencing gels contained 5% (final concentration) Long Ranger modified acrylamide (J. T. Baker, Phillipsburg, N.J.) and 8 M urea.

Large-scale isolation of cell-associated cyclic β-(1,2)-glucans and chromatographic analysis of glucan preparations.

Cell-associated cyclic β-(1,2)-glucans were extracted from 500-ml cultures by using 70% ethanol (8). Extracted glucans were pooled, concentrated, and analyzed by gel filtration, ion exchange, and thin-layer chromatography as previously described (8). The phosphorus content of cyclic β-(1,2)-glucan preparations was determined spectrophotometrically after digestion of samples with magnesium nitrate as previously described (37).

Chemicals and enzymes.

Restriction enzymes were purchased from U.S. Biochemical Corp. and Gibco BRL. T4 ligase was purchased from Gibco BRL. Biotin-labeled dUTP was purchased from Boehringer Mannheim (Indianapolis, Ind.). α-35S-labeled dATP and [γ-32P]ATP were purchased from New England Nuclear (Boston, Mass.).

Nucleotide sequence accession number.

Nucleotide sequence data have been submitted to GenBank under accession no. U67998.

RESULTS AND DISCUSSION

The Tn5 insertion within mutant S9 lies on a 3.2-kb EcoRI fragment.

Southern hybridization analyses of restricted genomic DNA revealed that the Tn5 insertion within mutant S9 was present within a 3.2-kb EcoRI fragment (Fig. 1). As expected, no hybridization signal was detected when the wild-type parent strain, S. meliloti 1021, was probed with the 1-kb biotinylated fragment internal to Tn5. However, when the 3.3-kb biotinylated probe (for sequences flanking the Tn5 insertion within mutant S9) was used, a single 3.2-kb EcoRI fragment was detected in S. meliloti 1021, and a single 9-kb EcoRI fragment was detected in mutant S9 (consistent with the presence of the 5.8-kb Tn5 insertion within mutant S9).

FIG. 1.

FIG. 1

Open in a new tab

Southern hybridization analysis of S. meliloti genomic DNA. (A) Blots of S. meliloti 1021 genomic DNA (lanes 1, 2, and 3) and S. meliloti mutant S9 genomic DNA (lanes 4, 5, and 6) were hybridized with a 1-kb probe corresponding to an internal fragment of Tn5. Genomic DNA was digested with BamHI and EcoRI (lanes 1 and 4); EcoRI (lanes 2 and 5); and BamHI (lanes 3 and 6). (B) Blots of EcoRI-digested genomic DNA from S. meliloti mutant S9 (lane 1) and S. meliloti 1021 (lane 2) were hybridized with a 3.3-kb probe corresponding to DNA sequences flanking the Tn5 insert within mutant S9; the 3.3-kb product of inverse PCR (see text) served as a positive control (lane 3).

Identification of two cosmids from the S. meliloti 1021 genomic library that complement mutant S9.

When the 3.3-kb biotinylated probe generated by inverse PCR was used to screen the S. meliloti 1021 cosmid library, a total of 13 clones that showed a positive hybridization signal were identified. All of these cosmids were subsequently mobilized into mutant S9 through triparental matings, and two (cosmids 12A7 and 13H9) were found to functionally complement this mutation. That is, mutant S9 carrying either cosmid 12A7 or 13H9 was shown to synthesize cyclic β-(1,2)-glucans containing phosphoglycerol substituents (Fig. 2).

FIG. 2.

FIG. 2

Open in a new tab

Thin-layer chromatogram of cyclic β-(1,2)-glucans isolated from S. meliloti mutant S9 complemented with cosmids 12A7 and 13H9. Periplasmic glucans were extracted from cells by using 70% ethanol, and half of each extract was treated with alkali before thin-layer chromatographic analysis as previously described (8). Alkali treatment results in the selective removal of succinyl substituents, whereas phosphoglycerol substituents are resistant to alkali treatment. Thus, cyclic β-(1,2)-glucans containing phosphoglycerol substituents are revealed by thin-layer chromatography after alkali treatment as anionic cyclic β-(1,2)-glucans. From left to right: S. meliloti S9 containing cosmid 13H9 (lanes 1 and 2); S. meliloti S9 containing cosmid 12A7 (lanes 3 and 4); S. meliloti 1021 (lanes 5 and 6); and S. meliloti S9 (lanes 7 and 8). +, extracts subjected to alkali treatment prior to thin-layer chromatography; −, extracts not subjected to alkali treatment. Anionic cyclic β-(1,2)-glucans containing either or both succinyl and phosphoglycerol substituents have the same relative mobility on this thin-layer chromatography system.

Restriction enzyme analysis of cosmids 12A7 and 13H9 revealed that both contained the same 25-kb insert of genomic DNA in the same orientation within the cosmid vector pLAFRI (Fig. 3A). The only difference between the two cosmids was that 13H9 contained an additional 2.5-kb EcoRI fragment at the end distal to the cos site of pLAFRI. When the 3.3-kb biotinylated probe for DNA sequences flanking the Tn5 insertion within mutant S9 was used in hybridization experiments with restriction enzyme-digested cosmid preparations, it was possible to determine the location of Tn5 within the genomic inserts of both cosmids (Fig. 3). Ultimately, sequencing the product of inverse PCR revealed the precise location of the 9-bp target sequence of the Tn5 insertion. Additional restriction enzymes were used to further map a 12-kb region surrounding the Tn5 insertion (Fig. 3B), thus providing a strategy for the subcloning experiments described below.

FIG. 3.

FIG. 3

Open in a new tab

Restriction maps of genomic DNA inserts within cosmids 12A7 and 13H9 and plasmid subclones. (A) EcoRI restriction maps of genomic DNA inserts within cosmids 12A7 and 13H9. The genomic DNA fragments are present in the same orientation in both cosmids, and the right ends of the inserts shown are near the cos site of the cosmid vector pLAFRI. (B) Enlarged view of an approximately 12-kb region within cosmids 12A7 and 13H9 (corresponding to the shaded area in A) that contains the Tn5 insert. Recognition sites for EcoRI, BamHI, SalI, and ApaI are shown. Plasmid pEW1 is a derivative of pTR100 containing a 2.9-kb BamHI fragment. Plasmid pEW2 is a derivative of pTR100 containing a 3.1-kb HindIII fragment. The 3.1-kb HindIII fragment was isolated from plasmid pFL6 and contains the 50-bp EcoRI-HindIII polylinker region from pUC19 as well as the 3.1-kb HindIII-EcoRI fragment derived from the 12-kb region within cosmids 12A7 and 13H9 (see Table 1). Note that not all HindIII recognition sites have been determined for this 12-kb region. Thus, the HindIII restriction map is not shown. The dark, horizontal bar indicates the 3,524 bp sequenced in the present study. (C) Further enlargement of sequenced region (3,524 bp). The horizontal arrows correspond to the location of the two open reading frames identified in the present study and also indicate the direction of transcription. A, ApaI; B, BamHI; E, EcoRI; H, HindIII; S, SalI. Vertical arrows indicate the location of the Tn5 insertion within mutant S9.

Subcloning experiments revealed that mutant S9 could be fully complemented by the 3.1-kb HindIII-EcoRI fragment within plasmid pEW2. That is, cyclic β-(1,2)-glucans synthesized by mutant S9 carrying plasmid pEW2 were found to contain wild-type levels of phosphoglycerol substituents (Fig. 4). Furthermore, cyclic β-(1,2)-glucans synthesized by R. leguminosarum bv. trifolii TA-1JH carrying plasmid pEW2 were also found to contain high levels of phosphoglycerol substituents (Fig. 4), consistent with the functional expression of an S. meliloti phosphoglycerol transferase gene. Additional subcloning experiments revealed that a partially overlapping 2.9-kb BamHI fragment (within plasmid pEW1) could not complement mutant S9 (data not shown) (the locations of the 3.1-kb HindIII-EcoRI fragment and the 2.9-kb BamHI fragment with respect to cosmids 13H9 and 12A7 are shown in Fig. 3B).

FIG. 4.

FIG. 4

Open in a new tab

DEAE-cellulose anion-exchange column chromatography profiles of the cell-associated cyclic β-(1,2)-glucans from S. meliloti mutant S9 and R. leguminosarum biovar trifolii TA-1JH. Cell-associated glucans were extracted from cells as described in the text and purified by gel filtration column chromatography on Sephadex G-50 followed by desalting on a Sephadex G-15 column as previously described (8). Prior to fractionation on DEAE-cellulose, extracts were treated with alkali to selectively remove succinyl substituents (as described in the legend to Fig. 2). These samples were again desalted on Sephadex G-15 prior to application to DEAE-cellulose. Because phosphoglycerol substituents are resistant to alkali treatment, cyclic β-(1,2)-glucans containing these substituents remain anionic after treatment and elute from the column only upon application of a salt gradient (beginning at approximately 40 ml). Neutral cyclic β-(1,2)-glucans elute at the void volume between 10 and 20 ml. (A) S. meliloti mutant S9. No anionic cyclic β-(1,2)-glucans were detected after alkali treatment. (B) S. meliloti mutant S9 carrying plasmid pEW2. Several peaks corresponding to anionic, glycerophosphorylated cyclic β-(1,2)-glucans were observed. The phosphorus content was determined to correspond to 1.7 mol per mol of total cyclic β-(1,2)-glucan, assuming an average size of 20 glucose residues per glucan molecule. (C) R. leguminosarum bv. trifolii TA-1JH. No anionic cyclic β-(1,2)-glucans were detected after alkali treatment. (D) R. leguminosarum bv. trifolii TA-1JH carrying plasmid pEW2. Several peaks corresponding to anionic, glycerophosphorylated cyclic β-(1,2)-glucans were observed. The phosphorus content was determined to correspond to 1.3 mol per mol of total cyclic β-(1,2)-glucan, assuming an average size of 20 glucose residues per glucan molecule. OD, optical density.

DNA sequence analysis.

Based on the results of subcloning and complementation experiments, the DNA sequence of the entire 3.1-kb HindIII-EcoRI fragment was determined. DNA sequence analysis revealed the presence of two major open reading frames (designated ORF1 and ORF2) within the 3.1-kb HindIII-EcoRI genomic fragment. These open reading frames are on opposite strands and are almost completely overlapping. ORF1 begins at nucleotide 711 and ends at nucleotide 2660, while ORF2 is carried on the cDNA strand beginning at nucleotide 2450 and ending at nucleotide 684. It is noted that the partially overlapping 2.9-kb BamHI fragment present in plasmid pEW1 (Fig. 3B), which did not complement mutant S9 (see above), contains only a portion of both open reading frames.

Additional sequencing studies using the inverse PCR probe as template revealed that the Tn5 insertion within mutant S9 lies within both ORF1 and ORF2 (Fig. 3C). The target site for the Tn5 insertion resides at nucleotides 2340 to 2348. In the case of ORF1, the Tn5 insertion is present near the region of the gene encoding the carboxyl terminus of the predicted protein. However, in the case of ORF2, the Tn5 insertion is present near the region of the gene encoding the amino terminus of the predicted protein. Because of the almost complete overlap of ORF1 and ORF2, it was not possible to determine from the complementation experiments whether Tn5 disruption of ORF1 or ORF2 is responsible for the mutant phenotype observed. Therefore, primer extension analyses were performed to determine if either or both of the transcripts were expressed. An ORF2 transcript could not be detected by using several different primers. However, a strong ORF1 transcript was clearly observed (see Fig. 6). The start site for this transcript is located at nucleotide 578 of Fig. 5. A sequence resembling the canonical −10 promoter element is not present. However, a 5′-GTGACA-3′ sequence, resembling the canonical −35 promoter element, is positioned between nucleotides −41 and −36 with respect to the transcription start site. Potential regulatory regions present upstream of the transcription start site are discussed below.

FIG. 6.

FIG. 6

Open in a new tab

Localization of the ORF1 transcription start site by primer extension analysis. Primer extension reactions were performed with RNA isolated from S. meliloti 1021 grown under low- (lanes 1 to 3) and high- (lanes 4 to 6) osmolarity conditions. Reaction mixtures contained 25 (lanes 1 and 4), 50 (lanes 2 and 5), or 100 μg (lanes 3 and 6) of RNA. The oligonucleotide primer used was 18-mer with the sequence 5′-CGGTTTGTGAACGGACTC-3′, corresponding to nucleotides 693 to 676 in Fig. 5. Dideoxy sequencing ladders were generated with the same primer used for the primer extension reaction. The nucleotide corresponding to the transcription start site is indicated by an asterisk. Cells were grown in GMS medium for low-osmolarity growth conditions and in GMS medium containing 0.4 M NaCl for high-osmolarity growth conditions.

FIG. 5.

FIG. 5

Open in a new tab

Nucleotide sequence of the region upstream of ORF1. Nucleotides 1 to 720 of the 3,524-nucleotide sequenced region are shown. Inverted repeat sequences are indicated by shading and dashed arrows and are numbered. Direct repeat sequences are shown by overlining. The transcription start site at nucleotide 578 is indicated by a boldface asterisk. The deduced amino acid sequence of the N terminus of the protein encoded by ORF1 is indicated with single-letter amino acid symbols aligned above the second nucleotide of each codon. A potential ribosome binding site is underlined and boldfaced.

The first possible start codon for ORF1 is an ATG codon located 133 nucleotides downstream of the transcription start site (nucleotide 711 of Fig. 5). A weak ribosome binding site (AAG) is positioned 7 to 9 nucleotides upstream of this start codon. The predicted protein contains 649 amino acids and has a calculated molecular mass of 71,600 Da. Hydrophobicity analysis using the TMpredict program did not predict any membrane-spanning segments within the deduced amino acid sequence encoded by ORF1. BLAST searches of the nonredundant and unfinished genome databases did not reveal any sequences with significant similarity to ORF1.

Although an ORF2 transcript was not detected, it is interesting to note that the deduced amino acid sequence encoded by ORF2 shows weak similarity to the amino acid sequences encoded by several open reading frames in the nonredundant sequence database at NCBI. These open reading frames include rkpI of S. meliloti (30), hypothetical open reading frames from Haemophilus influenzae (HI0275) (18) and Bacillus subtilis (yqgS, yflE, yfnI, and yvsF) (31), and an open reading frame (ORF7) of unknown function from Streptococcus mutans (48). The amino acid sequences encoded by these open reading frames have 19.1 (rkpI), 20.6 (HI0275), 19.6 (yqgS), 20.4 (yflE), 19.2 (yfnI), 18.3 (yvsF), and 18.1% (ORF7) identity to the deduced amino acid sequence encoded by ORF2. In addition, several translated sequences from the unfinished microbial genome database at NCBI were also found to have weak similarity to the deduced amino acid sequence encoded by ORF2. A PSI-BLAST search of the nonredundant sequence database detected similarity between the deduced amino acid sequence encoded by ORF2 and the arylsulfatase family of enzymes. Galperin et al. (21) have previously shown that the arylsulfatases are members of a metalloenzyme superfamily that includes alkaline phosphatases, phosphoglycerol mutases, and phosphopentomutases, as well as the MdoB phosphoglycerol transferase of E. coli (see below for more discussion of MdoB). Although transcription of ORF2 was not detected under the growth conditions tested, the similarity between the amino acid sequence encoded by ORF2 and the members of the metalloenzyme superfamily suggests that ORF2 encodes a protein that may be required under other growth conditions. The essentially complete overlap between ORF1 and ORF2 is an unusual situation and may play an interesting role in regulation of the expression of ORF1 and ORF2.

Transcriptional regulation of ORF1 expression.

Analysis of the region upstream of the translation start site of ORF1 revealed the presence of five inverted repeat elements. These inverted repeat elements have the sequence 5′-CCTGTG-(X5)-CACAGG-3′ (Fig. 5). The presence of the 5-bp spacer between the repeats places them approximately one helix turn apart. In addition, the beginning of each inverted repeat element is spaced 85 bp from the beginning of the next inverted repeat (except for the last inverted repeat, which is spaced 88 bp from the one before it). This spacing places the repeat elements approximately eight helix turns apart. The fifth inverted repeat element is positioned downstream of the transcription start site but is still located upstream of the translation start site.

Interestingly, the 5-bp spacers fall into two classes. The spacers for elements 1, 3, and 4 all have the sequence 5′-AC(A/G)(A/G)G-3′. The spacers for elements 2 and 5 have the sequence 5′-CTCGT-3′. The first four inverted repeat sequences are located within the nearly perfect 129-bp direct repeat sequences indicated by the overlined sequences in Fig. 5. The role that these inverted and direct repeat elements play in transcriptional regulation is unknown.

BLASTn searches with the nucleotide sequence upstream of ORF1 revealed the presence of an inverted repeat element upstream of the exoH (4) (GenBank accession no. Z17219) and lppB (45) (GenBank accession no. U81296) genes of S. meliloti and downstream of the acsA (GenBank accession no. AF080217) and rpsA (GenBank accession no. X07528) genes of S. meliloti and the abg gene (46) (GenBank accession no. M19033) of Agrobacterium sp. strain ATCC 21400. Interestingly, these inverted repeat elements are all identical to upstream element 2 of ORF1. Furthermore, both lppB and abg have a second inverted repeat similar to ORF1 upstream elements 1, 3, and 4. In addition to the inverted repeat elements, the regions upstream of lppB and exoH show an extended region (120 and 89 bp, respectively) of sequence identity (>85%) to the region upstream of the ORF1 beginning near inverted repeat element 2. These sequence similarities suggest that ORF1, the exoH and lppB genes, and the genes downstream of abg, acsA, and rpsA may be subject to a common form of regulation.

Since biosynthesis of cyclic β-glucans is osmotically regulated (9), it might be expected that ORF1 is also subject to osmotic regulation. Indeed, primer extension analysis revealed that transcription was strongly induced in S. meliloti 1021 cells grown under low-osmolarity conditions and was repressed in cells grown under high-osmolarity conditions (Fig. 6). These results are consistent with an earlier study from our laboratory which revealed that synthesis of glycerophosphorylated cyclic β-glucans is inhibited in S. meliloti 1021 when cells are grown at high osmolarity (10). However, this earlier study indicated that inhibition occurred at the level of enzyme activity and that the phosphoglycerol transferase is present constitutively (since synthesis of glycerophosphorylated cyclic β-glucans resumed upon hypo-osmotic shock even in the absence of protein synthesis). Therefore, we conclude that phosphoglycerol transferase activity in S. meliloti is regulated at both the transcriptional and posttranslational levels.

Expression of ORF1 in R. leguminosarum TA-1JH.

As shown above, R. leguminosarum bv. trifolii TA-1JH containing plasmid pEW2 produces cyclic β-(1,2)-glucans containing phosphoglycerol substituents, consistent with the functional expression of an S. meliloti phosphoglycerol transferase gene. Primer extension experiments using RNA preparations from R. leguminosarum bv. trifolii TA-1JH containing plasmid pEW2 confirmed the presence of the ORF1 transcript, while R. leguminosarum bv. trifolii TA-1JH containing the control plasmid (pTR100) did not contain this transcript (Fig. 7). Interestingly, the transcript detected in R. leguminosarum bv. trifolii TA-1JH (pEW2) did not have the same start nucleotide as that found for S. meliloti 1021. Instead, transcription began 13 nucleotides downstream of the transcription site used in S. meliloti.

FIG. 7.

FIG. 7

Open in a new tab

Transcription of the cgmB gene in R. leguminosarum bv. trifolii TA-1JH. Primer extension reactions were performed by using 15-μg samples of RNA isolated from R. leguminosarum bv. trifolii TA-1JH containing pTR100 (lane 1), R. leguminosarum bv. trifolii TA-1JH containing pEW2 (lane 2), and S. meliloti 1021 (lane 3). The oligonucleotide primer used in primer extension reactions and for generation of dideoxy sequencing ladders is the same as that used in Fig. 6.

ORF1 is designated cgmB for cyclic glucan modification.

Based on the phenotype of mutant S9, it appears likely that a gene encoding a phosphoglycerol transferase has been disrupted by the Tn5 insertion. We conclude this from our demonstration that this mutant is specifically impaired for the addition of phosphoglycerol substituents to the cyclic β-(1,2)-glucan backbone (8). Furthermore, both mutant S9 and R. leguminosarum bv. trifolii TA-1JH were shown to produce cyclic β-(1,2)-glucans containing high levels of phosphoglycerol substituents when ORF1 was expressed in these bacteria after transformation with plasmid pEW2. Based on these phenotypes, ORF1 is now designated cgmB for cyclic glucan modification.

Curiously, only limited similarity (17.5% identity) was found between CgmB and the product of the mdoB gene of E. coli (13, 33). MdoB is a phosphoglycerol transferase which mediates the transfer of sn-1-phosphoglycerol substituents from the head group of phosphatidylglycerol to the membrane-derived oligosaccharides (MDO) of E. coli (16, 27). The MDO of E. coli are periplasmic β-glucans that share many properties with the cyclic β-glucans of the Rhizobiaceae (29). Based on our earlier demonstration that the phosphoglycerol substituents on the cyclic β-(1,2)-glucans are also derived from the head group of phosphatidylglycerol (37), it would be predicted that an enzyme similar to MdoB should be present in S. meliloti. It must be noted, however, that hydrophobicity analysis by the method of Kyte and Doolittle (32) clearly predicts that the product of the mdoB gene (GenBank accession no. U14003) is a hydrophobic, transmembrane protein, consistent with its membrane localization (28), while CgmB is not predicted to have any transmembrane domains. This could be a major reason for the rather limited similarity between these two proteins.

Although the similarity between CgmB and MdoB is poor, it is still possible that both are phosphoglycerol transferases. In this regard, it should be noted that a second phosphoglycerol transferase (phosphoglycerol transferase II) that is also involved in the addition of phosphoglycerol substituents to the MDO β-glucan backbone has been detected in E. coli. This second phosphoglycerol transferase is localized within the periplasmic compartment and is believed to catalyze the transfer of phosphoglycerol moieties from nascent MDO (MDO bound to a lipid carrier) to periplasmic MDO (29). Thus, phosphoglycerol transferase II functions in secondary transfer reactions leading to MDO containing multiple phosphoglycerol substituents. The possibility exists that CgmB may be more homologous with phosphoglycerol transferase II. We note, however, that to our knowledge, the E. coli gene encoding phosphoglycerol transferase II has not yet been identified.

We have identified a gene, cgmB, from S. meliloti that encodes a phosphoglycerol transferase responsible for the addition of phosphoglycerol substituents to the cyclic β-(1,2)-glucan backbone. The cgmB open reading frame (ORF1) has the unusual feature that it completely overlaps an open reading frame (ORF2) on the opposite DNA strand encoding a sequence of 588 amino acids. It is especially intriguing that the deduced amino acid sequence of ORF2 shows weak similarity to the amino acid sequences of enzymes of the arylsulfatase superfamily, of which MdoB is a member. However, direct comparison of the amino acid sequence of the product of ORF2 to the sequence of MdoB reveals only 17.3% identity. Although we were unable to detect expression of ORF2 under the conditions tested, the possibility exists that this gene is expressed under different growth conditions. The complete overlap of these open reading frames may have implications in the regulation of one or both of these genes.

ACKNOWLEDGMENTS

This research was supported by National Science Foundation grant MCB-9505706.

REFERENCES

  • 1.Altschul S F, Gish W, Miller W, Myers E W, Lipman D J. Basic local alignment search tool. J Mol Biol. 1990;215:403–410. doi: 10.1016/S0022-2836(05)80360-2. [DOI] [PubMed] [Google Scholar]
  • 2.Altschul S F, Madden T L, Schäffer A A, Zhang J, Zhang Z, Miller W, Lipman D J. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997;25:3389–3402. doi: 10.1093/nar/25.17.3389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Anderson M L M, Young B D. Quantitative filter hybridization. In: Hames B D, Higgins S J, editors. Nucleic acid hybridization: a practical approach. Oxford, England: IRL Press; 1985. pp. 73–111. [Google Scholar]
  • 4.Becker A A, Kleickmann A A, Arnold W, Puhler A. Analysis of the Rhizobium meliloti exoH/exoK/exoL fragment: ExoK shows homology to excreted endo-beta-1,3-1,4-glucanases and ExoH resembles membrane proteins. Mol Gen Genet. 1993;238:145–154. doi: 10.1007/BF00279541. [DOI] [PubMed] [Google Scholar]
  • 5.Bhagwat A A, Gross K C, Tully R E, Keister D L. β-glucan synthesis in Bradyrhizobium japonicum: characterization of a new locus (ndvC) influencing β-(1→6) linkages. J Bacteriol. 1996;178:4635–4642. doi: 10.1128/jb.178.15.4635-4642.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Bhagwat A A, Keister D L. Site-directed mutagenesis of the β-(1,3), β-(1,6)-D-glucan synthesis locus of Bradyrhizobium japonicum. Mol Plant-Microbe Interact. 1995;8:366–370. [Google Scholar]
  • 7.Bhagwat A A, Tully R E, Keister D L. Identification and cloning of cyclic β-(1,3), β-(1,6)-d-glucan synthesis locus from Bradyrhizobium japonicum. FEMS Microbiol Lett. 1993;114:139–144. doi: 10.1111/j.1574-6968.1993.tb06564.x. [DOI] [PubMed] [Google Scholar]
  • 8.Breedveld M W, Hadley J A, Miller K J. A novel cyclic β-1,2-glucan mutant of Rhizobium meliloti. J Bacteriol. 1995;177:6346–6351. doi: 10.1128/jb.177.22.6346-6351.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Breedveld M W, Miller K J. Cyclic β-glucans of members of the family Rhizobiaceae. Microbiol Rev. 1994;58:145–161. doi: 10.1128/mr.58.2.145-161.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Breedveld M W, Miller K J. Synthesis of glycerophosphorylated cyclic (1,2)-β-glucans in Rhizobium meliloti strain 1021 after osmotic shock. Microbiology. 1995;141:583–588. doi: 10.1099/13500872-141-3-583. [DOI] [PubMed] [Google Scholar]
  • 11.Breedveld M W, Zevenhuizen L P T M, Zehnder A J B. Osmotically-induced oligo- and polysaccharide synthesis by Rhizobium meliloti SU-47. J Gen Microbiol. 1990;136:2511–2519. [Google Scholar]
  • 12.Breedveld M W, Zevenhuizen L P T M, Zehnder A J B. Synthesis of cyclic β-(1,2)-glucans by Rhizobium leguminosarum biovar trifolii TA-1: factors influencing excretion. J Bacteriol. 1992;174:6336–6342. doi: 10.1128/jb.174.20.6336-6342.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Burland V, Plunkett G, Sofia H J, Daniels D L, Blattner F R. Analysis of the Escherichia coli genome VI. DNA sequence of the region from 92.8 through 100 minutes. Nucleic Acids Res. 1995;23:2105–2219. doi: 10.1093/nar/23.12.2105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Cangelosi G A, Martinetti G, Leigh J A, Lee C C, Theines C, Nester E W. Role for Agrobacterium tumefaciens ChvA protein in export of β-1,2-glucan. J Bacteriol. 1989;171:1609–1615. doi: 10.1128/jb.171.3.1609-1615.1989. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Castro O A, Zorreguieta A, Ielmini V, Vega G, Ielpi L. Cyclic β-(1,2)-glucan synthesis in Rhizobiaceae: roles of the 319-kilodalton protein intermediate. J Bacteriol. 1996;178:6043–6048. doi: 10.1128/jb.178.20.6043-6048.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Fiedler W, Rotering H. Characterization of an Escherichia coli mdoB mutant strain unable to transfer sn-1-phosphoglycerol to membrane-derived oligosaccharides. J Biol Chem. 1985;260:4799–4806. [PubMed] [Google Scholar]
  • 17.Figurski D H, Helinski D R. Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc Natl Acad Sci USA. 1979;76:1648–1652. doi: 10.1073/pnas.76.4.1648. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Fleischmann R D, Adams M D, White O, Clayton R A, Kirkness E F, Kerlavage A R, Bult C J, Tomb J F, Dougherty B A, Merrick J M, McKenney K, Sutton G, FitzHugh W, Fields C A, Gocayne J D, Scott J D, Shirley R, Liu L-I, Glodek A, Kelley J M, Weidman J F, Phillips C A, Spriggs T, Hedblom E, Cotton M D, Utterback T R, Hanna M C, Nguyen D T, Saudek D M, Brandon R C, Fine L D, Fritchman J L, Fuhrman J L, Geoghagen N S M, Gnehm C L, McDonald L A, Small K V, Fraser C M, Smith H O, Venter J C. Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science. 1995;269:496–512. doi: 10.1126/science.7542800. [DOI] [PubMed] [Google Scholar]
  • 19.Franssen H J, Vijn I, Yang W C, Bisseling T. Developmental aspects of the Rhizobium-legume symbiosis. Plant Mol Biol. 1992;19:89–107. doi: 10.1007/BF00015608. [DOI] [PubMed] [Google Scholar]
  • 20.Friedman A M, Long S R, Brown S E, Buikema W J, Ausubel F M. Construction of a broad host range cosmid cloning vector and its use in the genetic analysis of Rhizobium mutants. Gene. 1982;18:289–296. doi: 10.1016/0378-1119(82)90167-6. [DOI] [PubMed] [Google Scholar]
  • 21.Galperin M Y, Bairoch A, Koonin E V. A superfamily of metalloenzymes unifies phosphopentomutase and cofactor-independent phosphoglycerate mutase with alkaline phosphatases and sulfatases. Protein Sci. 1998;7:1829–1835. doi: 10.1002/pro.5560070819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Gray J X, Rolfe B G. Exopolysaccharide production in Rhizobium and its role in invasion. Mol Microbiol. 1990;4:1425–1431. doi: 10.1111/j.1365-2958.1990.tb02052.x. [DOI] [PubMed] [Google Scholar]
  • 23.Hanahan D. Techniques for transformation of E. coli. In: Glover D M, editor. DNA cloning. Vol. 1. Oxford, England: IRL Press; 1985. pp. 109–135. [Google Scholar]
  • 24.Hofmann R, Stoffel W. TMBASE - a database of membrane-spanning protein segments. Biol Chem Hoppe-Seyler. 1993;347:166. [Google Scholar]
  • 25.Ielpi L, Dylan T, Ditta G S, Helinski D R, Stanfield S W. The ndvB locus of Rhizobium meliloti encodes a 319-kDa protein involved in the production of β-(1,2) glucan. J Biol Chem. 1990;265:2843–2851. [PubMed] [Google Scholar]
  • 26.Iñon de Iannino N, Ugalde R A. Biochemical characterization of avirulent Agrobacterium tumefaciens chvA mutants: synthesis and excretion of β-(1,2)glucan. J Bacteriol. 1989;171:2842–2849. doi: 10.1128/jb.171.5.2842-2849.1989. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Jackson B J, Bohin J-P, Kennedy E P. Biosynthesis of membrane-derived oligosaccharides: characterization of mdoB mutants defective in phosphoglycerol transferase I activity. J Bacteriol. 1984;160:976–981. doi: 10.1128/jb.160.3.976-981.1984. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Jackson B J, Kennedy E P. The biosynthesis of membrane-derived oligosaccharides: a membrane-bound phosphoglycerol transferase. J Biol Chem. 1993;258:2394–2398. [PubMed] [Google Scholar]
  • 29.Kennedy E P. Membrane-derived oligosaccharides (periplasmic beta-d-glucans) of Escherichia coli. In: Neidhardt F C, Curtiss III R, Ingraham J L, Lin E C C, Low K B, Magasanik B, Reznikoff W S, Riley M, Schaechter M, Umbarger H E, editors. Escherichia coli and Salmonella cellular and molecular biology. Washington, D.C: ASM Press; 1996. pp. 1064–1071. [Google Scholar]
  • 30.Kiss E, Reuhs B L, Kim J S, Kereszt A, Petrovics G, Putnoky P, Dusha I, Carlson R W, Kondorosi A. The rkpGHI and J genes are involved in capsular polysaccharide production by Rhizobium meliloti. J Bacteriol. 1997;179:2132–2140. doi: 10.1128/jb.179.7.2132-2140.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Kunst F, Ogasawara N, Moszer I, et al. The complete genome sequence of the Gram-positive bacterium Bacillus subtilis. Nature. 1997;390:249–256. doi: 10.1038/36786. [DOI] [PubMed] [Google Scholar]
  • 32.Kyte J, Doolittle R F. A simple method for displaying the hydropathic character of a protein. J Mol Biol. 1982;157:105–132. doi: 10.1016/0022-2836(82)90515-0. [DOI] [PubMed] [Google Scholar]
  • 33.Lanfroy E, Bohin J-P. Physical map location of the Escherichia coli gene encoding phosphoglycerol transferase I. J Bacteriol. 1993;175:5736–5737. doi: 10.1128/jb.175.17.5736-5737.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Leigh J A, Signer E R, Walker G C. Exopolysaccharide-deficient mutants of Rhizobium meliloti that form ineffective nodules. Proc Natl Acad Sci USA. 1985;82:6231–6235. doi: 10.1073/pnas.82.18.6231. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Leigh J L, Coplin D L. Exopolysaccharides in plant-bacterial interactions. Annu Rev Microbiol. 1992;46:307–346. doi: 10.1146/annurev.mi.46.100192.001515. [DOI] [PubMed] [Google Scholar]
  • 36.Miller J H. Experiments in molecular genetics. Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory Press; 1972. [Google Scholar]
  • 37.Miller K J, Gore R S, Benesi A J. Phosphoglycerol substituents present on the cyclic β-1,2-glucans of Rhizobium meliloti are derived from phosphatidylglycerol. J Bacteriol. 1988;170:4569–4575. doi: 10.1128/jb.170.10.4569-4575.1988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Moran C P. Measuring gene expression in Bacillus. In: Harwood C R, Cutting S M, editors. Molecular biological methods for bacillus. New York, N.Y: John Wiley and Sons; 1990. pp. 267–293. [Google Scholar]
  • 39.Rich J J, Willis D K. A single oligonucleotide can be used to rapidly isolate DNA sequences flanking a transposon Tn5 insertion by the polymerase chain reaction. Nucleic Acids Res. 1990;18:6673–6676. doi: 10.1093/nar/18.22.6673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Sambrook J, Fritsch E F, Maniatis T. Molecular cloning: a laboratory manual. 2nd ed. Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory Press; 1989. [Google Scholar]
  • 41.Sanger F, Nicklen S, Coulson A R. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA. 1977;74:5463–5467. doi: 10.1073/pnas.74.12.5463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Spaink H P. Rhizobial lipo-oligosaccharides—answers and questions. Plant Mol Biol. 1992;6:977–986. doi: 10.1007/BF00027167. [DOI] [PubMed] [Google Scholar]
  • 43.Stanfield S W, Ielpi L, O’Brochta D, Helinski D R, Ditta G S. The ndvA gene product of Rhizobium meliloti is required for β-(1,2)-glucan production and has homology to the ATP-binding export protein HlyB. J Bacteriol. 1988;170:3523–3530. doi: 10.1128/jb.170.8.3523-3530.1988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Streit W, Bjourson A J, Cooper J E, Werner D. Application of subtraction hybridization for the development of a Rhizobium leguminosarum biovar phaseoli and Rhizobium tropici group-specific DNA probe. FEMS Microbiol Ecol. 1993;13:59–68. [Google Scholar]
  • 45.Streit W R, Phillips D A. A biotin-regulated locus, bioS, in a possible survival operon of Rhizobium meliloti. Mol Plant-Microbe Interact. 1997;10:933–937. doi: 10.1094/MPMI.1997.10.7.933. [DOI] [PubMed] [Google Scholar]
  • 46.Wakarchuk W W, Greenberg N M, Kilburn D G, Miller R C, Warren R A J. Structure and transcription analysis of the gene encoding a cellobiase from Agrobacterium sp. strain ATCC 21400. J Bacteriol. 1988;170:301–307. doi: 10.1128/jb.170.1.301-307.1988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Weinstein M, Roberts R C, Helinski D R. A region of the broad-host-range plasmid RK2 causes stable in planta inheritance of plasmids in Rhizobium meliloti cells isolated from alfalfa root nodules. J Bacteriol. 1992;174:7486–7489. doi: 10.1128/jb.174.22.7486-7489.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Yamashita Y, Tsukioka Y, Tomihisa K, Nakano Y, Koga T. Genes involved in cell wall localization and side chain formation of rhamnose-glucose polysaccharide in Streptococcus mutans. J Bacteriol. 1998;180:5803–5807. doi: 10.1128/jb.180.21.5803-5807.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Yannisch-Perron C, Vieira J, Messing J. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene. 1985;33:103–119. doi: 10.1016/0378-1119(85)90120-9. [DOI] [PubMed] [Google Scholar]
  • 50.Zorreguieta A, Geremia R A, Cavaignac S, Cangelosi G A, Nester E A, Ugalde R A. Identification of the product of an Agrobacterium tumefaciens chromosomal virulence gene. Mol Plant-Microbe Interact. 1988;1:121–127. doi: 10.1094/mpmi-1-121. [DOI] [PubMed] [Google Scholar]
  • 51.Zorreguieta A, Ugalde R A. Formation in Rhizobium and Agrobacterium spp. of a 235-kilodalton protein intermediate in β-d(1,2) glucan synthesis. J Bacteriol. 1986;167:947–951. doi: 10.1128/jb.167.3.947-951.1986. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES