Skip to main content
NCBI home page
Journal of Bacteriology logoLink to Journal of Bacteriology
. 1998 Sep;180(17):4392–4400. doi: 10.1128/jb.180.17.4392-4400.1998

Molecular Cloning and Characterization of cgs, the Brucella abortus Cyclic β(1-2) Glucan Synthetase Gene: Genetic Complementation of Rhizobium meliloti ndvB and Agrobacterium tumefaciens chvB Mutants

Nora Iñón de Iannino 1, Gabriel Briones 1,2, Marcelo Tolmasky 3, Rodolfo A Ugalde 1,*
PMCID: PMC107446  PMID: 9721274

Abstract

The animal pathogen Brucella abortus contains a gene, cgs, that complemented a Rhizobium meliloti nodule development (ndvB) mutant and an Agrobacterium tumefaciens chromosomal virulence (chvB) mutant. The complemented strains recovered the synthesis of cyclic β(1-2) glucan, motility, virulence in A. tumefaciens, and nitrogen fixation in R. meliloti; all traits were strictly associated with the presence of an active cyclic β(1-2) glucan synthetase protein in the membranes. Nucleotide sequencing revealed the presence in B. abortus of an 8.49-kb open reading frame coding for a predicted membrane protein of 2,831 amino acids (316.2 kDa) and with 51% identity to R. meliloti NdvB. Four regions of the B. abortus protein spanning amino acids 520 to 800, 1025 to 1124, 1284 to 1526, and 2400 to 2660 displayed similarities of higher than 80% with R. meliloti NdvB. Tn3-HoHo1 mutagenesis showed that the C-terminal 825 amino acids of the Brucella protein, although highly conserved in Rhizobium, are not necessary for cyclic β(1-2) glucan synthesis. Confirmation of the identity of this protein as B. abortus cyclic β(1-2) glucan synthetase was done by the construction of a B. abortus Tn3-HoHo1 insertion mutant that does not form cyclic β(1-2) glucan and lacks the 316.2-kDa membrane protein. The recovery of this mutant from the spleens of inoculated mice was decreased by 3 orders of magnitude compared with that of the parental strain; this result suggests that cyclic β(1-2) glucan may be a virulence factor in Brucella infection.

Symbiotic nitrogen-fixing Rhizobium meliloti and crown gall tumor-inducing Agrobacterium tumefaciens have in common the ability to synthesize periplasmic cyclic β(1-2) glucan (2, 17, 26, 41, 42). This glucan is required, through a still-unknown mechanism, for effective nodule invasion or crown gall tumor induction (12, 14, 28). chvB in A. tumefaciens and ndvB in R. meliloti were identified as the genes coding for a high-molecular-weight inner membrane protein characterized as cyclic β(1-2) glucan synthetase (45). The ndvB and chvB genes are interchangeable, and mutations in one gene can be complemented by the other, indicating that their functions are highly conserved (12). Cyclic β(1-2) glucan synthetase contains all the enzymatic activities required for the synthesis of cyclic glucan, i.e., initiation, elongation, and cyclization (1, 4). The protein acts as an intermediate; it has been postulated that a linear polyglucose chain covalently bound through the reducing end to an as-yet-unidentified amino acid elongates until it reaches the appropriate degree of polymerization and thereafter is cyclized and released from the protein. All of these reactions are believed to take place on the cytoplasmic side of the inner membrane; it has been postulated that an inner membrane protein encoded by chvA in A. tumefaciens is responsible for secretion of the nonsubstituted cyclic glucan into the periplasm, where glycerol phosphate, succinate, and methyl malonate are added (21). Besides being defective in cyclic glucan synthesis, chvB and ndvB mutants are also defective in the assembly of flagella, having a nonmotile phenotype as a consequence (10, 14).

Brucella spp. are nonmotile gram-negative bacteria that cause a chronic disease in humans and other animals that results from the survival of the pathogens inside macrophages (34).

Brucella spp., which belong, according to 16S rRNA sequences, to the same (α 2) group of the Proteobactereaceae (7) as R. meliloti and A. tumefaciens, are also able to synthesize cyclic β(1-2) glucan (3). We recently studied the biosynthesis of cyclic β(1-2) glucan in Brucella abortus and B. ovis. A high-molecular-weight inner membrane protein similar to that observed in R. meliloti and A. tumefaciens was identified as cyclic β(1-2) glucan synthetase (3). Two differences were observed between the Brucella and Agrobacterium or Rhizobium cyclic β(1-2) glucan synthetases: (i) isolated Brucella membranes are defective in the cyclization reaction in vitro, accumulating a high-molecular-weight linear glucan on the protein intermediate after a chase with 2 mM UDP-glucose, and (ii) the accumulation of glucan in Brucella is not osmoregulated (3). The role of this glucan in Brucella-host interactions is not known yet. In the present work, we applied a functional complementation approach to clone a Brucella gene equivalent to chvB and ndvB from a B. abortus S19 gene library, based on the ability of recombinant cosmids to restore the motility of an R. meliloti ndvB mutant. We identified and characterized recombinant plasmids able to restore all the functions assigned to the ndvB locus of R. meliloti and the chvB locus of A. tumefaciens, i.e., the presence of a high-molecular-weight inner membrane protein, the synthesis of cyclic β(1-2) glucan, motility, and the ability to form normal nitrogen-fixing nodules in alfalfa roots and crown gall tumors in Kalanchoe daigremontiana leaves. The complete nucleotide sequence of the B. abortus S19 gene equivalent to ndvB and chvB is given.

(Part of this research was presented at the 96th General Meeting of the American Society for Microbiology, New Orleans, La., 19 to 23 May 1996.)

MATERIALS AND METHODS

Bacterial strains and plasmids.

R. meliloti, Escherichia coli, A. tumefaciens, and B. abortus strains and plasmids used are listed in Table 1. R. meliloti, E. coli, and A. tumefaciens strains were grown on yeast extract-mannitol medium (AMA) (22), in Luria broth (LB) (31), and on tryptone-yeast medium (45), respectively. B. abortus strains were grown in brucella broth (BB) (Difco Laboratories, Detroit, Mich.).

TABLE 1.

Bacteria and plasmids

Strain or plasmid Phenotype or genotypea Reference or source
Strains
 E. coli LE392 supE44 supF58 hsdR514 galK2 galT22 metB1 trpR55 lacY1 31
 E. coli HB101 thr leu thi recA hsdR hsdM pro Strr 31
 E. coli DH5α F endA1 hsdR17 (rK− mK−) supE44 thi-1 λ− recA1 gyrA96 relA1 φ80d lacZΔM15Δ(lacZYA-argF)U169/F′ proAB lacIqlacZΔM15 Tn5 (Kmr) 31
 E. coli C2110 NalrpolA1 rha his 25
 R. meliloti GR4 Nod+ Fix+ 38
 R. meliloti GRT21s ndvB Nod− Fix− 38
 A. tumefaciens A348 Wild type; Vir+ 11
 A. tumefaciens A1011 chvB Vir− Tn5::chromosome 11
 B. abortus S19 Vaccine strain 32
 B. abortus BAI129 cgs mutant; Tn3-HoHo1::chromosome This study
Plasmids
 pRK2013 Rep ColE1 Kmr 11
 pVK102 IncP Mob+ Tra− 11
 pHoHo1 Ampr ColE1 Tn3 LacZ 11
 pSshe Camr pACYC184 tnpA 11
 pPH1JI IncP Gmr Strr 16
 pBBR1-MCS-2 Kmr 24
 pBA19 Tcr Mot+ pVK102 (B. abortus S19 chromosome) This study
 pBA25 Tcr Mot+ pVK102 (B. abortus S19 chromosome) This study
 pI133 Tcr Cbr Mot− pBA19::Tn3-HoHo1 in cgs region This study
 pI132 Tcr Cbr Mot− pBA19::Tn3-HoHo1 in cgs region This study
 pI129 Tcr Cbr Mot− pBA19::Tn3-HoHo1 in cgs region This study
 pI128 Tcr Cbr Mot− pBA19::Tn3-HoHo1 in cgs region This study
 pI121 Tcr Cbr Mot− pBA19::Tn3-HoHo1 in cgs region This study
 pM47 Tcr Cbr Mot+ pBA19::Tn3-HoHo1 in cgs region This study
 pM14 Tcr Cbr Mot+ pBA19::Tn3-HoHo1 in cgs region This study
 pM202 Tcr Cbr Mot+ pBA19::Tn3-HoHo1 in cgs region This study
 pM20 Tcr Cbr Mot+ pBA19::Tn3-HoHo1 in cgs region This study
 pM30 Tcr Cbr Mot+ pBA19::Tn3-HoHo1 in cgs region This study
 pM432 Tcr Cbr Mot+ pBA19::Tn3-HoHo1 in cgs region This study
 pM145 Tcr Cbr Mot+ pBA19::Tn3-HoHo1 in cgs region This study
 pM21 Tcr Cbr Mot+ pBA19::Tn3-HoHo1 in cgs region This study
 pBBE52 pBBR1MCS2 (5.2-kb EcoRI fragment of pBA19) This study
 pBBE50 pBBR1MCS2 (5.0 kb EcoRI fragment of pBA19) This study
 pBBE45 pBBR1MCS2 (4.5-kb EcoRI fragment of PBA19) This study
 pBBC10 pBBR1MCS2 (10.6-kb ClaI fragment of pBA19) This study
 pVCH43 pVK102 (4.3-kb HindIII fragment of pBA19) This study
Open in a new tab
a

Ampr, ampicillin resistance; Cbr, carbenicillin resistance; Camr, chloramphenicol resistance; Gmr, gentamicin resistance; Kmr, kanamycin resistance; Nalr, nalidixic acid resistance; Strr, streptomycin resistance; Tcr, tetracycline resistance; Mot+, motile strain; Mot−, nonmotile strain; Vir+, virulent strain. 

DNA techniques, genetic complementation, and Tn3-HoHo1 mutagenesis.

A B. abortus S19 genomic library was prepared basically as previously described (37). Total genomic DNA was prepared as described previously (18) and partially digested with HindIII. Partially digested DNA was ligated to the HindIII- digested pVK102 vector (23) and packaged in vitro with a commercial in vitro packaging system according to the manufacturer’s specifications (Amersham, Arlington Heights, Ill.). Exponential-phase E. coli LE392 cells grown in LB containing 0.2% maltose and 10 mM MgSO4 were infected with λ phage particles at 30°C for 30 min, followed by the addition of 1 volume of LB. After 2 h of incubation at 37°C, cells were plated on LB agar containing 20 μg of tetracycline per ml. Tetracycline-resistant clones that were kanamycin sensitive (100 μg/ml) were candidates to harbor plasmids with inserted B. abortus DNA fragments. Twelve pools containing 100 individual clones from the B. abortus S19 gene bank (representing approximately five times the genome) were mass conjugated by triparental mating (9) into an R. meliloti GRT21s ndvB mutant (38).

Mating was carried out by patching similar amounts of donor cells (pool of 100 clones from the B. abortus gene bank), receptor cells (R. meliloti GRT21s), and helper cells (E. coli HB101 containing pRK2013) on AMA agar and allowing them to grow for 1 day. A mating patch was recovered by washing the agar with distilled water and was plated on AB-sucrose agar medium (11) with tetracycline (10 μg/ml). Tetracycline-resistant clones were recovered from the plate by washing the agar with distilled water and were centrifuged in an Eppendorf centrifuge; pellets were punched with a toothpick into AB-sucrose soft-agar medium (0.4% agar). From 12 conjugation events, five motile clones were recovered after 3 days of incubation at 28°C. The external part of the motility halo was punched into AB-sucrose soft-agar medium (0.4% agar), and the bacteria were grown for 5 days. The external part of the motility halo was streaked on AB-sucrose agar medium with tetracycline, and five individual motile clones were isolated.

Tn3-HoHo1 mutagenesis was carried out as described previously (35). Cosmids harboring a Tn3-HoHo1 insertion were isolated by antibiotic selection (resistance to carbenicillin and tetracycline). Cosmids with Tn3-HoHo1 insertions were introduced by conjugation into R. meliloti GRT21s, and the resulting strains were screened for motility on AB-sucrose soft-agar medium plates with the appropriate antibiotics. More than 100 individual events were screened and scored by this procedure. The orientations of Tn3-HoHo1 insertions were determined by screening for blue colonies on AMA plates with 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside.

Mapping of insertions affecting motility (motile and nonmotile) in pBA19 was carried out by digesting cosmids recovered from E. coli HB101 strains with HindIII and EcoRI restriction enzymes. Gel electrophoresis on 0.8% agarose was used to estimate the sizes of the generated fragments. Precise mapping of Tn3-HoHo1 insertions I129, I121, M47, and M14 was carried out by DNA sequencing with an oligonucleotide (5′TAAAAGAGGCGTCAGAGG3′) complementary to a region located 50 bp downstream of the end of the left inverted repeated of the transposon. For nucleotide sequencing, the DNA cloned in pBA19 was subcloned in pBCSK(+) (Stratagene, La Jolla, Calif.). DNA sequencing was carried out by the dideoxy method (33) with an automated model 373 DNA sequencer (Perkin-Elmer Applied Biosystems Division, Foster City, Calif.) according to the manufacturer’s instructions. Protein alignment and comparisons were carried out by the Clustal method (15). Prediction of transmembrane helix regions and topology was carried out as described previously (30).

Isolation of β(1-2) glucan from cells.

Cells from 100-ml cultures were harvested by centrifugation at 10,000 × g for 10 min. Pellets were extracted with 1% trichloroacetic acid (TCA), and the extracts were subjected to gel filtration on a Bio-Gel P4 column (78 by 1.8 cm) (Bio-Rad Laboratories, Richmond, Calif.) as previously described (8, 21). Alternatively, cells from 3-ml cultures were washed once with 1 ml of water and centrifuged in an Eppendorf centrifuge, and glucans were extracted from the pellets with 0.3 ml of 70% ethanol at 37°C over 1 h. The cells were centrifuged in an Eppendorf centrifuge, and the supernatants were dried in a Speed-Vac centrifuge. Extracted glucans were redissolved in 70% ethanol and subjected to thin-layer chromatography (TLC) on Silica Gel-60 plates (Merck KGaA, Darmstadt, Germany) developed with n-butanol–ethanol–water (5:5:4, vol/vol) as described previously (43). Glucans on the TLC plates were detected by spraying with 5% sulfuric acid in ethanol and heating for 5 min at 120°C.

Acid hydrolysis, paper chromatography, reduction with sodium borohydride, and paper electrophoresis.

Acid hydrolysis, paper chromatography, reduction with sodium borohydride, and paper electrophoresis were carried out as described previously (21, 22).

In vitro synthesis of cyclic β(1-2) glucans.

In vitro synthesis of cyclic β(1-2) glucans was carried out, depending on the experiment, with three different type of enzyme preparations: permeabilized cells, total membranes, or inner membranes. Inner and total membranes were prepared by a modification of the Osborn-Munson method (27) as previously described (45). The membranes were resuspended in 30 mM HCl-Tris buffer (pH 8.0)–3 mM EDTA at approximately 40 mg of protein per ml and stored at −20°C until used. For the preparation of permeabilized cells, cultures were harvested by centrifugation (10,000 × g for 20 min), and cell pellets were resuspended in 0.01 M EDTA–Tris buffer (pH 8.2) and subjected to four cycles of freezing and thawing in liquid nitrogen. Total membranes, inner membranes, or permeabilized cells were incubated with UDP-[14C]glucose (300,000 cpm; 300 μCi/μmol) in 50 mM HCl-Tris buffer (pH 8.2)–10 mM MgCl2 as previously described (3). Polyacrylamide gel electrophoresis (PAGE) of membrane proteins and fluorography were carried out as described previously (45).

Nodulation and virulence test.

Alfalfa seeds were surface sterilized with concentrated sulfuric acid for 30 s and washed several times with sterile distilled water until total removal of the acid. Seeds were germinated on wet filter paper in petri dishes. Two-day-old seedlings were planted in autoclaved modified Leonard jars filled with vermiculite and Jensen’s N-free solution (39). Seedlings were dipped into a 2-day-old Rhizobium culture immediately before planting. After 4 weeks, plants were removed, nitrogen fixation was evaluated by the acetylene reduction assay as described previously (40), and Rhizobium strains were isolated from nodules as described previously (20). Virulence assays were carried out on Kalanchoe leaves as previously described (13).

Construction of a B. abortus β(1-2) glucan synthetase mutant.

Mutagenesis was carried out by gene replacement of the B. abortus cyclic β(1-2) glucan synthetase wild-type gene with a Tn3-HoHo1-mutated gene. Plasmid pI129 (Tcr Ampr) with a Tn3-HoHo1 insertion in the ClaI-EcoRI 1.5-kb DNA fragment (Fig. 1) was conjugated into B. abortus S19. The transconjugant strain, B. abortus(pI129), was selected on BB agar with nalidixic acid (5 μg/ml), carbenicillin (100 μg/ml), and tetracycline (2 μg/ml). Homogenotization of the mutated gene into the B. abortus chromosome was carried out by conjugation into B. abortus(pI129) of the incompatible plasmid pPH1JI (Gmr) as described previously (16). Single-crossover integration events were selected on BB agar with gentamicin (4 μg/ml) and carbenicillin (100 μg/ml). Double crossover events were selected by streaking colonies in duplicate on BB agar with carbenicillin (100 μg/ml), nalidixic acid (5 μg/ml), and tetracycline (2 μg/ml) and BB agar with carbenicillin (100 μg/ml) and nalidixic acid (5 μg/ml). Colonies that were carbenicillin and nalidixic acid resistant and tetracycline sensitive were selected as possible double recombinants.

FIG. 1.

FIG. 1

Open in a new tab

Genetic and physical maps of the DNA insert in recombinant cosmid pBA19. The top segment represents the DNA fragment cloned in pBA19. The arrow represents the cgs gene encoding cyclic β(1-2) glucan synthetase. The stalks with a circle on top represent Tn3-HoHo1 insertions; plus and minus symbols above them indicate restoration (+) or no restoration (−) of the motility phenotype to R. meliloti GRT21s. The DNA fragments subcloned in other recombinant plasmids as well as the DNA fragment cloned in pBA25 are indicated by segments below the pBA19 segment. Their ability (+) or lack of ability (−) to complement the Ndv phenotype of R. meliloti GRT21s is indicated. Restriction sites: H, HindIII; E, EcoRI; C, ClaI; B, BamHI.

Experimental infection of mice.

Two groups of five female 9-week-old BALB/c mice were injected intraperitoneally, one with 5.1 × 108 ± 4.2 × 108 CFU of B. abortus S19 and the other with 3.4 × 108 ± 1.0 × 108 CFU of the cgs mutant B. abortus BAI129. At 30 days postinfection, mice were bled from the retrorbital sinus and sera were stored at −20°C until used. Animals were killed by cervical dislocation, spleens were weighed and homogenized in 1 ml of 150 mM NaCl, and homogenates were serially diluted and plated in duplicate on BB agar. Colonies were counted after 4 days of incubation at 37°C.

KELA.

Antibodies against B. abortus lipopolysaccharide (LPS) were measured in an indirect, computer-assisted kinetics-based enzyme-linked assay (KELA) as described previously (6). The rate, expressed as a slope, was proportional to the amount of antibody in the sample and was determined from linear regression analysis of absorbance versus time; slope values (103) were given as titers of antibodies.

Nucleotide sequence accession number.

The sequence of the B. abortus cyclic β(1-2) glucan synthetase gene (cgs) has been assigned GenBank accession no. AF047823.

RESULTS

Identification of a B. abortus genetic determinant that restores the motility of R. meliloti ndvB and A. tumefaciens chvB mutants.

Mass conjugation of the B. abortus S19 gene bank into the R. meliloti GRT21s ndvB mutant led to the isolation of recombinant clones with restored motility on soft-agar plates. Representative motile clones were selected and conjugated back into E. coli HB101 by the triparental mating method (9) in order to isolate the corresponding complementing cosmids. Digestion with restriction enzyme HindIII revealed the presence of two different cosmids with overlapping DNA fragments, pBA19 with a 19-kb insert and pBA25 with a 25-kb insert (Fig. 1). Conjugation of cosmids pBA19 and pBA25 into the A. tumefaciens A1011 chvB mutant restored the motility of the mutant, indicating that both cosmids contain the appropriate genetic information to complement the nonmotile phenotype of both A. tumefaciens chvB and R. meliloti ndvB mutants.

B. abortus cosmids pBA19 and pBA25 complement nodule invasion and virulence of R. meliloti ndvB and A. tumefaciens chvB mutants.

Cosmids pBA19 and pBA25 restored the formation of normal nodules by the R. meliloti GRT21s ndvB mutant. As shown in Table 2, R. meliloti GRT21s(pBA19) and R. meliloti GRT21s(pBA25) induced the formation of nodules with wild-type levels of nitrogen fixation activity. When both cosmids were introduced by conjugation into the avirulent A. tumefaciens A1011 chvB mutant, the resulting recombinant strains, A. tumefaciens A1011(pBA19) and A. tumefaciens A1011(pBA25), recovered virulence on Kalanchoe leaves (Table 2). These results demonstrated that both pBA19 and pBA25 contain a gene(s) able to complement the functions assigned to the R. meliloti ndvB and A. tumefaciens chvB genes.

TABLE 2.

Complementation of Rhizobium and Agrobacterium mutants

Strain Motilitya Nitrogen fixationb Virulencec Glucan synthetased β(1-2) glucane
R. meliloti
 GR4 (wild type) + 24.6 ± 8.6  NDf + 12.1
 GRT21s (ndvB) − 7.2 ± 2.9 ND − <0.1
 GRT21s(pBA19) + 20.9 ± 6.5 ND + 16.5
 GRT21s(pBA25) + 18.4 ± 4.2 ND + 18.3
A. tumefaciens
 A348 (wild type) + ND + + 6.5
 A1011 (chvB) − ND − − <0.1
 A1011(pBA19) + ND + + 5.2
 A1011(pBA25) + ND + + 5.4
Open in a new tab
a

Motility on 0.4% agar medium was studied as described in Materials and Methods. 

b

Determined by the acetylene reduction assay as described previously (40) and expressed in nanomoles of ethylene hour−1 plant−1. 

c

Determined on Kalanchoe leaves as described previously (13). 

d

The presence of cyclic β(1-2) glucan synthetase was determined by PAGE as described previously (45). 

e

The accumulation of cellular β(1-2) glucan was determined as described in Materials and Methods and expressed in milligrams of glucose equivalents per gram of cells (wet weight). 

f

ND, not determined. 

Mapping and nucleotide sequencing of the ndvB- and chvB-complementing function on cosmid pBA19.

Cosmid pBA19 was selected for further studies. It was recovered from R. meliloti GRT21s(pBA19) by back conjugation into E. coli HB101 by the triparental mating method and was subjected to restriction enzyme analysis and subcloning. The restriction maps of the DNA insert present in pBA19 as well as some subclones are shown in Fig. 1. Subclones pBBE52 (5.2-kb EcoRI fragment), pBBE45 (4.5-kb EcoRI fragment), pBBE50 (5.0-kb EcoRI fragment), pVKH43 (4.3-kb HindIII fragment), and pBBC10 (10.6-kb ClaI fragment), harboring overlapping DNA fragments, were analyzed to study their ability to restore motility or the nodulation phenotype to the R. meliloti GRT21s ndvB mutant. As indicated in Fig. 1, none of these recombinant plasmids could restore motility or the nodulation phenotype to the R. meliloti GRT21s ndvB mutant. Therefore, none of these plasmids contained the complete gene required for complementation.

To map the region harboring all the information needed for complementation, plasmid pBA19 was subjected to Tn3-HoHo1 mutagenesis as described in Materials and Methods. A collection of Tn3-HoHo1 insertions was generated and analyzed. Figure 1 shows that two classes of Tn3-HoHo1 insertions were obtained: class 1 did not restore motility to the R. meliloti GRT21s mutant, and class 2 retained complementing ability. These results defined a maximum region of 8 kb in the DNA insert in pBA19 that is required for restitution of motility.

The nucleotide sequence of this 8-kb region was determined. An open reading frame (ORF) encompassing an 8.49-kb fragment flanked by ClaI and BamHI sites was found (Fig. 1). This ORF encoded a high-molecular-mass protein of 316.2 kDa. In the promoter region at −45 bp from the putative ATG start codon there was a conserved −10 consensus sequence for a sigma 70 RNA polymerase; however, no −35 consensus sequence was observed. At −12 bp from the putative ATG start codon a conserved Shine-Dalgarno sequence was observed. At 144 bp downstream of a putative TAA stop codon a possible transcriptional terminator stem-loop of 11 inverted repeat bases flanked by T-rich regions was present.

Analysis of the amino acid sequence of the cyclic β(1-2) glucan synthetase.

Analysis of the 2,831-amino-acid-residue cyclic β(1-2) glucan synthetase protein indicated that it has the features of a membrane protein. Comparison with protein databases showed that this sequence has an overall identity of 51% with the R. meliloti NdvB protein sequence. Tn3-HoHo1 insertions that did not restore motility and cyclic β(1-2) glucan synthesis mapped in the region encoding the N-terminal half of the protein. However, at the 3′ end of the ORF, Tn3-HoHo1 insertions that had no effect on motility and cyclic glucan synthesis were found (see insertions M47, M14, M202, M20, M30, and M432 in Fig. 1), suggesting that the C-terminal region of the protein may be dispensable for cyclic glucan synthesis.

Different degrees of similarity exist at different regions of the proteins. Four regions of the B. abortus protein spanning from amino acids 520 to 800, 1025 to 1124, 1284 to 1526, and 2400 to 2660 display similarities of greater than 80% with R. meliloti NdvB. There is a strong divergence between the amino acid sequences at the N terminus, where the similarity is less than 30%. Prediction of helical transmembrane regions (30) showed that cyclic β(1-2) glucan synthetase has six transmembrane regions from positions 424 to 442, 455 to 472, 818 to 835, 840 to 858, 946 to 962, and 967 to 984. These regions may determine three cytoplasmic domains that share the highest sequence similarity with the NdvB protein.

The C-terminal region of the Brucella and the Rhizobium proteins shows overall identities of 27 and 28% with the complete cellobiose and cellodextrin phosphorylases of Clostridium stercorarium, respectively. It is remarkable that this region of the protein is highly conserved, although it is not required for glucan synthesis. It remains to be established if the β(1-2) glucan synthetase has cellobiose and/or cellodextrin phosphorylase enzymatic activity.

Motility of and synthesis of cyclic β(1-2) glucan by the R. meliloti ndvB mutant complemented with cosmids with different Tn3-HoHo1 insertions.

Since it was previously found that R. meliloti GRT21s ndvB and A. tumefaciens A1011 chvB mutants do not form cyclic β(1-2) glucans and display a pleiotropic nonmotile phenotype (10, 14, 28), we decided to determine if the restoration of motility correlates with the restitution of cyclic glucan synthesis in the R. meliloti GRT21s mutant harboring cosmids with insertions affecting motility (nonmotile and motile). Table 2 shows that the introduction of cosmids pBA19 and pBA25 restored in both bacteria motility and the synthesis of β(1-2) glucans. In order to study the correlation between the restoration of both phenotypes, R. meliloti GRT21s strains harboring cosmids with different Tn3-HoHo1 insertions (Fig. 1) were subjected to glucan extraction and characterization by TLC. As shown in Fig. 2, insertions that did not restore motility were negative for the synthesis of glucans (see insertions I133, I129, I132, and I121 in Fig. 2), whereas insertions that did not affect motility were positive for the synthesis of glucans (see insertions M47, M14, and M21 in Fig. 2). These results indicate that both functions mapped to the same DNA region of cosmid pBA19 and that there was a strict correlation between the restoration of motility and the restitution of cyclic β(1-2) glucan synthesis.

FIG. 2.

FIG. 2

Open in a new tab

TLC of cyclic β(1-2) glucans formed by B. abortus and R. meliloti GRT21s harboring plasmid pBA19 with different Tn3-HoHo1 insertions. Total cellular glucans were extracted and subjected to TLC as described in Materials and Methods. Lane 1, wild-type R. meliloti GR4; lane 2, R. meliloti GRT21s ndvB mutant; lane 3, wild-type B. abortus S19; lane 4, B. abortus BAI129; lane 5, R. meliloti GRT21s(pI133); lane 6, R. meliloti GRT21s(pI129); lane 7, R. meliloti GRT21s(pI132); lane 8, R. meliloti GRT21s(pI121); lane 9, R. meliloti GRT21s(pM47); lane 10, R. meliloti GRT21s(pM14); lane 11, R. meliloti GRT21s(pM21); lane 12, R. meliloti GRT21s(pBA19); lane 13, R. meliloti GRT21s(pBA25). Asterisks next to lanes 1 and 3 show the migratory positions of the main components of R. meliloti and B. abortus cyclic β(1-2) glucans.

Figure 2 shows that the R. meliloti GRT21s ndvB mutant complemented with cosmids with insertions M47 and M14 (mapped by DNA sequencing at positions 6019 and 6519 of the cgs gene, respectively) (lanes 9 and 10) produced larger amounts of glucans than did mutants complemented with wild-type cosmids pBA19 and pBA23 (lanes 12 and 13) or with a cosmid with an insertion downstream of the coding region of cgs (insertion M21, lane 11). Moreover, it seems that cosmids with insertions M47 and M14 produced glucans with TLC behavior intermediate between those observed for R. meliloti and B. abortus glucans (Fig. 2, lanes 1 and 3). These results were consistently observed in different experiments, and it remains to be established whether the C-terminal 825 amino acid residues of cyclic β(1-2) glucan synthetase have any role in regulating the synthesis and the type of glucans formed.

Characterization of cyclic β(1-2) glucans produced by ndvB mutants complemented with cosmid pBA19.

Membranes from B. abortus S19, wild-type R. meliloti GR4, and the R. meliloti GRT21s ndvB mutant complemented with cosmid pBA19, strain GRT21s(pBA19), were used as enzyme sources for the synthesis in vitro of cyclic β(1-2) glucans as described in Materials and Methods. The products were characterized as cyclic β(1-2) glucans by total and partial acid hydrolysis, reduction with borohydride, and TLC (data not shown). The apparent degree of polymerization of glucans formed in vitro was determined by chromatography on Bio-Gel P4 columns. R. meliloti GRT21s(pBA19) synthesized in vitro a cyclic β(1-2) glucan with a degree of polymerization identical to that of the glucan formed by B. abortus S19 membranes and slightly smaller than that of the glucan synthesized by wild-type R. meliloti GR4 membranes, indicating that complementation was achieved in a heterologous background. Experiments carried out with the A. tumefaciens A1011 chvB mutant complemented with plasmid pBA19 resulted in the formation of a glucan with the same degree of polymerization.

Identification by PAGE of the B. abortus cyclic β(1-2) glucan synthetase.

It was previously found that in the Rhizobiaceae, cyclic β(1-2) glucan synthetase is an inner membrane protein with apparent molecular masses of approximately 235 kDa as determined by PAGE (45) and of 319 kDa as determined by nucleotide sequence analysis (19). The enzyme can be pulse-labeled with UDP-[14C]glucose and chased with nonradioactive UDP-glucose, due to the fact that the protein itself is an intermediate during the synthesis of cyclic β(1-2) glucan (45). It was also found that in B. abortus, cyclic β(1-2) glucan synthetase has an apparent molecular mass similar to that of the A. tumefaciens enzyme (3). When Brucella membranes are used as an enzyme source, the cyclization reaction is defective, and no chase effect is observed upon the addition of nonlabeled UDP-glucose (3). The property of being labeled after a pulse incubation with UDP-[14C]glucose allows the rapid identification of the protein by PAGE after fluorography.

Membranes prepared from various strains were incubated with UDP-[14C]glucose and subjected to PAGE as described in Materials and Methods. As shown in Fig. 3, lanes 6 to 9, strains GRT21s(pBA19) and GRT21s(pBA25) contained a 14C-labeled membrane protein with an apparent molecular weight indistinguishable from that of the protein present in B. abortus membranes (lanes 1 to 3). Membranes contained, besides the high-molecular-weight proteins, at least three other labeled proteins having lower molecular weights; these probably represented products of proteolytic degradation, since their relative amounts changed in different membrane preparations (see Fig. 4). The protein produced by R. meliloti GR4, the wild-type parental strain of mutant GRT21s (Fig. 3, lane 4), had an apparent molecular weight higher than that of the B. abortus protein (lane 1). On the other hand, the R. meliloti 102F34 protein (Fig. 4, lanes 3) had a molecular weight similar to that of the B. abortus protein (lanes 1).

FIG. 3.

FIG. 3

Open in a new tab

PAGE of membrane proteins of B. abortus S19, R. meliloti GRT21s, and R. meliloti GRT21s complemented with plasmids pBA19 and pBA25. Total or inner membranes (0.2 mg of protein) were incubated with UDP-[14C]glucose. The reaction was stopped by the addition of 10% TCA, and the precipitates were subjected to gel electrophoresis as described in Materials and Methods. Radioactivity was detected by fluorography. For the chase experiments in lanes 2, 5, 7, and 9, 2 mM nonradioactive UDP-glucose was added after 15 min of incubation, and the mixture was further incubated for 20 min. In lane 3, 20 mM nonradioactive UDP-glucose was added for the chase experiment. Lanes 1, 2, and 3, total membranes of B. abortus S19; lanes 4 and 5, inner membranes of R. meliloti GR4; lanes 6 and 7, inner membranes of R. meliloti GRT21s(pBA19); lanes 8 and 9, inner membranes R. meliloti GRT21s(pBA25). Numbers on the left indicate molecular masses (in kilodaltons) of standards.

FIG. 4.

FIG. 4

Open in a new tab

PAGE of membrane proteins of B. abortus S19, B. abortus BAI129 cgs mutant, R. meliloti 102F34, R. meliloti GRT21s ndvB mutant, and R. meliloti GRT21s complemented with cosmids pBA19 and pM14 (pBA19 with Tn3-HoHo1 insertion M14). Membranes or permeabilized cells (0.2 mg of protein) were incubated with UDP-[14C]glucose as described in Materials and Methods. The reaction was stopped by the addition of 10% TCA, and the precipitates were subjected to gel electrophoresis. Proteins were stained with Coomassie blue (A), and radioactivity was detected by fluorography (B). Lane 1, B. abortus S19; lane 2, B. abortus BAI129 cgs mutant; lane 3, R. meliloti 102F34; lane 4, R. meliloti GRT21s ndvB mutant; lane 5, R. meliloti GRT21s(pBA19); lane 6, R. meliloti GRT21s(pM14). Numbers on the left indicate molecular masses (in kilodaltons) of standards.

Neither the radioactivity that accumulated on the B. abortus S19 protein nor that present on the proteins of R. meliloti GRT21s(pBA19) and R. meliloti GRT21s(pBA25) was chased after the addition of nonradioactive UDP-glucose (Fig. 3, lanes 2, 3, 7, and 9). During the chase, an increase in the apparent molecular mass became apparent and was more visible in the proteolytic fragments. This behavior was previously described for B. abortus (3) and was due to the accumulation of linear high-molecular-weight oligosaccharides on the protein (3). Figure 3, lane 5, shows that the radioactivity that accumulated on the wild-type R. meliloti GR4 protein was chased upon the addition of nonradioactive UDP-glucose; this result distinguished the behaviors of the Rhizobium and Brucella cyclic β(1-2) glucan synthetases. Figure 4, lane 2, shows that in the B. abortus BAI129 cgs mutant, the high-molecular-weight protein was not present and that no other protein became labeled upon incubation with UDP-[14C]glucose. The R. meliloti GRT21s ndvB mutant (Fig. 4A, lane 4) complemented with cosmid pBA19 (lane 5) overproduced a protein with the same apparent molecular weight as the wild-type B. abortus protein (lane 1). On the other hand, this Rhizobium mutant complemented with cosmid pM14 (Fig. 1) produced a shorter protein (Fig. 4A, lane 6) that became labeled after incubation with UDP-[14C]glucose (Fig. 4B, lane 6) and that was active in the synthesis of cyclic glucan (Fig. 2, lane 10). These results indicated that there is a region at the C terminus of cyclic β(1-2) glucan synthetase not required for the synthesis of cyclic glucan. These results demonstrated that restitution of the wild-type phenotype of the R. meliloti ndvB mutant was accomplished by the expression in the Rhizobium background of the B. abortus cyclic β(1-2) glucan synthetase protein.

The same results were obtained when plasmid pBA19 was used to complement the A. tumefaciens A1011 chvB mutant. Figure 5A, lanes 1 and 2, shows that membranes of wild-type A. tumefaciens A348 contained a high-molecular-weight protein which could be pulse-labeled with UDP-[14C]glucose and chased with 2 mM nonradioactive UDP-glucose (Fig. 5B, lanes 1 and 2). A1011 chvB mutant membranes lacked this protein (Fig. 5A, lanes 3 and 4), and no other protein became labeled upon incubation with UDP-[14C]glucose (Fig. 5B, lanes 3 and 4). The introduction of cosmid pBA19 in mutant strain A1011(pBA19) restored the presence of a high-molecular-weight membrane protein with an apparent molecular weight indistinguishable from that of the protein present in the wild-type A. tumefaciens strain (Fig. 5A, lanes 5 and 6). The protein became labeled upon incubation with UDP-[14C]glucose; however, no chase was observed after the addition of 2 mM nonradioactive UDP-glucose (Fig. 5B, lanes 5 and 6). The high-molecular-weight protein was overexpressed in strain A1011(pBA19) (Fig. 5A, lanes 5 and 6). Membranes of strain A1011(pBA19) contained, besides the high-molecular-weight protein, several labeled proteins having lower molecular weights (Fig. 5B, lanes 5 and 6). They probably represented products of proteolytic degradation. Figure 5B, lanes 7 and 8, shows that, as described for B. abortus (3), when permeabilized cells were used as an enzyme source, a normal chase effect was observed after the addition of nonradioactive UDP-glucose, and no proteolytic degradation products were observed. We concluded that pBA19 codes for the functional B. abortus cyclic β(1-2) glucan synthetase gene (cgs), which is functional in A. tumefaciens and R. meliloti.

FIG. 5.

FIG. 5

Open in a new tab

PAGE of membrane proteins of wild-type A. tumefaciens A348, A. tumefaciens A1011 chvB mutant, and A. tumefaciens A1011 complemented with plasmid pBA19. Inner membranes (lanes 1 to 6) or permeabilized cells (lanes 7 and 8) (0.2 mg of protein) were incubated with UDP-[14C]glucose as described in Materials and Methods. The reaction was stopped by the addition of 10% TCA, and the precipitates were subjected to gel electrophoresis. Proteins were stained with Coomassie blue (A), and radioactivity was detected by fluorography (B). For the chase experiment (lanes 2, 4, 6, and 8), nonradioactive UDP-glucose (2 mM) was added after 10 min of incubation, and the mixture was further incubated for 20 min. Lanes 1 and 2, A. tumefaciens wild-type A348; lanes 3 and 4, A. tumefaciens A1011 chvB mutant; lanes 5, 6, 7, and 8, A. tumefaciens A1011(pBA19). Numbers on the left indicate molecular masses (in kilodaltons) of standards.

Site-directed mutagenesis of the B. abortus cgs gene.

Further confirmation that the B. abortus cgs gene is responsible for the synthesis of cyclic β(1-2) glucan was obtained by constructing a mutant by Tn3-HoHo1 mutagenesis. Cosmid pBA19 harboring Tn3-HoHo1 insertion I129 (Fig. 1) was conjugated into wild-type B. abortus S19, and double recombination events and gene replacements were obtained as described in Materials and Methods. Tn3-HoHo1 insertion I129 in pBA19 was located, by DNA sequencing as described in Materials and Methods, 3,055 bp downstream of the putative ATG start codon of the cgs ORF. Mutant BAI129 was obtained, and the formation of cellular cyclic β(1-2) glucan and the presence of the high-molecular-weight membrane protein were studied. As shown in Fig. 2, lane 4, mutant BAI129 did not accumulate cellular glucan. PAGE of B. abortus BAI129 revealed that the high-molecular-weight protein was absent (Fig. 4A, lane 2); moreover, upon incubation with UDP-[14C]glucose, no labeled protein was detected (Fig. 4B, lane 2). Thus, the gene identified in pBA19 codes for the B. abortus cyclic β(1-2) glucan synthetase. Cosmid pBA19 harboring Tn3-HoHo1 insertion M47, located, by DNA sequencing as described in Materials and Methods, 6,019 bp downstream of the putative ATG start codon of the cgs ORF, restored the synthesis of cyclic β(1-2) glucan to wild-type levels in the B. abortus BAI129 cgs mutant (3.38 mg of glucose equivalents g of cell pellet [wet weight]−1). This result indicated that a truncated cyclic β(1-2) glucan synthetase protein is active in cyclic β(1-2) glucan synthesis in the Brucella background.

Persistence of the B. abortus S19 cgs mutant in mice.

In order to assess the possible role of cyclic β(1-2) glucan in the infectivity of B. abortus S19, experimental infections of mice were carried out as described in Materials and Methods. One group of BALB/c mice was injected with B. abortus S19 (5.1 × 108 ± 4.2 × 108 CFU), and the other was injected with the B. abortus BAI129 cgs mutant (3.4 × 108 ± 1.0 × 108 CFU). Titers of antibodies against B. abortus LPS, weights of spleens, and the persistence of live bacteria in spleens were estimated 30 days postinfection. From spleens of mice inoculated with B. abortus S19, 9.6 × 105 ± 2.7 × 105 CFU per spleen was recovered; on the other hand, from spleens of animals inoculated with the B. abortus BAI129 cgs mutant, only 1.5 × 103 ± 0.1 × 103 CFU was recovered. These results indicated that the level of persistence of the cgs mutant in the inoculated animals was approximately 3 orders of magnitude lower than that of parental strain B. abortus S19. Moreover, the spleen weights were 417 ± 20 mg with B. abortus S19, 191 ± 16 mg with the cgs mutant, and 83 ± 6 mg for the control, noninfected animals. LPS antibody titers, estimated by KELA, were 65 with B. abortus S19 and 37 with the cgs mutant. These results indicated that the cgs mutant induced an immunological response against LPS that was still good although 56.9% lower than that elicited by B. abortus S19. These data suggested that the cgs mutation reduced the virulence of B. abortus S19 in mice.

DISCUSSION

Two cosmids (pBA19 and pBA25) containing overlapping DNA inserts that restored the wild-type phenotype of nodulation to the R. meliloti GRT21s ndvB mutant were isolated from a gene bank of B. abortus S19. The complemented strains recovered both the synthesis of cyclic β(1-2) glucans and motility. Both traits were strictly associated with the presence of a high-molecular-weight inner membrane protein, similar to that encoded by the A. tumefaciens chvB and R. meliloti ndvB genes and identified as the cyclic β(1-2) glucan synthetase (19, 45). The properties of the cyclic β(1-2) glucan synthetase and the degree of polymerization of the cyclic β(1-2) glucans synthesized by the transconjugant strains are similar to those of the B. abortus S19 cyclic β(1-2) glucan synthetase described previously (3). We conclude that plasmids pBA19 and pBA25 contain a gene (cgs) which encodes the Brucella cyclic β(1-2) glucan synthetase, which is functional in R. meliloti and A. tumefaciens backgrounds, and which is able to complement ndvB and chvB mutants.

It is interesting that although Brucella is a nonmotile bacterium, due to a lack of flagella, restoration of the synthesis of cyclic β(1-2) glucan, nodule invasion, and virulence in R. meliloti and A. tumefaciens were associated with the recovery of motility. The pleotropic nonmotility effect observed for ndvB and chvB mutants has been assigned to a defect in flagellum assembly (10). Random Tn3-HoHo1 mutagenesis of plasmid pBA19 and screening for loss of the ability to restore motility in R. meliloti ndvB mutants led to the identification of an 8-kb DNA fragment containing the complementing function. However, the region defined by Tn3-HoHo1 mutagenesis was smaller than that expected for the codification of the complete cyclic β(1-2) glucan synthetase membrane protein, according to sequence data. This apparent paradox was also observed for A. tumefaciens chvB mutants, where it was observed that a truncated protein of 150 kDa was active in the synthesis of cyclic β(1-2) glucan; consequently, the virulence locus was smaller than the complete protein gene (44). Inner membranes prepared from R. meliloti GRT21s complemented with a plasmid containing Tn3-HoHo1 insertion M14 (position 6519 of the cgs gene) led to the synthesis of a shorter protein that was active in the synthesis of cyclic β(1-2) glucan and had restored motility. These results demonstrated that in Brucella, as in Agrobacterium, there is a region at the carboxy terminus of the protein that is not required for the synthesis of the cyclic glucan. The function of this highly conserved region of the protein remains to be determined. It is interesting, however, that this region shows high similarity to cellobiose and cellodextrin phosphorylases from C. stercorarium (29).

Cyclic β(1-2) glucans are involved in osmoregulation and play an important role in symbiosis and tumorigenesis. The finding that B. abortus cgs was able to complement R. meliloti ndvB mutants (defective in nodule invasion) and A. tumefaciens chvB mutants (defective in tumor induction) led to speculation about the role of cgs in Brucella-cell interactions. The attenuated strain B. abortus S19 is widely used as a live vaccine, although it conserves a low degree of virulence (36). B. abortus S19 replicates in the spleens of mice during the first 2 weeks postinfection and persists for 12 weeks (36). Cyclic β(1-2) glucan is produced by B. abortus and other species of Brucella; however, it does not induce in their hosts the formation of antibodies (5). To our knowledge, no mutants affected in the synthesis of cyclic β(1-2) glucan had been obtained so far for B. abortus or any other species of Brucella. This is the first report in which such a mutant was obtained and its virulence in mice was studied. We showed that the B. abortus BAI129 cgs mutant displayed reduced virulence in mice, according to the criteria of spleen weight and number of live bacteria recovered from the spleens 4 weeks postinfection. These results suggested that in B. abortus, as in A. tumefaciens and R. meliloti, the lack of cyclic β(1-2) glucan affects bacterium-host interactions. Therefore, it seems likely that cgs in B. abortus is a virulence gene, like chvB in A. tumefaciens. The fact that cyclic β(1-2) glucan does not induce the formation of antibodies may make it a very efficient factor for the survival of Brucella spp. inside the host. The effect of the cgs mutation in fully pathogenic strains of B. abortus and other species of Brucella remains to be studied.

ACKNOWLEDGMENTS

Nora Iñón de Iannino and Gabriel Briones contributed equally to this work.

We thank J. J. Cazzulo for critical reading of the manuscript, E. W. Nester for providing A. tumefaciens strains and plasmids for Tn3-HoHo1 mutagenesis, M. E. Kovach for providing the pBBR1 MCS2 plasmid, A. Vigliocco for providing Brucella strains and useful suggestions, Fernando Pieckenstain for helping in acetylene reduction assays, and Susana Raffo and María de los Angeles Curtó (from Consejo Nacional de Investigaciones Científicas y Técnicas de la República Argentina [CONICET]) for preparing UDP-[14C]glucose.

This work was supported in part by a grant from the Ministerio de Cultura y Educación, República Argentina, to the Instituto de Investigaciones Biotecnológicas de la Universidad Nacional de General San Martín. We acknowledge the financial support of the Comisión Nacional de Energía Atómica, República Argentina, and of the Universidad Nacional de General San Martín. N.I. and R.A.U. are members of the Research Career of CONICET.

REFERENCES

  • 1.Altabe S, Iñón de Iannino N, de Mendoza D, Ugalde R A. Expression of the Agrobacterium tumefaciens chvB virulence region in Azospirillum spp. J Bacteriol. 1990;172:2563–2567. doi: 10.1128/jb.172.5.2563-2567.1990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Be A, Amemura A, Higashi S. Studies on cyclic β(1-2) glucan obtained from periplasmic space of Rhizobium trifolii cells. Plant Soil. 1982;64:315–324. [Google Scholar]
  • 3.Briones G, Iñón de Iannino N, Steinberg M, Ugalde R A. Periplasmic cyclic 1,2-β-glucan in Brucella spp. is not osmoregulated. Microbiology. 1997;143:1115–1124. doi: 10.1099/00221287-143-4-1115. [DOI] [PubMed] [Google Scholar]
  • 4.Castro O A, Zorreguieta A, Ielmini V, Vega G, Ielpi L. Cyclic β-(1-2)-glucan synthesis in Rhizobiaceae: role 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]
  • 5.Cherwonogrodzky J W, Dubray G, Moreno E, Mayer H. Antigens of Brucella. In: Nielsen K, Duncan J R, editors. Animal brucellosis. Boca Raton, Fla: CRC Press, Inc.; 1990. pp. 19–64. [Google Scholar]
  • 6.Comerci, D. J., G. D. Pollevick, A. M. Vigliocco, A. C. C. Frasch, and R. A. Ugalde. Vector development for the expression of foreign proteins in the vaccine strain Brucella abortus S19. Infect. Immun. 66:3862–3866. [DOI] [PMC free article] [PubMed]
  • 7.De Lay J, Mannheim W, Segers P, Lievens A, Denijn M, Vanhoucke M, Gillis M. Ribosomal ribonucleic acid cistron similarities and taxonomic neighborhood of Brucella and CDC group Vd. Int J Syst Bacteriol. 1987;37:35–42. [Google Scholar]
  • 8.Dische Z. General color reactions. Methods Carbohydr Chem. 1962;1:478–492. [Google Scholar]
  • 9.Ditta G, Stanfield S, Corbin C, Helinski D R. Broad host range DNA cloning system for gram-negative bacteria: construction of a gene bank of Rhizobium meliloti. Proc Natl Acad Sci USA. 1980;77:7347–7351. doi: 10.1073/pnas.77.12.7347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Douglas C J, Halperin W, Nester E W. Agrobacterium tumefaciens mutants affected in attachment to plant cells. J Bacteriol. 1982;152:1265–1275. doi: 10.1128/jb.152.3.1265-1275.1982. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Douglas C J, Staneloni R J, Rubin R A, Nester E W. Identification and genetic analysis of an Agrobacterium tumefaciens chromosomal virulence region. J Bacteriol. 1985;161:850–860. doi: 10.1128/jb.161.3.850-860.1985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Dylan T, Ielpi L, Stanfield S, Kashyap L, Douglas C, Yanofsky M, Nester E W, Helinski D R, Ditta G. Rhizobium meliloti genes required for nodule development are related to chromosomal virulence genes in Agrobacterium tumefaciens. Proc Natl Acad Sci USA. 1986;83:4403–4407. doi: 10.1073/pnas.83.12.4403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Garfinkel D J, Nester E W. Agrobacterium tumefaciens mutants affected in crown gall tumorigenesis and octopine catabolism. J Bacteriol. 1980;144:732–743. doi: 10.1128/jb.144.2.732-743.1980. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Geremia R A, Cavaignac S, Zorreguieta A, Toro N, Olivares J, Ugalde R A. A Rhizobium meliloti mutant that forms ineffective pseudonodules in alfalfa produces exopolysaccharide but fails to form β(1-2)glucan. J Bacteriol. 1987;169:880–884. doi: 10.1128/jb.169.2.880-884.1987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Higgins D G, Sharp P M. Fast and sensitive multiple sequence alignments on a microcomputer. CABIOS. 1989;5:151–153. doi: 10.1093/bioinformatics/5.2.151. [DOI] [PubMed] [Google Scholar]
  • 16.Hirsch P, Beringer J. A physical map of pPH1JI and pJB4JI. Plasmid. 1984;12:139–141. doi: 10.1016/0147-619x(84)90059-3. [DOI] [PubMed] [Google Scholar]
  • 17.Hisamatsu M, Amemura A, Koizumi K, Utamura T, Okada Y. Structural studies on cyclic(1-2)-beta-d-glucans (cyclosophoraose) produced by Agrobacterium and Rhizobium. Carbohydr Res. 1983;121:31–40. [Google Scholar]
  • 18.Hull R A, Gill R E, Hsu O, Minshew B, Falkow S. Construction and expression of recombinant plasmids encoding type I or d-mannose-resistant pili from a urinary tract infection Escherichia coli isolate. Infect Immun. 1981;33:933–938. doi: 10.1128/iai.33.3.933-938.1981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.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]
  • 20.Iñón de Iannino N, Briones G, Kreiman G, Ugalde R A. Characterization of the biosynthesis of β(1-2) cyclic glucan in R. fredii. β(1-2) Glucan has no apparent role in nodule invasion of McCall and Peking soybean cultivars. Cell Mol Biol. 1996;42:617–629. [PubMed] [Google Scholar]
  • 21.Iñón 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]
  • 22.Iñón de Iannino N, Ugalde R A. Biosynthesis of cyclic β(1-3),β(1-6) glucan in Bradyrhizobium spp. Arch Microbiol. 1993;159:30–38. doi: 10.1007/BF00244260. [DOI] [PubMed] [Google Scholar]
  • 23.Knauf V C, Nester E W. Wide host range cloning vectors: a cosmid clone bank of an Agrobacterium Ti plasmid. Plasmid. 1983;8:45–54. doi: 10.1016/0147-619x(82)90040-3. [DOI] [PubMed] [Google Scholar]
  • 24.Kovach M E, Elzer P H, Hill D S, Robertson G T, Farris M A, Roop II R M, Peterson K M. Four new derivatives of the broad-range cloning vector pBBR1MCS, carrying different antibiotic-resistance cassettes. Gene. 1995;166:175–176. doi: 10.1016/0378-1119(95)00584-1. [DOI] [PubMed] [Google Scholar]
  • 25.Long S, Ditta G S, Helinski D R. Heme biosynthesis in Rhizobium. Identification of a cloned gene coding δ-aminolevulinic acid synthetase from Rhizobium meliloti. J Biol Chem. 1982;257:8724–8730. [PubMed] [Google Scholar]
  • 26.Miller K J, Kennedy E P, Reinhold V N. Osmotic adaptation by Gram-negative bacteria: possible role for periplasmic oligosaccharides. Science. 1986;231:48–51. doi: 10.1126/science.3941890. [DOI] [PubMed] [Google Scholar]
  • 27.Osborn M, Munson R. Separation of inner (cytoplasmic) and outer membranes of Gram-negative bacteria. Methods Enzymol. 1984;31A:642–653. doi: 10.1016/0076-6879(74)31070-1. [DOI] [PubMed] [Google Scholar]
  • 28.Puvanesarajah V, Schell F M, Stacey G, Douglas C J, Nester E W. Role for 2-linked β-d-glucan in the virulence of Agrobacterium tumefaciens. J Bacteriol. 1985;164:102–106. doi: 10.1128/jb.164.1.102-106.1985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Reichenbecher M, Lottspeich F, Bronnenmeier K. Purification and properties of a cellobiose phosphorylase (CepA) and a cellodextrin phosphorylase (CepB) from the cellulolytic thermophile Clostridium stercorarium. Eur J Biochem. 1997;247:262–267. doi: 10.1111/j.1432-1033.1997.00262.x. [DOI] [PubMed] [Google Scholar]
  • 30.Rost B, Fariselli P, Casadio R. Topology prediction for helical transmembrane proteins at 86% accuracy. Protein Sci. 1996;5:1704–1718. doi: 10.1002/pro.5560050824. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.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]
  • 32.Sangari F J, Garcia-Lobo J M, Agüero J. The Brucella abortus vaccine strain B19 carries a deletion in the erythritol catabolic genes. FEMS Microbiol Lett. 1994;121:337–342. doi: 10.1111/j.1574-6968.1994.tb07123.x. [DOI] [PubMed] [Google Scholar]
  • 33.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]
  • 34.Smith L D, Ficht T A. Pathogenesis of Brucella. Crit Rev Microbiol. 1990;17:209–229. doi: 10.3109/10408419009105726. [DOI] [PubMed] [Google Scholar]
  • 35.Stachel S E, An G, Flores C, Nester E W. A Tn3 lacZ transposon for the random generation of β-galactosidase gene fusions: application to the analysis of gene expression in Agrobacterium. EMBO J. 1985;4:891–898. doi: 10.1002/j.1460-2075.1985.tb03715.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Stevens M G, Olsen S C, Pugh G W, Jr, Brees D. Comparison of the immune responses and resistance to brucellosis in mice vaccinated with Brucella abortus S19 or RB51. Infect Immun. 1995;63:264–270. doi: 10.1128/iai.63.1.264-270.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Tolmasky M E, Gammie A E, Crosa J H. Characterization of the rec A gene of Vibrio anguillarum. Gene. 1992;110:41–48. doi: 10.1016/0378-1119(92)90442-r. [DOI] [PubMed] [Google Scholar]
  • 38.Toro N, Olivares J. Analysis of Rhizobium meliloti Sym mutants obtained by heat treatment. Appl Environ Microbiol. 1986;51:1148–1150. doi: 10.1128/aem.51.5.1148-1150.1986. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Vincent J M. A manual for practical study of root-nodule bacteria. Oxford, England: Blackwell Scientific Publications Ltd.; 1970. pp. 90–121. [Google Scholar]
  • 40.Wacek T J, Brill W J. Simple rapid assay for screening nitrogen fixing ability in soybean. Crop Sci. 1976;16:519–522. [Google Scholar]
  • 41.York W S, McNeil M, Darvill A G, Albersheim P. Beta-2-linked glucans secreted by fast-growing species of Rhizobium. J Bacteriol. 1978;142:243–248. doi: 10.1128/jb.142.1.243-248.1980. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Zevenhuizen L P T M, van Neerven A R W. (1-2)-β-d-Glucan and acyclic oligosaccharides produced by Rhizobium meliloti. Carbohydr Res. 1983;118:127–134. [Google Scholar]
  • 43.Zevenhuizen L P T M, van Veldhuizen A, Fokkens R H. Re-examination of cellular cyclic β1-2-glucans of Rhizobiaceae: distribution of ring sizes and degrees of glycerol-1-phosphate substitution. Antonie Leeuwenhoek. 1990;57:173–178. doi: 10.1007/BF00403952. [DOI] [PubMed] [Google Scholar]
  • 44.Zorreguieta A, Geremia R A, Cavaignac S, Cangelosi G A, Nester E W, Ugalde R A. Identification of the product of an Agrobacterium tumefaciens chromosomal virulence gene. Mol Plant Microbe Interact. 1988;3:121–127. doi: 10.1094/mpmi-1-121. [DOI] [PubMed] [Google Scholar]
  • 45.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