Identification and Characterization of the Brucella abortus Phosphoglucomutase Gene: Role of Lipopolysaccharide in Virulence and Intracellular Multiplication

INFECTION AND IMMUNITY,0019-9567/00/$04.0010

Oct. 2000, p. 5716–5723 Vol. 68, No. 10

Copyright © 2000, American Society for Microbiology. All Rights Reserved.

Identification and Characterization of the Brucella abortusPhosphoglucomutase Gene: Role of Lipopolysaccharide

in Virulence and Intracellular MultiplicationJUAN E. UGALDE,1 CECILIA CZIBENER,1,2 MARIO F. FELDMAN,1 AND RODOLFO A. UGALDE1*

Instituto de Investigaciones Biotecnologicas-Instituto Tecnologico de Chascomus, Universidad Nacional de GeneralSan Martın, Buenos Aires,1 and Centro de Virologıa Animal, CEVAN, Capital Federal,2 Argentina

Received 25 February 2000/Returned for modification 12 May 2000/Accepted 4 July 2000

Smooth lipopolysaccharide (LPS) of Brucella abortus has been reported to be an important virulence factor,although its precise role in pathogenesis is not yet clear. While the protective properties of LPS againstcomplement are well accepted, there is still some controversy about the capacity of rough mutants to replicateintracellularly. The B. abortus phosphoglucomutase gene (pgm) was cloned, sequenced, and disrupted. The genehas a high index of identity to Agrobacterium tumefaciens pgm but is not part of the glycogen operon. A B. abortusnull mutant lacks LPS O antigen but has an LPS core with an electrophoretic profile undistinguishable fromthat of the wild-type core, suggesting that glucose, galactose, or a derivative of these sugars may be part of thelinkage between the core and the O antigen. This mutant is unable to survive in mice but replicates in HeLacells, indicating that the complete LPS is not essential either for invasion or for intracellular multiplication.This behavior suggests that the LPS may play a role in extracellular survival in the animal, probably protectingthe cell against complement-mediated lysis, but is not involved in intracellular survival.

Brucella spp. are gram-negative, facultative intracellular bac-teria that cause a chronic zoonotic disease worldwide. Six spe-cies of Brucella with different host specificies and pathogeneseshave been described (11, 35). Brucella abortus is the etiologicalagent of bovine brucellosis, but it can also affect humans,causing undulant fever; this disease is caused also by Brucellamelitensis, Brucella suis, and Brucella canis. Brucellae prolifer-ate within host macrophages, and virulence is associated withthe ability to multiply intracellularly. Once inside the cells,Brucella avoids the fusion of the phagosome with the lysosomeby altering the intracellular traffic of the early phagosomevesicle. It has recently been demonstrated that brucellae rep-licate in a vesicle compartment containing reticuloendoplasmicmarkers reached after preventing the fusion between phago-somes and lysosomes (19–21).

As in many other gram-negative bacteria, lipopolysaccharide(LPS) is an important component of the outer membrane. LPShas three domains: lipid A, the core oligosaccharide, and the Oantigen or O side chain. The complete structure of BrucellaLPS has not yet been elucidated, but it was reported that lipidA is composed of glucosamine, n-tetradecanoic acid, n-hexade-canoic acid, 3-hydroxytetradecanoic acid, and 3-hydroxyhexa-decanoic acid (5). The O side chain is a linear homopolymer ofa-1,2-linked 4,6-dideoxy-4-formamido-a-D-mannopyranosyl(perosamine) subunits, usually with a degree of polymerizationaveraging between 96 to 100 subunits (5). The complete struc-ture of B. abortus core LPS has not yet been determined.Previous reports have shown that is formed by 2-keto-3-de-oxy-D-manno-2-octulosonic acid, glucosamine, and glucose, al-though the exact amounts of these sugars have not been de-termined (17).

The absence of the O side chain from LPS determines therough phenotype. Generally these mutants are less virulent

than the wild type, with the exception of those of Brucella ovisand B. canis, which are rough but virulent (24). It is acceptedthat rough mutants are more sensitive to lysis mediated bycomplement, and probably this is the main reason why mostrough variants have an avirulent phenotype in animal models.To date the question about the capacity of rough mutants toreplicate intracellularly is not solved. Some authors have re-ported that smooth LPS is essential for intracellular survival(22, 23), for example, the vaccine strain RB51 exhibits loss ofvirulence and cannot replicate within macrophages (27). Onthe other hand, there are some reports in which geneticallycharacterized rough mutants did not loose the capacity toreplicate intracellularly despite the total absence of the Oantigen (1). In a recent search for rough mutants of B. meliten-sis, the gene coding for the perosamine synthetase was isolated.A mutant with a mutation in this gene has a rough phenotypeand is unable to survive in mice but can replicate in bovinemacrophages (9).

One possible explanation for these discrepancies may bethat many of the experiments carried out to understand therole of the O antigen were performed using mutants fortu-itously isolated by screening for the rough phenotype. As aresult of this procedure, the isolated mutants lack genetic def-inition, and in consequence a relation between the rough phe-notype and defective intracellular replication has not beendirectly confirmed. It is interesting that the only two roughmutants with a genetic characterization are able to replicate inmacrophages (1, 9). Increasing knowledge on the genetic lociinvolved in LPS biosynthesis will allow studies of the role ofLPS in intracellular survival. With this information available,the idea that rough phenotypes are always associated withdeficient intracellular replication may no longer be the rule.These studies must be done by altering one gene at a time andanalyzing the generated phenotype.

In Agrobacterium tumefaciens, a member of the alpha sub-group of the proteobacteria closely related to Brucella spp., thegene that codes for phosphoglucomutase was extensively stud-ied at the biochemical and molecular levels by our group (29,

* Corresponding author. Mailing address: IIB-UNSAM, Av GeneralPaz entre Constituyentes y Albarellos, P.O. Box 30 (1650) San Martın,Provincia de Buenos Aires, Argentina. Phone: 54-11-4580-7255. Fax:54-11-4752-9639. E-mail: [email protected].

5716

31–33). We found that this gene is absolutely necessary for thebiosynthesis of ADP-glucose, UDP-glucose, and UDP-galac-tose, the donors of glucose or galactose for the biosynthesisof molecules containing these sugars. An A. tumefaciens pgmmutant is avirulent and cannot synthesize exopolysaccharide,b(1,2) cyclic glucan, glycogen, and LPS (4, 32). In view of theseresults and the close relationship between agrobacteria andbrucellae, we studied the effect of pgm mutation on the viru-lence and intracellular multiplication of B. abortus.

We describe in this report the cloning, nucleotide sequence,and insertional mutagenesis of a gene (pgm) encoding thephosphoglucomutase of B. abortus. The characterization of theLPS, the virulence of the mutant in the mouse model, and theintracellular multiplication of the mutant were analyzed.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions. The bacterial strains andplasmids used in this work are listed in Table 1. Escherichia coli was grown at37°C in Luria-Bertani broth (25) or Terrific broth (28). Brucella strains weregrown at 37°C in tryptic soy broth (TSB). If necessary, the medium was supple-mented with appropriate antibiotics as follows: ampicillin, 100 mg/ml for E. coliand 50 mg/ml for B. abortus; gentamicin, 20 mg/ml for E. coli and 2.5 mg/ml forB. abortus; and tetracycline, 10 mg/ml for E. coli.

Cloning, DNA sequencing, and gene disruption. To isolate the pgm gene of B.abortus, a clone, named H7 (30) (accession number AQ752933), with high ho-mology to the pgm gene of A. tumefaciens, was used as a probe to screen agenomic library of B. abortus strain 2308 (12). The screening was carried out asdescribed previously (25) with filters washed at high stringency. Three cosmids,pBR261, pBR262, and pBR263, with different restriction enzyme patterns wereisolated. Southern blot analysis was performed with the three cosmids digestedwith EcoRI as described previously (25), using the H7 clone as a probe. Afragment of 4.3 kb was identified in cosmid pBR262; it was eluted from the geland ligated into pBluescript KS II(1) (Stratagene, La Jolla, Calif.) digested withEcoRI. This plasmid was named pBE39. Plasmid pBE39 was further digestedwith HindIII, and three fragments of approximately 3, 1, and 0.8 kb were gen-erated. These fragments were subcloned in pBluescript KS II(1) digested withHindIII, and they were sequenced by the dideoxy terminator method using theT7 Sequenase version 2.0 DNA sequencing kit (Amersham Life Science).

In order to mutagenize the pgm gene of B. abortus 2308, plasmid pBE39 wasdigested with EcoRI, and the DNA fragment of 4.3 kb was eluted from the geland ligated into pUC19 digested with EcoRI. The resulting plasmid was namedpUB22. Since pUC19 lacks an EcoRV restriction enzyme site, a gentamicin-resistant nonpolar cassette was introduced in a unique EcoRV site of the pgmgene by digesting the cassette with SmaI and ligating it into pUB22 digested withEcoRV. The resulting plasmid was named pUB22G, and it was introduced in B.abortus by electroporation. Double recombination events (Gmr Amps) wereselected and confirmed by PCR with a set of primers that amplified a 442-bpfragment from the wild-type gene and a 1,193-bp fragment from the gentamicin-interrupted gene.

For complementation experiments, the 4.3-kb fragment was ligated inpBBR1MCS-4 (13) digested with EcoRI; the resulting plasmid was namedpBBE30.

Construction of a nonpolar gentamicin resistance cassette. The gene accI,coding for gentamicin resistance, was amplified with oligonucleotides 59-TAGGATCCTTGACATAAGCCTGTTCG-39 and 59-TAGGATCCTTAGGTGGCGGTACTTGG-39 from the promoter to the stop codon. This fragment, lacking the

termination stem-loop of the gene, was cloned in pBluescript KS II(1) digestedwith BamHI. The resulting plasmid was digested with HindIII and NotI andligated into pSport1 (Gibco, Paisley, Scotland) digested with the same restrictionenzymes. The resulting plasmid, named pSPG1, has, at both flanking sides of theaccI gene, the following restriction sites: BamHI, SmaI, PstI, and EcoRI. It alsohas other nonsymmetrical restriction sites.

LPS purification and analysis. LPS was isolated by the hot-phenol-waterextraction procedure (36) from 100 ml of cells from overnight cultures. Theconcentration of LPS was measured by the 2-keto-3-deoxy-D-manno-2-octu-losonic acid assay (18) and analyzed by sodium dodecyl sulfate-polyacrylamidegel electrophoresis (SDS-PAGE) in 14% gels containing 4 M urea. LPS wasdetected by silver staining as described elsewhere (15). Tricine gel analysis wasperformed as described previously (26).

Intracellular replication in nonprofessionalphagocytes. Intracellular replica-tion was evaluated in HeLa cells as described elsewhere (19). Briefly, Brucellastrains and mutants were grown in liquid medium for 24 h and resuspended inminimal essential medium (Gibco) complemented with 5% fetal calf serum and2 mM glutamine without antibiotics (complete culture medium) at 107 CFU perml. This suspension was added to HeLa cells at a multiplicity of infection of 500:1and centrifuged at 180 3 g for 10 min. After 1 h of incubation at 37°C, freshcomplete culture medium with 100 mg of gentamicin per ml and 50 mg of strep-tomycin per ml was added to the monolayers. At 4, 24, and 48 h, the monolayerswere washed five times with phosphate-buffered saline (pH 7.4) and lysed with0.1% Triton X-100. The Triton lysates were then diluted serially and plated onTSB agar with the appropriate antibiotics to determine the number of CFUrecovered per milliliter.

Virulence in mice. Virulence was determined by quantitating the survival ofthe strains in the spleen after 2 weeks. Nine-week-old female BALB/c mice wereinjected intraperitoneally with approximately 105 CFU of brucellae in 0.1 ml of150 mM NaCl. Groups of five mice were injected with either B. abortus 2308,B. abortus B2211, or B. abortus B2211 complemented with plasmid pBBE30.

At 15 days postinfection animals were sacrificed by decapitation, and spleenswere removed, weighed, and homogenized in 150 mM NaCl. Tissue homoge-nates were serially diluted with phosphate-buffered saline and plated on TSBagar with the appropriate antibiotics to determine the number of CFU perspleen.

PmB assays. The bactericidal effect of polymyxin B (PmB) was tested asdescribed elsewhere (1). Overnight cultures of both strains were centrifuged andresuspended with HEPES (1 mM, pH 8). Approximately 103 CFU was incubatedat 37°C for 1 h over a range of PmB concentrations. Following the 1-h incubationperiod, 10-ml portions of the cell suspensions were spotted quadruplicated onTSB agar plates. This assay was performed in duplicate. The percentage ofsurviving bacteria was calculated according to the CFU recovered.

Inhibition of growth was calculated by plating a known number of CFU onTSB agar plates with increasing concentrations of PmB.

Nucleotide sequence accession number. The sequence of the pgm gene of B.abortus has been assigned GenBank accession number AF232056.

RESULTS

Identification, cloning, and nucleotide sequence of a geneencoding the phosphoglucomutase of B. abortus. As a result ofa B. abortus genome project (30), a clone named H7 (accessionnumber AQ752933) with high homology to the phosphogluco-mutase of A. tumefaciens was identified (33). Using this cloneas a probe, a library of cosmids of B. abortus 2308 was screenedand three cosmids with different restriction enzyme patternswere isolated. In order to verify the presence of pgm in the

TABLE 1. Bacterial strains and plasmids used in this study

Strain and plasmid Characteristics Reference or source

E. coli K-12 DH5a-F9IQ F9 f80dlacZDM15 D(lacZYA-argF)U169 deoR recA1 endA1 hsdR17 (rK2 mK

1) phoA supE44l2 thi-1 gyrA96 relA1/F9 proAB1 lacIqZDM15 zzf::Tn5(Kmr)

37

A. tumefaciens A5129 pgm::Tn5 33

PlasmidspBR262 20-kb fragment of genomic DNA of B. abortus 2308 in pVK102 that complements A5129 This studypBE39 4.3-kb EcoRI fraghment of pBR262 containing pgm of B. abortus in pBluescript KSII(1) This studypUB22 4.3-kb EcoRI fragment of pBR262 containing pgm of B. abortus in pUC19 This studypUB22G pUB22, pgm::Gmr This studypBBE30 4.3-kb EcoRI fragment of pBR262 containing pgm of B. abortus in pBBR1MCS-4 This studypSPG1 Nonpolar cassette accI (Gmr) in pSportI This studypBBR1MCS-4 Broad-host-range cloning vector (Ampr) 13

VOL. 68, 2000 PHOSPHOGLUCOMUTASE OF B. ABORTUS 5717

cosmid inserts, they were introduced by electroporation intothe dark-phenotype A. tumefaciens pgm mutant A5129 (33)(Table 1). The resulting transformants were plated on Luria-Bertani agar with 0.02% Calcofluor (Sigma Chemical Co.) andobserved under UV light. The three cosmids complementedthe dark phenotype, thus indicating that the pgm gene waspresent in the cosmids and correctly expressed in the Agrobac-terium background. The cosmids were digested with the EcoRIrestriction enzyme, and a Southern blot analysis was per-formed using clone H7 as a probe. A DNA fragment of ap-proximately 4,300 bp was identified in one of the cosmids, andthis fragment was cloned in pBluescript KS II(1). Both strands

of the DNA insert of the resulting plasmid, named pBE39,were sequenced. Analysis of the sequence revealed that the4.3-kb DNA fragment contained an open reading frame of1,635 bp coding for a protein of 544 amino acids which is74.7% identical to the phosphoglucomutase of A. tumefaciens(33) and 50% identical to the protein of Arabidopsis thaliana(accession number AAC00601) (Fig. 1). Sequence analysis ofthe regions upstream and downstream of pgm revealed nosignificant homology to any gene in the database, which wassurprising since in A. tumefaciens and Rhizobium loti, pgm ispart of the glycogen operon (29). This result indicates that inB. abortus, pgm is not part of the glycogen operon.

FIG. 1. Comparison of the B. abortus phosphoglucomutase protein with the A. tumefaciens and A. thaliana proteins. Conserved amino acids are indicated by blackboxes. The alignment was performed with the MegAlign program.

5718 UGALDE ET AL. INFECT. IMMUN.

A pgm mutant strain lacks the O antigen but has a completecore. Insertional mutagenesis of pgm was carried out, intro-ducing a gentamicin-resistant nonpolar cassette in a uniqueEcoRV site of the B. abortus pgm gene (see Materials andMethods); the resulting mutant strain, named B. abortus B2211,was used for further studies. As can be seen in Fig. 2A, theelectrophoretic profile of the LPS extracted from mutantB2211 indicates that it lacks the O antigen. However, the coreregion of the mutant LPS migrated in Tricine-PAGE electro-phoresis in a position that was indistinguishable from that ofthe wild type core (Fig. 2B), thus indicating that there wereminor differences between them. These results suggest that theB. abortus core LPS contains glucose or a derivative of glucosesynthesized through a sugar nucleotide intermediate, and itcan be speculated that the amount of glucose or derivativespresent in the core is minimal in relation to the total amountof sugars (one or two units). When an immunoblot analysis wasperformed, anti-LPS antibodies reacted with both prepara-tions, indicating that the compounds detected by silver staining

corresponded to LPS core and retained antigenic determinants(data not shown).

To corroborate the rough phenotype of strain B2211, thephage Tbilisi, which infects smooth strains, was used to infectthe B. abortus 2308 wild-type strain, the B. abortus B2211 pgmmutant, and the B. abortus B2211 pgm mutant complementedwith plasmid pBBE30. It was observed that the wild-type strainand the complemented pgm strain were susceptible but themutant strain B2211 was resistant, again indicating that themutant completely lacks the LPS O antigen.

Mutant strain B2211 is able to multiple within HeLa cells.To evaluate the importance of B. abortus smooth LPS in in-tracellular survival, the multiplication of the B. abortus wild-type strain 2308, the B. abortus B2211 pgm mutant, and theB. abortus B2211 pgm mutant complemented with plasmidpBBE30 was studied in HeLa cells. The number of viablebacteria was estimated at 4, 24, and 48 h after infection. At 4 hafter infection, the numbers of intracellular bacteria of thethree strains showed no significant differences, indicating thatthe rough mutant strain was internalized at the same rate asthe parental strain (Fig. 3). As it can be seen in Fig. 3, at 24 and48 h postinfection the numbers of bacteria recovered fromHeLa cells infected with the B. abortus 2308 wild-type strainor with the B. abortus B2211 mutant strain complementedwith plasmid pBBE30 were the same. However, the numberof bacteria recovered at 48 h from cells infected with theB. abortus B2211 pgm mutant strain was approximately 1 log10unit lower. Although the exponential intracellular replicationof the pgm mutant was delayed by approximately 20 h withrespect to that of the wild type, the high number of recoverablebacteria at 48 h postinfection indicates that mutant strainB2211 replicates inside HeLa host cells.

These results indicate that smooth LPS is not essential, ei-ther for invasion of the host cell or for intracellular replication.The fact that the mutant replicates at a lower rate is notnecessarily a consequence of the rough phenotype, since theabsence of pgm affects many other components of the cell wall,such as, for example, the synthesis of b(1,2) cyclic glucan. Itrecently has been demonstrated that a B. abortus cgs (cyclicglucan synthetase) mutant has a phenotype in HeLa cells sim-ilar to that of mutant B2211 (C. G. Briones, unpublished re-sults).

FIG. 2. PAGE of B. abortus LPS. (A) PAGE with 12% acrylamide. Lane 1,wild-type LPS; lane 2, B. abortus B2211 pgm mutant LPS. (B) Tricine-PAGE.Lane 1, wild-type LPS; lane 2, B. abortus B2211 pgm mutant LPS. Gels weresilver stained.

FIG. 3. Intracellular multiplication of the B. abortus wild-type strain and the B. abortus B2211 pgm mutant in HeLa cells. Cells were infected, and at the indicatedtimes postinfection (p.i.) the number of intracellular bacteria was determined as described in Materials and Methods.

VOL. 68, 2000 PHOSPHOGLUCOMUTASE OF B. ABORTUS 5719

Survival of the B. abortus B2211 pgm mutant in the mousemodel. Groups of five mice were injected intraperitoneallywith 105 CFU of the B. abortus B2211 pgm mutant strain, theB. abortus 2308 wild-type strain, or the B. abortus B2211 pgmmutant strain complemented with pBBE30. At 2 weeks post-inoculation mice were sacrificed, and spleens were weighedand examined for Brucella proliferation. As shown in Fig. 4A,the numbers of viable bacteria recovered from spleens of miceinjected with the B. abortus 2308 wild-type strain and theB. abortus B2211 pgm mutant complemented with plasmidpBBE30 were 7 3 105 and 1 3 106 CFU/ml, respectively. Onthe other hand, no viable bacteria were recovered from miceinjected with the B. abortus B2211 pgm mutant strain, withinthe detection limits of the method. In Fig. 4B it can also be

seen that the weights of the spleens of mice injected with theB. abortus B2211 pgm mutant strain were significantly lowerthan those of mice injected with the wild-type strain or with thepgm mutant complemented with plasmid pBBE30. This indi-cates that the inflammatory response generated by the pgmmutant was also lower than the one generated by the parentalwild-type strain.

Sensitivity to PmB. Defensins are cationic, amphipathic,low-molecular-weight peptides thought to permeabilize mem-branes of gram-negative organisms inside phagocytic cells (34).PmB, an amphipathic peptide that forms ionic interactionswith saccharide components of LPS, including 2-keto-3-de-oxyoctulonic acid and phosphate, has the greatest bactericidaleffect on B. abortus (16). PmB was used as a model defensin to

FIG. 4. Virulence in mice. Mice were infected as described in Materials and Methods. (A) Recovery of viable bacteria from spleens at 15 days postinfection. (B)Weights of spleens of infected mice at 15 days postinfection. Error bars indicate standard deviations.

5720 UGALDE ET AL. INFECT. IMMUN.

study the survival of the B. abortus B2211 pgm mutant. Figure5 shows the effect of PmB on strains 2308 and B2211 tested bytwo different assays, as described in Materials and Methods. Itcan be observed that in the concentration range of 0 to 10

mg/ml, PmB has a higher bactericidal effect on strain B2211than on the wild-type parental strain (Fig. 5A). These resultsindicated that, as was observed with other Brucella rough mu-tants, the lack of LPS O antigen increases the bactericidal

FIG. 5. Effect of PmB on B. abortus strains B2211 (pgm mutant) and 2308. (A) Bactericidal effect. PmB-mediated killing was performed as described in Materialsand Methods, and results are expressed as percentages of survival. (B) Inhibition of growth. The assay was carried out as described in Materials and Methods. Resultsare expressed as percentages of CFU recovered at the indicated concentrations of PmB. Error bars indicate standard deviations.

VOL. 68, 2000 PHOSPHOGLUCOMUTASE OF B. ABORTUS 5721

effect of cationic peptides. The inhibition of growth by PmB onboth strains is shown in Fig. 5B. It can be seen that PmBinhibited the growth of strain B2211 at concentration as low as0.25 mg/ml. On the other hand, the wild-type parental strain2308 grew at concentrations as high as 1 mg/ml. These resultsindicate that the concentration of PmB required for inhibitionof growth is approximately 1 order of magnitude lower thanthe concentration required for a bactericidal effect. However,both assays clearly showed a significant difference between theB. abortus pgm mutant and the wild-type parental strain.

DISCUSSION

In this study we have identified, sequenced, and disruptedthe pgm gene of B. abortus. The predicted protein is 74.7%identical to its homologue in A. tumefaciens but is not part ofthe glycogen operon as it is in Agrobacterium (29, 33). This typeof Pgm is specific for the synthesis of glucose sugar nucleotidederivatives like UDP-glucose or ADP-glucose, and it has nophosphomannomutase (Pmm) activity. In Agrobacterium andBrucella, Pmm is a separate protein (1, 8, 31). B. abortus LPSO antigen is a homopolymer of perosamine, a derivative ofmannose that is synthesized through GDP-mannose, thus, apgm mutant of this species would not be impaired in the syn-thesis of GDP-perosamine, the sugar donor of O-antigen sub-units. The lack of O antigen in a B. abortus pgm mutant mustbe the result of incomplete synthesis of the LPS core, whichwas described to contain glucose. As deduced from the elec-trophoretic profile of LPS in SDS-PAGE and Tricine gels, themutant has an almost complete core, which indicates that it isnot formed primarily by glucose as has been previously re-ported (17). One possible explanation for these differences isthat the preparations used to characterize the O antigen andthe core were contaminated with b(1,2) cyclic glucan, since ithas been reported that this polysaccharide strongly interactswith B. abortus LPS (3, 6, 14).

In mice the mutant was completely cleared from the spleensat 15 days postinoculation, thus indicating that it had lostvirulence. However, the ability to multiply intracellularly inHeLa cells, although delayed compared to that of the wildtype, was not abolished, indicating that intracellular multipli-cation is necessary but not sufficient for virulence. The impair-ment of the mutant in the synthesis of UDP-glucose results inthe inability to produce other polymers containing glucose,including b(1,2) cyclic glucan. In our laboratory we have re-cently observed that a B. abortus b(1,2) cyclic glucan synthetase(cgs) mutant also showed a delay in exponential intracellularreplication (C. G. Briones, unpublished results). Thus, it ispossible that this phenotype might be the result of the absenceof b(1,2) cyclic glucan rather than of the lack of O antigen.Despite this delay, the B. abortus pgm mutant replicated inHeLa cells, reaching 7 3 105 bacteria/ml, which indicates thatthe O antigen is not essential either for invasion or for intra-cellular multiplication.

Many reports have analyzed the phenotypes of rough mu-tants that were fortuitously isolated. Two important featuresoften analyzed in rough mutants are survival in mice andintracellular replication in professional or nonprofessionalphagocytes. It has been proposed that rough mutants are lessvirulent due to two causes: high sensitivity to lysis mediated bycomplement and inability to replicate intracellularly. It hasbeen demonstrated that the lack of O antigen in Brucella in-creases bacterial sensitivity to the killing activities of comple-ment (7). On the other hand the inability of rough strains toreplicate inside the cell is not a solved issue. Some roughmutants have been reported to be unable to replicate inside

professional phagocytes (22, 23, 27), but there are some re-ports that have demonstrated that rough mutants with well-characterized genetic backgrounds and parental strains haveno significant differences in their ability to replicate in macro-phages (1, 9).

The fact that a mutant lacking the O antigen did not loosethe ability to replicate inside the cells indicates that a completeLPS is not essential for either invasion or protection of Brucellaagainst the cellular defenses of the host. One possible expla-nation for the discrepancies among different reports is thatmany of the rough mutants used to study this phenotype werefortuitously isolated and are not genetically characterized, thuspossibly leading to incorrect conclusions.

Historically the search for rough mutants was done in twoways: by a spontaneous phenomenon known as phase varia-tion, in which rough variants appear in culture by a mechanismnot well understood (2, 10), or by successive passages in dif-ferent culture media with searching for spontaneously roughphenotypes. These procedures gave a high number of roughmutants. All of them are avirulent in animal models, but inmost cases the parental strain is not available. When the ca-pacity of these mutants to replicate in professional or nonpro-fessional cells is analyzed, the results must be interpreted verycarefully. We conclude that the incapacity to multiply insidethe cell is not a consequence of the rough phenotype but ratheris determined by a complex interaction between the bacteriaand the cells.

The main drawback a of live B. abortus S19 vaccine is theinduction of antibodies against the O antigen, which causesdifficulties in diagnosis in eradication campaigns carried outwith vaccination. The intracellular multiplication of this mu-tant, associated with the inability to survive in mice and thelack of O-antigen determinants, might be important in consid-eration of this mutant as a potential live vaccine for cattle.Protection induced by this mutant strain is under study in ourlaboratory.

ACKNOWLEDGMENTS

This work was supported by grants from the Agencia Nacional dePromocion Cientıfica y Tecnica, SECyT, Argentina, (PICT98 no. 01-04180 and PICT97 no. 01-00080-01767). J.E.U. and C.C. are fellowsof the Consejo Nacional de Investigaciones Cientıficas y Tecnicas,CONICET, Argentina. M.F.F. is a fellow of the Agencia Nacional dePromocion Cientıfica y Tecnologica. R.A.U. is a member of the re-search carrier of the CONICET, Argentina.

We acknowledge Diego Comerci and Rodrigo Sieira, University ofGeneral San Martın, for useful suggestions; Fabio Fraga, University ofGeneral San Martın, for technical assistance; Patricia Silvapaulo forkindly providing PmB; and J. J. Cazzulo, University of General SanMartın, for critical reading of the manuscript and useful suggestions.

REFERENCES

1. Allen, C. A., L. G. Adams, and T. A. Ficht. 1998. Transposon-derived Brucellaabortus rough mutants are attenuated and exhibit reduced intracellular sur-vival. Infect. Immun. 66:1008–1016.

2. Brawn, W. 1947. Bacterial dissociation. Bacteriol. Rev. 11:75–114.3. Bundle, D. R., J. W. Cherwonogrodzky, and M. B. Perry. 1988. Character-

ization of Brucella polysaccharide B. Infect. Immun. 56:1101–1106.4. Cangelosi, G. A., L. Hung, V. Puvanesarajah, G. Stacey, D. A. Ozga, J. A.

Leigh, and E. W. Nester. 1987. Common loci for Agrobacterium tumefaciensand Rhizobium meliloti exopolysaccharide synthesis and their roles in plantinteractions. J. Bacteriol. 169:2086–2091.

5. Caroff, M., D. R. Bundle, M. B. Perry, J. W. Cherwonogrodzky, and J. R.Duncan. 1984. Antigenic S-type lipopolysaccharide of Brucella abortus1119-3. Infect. Immun. 46:384–3828.

6. Cherwonogrodzky, J. W., and K. H. Nielsen. 1988. Brucella abortus 1119-3O-chain polysaccharide to differentiate sera from B. abortus S-19-vaccinatedand field-strain-infected cattle by agar gel immunodiffusion. J. Clin. Micro-biol. 26:1120–1123.

7. Corbeil, L. B., K. Blau, T. J. Inzana, K. H. Nielsen, R. H. Jacobson, R. R.

5722 UGALDE ET AL. INFECT. IMMUN.

Corbeil, and A. J. Winter. 1988. Killing of Brucella abortus by bovine serum.Infect. Immun. 56:3251–3261.

8. Foulongne, V., G. Bourg, C. Cazevieille, S. Michaux-Charachon, and D.O’Callaghan. 2000. Identification of Brucella suis genes affecting intracellu-lar survival in an in vitro human macrophage infection model by signature-tagged transposon mutagenesis. Infect. Immun. 68:1297–1303.

9. Godfroid, F., B. Taminiau, I. Danese, P. Denoel, A. Tibor, V. Weynants, A.Cloeckaert, J. Godfroid, and J. J. Letesson. 1998. Identification of the pero-samine synthetase gene of Brucella melitensis 16M and involvement of lipo-polysaccharide O side chain in Brucella survival in mice and in macrophages.Infect. Immun. 66:5485–5493.

10. Henry, B. S. 1933. Dissociation in the genus Brucella. J. Infect. Dis. 52:374–402.

11. Hoyer, B. H., and N. B. McCullough. 1968. Homologies of deoxyribonucleicacids from Brucella ovis, canine abortion organisms, and other Brucellaspecies. J. Bacteriol. 96:1783–1790.

12. Inon de Iannino, N., G. Briones, M. Tolmasky, and R. A. Ugalde. 1998.Molecular cloning and characterization of cgs, the Brucella abortus cyclicbeta(1-2) glucan synthetase gene: genetic complementation of Rhizobiummeliloti ndvB and Agrobacterium tumefaciens chvB mutants. J. Bacteriol.180:4392–4400.

13. Kovach, M. E., P. H. Elzer, D. S. Hill, G. T. Robertson, M. A. Farris, R. M.Roop, 2nd, and K. M. Peterson. 1995. Four new derivatives of the broad-host-range cloning vector pBBR1MCS, carrying different antibiotic-resis-tance cassettes. Gene 166:175–176.

14. L’vov, V. I., G. N. Pluzhnikova, E. B. Lapina, A. S. Shashkov, S. A. Askerova,V. E. Malikov, E. A. Dranovskaya, and B. A. Dmitriev. 1987. Molecularnature of the polysaccharide B antigen from Brucella (poly-B). Bioorgan.Khim. 13:4070.

15. Marolda, C. L., J. Welsh, L. Dafoe, and M. A. Valvano. 1990. Geneticanalysis of the O7-polysaccharide biosynthesis region from the Escherichiacoli O7:K1 strain VW187. J. Bacteriol. 172:3590–3599.

16. Martinez de Tejada, G., J. Pizarro-Cerda, E. Moreno, and I. Moriyon. 1995.The outer membranes of Brucella spp. are resistant to bactericidal cationicpeptides. Infect. Immun. 63:3054–3061.

17. Moreno, E., S. L. Speth, L. M. Jones, and D. T. Berman. 1981. Immuno-chemical characterization of Brucella lipopolysaccharides and polysaccha-rides. Infect. Immun. 31:214–222.

18. Osborn, M. J. 1963. Studies in the Gram-negative cell wall. I. Evidence forthe role of 2-keto-3-deoxyoctonoate in the lipopolysaccharide of Salmonellatyphimurium. Proc. Natl. Acad. Sci. USA 50:499–506.

19. Pizarro-Cerda, J., S. Meresse, R. G. Parton, G. van der Goot, A. Sola-Landa,I. Lopez-Goni, E. Moreno, and J. P. Gorvel. 1998. Brucella abortus transitsthrough the autophagic pathway and replicates in the endoplasmic reticulumof nonprofessional phagocytes. Infect. Immun. 66:5711–5724.

20. Pizarro-Cerda, J., E. Moreno, V. Sanguedolce, J. L. Mege, and J. P. Gorvel.1998. Virulent Brucella abortus prevents lysosome fusion and is distributedwithin autophagosome-like compartments. Infect. Immun. 66:2387–2392.

21. Pizarro-Cerda, J., E. Moreno, and J. P. Gorvel. 1999. Brucella abortus inva-sion and survival within professional and non-professional phagocytes, p.201–232. Advances in cell and molecular biology of membranes and or-

ganelles, vol. 6. JAI Press Inc., New York, N.Y.22. Riley, L. K., and D. C. Robertson. 1984. Brucellacidal activity of human and

bovine polymorphonuclear leukocyte granule extracts against smooth andrough strains of Brucella abortus. Infect. Immun. 46:231–236.

23. Riley, L. K., and D. C. Robertson. 1984. Ingestion and intracellular survivalof Brucella abortus in human and bovine polymorphonuclear leukocytes.Infect. Immun. 46:224–230.

24. Roantree, R. J. 1971. The relationship of lipopolysaccharide to bacterialvirulence, p. 1–37. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbialtoxins: a comprehensive treatise, vol. 5. Academic Press, Inc., New York,N.Y.

25. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

26. Schagger, H., and G. von Jagow. 1987. Tricine-sodium dodecyl sulfate-poly-acrylamide gel electrophoresis for the separation of proteins in the rangefrom 1 to 100 kDa. Anal. Biochem. 166:368–379.

27. Schurig, G. G., R. M. d. Roop, T. Bagchi, S. Boyle, D. Buhrman, and N.Sriranganathan. 1991. Biological properties of RB51; a stable rough strainof Brucella abortus. Vet. Microbiol. 28:171–188.

28. Tarof, C. W., and C. A. Hobbs. 1987. Improved media for growing plasmidand cosmid clones. Focus 9:12.

29. Ugalde, J. E., V. Lepek, A. Uttaro, J. Estrella, A. Iglesias, and R. A. Ugalde.1998. Gene organization and transcription analysis of the Agrobacteriumtumefaciens glycogen (glg) operon: two transcripts for the single phospho-glucomutase gene. J. Bacteriol. 180:6557–6564.

30. Ugalde, R. A. 1999. Intracellular lifestyle of Brucella spp.: common geneswith other animal pathogens, plant pathogens, and endosymbionts. MicrobesInfect. 1:1211–1219.

31. Uttaro, A., L. Ielpi, and R. A. Ugalde. 1993. Galactose metabolism in Rhi-zobiaceae. Characterization of exoB mutants. J. Gen. Microbiol. 139:1055–1062.

32. Uttaro, A. D., G. A. Cangelosi, R. A. Geremia, E. W. Nester, and R. A.Ugalde. 1990. Biochemical characterization of avirulent exoC mutants ofAgrobacterium tumefaciens. J. Bacteriol. 172:1640–1646.

33. Uttaro, A. D., and R. A. Ugalde. 1994. A chromosomal cluster of genesencoding ADP-glucose synthetase, glycogen synthase and phosphogluco-mutase in Agrobacterium tumefaciens. Gene 150:117–122.

34. Vaara, M. 1992. Agents that increase the permeability of the outer mem-brane. Microbiol. Rev. 56:395–411.

35. Verger, J. M., F. Grimont, P. A. D. Grimont, and M. Grayon. 1985. Brucella,a monospecific genus as shown by deoxyribonucleic acid hybridization. Int. J.Syst. Bacteriol. 35:292–295.

36. Westphal, O., and K. Jann. 1965. Bacterial lipopolysaccharides. MethodsCarbohydr. Chem. 5:83–91.

37. Woodcock, D. M., P. J. Crowther, J. Doherty, S. Jefferson, E. DeCruz, M.Noyer-Weidner, S. S. Smith, M. Z. Michael, and M. W. Graham. 1989.Quantitative evaluation of Escherichia coli host strains for tolerance to cy-tosine methylation in plasmid and phage recombinants. Nucleic Acids Res.17:3469–3478.

Editor: W. A. Petri, Jr.

VOL. 68, 2000 PHOSPHOGLUCOMUTASE OF B. ABORTUS 5723