Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Rajashekara, G. Articles by Splitter, G. Search for Related Content PubMed PubMed Citation Articles by Rajashekara, G. Articles by Splitter, G.

 Previous Article  |  Next Article 

Journal of Bacteriology, September 2008, p. 6243-6252, Vol. 190, No. 18
0021-9193/08/$08.00+0     doi:10.1128/JB.00520-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Genomic Island 2 of Brucella melitensis Is a Major Virulence Determinant: Functional Analyses of Genomic Islands ^,

Gireesh Rajashekara,^1* Jill Covert,² Erik Petersen,² Linda Eskra,² and Gary Splitter^2*

Food Animal Health Research Program, Ohio Agricultural Research Development Center, Department of Veterinary Preventive Medicine, The Ohio State University, Wooster, Ohio 44691,¹ Department of Animal Health and Biomedical Sciences, University of Wisconsin, 1656 Linden Drive, Madison, Wisconsin 53706²

Received 16 April 2008/ Accepted 6 July 2008

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Brucella genomic islands (GIs) share similarities in their genomic organization to pathogenicity islands from other bacteria and are likely acquired by lateral gene transfer. Here, we report the identification of a GI that is important for the pathogenicity of Brucella melitensis. The deletion of GI-1, GI-5, or GI-6 did not affect bacterial growth in macrophages as well as their virulence in interferon regulatory factor 1-deficient (IRF-1^–/–) mice, suggesting that these islands do not contribute to Brucella virulence. However, the deletion of GI-2 resulted in the attenuation of bacterial growth in macrophages and virulence in IRF-1^–/– mice. The GI-2 mutant also displayed a rough lipopolysaccharide (LPS) phenotype indicated by acriflavin agglutination, suggesting that in vitro and in vivo attenuation is a result of LPS alteration. Further, systematic analysis of the entire GI-2 revealed two open reading frames (ORFs), BMEI0997 and I0998, that encode hypothetical sugar transferases and contribute to LPS alteration, as the deletion of either of these ORFs resulted in a rough phenotype similar to that of the GI-2 mutant. Complementation analyses indicated that in addition to I0997 and I0998, I0999 is required to restore the smooth LPS in the GI-2 mutant as well as its full in vitro and in vivo virulence. The I0999 sequence analysis suggested that it might function as a transporter to help facilitate the transport or linking of the O antigen to the LPS. Our study also indicated that the rough LPS resulting from the GI-2 deletion may affect pathogen-associated molecular pattern recognition by Toll-like receptors.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Brucellosis caused by Brucella species is a zoonotic disease with a serious impact on public health and the livestock industry (13). Humans are targeted mainly by Brucella melitensis, although Brucella suis and Brucella abortus can also cause brucellosis in humans. This gram-negative, broad-host-range, facultative pathogen causes a systemic, febrile illness in humans and, while rarely fatal, can lead to chronic debilitating diseases such as endocarditis, osteoarthritis, and neurological impairment (13, 45). In animals, Brucella infection can lead to abortions in females and sterility in males (13).

Brucella, during entry into macrophages, interacts with lipid rafts. Lipid raft association is essential for Brucella to evade the degradative endocytic pathway and permit intracellular replication (24, 25, 33). Brucella lipopolysaccharide (smooth LPS), an important virulence factor, is required for lipid raft-mediated entry that allows intracellular survival (20, 35). Brucella species, unlike other gram-negative bacteria, possess nonclassical LPS with altered pathogen-associated molecular patterns (PAMP) leading to reduced endotoxicity. Toll-like receptors (TLRs) constitute a family of host pattern recognition molecules that respond to many PAMPs. TLR2 and TLR4 recognize lipoproteins and LPS of gram-negative bacteria (1, 40), resulting in the expression of proinflammatory cytokines, including tumor necrosis factor alpha (TNF-), interleukin-6 (IL-6), IL-12, macrophage inflammatory protein 2, and IL-10, that are critical for an effective immune response (3, 15). Brucella signals through TLR2, TLR4, and Myd88 (44); however, the activation of TLR4 signaling via smooth LPS is weak, which suggests that the bacteria have evolved to adapt to an intracellular lifestyle. The minimal activation of innate immune mechanisms aids early immune evasion and may favor successful trafficking and the establishment of an intracellular niche for replication within host cells. In contrast, when host cells are infected with Brucella species that possess rough LPS, a strong induction of proinflammatory cytokines (TNF-, IL-1, MIP2, and IL-10) occurs (23, 38). This strong activation of host innate immune mechanisms results in faster bacterial clearance; thus, rough Brucella strains are less virulent. The unusual structure of smooth LPS, consisting of a reduced number of anionic groups in the core oligosaccharide, the diaminoglucose backbone, and the presence of very long acyl chains in the amide and acyloxyacyl linkages of lipid A (20), largely accounts for the low bioactivity. In addition, smooth LPS may hamper the interaction of other PAMP-bearing surface molecules from host pattern recognition molecules interfering with cellular events that lead to pathogen clearance.

Because the genomes of Brucella species are highly conserved, host preference and virulence difference must stem from the limited genome diversity. Our previous comparative genomic analysis revealed six islands missing in Brucella species that are not pathogenic to humans (B. neotomae and B. ovis) yet are present in species that are able to effectively infect human hosts (B. melitensis, B. suis, and B. abortus) (36). These islands range in size from approximately 7 kb to 45 kb and display characteristics typical of horizontally transferred DNA. Genomic islands (GIs) constitute major elements in the evolution of bacterial pathogens, because the acquisition of many GIs converts innocuous strains into pathogenic strains (21). The presence of GIs in most Brucella species suggests that many of these islands were acquired before Brucella diverged from its ancestor. Brucella belongs to the -2 subclass of Proteobacteria that includes plant symbionts and soil inhabitants. Therefore, such DNA acquisition or loss may have provided a selective advantage to this bacterium for adaptation to its current intracellular lifestyle in mammalian hosts. Most islands contain a number of hypothetical proteins with unidentified functions. In addition, a preliminary examination of open reading frames (ORFs) within GI-2 and GI-5 revealed potential virulence factors. To gain a comprehensive understanding of the contribution of GIs to Brucella pathogenicity, we undertook a systematic approach by generating a full set of deletion mutants for these GIs in a virulent B. melitensis 16M strain to dissect their role in virulence. Our systematic deletion of genes within GI-2 identified genes essential for the virulence of B. melitensis. The deletion of GI-2 altered the LPS, leading to a rough phenotype, and may have altered TLR signaling.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Bacterial strains, plasmids, and growth conditions. Bacterial strains and plasmids used in this study are listed in Tables S1 and S2 in the supplemental material. The B. melitensis 16M strain (ATCC 23456), B. ovis REO198, and B. canis RM666 (ATCC 23365) from our laboratory collection were used in this study. All Brucella strains were grown in brucella broth (Difco). Brucella strains were grown at 37°C with shaking unless otherwise stated. Escherichia coli strain DH5 (Invitrogen) was used in cloning experiments and was grown in LB broth (Difco) at 37°C on a rotary shaker. Ampicillin (100 µg/ml), chloramphenicol (20 µg/ml), kanamycin (50 µg/ml), and zeocin (50 µg/ml for E. coli and 250 µg/ml for Brucella) were added to the medium as necessary.

Mice. Six- to eight-week-old C57BL/6 wild-type (WT) and C57BL/10ScNJ TLR4-deficient mice were purchased from Jackson Laboratories (Bar Harbor, ME). C57BL/6 TLR2-deficient mice were a gift from M. J. Fenton (University of Maryland—Baltimore) and originally developed by S. Akira. The mice were housed in AAALAC-approved facilities under pathogen-free conditions using standard protocols. Interferon regulatory factor 1-deficient (IRF-1^–/–) mice were bred and maintained in University of Wisconsin—Madison animal care facilities, and 6- to 9-week-old IRF-1^–/– mice or 10- to 12-week-old TLR-deficient mice were used for the experiments.

Plasmid construction. Suicide vectors for generating deletions in the GIs or individual genes were created using pZErO-1. Oligonucleotides used for deleting Brucella GIs and the individual genes or operons within the island are provided in Table S3 in the supplemental material. Oligonucleotides were designed using the available B. melitensis genomic sequences (NCBI; GenBank accession numbers are AE008917 and AE008918 for chromosomes I and II, respectively). Suicide vectors for generating specific deletions were constructed using two approaches. Briefly, for larger deletions (more than 4 kb), an approximately 1-kb DNA fragment both upstream and downstream of the desired deletion target was cloned into pZErO-1 (Invitrogen) in two steps. The primers for amplifying the flanking regions were designed with appropriate restriction sequences added to the 5' end to facilitate the insertion of a kanamycin resistance gene, aph(3')-Ia (Kan^r), from pUC4K between the two fragments to create the final suicide vector. Alternatively, for a smaller deletion, the desired deletion target was amplified along with approximately 1-kb upstream and downstream sequences and cloned into pZErO-1. Then inverse PCR was performed, using primers designed to amplify all but the deletion target. Appropriate restriction sites were included in the inverse PCR primers to facilitate the insertion of the Kan^r gene. The inverse PCR products were digested with restriction enzymes and ligated to the Kan^r gene fragment to generate the final suicide vector.

To construct the complementation plasmids, DNA sequences encoding the respective ORFs plus the ribosome binding site but lacking their promoter sequences were amplified using specific primers with appropriate restriction sites to aid cloning. Since the promoters for these genes are unknown, the expression of these ORFs with their own ribosome binding site under the constitutively active pBBRMCS4 lac promoter is an effective way to perform complementation analysis in B. melitensis. Earlier studies using a promoterless Lux operon have suggested that the lac promoter in pBBRMCS1 is constitutive in Brucella (data not shown). The PCR products were digested and ligated with similarly digested pBBR-MCS4 to generate the complementation plasmids. The genes of interest were directionally cloned into pBBR1-MCS4 to ensure that these genes were transcribed from the lac promoter present in the plasmid.

Generation of deletion mutants. To generate specific deletions, suicide vectors were electroporated into B. melitensis 16M. Following electroporation, the bacteria were plated on brucella agar containing kanamycin. To select for double recombinants, the Kan^r colonies were checked for sensitivity to zeocin (Zeo^s). The resulting Kan^r and Zeo^s clones were streak purified, and one such purified clone was used for further study.

PCR and Southern hybridization. The deletions of GIs were confirmed by PCR and Southern hybridization. A primer internal to the Kan^r gene and a primer external to the flanking fragment were used to confirm the deletions. Colonies that tested positive were further confirmed by Southern blotting. For Southern hybridization, 10 µg of genomic DNA was digested with HindIII and separated on a 0.7% agarose gel by electrophoresis. The deletion of the GIs with the insertion of the kanamycin gene at the expected location was detected using the Kan^r gene as a probe as previously described (37).

RAW 264.7 macrophage infection. Macrophage-like RAW 264.7 cells were cultured in RPMI supplemented with 10% heat-inactivated fetal calf serum. For the macrophage growth assays, 24-well plates were seeded with 5 x 10⁵ macrophages/well and infected with different B. melitensis strains at a multiplicity of infection (MOI) of 100. The cells were incubated for 1 h at 37°C in 5% CO₂, and extracellular bacteria were removed with three washes of phosphate-buffered saline (PBS), followed by gentamicin treatment (25 µg/ml) for 30 min. Then the cells were maintained with medium containing 5 µg of gentamicin/ml. At specified times, the cells were washed with PBS three times, lysed with 0.1% Triton X-100, and plated on brucella agar to determine the number of intracellular bacteria. All experiments were performed at least in duplicate.

BMDM infections. Bone marrow-derived macrophage (BMDM) cells were flushed from the femurs and tibiae of 10- to 12-week-old WT C57BL/6, TLR2^–/–, or TLR4^–/– mice using 5 ml of RPMI through a 25-gauge needle. The cells were grown for 5 to 8 days in RPMI 1640 (Gibco, Grand Island, NY), 10% fetal bovine serum (Equitech-Bio, Inc., Kerrville, TX), 25 µg/ml gentamicin, and 20 to 30 ng/ml macrophage colony stimulating factor (M-CSF; R&D Systems, Minneapolis, MN). The cells were harvested and plated at a density of 1 x 10⁶ cells per well in six-well plates (Corning, Costar, Corning, NY) in antibiotic-free RPMI and 20 ng/ml M-CSF for 12 to 24 h. M-CSF was removed from the medium 10 to 12 h before infection. BMDM cells were infected with WT B. melitensis or the GI-2 mutant at an MOI of 100 and incubated at 37°C for up to 72 h.

Fluorescence-activated cell sorter analysis. Adherent BMDMs cultured for 5 to 8 days in the presence of M-CSF were harvested for flow cytometric analysis concurrently when cells were plated for infections. The cells were fixed with 4% paraformaldehyde (Electron Microscopy Sciences, Hatfield, PA) and blocked with 1% bovine serum albumin in PBS before staining with phycoerythrin anti-mouse F4/80 antigen pan macrophage marker BM8 (eBioscience, San Diego, CA) or phycoerythrin rat immunoglobulin G2a, isotype control (BioLegend, San Diego, CA). Using a FACScan (Becton Dickinson Immunocytometer Systems, Palo Alto, CA), 10,000 events were collected. The analyzed cells consistently averaged >91% F4/80⁺ (data not shown) for BMDM cells from WT, TLR2^–/–, or TLR4^–/– mice, as analyzed by FloJo software (Tree Star, Ashland, OR).

Macrophages infected with B. melitensis strains 16M and GI-2 knockout containing plasmid constructs of green fluorescent protein (GFP_UV) on plasmid pMC221 were analyzed by flow cytometry. The number of BMDM cells infected was determined when the supernatants were collected for cytokine analysis. B. melitensis strains containing plasmid constructs replicated similarly to WT bacteria in macrophages.

Cytokine measurements from BMDM supernatants. Supernatants were collected at 4, 12, or 24 h from cells infected with the GI-2 knockout strain or WT B. melitensis. The early infection times for each bacterial strain were chosen when 70% to 90% of the cells were infected based on preliminary data with RAW 264.7 cell infections (data not shown). The late infection times for each bacterial strain were chosen to maintain similar cell viability, based on crystal violet cell staining (data not shown). In addition, the supernatants for the cytokine assays were collected when an approximately similar percentage of BMDMs were infected with either the 16M or GI-2 knockout strains as described above. The supernatants were filtered through a 0.22-µm polyethersulfone Millipore filter (Millipore, Billerica, MA) and assayed for TNF-, IL-10, or IL-12 using Ready-Set-Go enzyme-linked immunosorbent assay kits (eBioscience, San Diego, CA).

IRF-1^–/– mouse virulence assay. Groups of 6- to 9-week-old IRF-1^–/– mice were infected intraperitoneally with 1 x 10⁷ B. melitensis strains. The infected mice were housed in biosafety level 3 facilities and monitored for survival. Following infection, the mice were monitored for 14 days, as virulent Brucella kills these mice within 14 days (26). From the surviving mice, the livers and spleens were collected aseptically, homogenized in PBS, and plated on brucella agar. The plates were incubated at 37°C for 4 days and the CFU determined.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
GI-2 but not the GI-1, GI-5, or GI-6 mutant is attenuated in RAW macrophages. Previously, we described the identification of several GIs that are specifically present in virulent Brucella but absent in nonpathogenic Brucella species (36). These islands include a number of ORFs with no annotated functions, and therefore, identifying the contributions of these GIs to Brucella virulence and pathogenicity would reveal novel virulence mechanisms. We created deletions in GI-1 (8.1 kb), GI-2 (15.1 kb), GI-5 (45 kb), and GI-6 (17.5 kb). GI-1, -2, and -5 are absent in B. ovis, and GI-6 is absent in B. neotomae. Because of the large size of GI-5, the deletion of this island was accomplished by deleting the entire region in four 11- to 12-kb segments. In addition, we also created deletions in two ORFs, BMEII0712 and BMEII0986. Both ORFs encode putative transcriptional regulators; BMEII0712 is part of GI-6, whereas BMEII0986 is a cyclic AMP receptor protein family transcriptional regulator involved in denitrification (5). Previously, we have shown that B. neotomae that naturally lacks BMEII0986 is less virulent in IRF-1^–/– mice compared to B. neotomae complemented with BMEII0986 in trans (5). Deletion mutants were created by homologous recombination, replacing the desired target with the Kan^r marker, and the deletion was confirmed by PCR using primers external to the flanking regions and by Southern blot analysis (data not shown).

To test whether the GI mutants were attenuated in macrophages, we examined their growth in RAW macrophage-like cells. All GI mutants grew similarly to 16M in brucella broth with a duplication time of 2 h, suggesting no general growth defect (Fig. 1). The GI-1, GI-5, GI-6, BMEII0712, and BMEII0986 mutants showed infection patterns similar to those of WT 16M, with an initial drop in intracellular bacteria of 1 to 2 logs between 0 and 4 h after infection (Fig. 1), followed by replication similar to 16M at 72 h. However, the GI-2 mutant exhibited a different pattern. The GI-2 mutant was phagocytosed more with a slight initial drop in intracellular bacteria with no apparent intramacrophage replication as bacterial levels remained constant through the 72-h period (Fig. 1). Since RAW macrophages may have lower bactericidal activities than primary macrophages (12), we tested mutant strains in BMDMs. BMDMs were infected with the mutant strains and 16M similarly to RAW cell infection. Similarly to that observed with the RAW cells, the GI-2 mutant exhibited a slight initial drop with no apparent intracellular replication, whereas the other GI mutants followed the same pattern as 16M (data not shown). Because the results from the two macrophage growth assays were similar, RAW cells were used for the remainder of the experiments. Since growth curves for these mutants were similar in brucella broth (Fig. 1), the phenotypic difference in bacterial numbers observed in the macrophages is likely an effect of GI-2 deletion.

View larger version (31K): [in this window] [in a new window]

FIG. 1. Replication kinetics of B. melitensis GI mutants. (A and B) Stationary-phase grown cultures (30 µl) were inoculated into 30 ml of brucella broth and grown at 37°C with shaking, and the optical density at 600 nm (OD600) was determined. (C and D) RAW 264.7 macrophages were inoculated with a standardized bacterial suspension of different strains, and growth was monitored at specified times. CFU are presented as the geometric means ± standard errors of the data from two independent experiments.

Growth of the GI-2 mutant in macrophages was similar to that of rough strains of Brucella (38). However, B. ovis, which is naturally rough and lacks GI-2, was attenuated in growth in macrophages with a significant decrease in intracellular bacteria by 24 h postinfection (Fig. 1), suggesting that LPS from these two species may contribute differently to their survival in the macrophages. Rough strains of Brucella are phagocytosed more efficiently than smooth strains and persist intracellularly at higher levels though they are defective for intramacrophage replication (23, 34). To confirm that GI-2 was internalized in higher numbers than WT Brucella, we evaluated the GI-2 mutant constitutively expressing GFP by fluorescent microscopy and flow cytometry. Consistent with the results of the macrophage growth assay, macrophages infected with the GI-2 mutant were highly fluorescent compared to the WT Brucella-infected macrophages, and this was further supported by flow cytometry (Fig. 2). These findings led us to hypothesize that the GI-2 mutant possessed an LPS defect. To analyze for an LPS defect, the GI-2 mutant was tested for sedimentation and agglutination in the presence of acriflavin (9). Cultures of WT Brucella 16M and the GI-1, GI-5, and GI-6 mutants grown to stationary phase for 72 h without shaking remained opaque, indicating a smooth phenotype. In contrast, cultures of the Brucella GI-2 mutant appeared clear, as bacteria sedimented to the bottom of the test tubes, indicating a rough phenotype (see Table 2). Further, the agglutination assay in the presence of acriflavin resulted in the clumping of the GI-2 mutant but not 16M or the other GI mutants (see Table 2). These findings suggest that the GI-2 mutant is unable to produce a smooth LPS.

View larger version (29K): [in this window] [in a new window]

FIG. 2. The GI-2 mutant is internalized in higher numbers by macrophages than WT 16M. RAW 264.7 macrophages were infected for 6 h with B. melitensis 16M, B. melitensis GI-2, B. canis, or B. suis containing GFPUV-expressing plasmid pMC221 at an MOI of 100. The cells were washed two times with PBS before fixing in 4% paraformaldehyde for 30 min, then washed twice, resuspended in PBS plus 1% bovine serum albumin, and filtered through nylon mesh. (A) Fluorescent images were digitally captured at x63 magnification. (B) The percentage of infected cells was determined by flow cytometry (FACScan; BD Biosciences) and analyzed using FlowJo version 7.2 (Tree Star, Inc.).

View this table: [in this window] [in a new window]

TABLE 2. Virulence in IRF-1–/– mice and the LPS phenotype of mutants within GI-2 and complemented strains

Pathogenicity of B. melitensis GI mutants. GI mutants were evaluated in IRF-1^–/– mice to determine their virulence. IRF-1^–/– mice serve as a rapid in vivo indicator system to assess the virulence of different Brucella strains (26). In addition, we also tested the less- or nonpathogenic B. canis, B. ovis, and B. neotomae that naturally lacked one or more of these GIs (36). Both B. canis and B. ovis are naturally occurring rough strains, whereas B. neotomae has smooth LPS. IRF-1^–/– mice infected with all the GI mutants except the GI-2 mutant died by 7 to 10 days postinfection, similarly to 16M-, B. canis-, and B. neotomae-infected mice. Spleens collected from the moribund animals contained 10⁶ to 10⁸ CFU (Table 1). In contrast, mice infected with the GI-2 mutant appeared healthy and survived longer than 24 days, suggesting attenuation. Spleens from these mice had less than 10³ CFU. Similarly to the GI-2 mutant, the B. ovis-infected mice also survived greater than 24 days, and bacterial CFU counts from their spleens were similar to those from the GI-2 infected mice (Table 1).

View this table: [in this window] [in a new window]

TABLE 1. Virulence of GI mutants in IRF-1–/– mice

Multiple genes within GI-2 are required for virulence. GI-2 encodes 20 ORFs, with a majority of these ORFs having no known function. The island includes two hypothetical sugar transferases, BMEI0997 and BMEI0998, that are likely involved in LPS biogenesis (27, 41, 43), a hypothetical secretion activator protein (BMEI0995), a 25-kDa outer membrane precursor protein implicated in Brucella virulence (7, 14), and a tRNA-ribosyltransferase (BMEI1003) (Fig. 3). To determine the contribution of genes within this island to pathogenesis, we created nonpolar mutants by systematically deleting individual ORFs or potential operons. These mutants were analyzed in RAW macrophages to identify their role in Brucella replication. The deletion of I0995, I1007, and I1007-1010 resulted in infection patterns similar to those of 16M and smooth LPS (data not shown), which suggests that these proteins may not play a role in Brucella virulence. However, the I0997, I0998, I0997-0998, and I0997-1003 deletion mutants showed an infection pattern similar to that of the GI-2 mutant (Fig. 4). Additionally, sedimentation and agglutination analyses suggested that these later mutants possessed rough LPS similar to those of the GI-2 mutant (Table 2). Consistent with macrophage infection data, mice infected with the I0995, I1007, and I1007-1010 deletion mutants all died within 7 to 10 days postinfection (Table 2). However, mice infected with I0997, I0998, I0997-998, and I0997-1003 survived the length of the experiment, indicating that one or more of these genes is necessary for virulence (Table 2).

View larger version (19K): [in this window] [in a new window]

FIG. 3. Genetic organization of the ORFs within GI-2. The arrows indicate the direction of transcription with numbers corresponding to the B. melitensis 16M genome sequence. Only relevant features are shown; the picture is not drawn to scale. OMP, outer membrane protein; HP, hypothetical protein; Tnpase, transposase.

View larger version (24K): [in this window] [in a new window]

FIG. 4. Replication kinetics of mutants within GI-2. All mutants (I0997, I0998, I0997-0998, and I0997-1003) display an infection pattern comparable to that of the GI-2 mutant. RAW 264.7 macrophages were inoculated with a standardized bacterial suspension of different strains, and growth was monitored at specified times. CFU are the geometric means ± standard errors of the data from two independent experiments.

Restoration of virulence to the GI-2 mutant requires BMEI0999 in addition to BMEI0997 and BMEI0998. To confirm the genes responsible for the observed attenuation of the GI-2 mutant, several complementation plasmids within the BMEI0997-1003 region were constructed. Complementation was tested either by using BMEI0997 or I0998 alone or I0997-0998 together or by adding an additional ORF to the I0997-0998 construct, thus representing the entire BMEI0997-1003 region. RAW macrophages were infected with the WT 16M, GI-2 mutant, or GI-2 mutant-complemented strains. Except for the GI-2 mutant complemented with I0997, I0998, or I0997-0998, all other complemented strains had an infection curve similar to that of WT 16M (Fig. 5), suggesting that BMEI0999 in addition to I0997-0998 is sufficient for the restoration of the virulent phenotype. To further confirm this conclusion, these complemented mutants were tested for sedimentation and agglutination to verify the LPS phenotype. The GI-2 mutant complemented with BMEI0997-999 or upstream genes had sedimentation and agglutination profiles similar to those of WT 16M. In contrast, complementation with BMEI0997, I0998, or I0997-998 retained the rough phenotype or displayed sedimentation and agglutination (Table 2). These results also indicate that while both I0997 and I0998 are necessary for LPS biosynthesis, all three genes, including I0999, are required for smooth LPS formation, and, therefore, fully virulent bacteria.

View larger version (25K): [in this window] [in a new window]

FIG. 5. Complementation with I0997-0999 restores the WT growth kinetics of the GI-2 mutant. RAW 264.7 macrophages were inoculated with a standardized bacterial suspension of mutant and complemented strains, and the growth was monitored at specified times. The CFU counts were log transformed, and values are the averages ± standard errors of the data from two independent experiments.

To test whether complementation of the GI-2 mutant with ORFs within BMEI0997-1003 would restore the in vivo virulence, IRF-1^–/– mice were infected with the GI-2 mutant complemented with the above constructs. All mice infected with the GI-2 mutant complemented with either I0997-999c, I0997-1000c, I0997-1001c, or I0997-1003c died by 7 to 10 days postinfection similarly to the WT 16M-infected mice, whereas the mice infected with the GI-2 mutant or the GI mutant complemented with I0997-998 survived greater than 24 days (Table 2). These results further support that I0999, in addition to I0997 and I0998, is required to restore the virulence of the GI-2 mutant in vitro and in vivo.

BMEI0999 is transcribed as an independent unit. The BMEI0997-1003 region has seven ORFs that are annotated to be transcribed in the same orientation. To determine the transcriptional organization of the region, we performed overlapping reverse transcriptase PCR (RT-PCR). Total RNA was extracted from the WT 16M strain grown to log phase in brucella broth. RT-PCR analysis of the total RNA using several sets of overlapping primers suggested that this region is organized into four transcriptional units: I0997-998, I0999, I1000-1002, and I1003 (Fig. 6). RT-PCR with primers c/d and e/f did not yield any overlapping products, suggesting that I0999 is likely transcribed as an independent unit (Fig. 6). Closer analysis of the sequence in this region identified a 331-bp intergenic region between I0999 and I1000 that may potentially contain the promoter elements consistent with the RT-PCR data (Fig. 6). BMEI0999 encodes a hypothetical protein that has similarity to an outer membrane autotransporter barrel domain protein of Rhodopseudomonas palustris (33% identity and 44% positive) that is closely related to Brucella and Chlorobium phaeobacteroides (25% identity and 41% positive).

View larger version (63K): [in this window] [in a new window]

FIG. 6. Transcriptional organization of the BMEI0997-03 region within GI-2. (A) Organization of the BMEI0997-03 region showing the primer locations indicated by arrows designated by a letter. The intergenic spaces are indicated by a horizontal line with a number indicating the space between genes. The potential promoters in the intergenic spaces are indicated by the dark boxes. (B) Agarose gel of the RT-PCR amplification products. RNA was extracted from 16M grown in brucella broth using a Masterpure RNA kit (Epicentre). The primers used are indicated above each lane.

Rough LPS of the GI-2 mutant may alter TLR signaling. WT B. melitensis possesses a nonclassical smooth LPS that is weakly endotoxic and poorly signals through TLRs, suggesting Brucella evolved to adapt to an intracellular environment/lifestyle (6). However, Brucella species with rough LPS can induce a more pronounced proinflammatory cytokine response, and this rapid response is implicated in the faster clearance of rough Brucella. This observation has led us to hypothesize that the unusual structure of smooth LPS may interfere with other PAMP-associated structures from host TLR recognition. Thus, smooth LPS would interfere with host cellular events that lead to pathogen clearance. To determine whether rough LPS in the GI-2 mutant have altered the presentation of bacterial surface structures recognized by TLR2 and TLR4, we analyzed the cytokine profiles of BMDMs from TLR2^–/–, TLR4^–/–, and WT mice infected with the GI-2 mutant. Supernatants from BMDMs infected with the GI-2 mutant were assayed for TNF-, IL-12, and IL-10 at 4 and 12 h postinfection. The assay times were chosen when most macrophages were infected and viable (data not shown). TNF-, Il-12, and IL-10 levels were similar in the supernatants collected from WT BMDMs or TLR4^–/–-infected cells at 4 as well as 12 h postinfection (Fig. 7). However, supernatants from the TLR-2^–/– BMDMs had significantly lower levels of TNF- (P < 0.05) and IL-10 (P < 0.05) at 4 or 12 h postinfection (Fig. 7). Similar results for TNF-, IL-12, and IL-10 production were observed in the supernatants of BMDMs from TLR-4^–/–, TLR-2^–/–, or WT mice infected with 16M. In addition, the supernatants from TLR2^–/– BMDMs infected with 16M had significantly lower levels of TNF- and IL-10 at 24 h postinfection compared to supernatants from WT BMDMs infected with 16M (Fig. 7). These results suggest the delayed signaling of B. melitensis through TLR-2 and TLR-4 consistent with previous findings that PAMP signaling is delayed, likely permitting the bacteria to establish an intracellular niche (29). Interestingly, there was a significant induction of all three cytokines in the TLR2^–/– BMDMs infected with the GI-2 mutant compared to those in 16M at 12 h postinfection, suggesting that rough LPS of the GI-2 mutant may have altered the recognition of bacterial surface structures by other TLRs. In addition, the infection levels of the GI-2 mutant as well as 16M were similar at the experimental time points chosen in these BMDMs (data not shown), suggesting that the cytokine induction observed in macrophages was not due to a difference in bacterial infection.

View larger version (41K): [in this window] [in a new window]

FIG. 7. Induction of proinflammatory cytokines by the GI-2 mutant in BMDMs from TLR2–/–, TLR4–/– knockout, and WT mice. The cytokine levels were assayed from the supernatants of the GI-2 mutant or B. melitensis 16M-infected C57BL/10ScNJ (TLR4–/–), C57BL/6 (TLR2–/–), and C57BL/6 WT BMDMs. BMDMs were cultured 5 to 7 days preinfection. Cytokines were measured in pg/ml by enzyme-linked immunosorbent assay at 4, 12, or 24 h postinfection. TNF-, IL-12, and IL-10 levels from GI-2 mutant-infected or 16M-infected TLR4–/–, TLR2–/–, and WT BMDMs. Data are from three to seven mice/experiment. Significant differences ( *, P < 0.05; * *, P < 0.005) between TLR2–/– or TLR4 –/– BMDM versus WT BMDM cytokine levels were determined by the Student t test.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Brucella GIs have striking similarities in genomic organization to pathogenicity islands from other bacteria. Pathogenicity islands are mostly found adjacent to or integrated into tRNA genes and flanked by insertion sequences. All four GIs, GI-1, -2, -5, and -6, examined in the present study were found either within or adjacent to tRNA genes with insertion sequences flanking one or both ends and have direct or inverted repeats at their ends, suggesting that these GIs were acquired by lateral gene transfer (36). These findings lead us to hypothesize that these GIs may represent Brucella pathogenicity islands and a relatively small number of differences are responsible for the host preference and virulence difference of Brucella species. In the present study, we describe the contribution of different GIs in B. melitensis virulence.

GI-1 contains mostly ORFs encoding hypothetical proteins and GI-5 includes 19 ORFs encoding peptide ABC-type transporters, such as Dpp, Opp, and Pot systems. The deletion of either GI-1 or GI-5 did not affect the ability of B. melitensis to grow in macrophages or its virulence in IRF-1^–/– mice. Further, GI-5 mutants, unlike B. ovis, showed no difference in their susceptibility to cationic peptides compared to that of WT 16M (data not shown). Homologs of GI-5-encoded transporters in other bacteria are important for root colonization, intracellular survival, attachment to the host cell, and virulence (8, 28, 31), and the absence of GI-5-encoded transporters may contribute to the susceptibility of B. ovis to cationic antimicrobial peptides (16; unpublished data), making them more readily destroyed within professional as well as nonprofessional phagocytes compared to other Brucella spp. (16). In the present study, the lack of a discernible phenotype for GI-5 mutants may be the result of our inability to delete the whole GI as a single entity or of B. melitensis using alternative factors to overcome host innate responses.

The deletion of GI-2 resulted in a rough LPS phenotype and the attenuation of growth in macrophages as well as virulence in IRF-1^–/– mice. Since many ORFs within GI-2 have been implicated in Brucella virulence (27, 41), we systematically deleted several ORFs and analyzed their contributions to virulence. A macrophage growth assay and an in vivo IRF-1^–/– mice virulence assay revealed a seven-ORF cluster (BMEI0997-1003) that is responsible for attenuation. This cluster includes two hypothetical sugar transferases, mannosyltransferase (BMEI0997 [wbdA]) and glycosyltransferase (BMEI0998 [wboA]). The deletion of wboA has been shown to result in the rough LPS phenotype, suggesting that it is involved in the biosynthesis of the LPS O chain (41, 42). The inactivation of B. suis wbdA, a homolog of I0997, has been shown to attenuate replication in macrophages (27); however, it was not determined whether attenuation was a result of LPS alteration. Our study revealed that the deletion of I0997 also led to a rough LPS phenotype similar to I0998 deletion (Table 2), suggesting that both are involved in the biosynthesis of the LPS O chain. However, complementation of the GI-2 mutant with either I0997 or I0998 alone or together did not restore the smooth LPS phenotype and did not restore virulence in macrophages or mice. These findings suggest that an additional gene(s) within the I0997-1003 cluster is responsible for the smooth LPS phenotype and, by extension, the virulence. Further, our complementation analysis indicated that I0999 is required in addition to I0997-I0998 to restore the smooth LPS phenotype as well as full virulence. Previous complementation studies with a spontaneous rough mutant of B. abortus RB51 that has an insertion element inserted within wboA (I0998) also suggested that an additional protein(s) other than WboA is necessary to restore smooth LPS as well as virulence (41, 42). This additional protein(s) may serve to mediate the transport of the O antigen to the surface of the bacteria or to participate in the proper attachment of the O antigen to the LPS core. BMEI0999 does not share any sequence homology to many of the proteins that are involved in O antigen flipping in gram-negative bacteria (2). However, I0999 has homology to an outer membrane autotransporter protein of R. palustris that is closely related to Brucella and may participate in O antigen transport. Previous studies have suggested that the translocation of lipid-linked oligosaccharides across membranes can be catalyzed by proteins that do not share any sequence similarity (2).

Brucella possesses nonclassical LPS, and it has been speculated that this may mask the PAMPs, thereby delaying recognition by TLRs aiding pathogen survival (6). However, rough strains of Brucella induce an enhanced cytokine response in cultured macrophages (23, 38), suggesting an early recognition by TLRs. Brucella signals through both TLR2 and TLR4 (6, 18, 44); however, which Brucella components activate TLR2 and TLR4 is not clearly understood. Macrophages from TLR4^–/– mice responded similarly as the WT BMDMs to 16M or the GI-2 mutant, indicating that TLR4 does not mediate the recognition of Brucella early in infection. However, there was reduced cytokine production by TLR4^–/– macrophages 24 h following infection with 16M, suggesting a delayed recognition that is consistent with previous studies (6, 44). The impaired production of inflammatory cytokines by macrophages from TLR2^–/– mice after 16M stimulation highlights the fundamental role of TLR2 in Brucella recognition. The significant induction of all three cytokines was observed in GI-2 mutant-stimulated TLR2^–/– BMDMs at 12 h compared to that observed in 16M-stimulated BMDMs despite their similar infection levels. This observation suggests that structural changes in the LPS due to the loss of GI-2 may have facilitated the recognition of bacterial PAMPs by other TLRs. A severe defect in cytokine production after 16M infection was observed for macrophages lacking only TLR2, suggesting that detection by other TLRs is insufficient for a robust host response. A lack of response by other TLRs may be related to the lack of targets and the intracellular lifestyle of Brucella. Brucella degradation leading to the release and TLR9 recognition of CpG DNA is less likely because Brucella bacteria are found in an acidic phagosomal vacuole (30). Further, Brucella bacteria have not been shown to produce flagella, except under artificial conditions (17), making recognition by TLR5 unlikely (22). Our results suggest that Brucella evasion of host immunity during the initial stages of infections and disease development may be related to a lack of integration of multiple TLR-mediated signaling, and a lack of TLR signaling may facilitate the high infectivity of Brucella. A better understanding of the biochemical interactions between Brucella PAMPs and specific TLRs may provide new insights into the complex mechanisms of pathogenesis.

The O antigen of Brucella is a homopolymer of 4,6-dideoxy-4-formamido--D-mannopyranosyl residues linked by either an -1,2 linkage in A antigen-positive strains or an -1,3 linkage in M antigen-positive strains (10, 11). Several genes have been implicated in the synthesis of the Brucella LPS O chain (4, 19, 32); however, the biochemical steps involved in the synthesis of Brucella smooth LPS are not known. Thus, the specific roles of the I0997 and I0998 sugar transferases in the physiology of O-polysaccharide synthesis are presently unclear. In addition, our studies cannot rule out the possibility that I0999 also contributes directly to O chain synthesis. Further studies are needed to determine the specific enzymatic modifications mediated by these sugar transferases and how I0999 participates in the formation of the Brucella O-polysaccharide leading to smooth LPS. Such studies will be highly valuable and would lead to the development of effective vaccine and therapeutic targets for Brucella. B. abortus RB51, which lacks sugar transferase I0998, is an attenuated, stable, spontaneous rough mutant (39) that is currently used as a vaccine for animal brucellosis in the United States and other countries. Therefore, understanding the biosynthetic steps mediated by these proteins in the formation of Brucella smooth LPS would lead to designing novel therapies for brucellosis.

 
This work was supported by the NIH/NIAID RCE for Biodefense and Emerging Infectious Diseases Research grants 1-U54-AI-057153, R01-AI-073558, R21-AI-070229, and BARD-US-3829-06.

We thank Elizabeth Rondon, Ashley Shade, Petra Kohler, and Matthew Frankel for assistance with plasmid constructions.

 
* Corresponding author. Mailing address for Gireesh Rajashekara: Food Animal Health Research Program, Ohio Agricultural Research Development Center, Department of Veterinary Preventive Medicine, The Ohio State University, Wooster, OH 44691. Phone: (330) 263-3745. Fax: (330) 263-3677. E-mail: rajashekara.2{at}osu.edu. Mailing address for Gary A. Splitter: Department of Animal Health and Biomedical Sciences, University of Wisconsin, 1656 Linden Drive, Madison, WI 53706. Phone: (608) 262-1837. Fax: (608) 262-7420. E-mail: splitter{at}svm.vetmed.wisc.edu

Published ahead of print on 18 July 2008.

Supplemental material for this article may be found at http://jb.asm.org/.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1
1. Akira, S., and K. Takeda. 2004. Toll-like receptor signaling. Nat. Rev. Immunol. 4:499-511.[CrossRef][Medline]
2
2. Alaimo, C., I. Catrein, L. Morf, C. L. Marolda, N. Callewaert, M. A. Valvano, M. F. Feldman, and M. Aebi. 2006. Two distinct but interchangeable mechanisms for flipping of lipid-linked oligosaccharides. EMBO J. 25:967-976.[CrossRef][Medline]
3
3. Alexander, C., and E. T. Rietschel. 2001. Bacterial lipopolysaccharides and innate immunity. J. Endotoxin Res. 7:167-202.[CrossRef][Medline]
4
4. Allen, C. A., L. G. Adams, and T. A. Ficht. 1998. Transposon-derived Brucella abortus rough mutants are attenuated and exhibit reduced intracellular survival. Infect. Immun. 66:1008-1016.[Abstract/Free Full Text]
5
5. Baek, S. H., G. Rajashekara, G. A. Splitter, and J. P. Shapleigh. 2004. Denitrification genes regulate Brucella virulence in mice. J. Bacteriol. 186:6025-6031.[Abstract/Free Full Text]
6
6. Barquero-Calvo, E., E. Chaves-Olarte, D. S. Weiss, C. Guzmán-Verri, C. Chacón-Díaz, A. Rucavado, I. Moriyón, and E. Moreno. 2007. Brucella abortus uses a stealthy strategy to avoid activation of the innate immune system during the onset of infection. PLoS ONE 2:e631.[CrossRef]
7
7. Boigegrain, R. A., I. Salhi, M. T. Alvarez-Martinez, J. Machold, Y. Fedon, M. Arpagaus, C. Weise, M. Rittig, and B. Rouot. 2004. Release of periplasmic proteins of Brucella suis upon acidic shock involves the outer membrane protein Omp25. Infect. Immun. 72:5693-5703.[Abstract/Free Full Text]
8
8. Borezee, E., E. Pellegrini, and P. Berche. 2000. OppA of Listeria monocytogenes, an oligopeptide-binding protein required for bacterial growth at low temperature and involved in intracellular survival. Infect. Immun. 68:7069-7077.[Abstract/Free Full Text]
9
9. Braun, W., and A. E. Bonestell. 1947. Independent variation of characteristics in Brucella abortus variants and their detection. Am. J. Vet. Res. 8:386-390.
10
10. Bundle, D. R., J. W. Cherwonogrodzky, and M. B. Perry. 1987. Structural elucidation of the Brucella melitensis M antigen by high-resolution NMR at 500 MHz. Biochemistry 26:8717-8726.[CrossRef][Medline]
11
11. Caroff, M., D. R. Bundle, M. B. Perry, J. W. Cherwonogrodzky, and J. R. Duncan. 1984. Antigenic S-type lipopolysaccharide of Brucella abortus 1119-3. Infect. Immun. 46:384-388.[Abstract/Free Full Text]
12
12. Celli, J., C. de Chastellier, D. M. Franchini, J. Pizarro-Cerda, E. Moreno, and J. P. Gorvel. 2003. Brucella evades macrophage killing via VirB-dependent sustained interactions with the endoplasmic reticulum. J. Exp. Med. 198:545-556.[Abstract/Free Full Text]
13
13. Corbel, M. J. 1997. Brucellosis: an overview. Emerg. Infect. Dis. 3:213-221.[Medline]
14
14. Edmonds, M. D., A. Cloeckaert, S. D. Hagius, L. E. Samartino, W. T. Fulton, J. V. Walker, F. M. Enright, N. J. Booth, and P. H. Elzer. 2002. Pathogenicity and protective activity in pregnant goats of a Brucella melitensis Omp 25 deletion mutant. Res. Vet. Sci. 72:235-239.[CrossRef][Medline]
15
15. Erridge, C., E. Bennett-Guerrero, and I. R. Poxton. 2002. Structure and function of lipopolysaccharides. Microbes Infect. 4:837-851.[CrossRef][Medline]
16
16. Freer, E., J. Pizarro-Cerda, A. Weintraub, J. A. Bengoechea, I. Moriyon, K. Hultenby, J. P. Gorvel, and E. Moreno. 1999. The outer membrane of Brucella ovis shows increased permeability to hydrophobic probes and is more susceptible to cationic peptides than are the outer membranes of mutant rough Brucella abortus strains. Infect. Immun. 67:6181-6186.[Abstract/Free Full Text]
17
17. Fretin, D., A. Fauconnier, S. Köhler, S. Halling, S. Léonard, C. Nijskens, J. Ferooz, P. Lestrate, R. M. Delrue, I. Danese, J. Vandenhaute, A. Tibor, X. DeBolle, and J. J. Letesson. 2005. The sheathed flagellum of Brucella melitensis is involved in persistence in a murine model of infection. Cell. Microbiol. 7:687-698.[CrossRef][Medline]
18
18. Giambartolomei, G. H., A. Zwerdling, J. Cassataro, L. Bruno, C. A. Fossati, and M. T. Philipp. 2004. Lipoproteins, not lipopolysaccharide, are the key mediators of the proinflammatory response elicited by heat-killed Brucella abortus. J. Immunol. 173:4635-4642.[Abstract/Free Full Text]
19
19. Godfroid, F., B. Taminiau, I. Danese, P. Denoel, A. Tibor, V. Weynants, A. Cloeckaert, J. Godfroid, and J.-J. Letesson. 1998. Identification of the perosamine synthetase gene of Brucella melitensis 16M and involvement of lipopolysaccharide O side chain in Brucella survival in mice and in macrophages. Infect. Immun. 66:5485-5493.[Abstract/Free Full Text]
20
20. Gorvel, J. P., and E. Moreno. 2002. Brucella intracellular life: from invasion to intracellular replication. Vet. Microbiol. 90:281-297.[CrossRef][Medline]
21
21. Hacker, J., and J. B. Kaper. 2000. Pathogenicity islands and the evolution of microbes. Annu. Rev. Microbiol. 54:641-679.[CrossRef][Medline]
22
22. Hayashi, F., K. D. Smith, A. Ozinsky, T. R. Hawn, E. C. Yi, D. R. Goodlett, J. K. Eng, S. Akira, D. M. Underhill, and A. Aderem. 2001. The innate immune response to bacterial flagellin is mediated by Toll-like receptor 5. Nature 410:1099-1103.[CrossRef][Medline]
23
23. Jimenez de Bagues, M. P., A. Gross, A. Terraza, and J. Dornand. 2005. Regulation of the mitogen-activated protein kinases by Brucella spp. expressing a smooth and rough phenotype: relationship to pathogen invasiveness. Infect. Immun. 73:3178-3183.[Abstract/Free Full Text]
24
24. Kim, S., M. Watarai, S. Makino, and T. Shirahata. 2002. Membrane sorting during swimming internalization of Brucella is required for phagosome trafficking decisions. Microb. Pathog. 33:225-237.[CrossRef][Medline]
25
25. Kim, S., M. Watarai, H. Suzuki, S. Makino, T. Kodama, and T. Shirahata. 2004. Lipid raft microdomains mediate class A scavenger receptor-dependent infection of Brucella abortus. Microb. Pathog. 37:11-19.[CrossRef][Medline]
26
26. Ko, J., A. Gendron-Fitzpatrick, T. A. Ficht, and G. A. Splitter. 2002. Virulence criteria for Brucella abortus strains as determined by interferon regulatory factor 1-deficient mice. Infect. Immun. 70:7004-7012.[Abstract/Free Full Text]
27
27. Kohler, S., V. Foulongne, S. Ouahrani-Bettache, G. Bourg, J. Teyssier, M. Ramuz, and J. P. Liautard. 2002. The analysis of the intramacrophagic virulome of Brucella suis deciphers the environment encountered by the pathogen inside the macrophage host cell. Proc. Natl. Acad. Sci. USA 99:15711-15716.[Abstract/Free Full Text]
28
28. Kuiper, I., G. V. Bloemberg, S. Noreen, J. E. Thomas-Oates, and B. J. Lugtenberg. 2001. Increased uptake of putrescine in the rhizosphere inhibits competitive root colonization by Pseudomonas fluorescens strain WCS365. Mol. Plant-Microbe Interact. 14:1096-1104.[Medline]
29
29. Lapaque, N., I. Moriyon, E. Moreno, and J. P. Gorvel. 2005. Brucella lipopolysaccharide acts as a virulence factor. Curr. Opin. Microbiol. 8:60-66.[CrossRef][Medline]
30
30. Latz, E., A. Schoenemeyer, A. Visintin, K. A. Fitzgerald, B. G. Monks, C. F. Knetter, E. Lien, N. J. Nilsen, T. Espevik, and D. T. Golenbock. 2004. TLR9 signals after translocating from the ER to CpG DNA in the lysosome. Nat. Immunol. 5:190-198.[CrossRef][Medline]
31
31. Matthysse, A. G., H. A. Yarnall, and N. Young. 1996. Requirement for genes with homology to ABC transport systems for attachment and virulence of Agrobacterium tumefaciens. J. Bacteriol. 178:5302-5308.[Abstract/Free Full Text]
32
32. McQuiston, J. R., R. Vemulapalli, T. J. Inzana, G. G. Schurig, N. Sriranganathan, D. Fritzinger, T. L. Hadfield, R. A. Warren, L. E. Lindler, N. Snellings, D. Hoover, S. M. Halling, and S. M. Boyle. 1999. Genetic characterization of a Tn5-disrupted glycosyltransferase gene homolog in Brucella abortus and its effect on lipopolysaccharide composition and virulence. Infect. Immun. 67:3830-3835.[Abstract/Free Full Text]
33
33. Naroeni, A., and F. Porte. 2002. Role of cholesterol and the ganglioside GM₁ in entry and short-term survival of Brucella suis in murine macrophages. Infect. Immun. 70:1640-1644.[Abstract/Free Full Text]
34
34. Pei, J., and T. A. Ficht. 2004. Brucella abortus rough mutants are cytopathic for macrophages in culture. Infect. Immun. 72:440-450.[Abstract/Free Full Text]
35
35. Porte, F., A. Naroeni, S. Ouahrani-Bettache, and J. P. Liautard. 2003. Role of the Brucella suis lipopolysaccharide O antigen in phagosomal genesis and in inhibition of phagosome-lysosome fusion in murine macrophages. Infect. Immun. 71:1481-1490.[Abstract/Free Full Text]
36
36. Rajashekara, G., J. D. Glasner, D. A. Glover, and G. A. Splitter. 2004. Comparative whole-genome hybridization reveals genomic islands in Brucella species. J. Bacteriol. 186:5040-5051.[Abstract/Free Full Text]
37
37. Rajashekara, G., D. A. Glover, M. Krepps, and G. A. Splitter. 2005. Temporal analysis of pathogenic events in virulent and avirulent Brucella melitensis infections. Cell. Microbiol. 7:1459-1473.[CrossRef][Medline]
38
38. Rittig, M. G., A. Kaufmann, A. Robins, B. Shaw, H. Sprenger, D. Gemsa, V. Foulongne, B. Rouot, and J. Dornand. 2003. Smooth and rough lipopolysaccharide phenotypes of Brucella induce different intracellular trafficking and cytokine/chemokine release in human monocytes. J. Leukoc. Biol. 74:1045-1055.[Abstract/Free Full Text]
39
39. Schurig, G. G., R. M. Roop II, T. Bagchi, S. Boyle, D. Buhrman, and N. Sriranganathan. 1991. Biological properties of RB51; a stable rough strain of Brucella abortus. Vet. Microbiol. 28:171-188.[CrossRef][Medline]
40
40. Takeda, K., T. Kaisho, and S. Akira. 2003. Toll-like receptors. Annu. Rev. Immunol. 21:335-376.[CrossRef][Medline]
41
41. Vemulapalli, R., J. R. McQuiston, G. G. Schurig, N. Sriranganathan, S. M. Halling, and S. M. Boyle. 1999. Identification of an IS711 element interrupting the wboA gene of Brucella abortus vaccine strain RB51 and a PCR assay to distinguish strain RB51 from other Brucella species and strains. Clin. Diagn. Lab. Immunol. 6:760-764.[Medline]
42
42. Vemulapalli, R., Y. He, L. S. Buccolo, S. M. Boyle, N. Sriranganathan, and G. G. Schurig. 2000. Complementation of Brucella abortus RB51 with a functional wboA gene results in O-antigen synthesis and enhanced vaccine efficacy but no change in rough phenotype and attenuation. Infect. Immun. 68:3927-3932.[Abstract/Free Full Text]
43
43. Vizcaíno, N., P. Caro-Hernández, A. Cloeckaert, and L. Fernández-Lago. 2004. DNA polymorphism in the omp25/omp31 family of Brucella spp.: identification of a 1.7-kb inversion in Brucella cetaceae and of a 15.1-kb genomic island, absent from Brucella ovis, related to the synthesis of smooth lipopolysaccharide. Microbes Infect. 6:821-834.[CrossRef][Medline]
44
44. Weiss, D. S., K. Takeda, S. Akira, A. Zychlinsky, and E. Moreno. 2005. MyD88, but not Toll-like receptors 4 and 2, is required for efficient clearance of Brucella abortus. Infect. Immun. 73:5137-5143.[Abstract/Free Full Text]
45
45. Young, E. J. 1995. An overview of human brucellosis. Clin. Infect. Dis. 21:283-289.[Medline]

---

Journal of Bacteriology, September 2008, p. 6243-6252, Vol. 190, No. 18
0021-9193/08/$08.00+0     doi:10.1128/JB.00520-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

This article has been cited by other articles:

• Wattam, A. R., Williams, K. P., Snyder, E. E., Almeida, N. F. Jr., Shukla, M., Dickerman, A. W., Crasta, O. R., Kenyon, R., Lu, J., Shallom, J. M., Yoo, H., Ficht, T. A., Tsolis, R. M., Munk, C., Tapia, R., Han, C. S., Detter, J. C., Bruce, D., Brettin, T. S., Sobral, B. W., Boyle, S. M., Setubal, J. C. (2009). Analysis of Ten Brucella Genomes Reveals Evidence for Horizontal Gene Transfer Despite a Preferred Intracellular Lifestyle. J. Bacteriol. 191: 3569-3579 [Abstract] [Full Text]  

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Rajashekara, G. Articles by Splitter, G. Search for Related Content PubMed PubMed Citation Articles by Rajashekara, G. Articles by Splitter, G.