Release of Periplasmic Proteins of Brucella suis upon Acidic Shock ...

INFECTION AND IMMUNITY, Oct. 2004, p. 5693–5703 0019-9567/04/$08.00⫹0 DOI: 10.1128/IAI.72.10.5693–5703.2004 Copyright © 2004, American Society for Microbiology. All Rights Reserved.

Vol. 72, No. 10

Release of Periplasmic Proteins of Brucella suis upon Acidic Shock Involves the Outer Membrane Protein Omp25 Rose-Anne Boigegrain,1 Imed Salhi,1† Maria-Teresa Alvarez-Martinez,1 Jan Machold,1‡ Yann Fedon,2 Martine Arpagaus,2 Christoph Weise,3 Michael Rittig,4 and Bruno Rouot1* INSERM U4311 and EMIP INRA 1133,2 Universite´ de Montpellier 2, Montpellier, France; Institut fu ¨r Chemie-Biochemie Freie Universita ¨t Berlin, Berlin, Germany3; and School of Biomedical Sciences, Queen’s Medical Centre, Nottingham, United Kingdom4 Received 10 February 2004/Returned for modification 5 April 2004/Accepted 7 June 2004

The survival and replication of Brucella in macrophages is initially triggered by a low intraphagosomal pH. In order to identify proteins released by Brucella during this early acidification step, we analyzed Brucella suis conditioned medium at various pH levels. No significant proteins were released at pH 4.0 in minimal medium or citrate buffer, whereas in acetate buffer, B. suis released a substantial amount of soluble proteins. Comparison of 13 N-terminal amino acid sequences determined by Edman degradation with their corresponding genomic sequences revealed that all of these proteins possessed a signal peptide indicative of their periplasmic location. Ten proteins are putative substrate binding proteins, including a homologue of the nopaline binding protein of Agrobacterium tumefaciens. The absence of this homologue in Brucella melitensis was due to the deletion of a 7.7-kb DNA fragment in its genome. We also characterized for the first time a hypothetical 9.8-kDa basic protein composed of five amino acid repeats. In B. suis, this protein contained 9 repeats, while 12 were present in the B. melitensis orthologue. B. suis in acetate buffer depended on neither the virB type IV secretory system nor the omp31 gene product. However, the integrity of the omp25 gene was required for release at acidic pH, while the absence of omp25b or omp25c displayed smaller effects. Together, these results suggest that Omp25 is involved in the membrane permeability of Brucella in acidic medium. Bacteria of the genus Brucella are gram-negative facultative intracellular pathogens of various wild and domestic mammals and are able to cause severe zoonotic infections in humans. Traditionally, three major species are distinguished by their predilections for certain animal hosts: Brucella abortus for cattle, Brucella melitensis for caprines, and Brucella suis for hogs. Whereas B. abortus is the livestock pathogen with the greatest economic impact, B. melitensis and B. suis account for most clinical cases in humans (1, 2, 11). To evade host defenses, Brucella can inhibit neutrophil degranulation and block tumor necrosis factor (TNF) production by macrophages. It has been shown that membrane integrity, in terms of both smooth lipopolysaccharide and outer membrane proteins, is required for such virulent behavior. Furthermore, studies using transposon or signature-tagged mutagenesis have unraveled, with respect to Brucella virulence, the crucial role of an operon homologous to the virB operon of Agrobacterium tumefaciens encoding a type IV secretion system (16, 21, 28). The virulence regulon of A. tumefaciens is triggered in response to chemical signals released at the plant wound site, such as acetosyringone and low pH. Type IV secretion system production is potentiated by monosaccharides (galactose and

arabinose) through binding to the periplasmic multiple sugar binding protein ChvE, as well as by low pH (6). However, it was found that under neutral conditions, this secretory system is already produced in B. melitensis or B. abortus, while it is undetectable in B. suis and Brucella canis (33). Moreover, a gene homologue of the chvE gene of A. tumefaciens was cloned in B. suis, and it was shown that, although its gene product is specifically required for D-(⫹)-galactose utilization in B. suis, it is not required for intracellular survival and multiplication within macrophage-like cells (3). A. tumefaciens strains differ in their abilities to induce in plants the production of one defined opine (nopaline, argopine, leucinopine, etc.), which is then taken up as a nutrient by a specific periplasmic binding protein. For example, Agrobacterium sp. strain C58, which triggers nopaline production in plants, utilizes the nopaline binding protein for its cellular uptake. As for Brucella, Rocha et al. suggested that a ribose binding protein linked to the outer membrane might be involved in virulence, since it participates in the attachment of Brucella to sialic acid residues of eukaryotic cells (35). The question of whether the various Brucella species possess one or several homologues of these specific Agrobacterium nutrient receptors has not yet been specifically addressed. Concerning the importance of the virB operon for Brucella virulence, reports have indicated that a complete Brucella virB operon is required for virulence in mice (36) and for replication in macrophage-like cells (43, 44) or HeLa cells (10, 12). Carefully performed intracellular pH measurements using pHsensitive fluorochromes in confocal microscopy showed that internalized B. suis organisms were located in acidic compart-

* Corresponding author. Present address: CNRS UMR, CC107, Universite´ Montpellier II, 34095 Montpellier Cedex 05, France. Phone: (33) 467 144 725. Fax: (33) 467 144 727. E-mail: rouot@crit .univ-montp2.fr. † Present address: INSERM, EMI 0227, CRLC Val d’Aurelle-Paul Lamarque, F-34298 Montpellier Cedex 5, France. ‡ Present address: Amersham Pharmacia Biotech, D-79111 Freiburg, Germany. 5693

5694

BOIGEGRAIN ET AL.

ments of host macrophages (29, 30). When lysosomotropic agents were used soon after phagocytosis to raise the intraphagosomal pH, subsequent CFU numbers were reduced, whereas a later rise in pH did not affect the intracellular survival of the brucellae. This suggests that an acidic intraphagosomal pH is a crucial initial trigger for successful intracellular parasitism, even for B. suis, which already produces proteins encoded by the virB operon. However, acidic shock of brucellae prior to phagocytosis did not improve intracellular survival (30), indicating that changes induced by acidic pH need to take place in the right environment in order to be effective. Even so, CFU counts early after infection have shown that very few brucellae survive the first hours within the host cell (30). As it is now becoming clear that many intracellular pathogens need to export regulatory bacterial proteins, it was of interest to investigate whether, in a nutrient-deprived acidic medium, Brucella could release specific entities that are involved in its early survival mechanism. Although the lipopolysaccharide composition is known to contribute to Brucella virulence, TNF-␣ inhibition was attributed to the well-studied 25-kDa outer membrane protein (Omp25) (22). Furthermore, B. melitensis, B. abortus, and B. ovis omp25 knockout mutants are attenuated in animal models of infection and also provide levels of protection similar to or better than the currently used attenuated vaccine strain B. melitensis Rev. 1 (13–15). The idea that the Omp25 protein is a virulence factor is indirectly supported by studies of B. abortus mutants in which insertion of the Tn5 transposon in either gene of the two-component bvrR/bvrS system leads to attenuated strains (20). It was found that the strains were more susceptible to bactericidal polycationic substances due to alteration of the outer membrane integrity (37). In fact, impairment of the bvrR/bvrS system in addition to modification of lipid A hinders production of at least two members of the Omp25/Omp31 family, namely, Omp25 and Omp22 (previously called Omp3b) (20). It was suggested that attenuation of the virulence of the bvrR/bvrS mutants also resulted from the absence of Omp25 (27). Interestingly, an analysis of the B. suis and B. melitensis genomes revealed that Brucella species possess seven omp25-related genes (20), whose products have been classified in four groups: Omp25 itself; a group closely related to Omp25, composed of Omp25b, Omp25c, and Omp25d; the Omp31 group, including Omp31 and Omp31b; and finally, a less closely related protein, Omp22 (33). All of the corresponding gene products were identified in either B. suis or B. abortus, but it is presently not known whether the functions of the proteins within a subgroup are the same or whether there are slight differences. Thus, the aim of the present study was to identify the virulence factor(s) of Brucella responsible for its survival within the acidic phagosomal environment. We therefore analyzed the proteome of proteins released by B. suis in various media upon incubation at different pH levels. Bacterial mutants with the virB operon or the omp25/omp31 gene family impaired were included in this study in order to gain further insight into the mechanism by which proteins could be released by Brucella in acidic medium. MATERIALS AND METHODS Bacterial strains and growth conditions. All strains were grown for 16 to 24 h at 37°C to the stationary phase in tryptic soy (TS) broth for the various Brucella

INFECT. IMMUN. TABLE 1. Bacterial strains used in this study Strain

B. suis 1330 1330 omp25 mutant 1330 omp25b mutant 1330 omp25c mutant 1330 ⌬omp25d mutant 1330 omp25 mutant/ omp25 1330 omp31 mutant

Relevant genotype and/or description

Reference or sourcea

Biotype 1; ATCC 23444 B. suis ⌬omp25::kan B. suis omp25b::kan

ATCC 22 35

B. suis omp25c::kan

35

B. suis omp25d::kan

35

B. suis ⌬omp25/omp25⫹::kan/Cm B. suis omp31::Cm

22

Biotype 1; ATCC 23456T

ATCC

Biotype 1; wild type, smooth, virulent Biotype 1; ATCC 23448T Biotype 3; ATCC 23450T

Laboratory stock ATCC ATCC

B. canis

ATCC 23365T

ATCC

B. ovis

ATCC 25840 (Reo 198)

ATCC

B. melitensis 16M B. abortus 2308 A1 A3

22

a

Due to antiterrorism measures, virulent brucellae might no longer be available from the American Type Culture Collection (ATCC).

species (Table 1) or Luria-Bertani broth for Escherichia coli. For the growth of B. suis 1330-derived mutants with the omp31 or omp25 gene impaired and the complemented omp25 mutant strain, TS broth was supplemented with kanamycin and/or chloramphenicol at 50 and 25 ␮g ml⫺1, respectively. Preparation of B. suis conditioned medium. Five milliliters of TS medium primary culture was inoculated with the different bacterial species and incubated at 37°C for 16 h (optical density at 600 nm, ⬃1.5). The secondary culture was prepared by a 200-fold dilution of the primary cultures in 1 liter of TS medium. After incubation at 37°C for 24 h with shaking, the bacterial cells were sedimented at 9,500 ⫻ g for 6 min at 4°C. The supernatants were discarded, and the pellets were resuspended, washed in 400 ml of phosphate-buffered saline (PBS), and centrifuged as before. The resulting pellets were resuspended in 400 ml of 25 mM ammonium acetate (pH 4) containing 120 mM NaCl or other buffers (PBS or RPMI 1640 [Life Technology], minimal medium A [5], or 25 mM sodium citrate [pH 4] containing 120 mM NaCl) and incubated at 37°C for 150 min with shaking. After centrifugation at 9,500 ⫻ g for 15 min at 4°C, the supernatant (supernatant 1 [S1]; 400 ml) was filtered through a 0.22-␮m-pore-size filter (Steritop; Millipore) and concentrated to 20 ml (S2) by ultrafiltration (Amicon concentration cell) through a 10-kDa-cutoff membrane. The volume was further reduced to ⬍1 ml (S3) using a 10-kDa-cutoff Centriprep device (Amicon) by centrifugation at 2,200 ⫻ g (4°C). About 120 to 150 ␮g of protein was usually obtained from 1 liter of B. suis 1330 culture, as measured by the method described by Bradford using bovine serum albumin as a standard. For some experiments, the S3 fraction was ultracentrifuged with an airfuge (Beckman) at room temperature at 126,000 ⫻ g (20 lb/in2) for 30 min, leading to the S4 soluble fraction and the pellets. Electron microscopy preparations. The ultracentrifuged pellets obtained concomitantly with the S4 fractions were fixed in a precooled mixture of 2.5% glutaraldehyde and 1% osmium tetroxide in 0.1 M cacodylate buffer, pH 7.4, on ice for 30 min, washed twice with cold 0.9% saline solution for 30 min each, transferred to warm 2% low-melting-point agarose (Gibco-BRL), and immediately centrifuged at 120,000 ⫻ g (in an airfuge) for 10 min. Having solidified the agar on ice, the part of the gel containing the pellet was removed and processed for electron microscopy via the following steps: (i) “en bloc” staining with an alcoholic mixture of 0.5% phosphotungstic acid and 0.25% uranyl acetate, (ii) physical dehydration with a graded series of alcohol solutions followed by pure acetone, and (iii) embedding in Epon 812 resin. Ultrathin sections cut from the Epon blocks were stained on a grid with a mixture of 10% uranyl acetate and 2% lead citrate and viewed with a Zeiss type 906 transmission electron microscope. Peripheral blood monocytes from buffy coats were infected with B. suis cells and processed for electron microscopy as previously described in detail (30, 31).

VOL. 72, 2004

B. SUIS RELEASES PERIPLASMIC PROTEINS AT ACIDIC pH

Ultrathin sections were screened for the presence of material consistent with bacterial outer membrane blebs. Western blot analysis. Bacterial lysates were prepared by resuspension of the various Brucella strain pellets directly in Laemmli sample buffer and being heated at 100°C for 15 min. For supernatants, samples containing 2 to 5 ␮g of protein were used. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) was performed as described by Laemmli (26), using a 12.5 or 15% (wt/vol) separating gel. The proteins were transferred onto Immobilon polyvinylidene difluoride membranes (Millipore) using a semidry transfer procedure and stained with Coomassie blue. Immunodetection in total cell lysates and in the different soluble fractions was performed with monoclonal antisera (used at 1/10,000) raised against Omp25 (A19 12B10 F4 or A59 5F1 C9) and Omp31 (A59 10F9 G10) and with polyclonal antisera (used at 1/5,000) raised against Rib and Nop. Horseradish peroxidase-conjugated goat anti-mouse and anti-rabbit antibodies (Jackson Immunoresearch Laboratories) were used, along with the ECL system (Amersham Pharmacia Biotech, Orsay, France), to develop blots for chemiluminescence before visualization on film (Kodak X-AR) Chromatographic separations of the supernatant proteins for microsequencing. Initial separation by gel filtration chromatography of the S3 or S4 fraction (40 ␮g of protein per run) was performed on an SEC 3000 column (0.75 by 30 cm; Beckman) with 100 mM sodium phosphate buffer, pH 7.5, containing 0.3 M NaCl, 0.3% CHAPS {3-[(-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}, and 0.02% NaN3 at a flow rate of 1 ml per min; UV absorption was monitored at 280 nm. Two protein fractions were collected, GF-A (mass ⬎ 1,000 kDa) and GF-B (70 ⬎ mass ⬎ 15 kDa). For subsequent ion-exchange chromatography fractionation, the GF-B protein fraction was applied directly to a DEAE anion exchanger (0.75 by 7 cm) column (TSK-5PW; Beckman) equilibrated with 10 mM sodium phosphate buffer, pH 8.0. Elution was carried out at 1 ml/min with a 0 to 0.5 M NaCl gradient. Fifteen fractions were collected. The DEAE fractions obtained, starting from a 4-liter initial bacterial supernatant preparation, were pooled and precipitated with chloroform-methanol and then separated on an SDS–15% PAGE gel. The proteins were blotted onto a polyvinylidene difluoride membrane (Immobilon-PSeq; Millipore) and stained with Coomassie blue, and the 14 major spots were subjected to Edman degradation (Table 2). Protein purification and antibody preparation. The protein purification and antibody preparation step was performed for purification of the major proteins for antibody production. DEAE high-performance liquid chromatography (HPLC) fractions were applied directly to a Vydac C4 reverse-phase HPLC column equilibrated with 10% (vol/vol) acetonitrile in water containing 0.1% trifluoroacetic acid. The proteins were eluted in an increasing linear gradient of acetonitrile (10 to 50%, with a slope of 1% per min) at a flow rate of 1 ml per min, and UV absorption was monitored at 215 nm. Fractions corresponding to proteins 3 (Rib) and 4 (Nop) were collected manually directly at the detector outlet. Corresponding fractions from different separation runs were pooled, and the acetonitrile was evaporated in a vacuum centrifuge (Speed-Vac; Savant). The pH of the fractions was adjusted to 7.4 for injection into New Zealand White rabbits, and antibody production was carried out as previously described (34). Sera were subsequently collected and stored frozen. Cloning of the major soluble ribose binding protein. Trypsinolysis of purified protein 3 was performed overnight at 37°C, with the use of L-(tosylamido-2phenyl)ethylchloromethyl ketone-treated trypsin (Sigma) at a 1:50 trypsin/substrate ratio. Proteolysis was terminated by injecting the mixture directly onto a Beckman C18 reverse-phase HPLC column (0.46 by 25 cm) equilibrated with 2% acetonitrile in water containing 0.1% trifluoroacetic acid. Tryptic peptides were eluted from the column by increasing the acetonitrile concentration in a shallow gradient from 2 to 32% (0.6 ml per min). The peptides were collected immediately at the detector outlet, dried under vacuum, and subjected to Edman degradation. The degenerated oligomers 5⬘-ATGACIAT(A/T/C)AT(A/T/C)GTIA A(C/T)GA(C/T)CC-3⬘ and 5⬘-CG(A/G)TCCCA(A/G)TT(A/G/C/T)GC (A/ G/C/A)AC(A/G)TC-3⬘ were deduced from the N-terminal sequence (MT IIVND [Table 2]) and one of the internal sequences (DVANWDR), respectively. A 500-bp DNA fragment was amplified by PCR from B. suis DNA with these oligonucleotides. B. suis genomic DNA was digested with various restriction enzymes, electrophoresed in a 1% agarose gel, and transferred to a nylon membrane. The blot was then hybridized with the 500-bp PCR product labeled with digoxigenin. Under mild conditions (55°C), the probe bound to a 1.6-kb HindIII fragment, which was purified from agarose gel and cloned into the plasmid Bluescript SK(⫹/⫺) (Stratagene). This fragment was sequenced using the ABI dye terminator kit and a model 373 Applied Biosystems DNA sequencer and corresponded to protein 3.

5695

RESULTS Release of B. suis proteins in acidic medium. To examine the ability of Brucella to release selective proteins in phagosomelike acidic medium, B. suis organisms were incubated for 2.5 h at 37°C under various culture conditions and at different pH levels. The supernatants were concentrated and analyzed by SDS-PAGE (Fig. 1A). About 150 ␮g of protein per liter of stationary-phase initial culture was obtained when the bacteria were incubated at pH 4.0 in 25 mM either ammonium acetate (Fig. 1A, lane 2) or sodium acetate (not shown) buffer containing 120 mM NaCl. The absence in the electrophoretic pattern of the acidic supernatant of proteins in the range of 20 to 65 kDa compared to that of the crude B. suis lysate (Fig. 1A, lane 1) suggested that the released proteins did not simply originate from bacterial lysis. This conclusion was further confirmed with a monoclonal antibody (A66 5H1 E9) directed against a cytoplasmic p17 protein, which detected such a protein in the lysate but not in the released proteins (data not shown). At neutral pH, the amount of proteins released by B. suis was small in the eukaryotic cell culture medium RPMI 1640 and even smaller in PBS (Fig. 1, lanes 3 and 4) or saline Tris-HCl buffer at pH 7.5 (data not shown). Since the outermembrane-bound proteins from various Brucella strains were reported to be present in culture medium (4), we used Western blotting to investigate the Omp25 contents in the various incubation supernatants. We found that the amount of Omp25 was also much greater in acidic medium than in PBS buffer or RPMI medium (Fig. 1B). The immunostaining detected in lanes 2 and 4, with an apparent molecular mass of ⬎31 kDa, most probably corresponded to the Omp25 paralogue previously identified and named Omp25b (35). Evidence for B. suis blebbing in acetate buffer at pH 4.0. To characterize the proteome of the acidic conditioned medium of B. suis, the released proteins were separated by gel filtration chromatography (Fig. 2A, trace a). The results indicated the presence of two distinct sets of proteins, i.e., a fraction of large entities (GF-A) with molecular masses of ⬎1,000 kDa and a second fraction (GF-B) that corresponded to proteins with apparent molecular masses between 10 and 50 kDa. The B. suis acidic conditioned medium was subjected to ultracentrifugation to assess the relative compositions of the GF-A and GF-B fractions. The chromatogram of the resulting supernatant indicated that the GF-A fraction was reduced by ⬃50% while almost all the proteins of the GF-B fraction remained in solution (Fig. 2A, trace b). Conversely, the chromatogram of the resuspended pellet (trace c) indicated a dramatic reduction in the quantity of GF-B proteins while it still contained half the starting amount of GF-A. These data suggested that the GF-B fraction was composed mainly of soluble proteins, whereas high-molecular-mass entities present in GF-A might contain membrane-bound proteins or insoluble material. Further separation by SDS-PAGE revealed that the lower-molecular-mass GF-B fraction (Fig. 2B, lane 3) contained most of the proteins detected by Coomassie blue staining in the acidic supernatant (lane 1), whereas no proteins were stained in the GF-A fraction (lane 2). However, immunoblotting analysis using antiOmp25 and -Omp31 antibodies (Fig. 2C) (7) revealed that these major outer membrane proteins were present in the

14 13

RDQIQIAGSSTVLPYAKIV

SEITIKLPDSV EDAMKTDHMKPDA

11

12 13

Mass

10.4 9.8

34.9

45.7

56.3

44.6

34.4

35.8 30.6 30.8 26.5 41.1

16.1

Th.

17/25 27

24

23

21

25

23

21 25 25 22/37g 29

19

Signal peptidea

0937 0858 0632 0960 1193

I 0440 II 0020

I 2138

II 0304

II 0538

II 0655

II 0265

II II II II II

II 0703

B. suis

I 1494 II 0073

I 1989

II 0945

II 0734

II 0625

II 0983

II 0338 II 0103

II 359/360 II 0435

II 0581

B. melitensis

B. melitensis homologued (chromosome and no.)

Hypothetical Hypothetical

Phosphate

Maltose

Oligopeptide

Glycerol-3-phosphate

D-Galactose

Multisugar D-Ribose Arginine/ornithine ABC transporter Leu/Ile/Val

Cu/Zn SOD

Putative function

e

Salmonella enterica serovar Typhimurium A. tumefaciens Y. pestis Mesorhizobium loti Mesorhizobium loti Pyrobaculum aerophilum Streptococcus coelicolor Streptococcus meliloti Streptococcus meliloti Streptococcus meliloti Streptococcus meliloti A. tumefaciens A. tumefaciens

Species

46 50

75

78

47

88

37

80 80 62 64 30

62

Identity (%)

Other bacterial homologuef

b

Amino acids in italics, questionable after microsequencing data, were deduced from the TIGR B. suis genome. Experimental (Exp.) or theoretical (Th.) molecular mass in kilodaltons. c Amino- acid number of the putative signal sequence. d The amino acid identity between B. suis and B. melitensis ranged from 99 to 100%, except for Nop (protein 4) (50%), Rib (protein 3) (97%), and the repeat-containing protein 13 (88%). The roman numerical I or II before the gene number refers to the large or small chromosome, respectively. (In the B. suis genome, the gene numbers of the large and small chromosomes are normally preceded by BR or BRA, respectively). e Putative function or substrate of the binding protein. f Best identity with homologues of other bacteria after tBLASTN search. g Since the first methionine is 37 amino acids upstream of the N terminus of the mature protein, CTG might also be an initiation codon, which would then lead to a 22-amino-acid-long signal sequence.

a

34

VEISIAANSTGDNI

10

45

39 50

QTELAWWHGMTGANNEM

32

EAVLNRGNDTD

EVKDAVVGFLMPDQASTRYEERDYPGFAA

7

35 31 30 28 40

8

EDKGLVAVAMPTKSSA EGLMTIIVNDPSNPYWFTEGEVAKKTAEGLGY KEWKEIRIASEGAYPPFNYM EDKTIKVGIMGGEDEDV EDVITLGASVQLSGPVANT

2/ChvE 3/Rib 4/Nop 5 6

18

Exp

b

B. suis protein characteristics

9

ESTTVKMYEALPTGPGKEV

N-terminal sequence

1

No.

a

TABLE 2. Brucella suis proteins released in acidic supernatant

5696 BOIGEGRAIN ET AL. INFECT. IMMUN.

VOL. 72, 2004


5697

FIG. 1. Release of proteins into bacterial supernatant increases in acidic acetate buffer (pH 4). Wild-type B. suis 1330 cells were incubated for 150 min in various media before the bacteria were centrifuged and the different supernatants were concentrated 1,000-fold. Brucella proteins from the lysate or the concentrated supernatants were separated by SDS-PAGE and stained with Coomassie blue (A) or blotted and immunostained with an anti-Omp25 antibody (B). Lanes: 1, bacterial lysate; 2, supernatants composed of ammonium acetate containing 120 mM NaCl, pH 4; 3, PBS, pH 7; 4, RPMI, pH 7.5; Markers, molecular mass markers.

GF-A fraction and in the pellet (Fig. 2C, lanes 2 and 4) but not in the GF-B soluble fraction (lane 3). These proteins were also present at neutral pH, showing that the release of outer membrane proteins containing structures (membrane blebs) is spontaneous and independent of the pH of the medium (data not shown). We investigated by transmission electron microscopy the contents of a pellet obtained after high-speed centrifugation of the Brucella supernatant in order to determine whether the GF-A fraction was composed of aggregated hydrophobic proteins or outer membrane proteins bound to membrane lipid fragments or vesicles. The electron microscopy pictures (Fig. 3A) show rounded vesicles ⬃80 nm in diameter, compatible with outer membrane blebbing from a Brucella cell ⬃1 ␮m in length. Structures of similar appearance were found in the close vicinity of bacteria during monocyte challenge, strongly suggesting that blebbing does not occur only under specific bacterial culture conditions (Fig. 3B). Our data implied that the blebs, which are likely composed of bacterial outer membranes including the Omp31 and Omp25 proteins, did not contain detectable amounts of soluble proteins, since no GF-B fraction entities were detected by Coomassie blue staining. Soluble proteins released by B. suis in acidic acetate buffer are of periplasmic origin. Proteins present in the acidic-supernatant GF-B fraction were further purified using anion-exchange chromatography (data not shown) and were microsequenced by Edman degradation after SDS-PAGE separation. Among the 13 N-terminal sequences obtained (Table 2), only that of protein 1 (18 kDa) exactly matched the sequence of periplasmic Cu-Zn superoxide dismutase (SOD) cloned from B. abortus (8) prior to systematic genome sequencing. When the amino acid microsequences were BLAST searched against

FIG. 2. Two groups of proteins (GF-A and GF-B) were separated by gel filtration of wild-type B. suis acidic supernatant. Proteins from wild-type Brucella incubated in ammonium acetate containing 120 mM NaCl, pH 4 (supernatant 1 [SN1]) were concentrated (SN3) and subjected to ultracentrifugation at 126,000 ⫻ g. Proteins from the various fractions were loaded into a gel filtration column, and separation was monitored at 280 nm and quantified as arbitrary units ⫻ 10⫺3 (mAu) (A) Initial supernatant (SN3) (trace a), ultracentrifuged supernatant (SN4) (trace b), and the resuspended pellet resulting from the ultracentrifugation (trace c). Proteins from the GF-A fraction had an apparent molecular mass of ⬎2,000 kDa, while proteins of the GF-B group exhibited an apparent molecular mass of ⬍100 kDa. (B and C) Proteins from the various fractions were further separated by SDSPAGE and stained with Coomassie blue (B) or subjected to immunoblotting with monoclonal antibodies directed against Omp31 and Omp25 (C). Lanes: M, molecular mass markers; 1, initial bacterial supernatant (SN3); 2, GF-A fraction; 3, GF-B fraction; 4, resuspended pellet.

the B. suis genome from The Institute for Genomic Research (TIGR), they all matched B. suis genes encoding proteins. An interesting finding was that all of the microsequenced N-terminal fragments were not situated at the very beginning of the open reading frame (ORF) but 19 to 29 amino acids downstream of the putative initial methionine (Table 2). This analysis revealed that the identified proteins possessed a peptide signal which was cleaved off during their translocation through the inner bacterial membrane and further reinforced the as-

5698

BOIGEGRAIN ET AL.

INFECT. IMMUN.

FIG. 3. Electron micrographs reveal the production of Brucella blebs in various media and within phagosomes. (A) Vesicles present in the pellet obtained after ultracentrifugation of the GF-A fraction. (B) Spontaneous blebbing of wild-type B. suis in RPMI medium used for macrophage infections.

sumption that all of them originated from the periplasmic space and not from the cytoplasm. Accordingly, the apparent molecular masses of the released proteins, as determined by SDS-PAGE, closely fitted the theoretical masses of the truncated protein (Table 2, compare experimental and theoretical masses). Comparison of the full-length sequence with that of the N-terminal sequence depicted in Table 2 indicated that the cleavage occurred after an AXA consensus sequence, except for proteins 4 and 5, in which the similarity between the sequences preceding the cleavage site was APXXA. The identities between B. suis amino acid sequences that released mature proteins and their corresponding B. melitensis homologues ranged from 98 to 100%, with the notable exception of proteins 3, 4, and 13, which displayed only 92, 50, and 88% identity, respectively (Table 2). This finding indicated that most proteins produced in B. suis are also potentially expressed in B. melitensis. Compared to other microbial genomes (Table 2, other bacterial homologues), the percent amino acid identities of the whole mature amino acid sequence deduced from the TIGR B. suis genome ranged from 37 to 88%. As expected, higher similarities were generally obtained with ␣-proteobacteria, either plant pathogens (A. tumefaciens) or plant symbionts (Mezorhizobium and Sinorhizobium), or with ␤-proteobacteria (Burkholderia). Most B. suis protein orthologues are putatively sugar, amino acid, peptide, or phosphate binding proteins, with the exception of SOD (protein 1) and proteins 12 and 13 of unknown function. The fact that their closest homologues in other species are substrate binding proteins known to be periplasmically located further agreed with the periplasmic origin of the B. suis proteins released under acidic conditions (Table 2, putative function). The released mature Brucella proteins were mainly acidic. Their theoretical isoelectric points (pIs) deduced from the B. suis genome thus ranged from 4.65 to 5.34 for the substrate binding proteins and protein 12, while that for SOD was 6.11 (data not shown). One striking exception was the very basic theoretical pI of 9.44 for the highly charged protein 13. This might explain why its B. melitensis orthologue, unlike that of protein 12, could not be detected by two-dimensional gel electrophoresis proteomic analysis (42).

FIG. 4. Comparison of the constitutive production of Rib- and Nop-like periplasmic proteins in various Brucella species. After overnight culture, the different Brucella species were centrifuged and the resulting bacterial pellets were lysed in Laemmli electrophoretic sample buffer. Brucella proteins were separated by SDS-PAGE and blotted, and the Nop (A) and Rib (B) proteins were detected using our rabbit polyclonal antibodies.

Characterization of B. suis nopaline-like (Nop) and D-riboselike (Rib) binding proteins. The most abundant proteins obtained in the B. suis acidic supernatant were those referred to as 1, 3, and 4. Since the last two proteins were newly described Brucella proteins, they were further characterized in various Brucella species by using specific antibodies raised against purified proteins. Protein 4 exhibited 52% identity with the nopaline binding protein of A. tumefaciens involved in the virulence process of this plant pathogen and is thus also referred to as Nop here. Surprisingly, the highest identity between protein 4 and proteins deduced from the B. melitensis genome was only 50%, while identities between B. suis and B. melitensis orthologues were generally as high as 98%. This very low identity within the Brucella genus suggested that B. melitensis did not contain any orthologue of the gene encoding a homologue of a periplasmic binding protein for nopaline present and expressed in B. suis. Examination of Nop immunoreactivity in the various Brucella species (Fig. 4A) indicated that the Nop protein was also produced in B. canis, in addition to B. suis, but was not detected in B. melitensis or in the various B. abortus or B. ovis strains tested under similar culture conditions. Comparison of completed Brucella genomes indicated that a DNA fragment of 7,737 bp containing the nop region of B. suis was missing in B. melitensis. Furthermore, this DNA fragment could not be found in other parts of the B. melitensis genome. In addition to loss of the nopaline-like gene, deletion of this DNA fragment also suppressed the ability of B. melitensis to express at full length six putative genes of B. suis, as indicated in Fig. 5A. However, B. melitensis could theoretically express a

VOL. 72, 2004


5699

FIG. 5. Comparison of schematic diagram of ORF region encoding B. suis Nop protein with that of corresponding B. melitensis DNA sequence. (A) In B. melitensis, deletion of the 7,737-bp fragment present in B. suis resulted in the loss of five full-length putative genes (BRA0631 to BRA0635) and truncated two genes (BRA0630 and BRA0636) located at the extremities of the deleted DNA fragment. (B) Close-up view of nucleotide sequences within which deletion occurred. Note that the unique B. melitensis sequence could be derived from the B. suis sequence by deletion of a 7,737-bp DNA fragment, which could occur at several different positions provided that the deleted part included the extra G nucleotide (-ccttG-).

chimeric protein, annotated as BME II0646, which is composed of the beginning of BRA 0636 and the end of BRA 0630. Interestingly, the extremities of the deleted B. melitensis fragment originated from two almost identical 14-nucleotide stretches except, in the second one, for an additional guanyl nucleotide after the first 4 nucleotides (Fig. 5B, top). This suggested that the deletion occurred during bacterial replication, where the transcription shifted from the first to the second conserved sequence. A similar hypothesis has been put forward to explain the deletion in B. abortus of the omp31containing DNA fragment present in other species between two identical 4-nucleotide stretches (40, 41). PCR analysis using primers (within and outside the deleted region in B. melitensis) on genomic DNAs of B. abortus A1 and A3 revealed that in these strains no deletion similar to that in B. melitensis has occurred (Y. Fedon, unpublished results). Protein 3 of B. suis is a homologue of a Yersinia pestis ribose binding protein and therefore is also referred to as Rib here. The first 32 amino acids of the mature Rib obtained by microsequencing perfectly matched the corresponding sequence deduced from the B. suis genome. Moreover, our previous microsequencing of four tryptic Rib fragments, which allowed us to clone the corresponding B. suis gene (see Materials and Methods), confirmed the whole rib nucleotide sequence of B. suis from TIGR. The B. melitensis orthologue deduced from the genome sequencing differed totally in the N-terminal part of the protein. This apparent frameshift most probably resulted from the omission of 1 nucleotide in the sequencing process. The electrophoretic mobility was slightly enhanced in the B. melitensis homologue compared to those of B. suis and B. canis. This higher mobility was very similar to those of proteins of various B. abortus strains and might have simply resulted from two point mutations (E119K and T303N).

Characterization of two previously unidentified proteins released by B. suis which have homologues in the genomes of various organisms. B. suis proteins 12 and 13 and their homologues in other bacteria were annotated as hypothetical proteins, and their functions were indeed unknown. BLAST analysis revealed the existence of homologues of protein 12 with similar molecular weights and with amino acid identities above 45% in A. tumefaciens, Pseudomonas aeruginosa, and Ralstonia solanacearum. Sequence alignments of these orthologue proteins indicated, besides the conserved positions of single amino acids (P, Y, C, G, or L) throughout the sequence, the presence of a well-preserved stretch with the following consensus sequence: NVISXSGAXY(X)4YXWWXKGXXAXL. This conserved domain is likely to represent a new family of proteins that is present in some bacteria independently of the type III or IV virulence secretion system. The very basic protein 13 was only 52 amino acids long after cleavage of the signal peptide. It was characterized by an almost perfect ninefold repeat of the short sequence D(A/ X)M(K/X)X, in which the fourth amino acid is a fivefold lysine (K). Interestingly, the B. melitensis orthologue exhibited three additional DAMKG repeats inserted at two different positions in the B. suis amino acid sequence (Fig. 6). BLAST searches clearly indicated that Brucella protein 13 exhibited similarities with proteins in various animals and plant pathogens (A. tumefaciens and Ralstonia), as well as in parasites, such as Plasmodium. However, unlike with other homologous proteins, the 15 DEXKX (mainly DEVKN) repeats of the membrane protein Ag-1 at the blood stage of Plasmodium yoelii are preceded by a 345-amino-acid sequence with homology with the Rho GAP protein (accession number AF103869.1). Characterization of the environmental and bacterial requirements for the release of soluble proteins under acidic

5700

BOIGEGRAIN ET AL.

INFECT. IMMUN.

FIG. 6. Sequence comparison of bacterial orthologues of B. suis protein 13. Alignment of the full-length mature orthologues of B. suis protein 13 clearly indicates the repetitive conserved stretches of five amino acids. Shaded columns indicate identical (*), strongly similar (:), or weakly similar (.) amino acids.

conditions. To determine if, in addition to acidic pH, the medium composition was essential for the release of soluble proteins, we incubated B. suis at pH 4.0 in ammonium and sodium acetate and also with sodium citrate, alone or in the presence of glucose, as in minimal medium A. While medium modifications had no effect on the Omp25/Omp31 protein content in the supernatants (data not shown), the quantities of the released periplasmic proteins were dramatically reduced, as revealed by immunostaining with polyclonal antibodies against either the Rib or Nop protein (Fig. 7). In acetate buffer, replacement of ammonium by sodium had no effect on the amounts of proteins released, while replacement of acetate by citrate completely eliminated the presence of soluble proteins in the medium. Several Brucella mutants were incubated in acetate buffer at pH 4.0 to investigate which secretory systems or proteins might be involved in the release mechanism in acidic medium. When mutants with the virB operon encoding the type IV secretory system impaired were evaluated, no significant difference in the amounts of released proteins was noted compared to the wild-type strain (data not shown). On the other hand, the conditioned-medium content of the omp25 mutant did not contain the classical electrophoretic pattern, except for SOD and the Omp31 outer membrane protein. This is exemplified in Fig. 8 using Rib and Nop antibodies. Although these proteins were present in the lysate of wild-type B. suis and some other

FIG. 7. Wild-type B. suis periplasmic proteins were not released in acidic media containing assimilable citrate buffer. Wild-type B. suis cells were incubated at pH 4 in different media for 150 min at 37°C. Proteins were separated by SDS-PAGE and analyzed by Western blotting with an anti-Omp25, anti-Rib, or anti-Nop antibody. Lanes: 1, ammonium acetate and NaCl; 2, sodium acetate and NaCl; 3, sodium citrate and NaCl; 4, minimum medium containing citrate (minimal medium A [MMA]) (see Materials and Methods).

omp25/omp31 mutants (Fig. 8A, top), no immunoreactivity could be detected in the supernatant from the strain with Omp25 impaired (Fig. 8B, lanes 2). This effect appeared to be specific to the omp25 mutant, since the supernatant of the omp31 mutant exhibited quantities similar to or higher than that of wild-type B. suis (Fig. 8B, lanes 4). However, our attempts to restore the protein release even partially with the complemented ⌬omp25/omp25* strain were unsuccessful (Fig. 8B, lanes 3), despite the clear production of Omp25 in this strain (22). To gain further insight into the dependence on Omp25 in the release phenomenon, we investigated whether the lack of Omp25 paralogues, recently described in Brucella species (35), could similarly affect the release of periplasmic proteins at

FIG. 8. Knockout of omp25 prevents the release of periplasmic proteins Rib and Nop. Wild-type (WT) B. suis organisms or B. suis mutants (2.5 ⫻ 1010/ml) were incubated for 150 min at 37°C in 25 mM ammonium acetate buffer containing 120 mM NaCl, pH 4. Proteins from lysates obtained after centrifugation (A) and from concentrated supernatants (B) were separated by SDS-PAGE and analyzed by Western blotting with an anti-Rib (␣-Rib) or anti-Nop antibody. Lanes: 1, WT B. suis; 2, ⌬omp25 B. suis; 3, ⌬omp25/omp25⫹ B. suis; 4, omp31 mutant B. suis. Arrows indicate Rib (left panels) or Nop (right panels) immunoreactivities.


VOL. 72, 2004

FIG. 9. Effects of Omp25/Omp31 protein impairment on release of the periplasmic protein Rib. Wild-type (WT) B. suis (lane 1) and the B. suis strains with omp25b (lane 2), omp25c (lane 3), omp25d (lane 4), and omp31 (lane 5) knocked out were incubated in ammonium acetate as described in Materials and Methods. The amounts of periplasmic proteins released by each mutant were assessed in the supernatant using anti-Rib antibody.

acidic pH. Figure 9 shows that the amounts of released proteins in acidic medium were reduced when Omp25b, and also to a lesser extent Omp25c, were not produced. On the other hand, the knockout of the omp25d gene was without significant effect (Fig. 9, lane 5). This indicated that among the outer membrane proteins of the Omp25/Omp31 family, three members of the Omp25 family were able to influence the release of periplasmic proteins. Complete elimination of the release mechanism was only obtained when the most abundant Omp25 protein was absent from the wild-type B. suis (Fig. 8 and 9). Furthermore, the B. suis mutant corresponding to the B. abortus bvrS mutant, which is also devoid of Omp25 production, was similarly unable to release periplasmic protein in acidic acetate buffer (data not shown). DISCUSSION In an attempt to detect Brucella virulence factors possibly released in the acidic phagosomal environment, we found that in acetate buffer at pH 4.0, two subsets of proteins were released: (i) soluble proteins with an apparent low molecular mass ranging from 10 to 50 kDa and (ii) high-molecular-mass complexes that were easily separated by ultracentrifugation or gel filtration chromatography. These complexes, enriched in outer membrane proteins of the Omp25/Omp31 family, were identified as vesicular structures referred to as blebs. Blebbing is a known phenomenon that has been described in various bacteria (39), including Brucella (17–19), and the participation of these released vesicles in bacterial virulence is a recurrent matter of debate (24, 38). In the present work, we demonstrated that the blebbing phenomenon occurred for B. suis at acidic pH, and electron microscopy analysis indicated that it also takes place in neutral cell culture medium during infection of monocytes/macrophages. Our study further showed that Brucella vesicles are composed mainly of outer membrane proteins (Omp25/Omp31 family) and that they did not contain

5701

substantial amounts of the soluble proteins initially located in the bacterial intermembrane space. Besides the release of outer membrane vesicles, the identified proteins of periplasmic origin were released in acidic acetate buffer, a phenomenon which cannot be solely attributed to a loss of Brucella viability (reference 25 and data not shown). Surprisingly, we found that replacement of the acetate anion by citrate strongly attenuated this release of periplasmic proteins. The fact that citrate can be used as a source of carbon by some bacteria might explain this observation. For example, transformation of E. coli with the citW gene of Klebsiella pneumoniae encoding a citrate-acetate transporter permitted the growth of E. coli on citrate as a sole carbon source (23). This citrate uptake might also occur in B. suis and B. melitensis, since citW homologues are present in B. suis (BRA0457) and B. melitensis (BMEII 0809). If they are functional in Brucella, these transporters, in the presence of citrate, could allow Brucella to actively retain periplasmic proteins in the intermembrane space. On the other hand, acidification in the presence of acetate might alter Brucella transporters and exchangers or pore proteins, leading to the release of periplasmic proteins into the culture medium. Regarding bacterial secretion systems, our results excluded the involvement of the type IV VirB secretion system in periplasmic-protein release. Indeed, we found that the strain with the virB gene cluster impaired released amounts of the periplasmic proteins similar to those released by wild-type B. suis (data not shown). Extensive investigation of B. suis mutants with various outer membrane proteins impaired indicated that alteration of the omp31 gene never reduced the acidic release phenomenon. On the contrary, impairment of Omp25 production dramatically reduced the protein release process. Plasmid complementation of the deleted omp25 mutant was, however, not able to restore the releasing capacity of the wild-type strain. This was previously mentioned for Brucella-induced inhibition of TNF-␣ production from macrophages (22). Thus, while the mutant with omp25 deleted prevents this inhibition, the complemented strain, identical to that used in this study, could only partially restore the TNF-␣-inhibitory effect of the wild-type strain. Because of the partial restoration of the protein release phenomenon, it cannot be definitely concluded that omp25 is solely responsible for the acetate-acidic release mechanism of periplasmic proteins. However, two pieces of evidence strongly support the assumption that Omp25, but not Omp31, is directly or indirectly involved in the control of Brucella outer membrane permeability. Firstly, B. suis strains with closely related Omp25 paralogues (Omp25b and Omp25c) impaired also showed reduced release of soluble proteins. Secondly, indirect impairment of Omp25 production through knockout of the bvrS gene resulted in a B. suis strain unable to release periplasmic proteins. The present study also led to the characterization of several periplasmic proteins. Specifically, the N-terminal sequence of each protein present in acidic B. suis supernatant was unambiguously determined by Edman microsequencing. A comparison of the amino acid sequences obtained with those deduced from the genome indicated the existence, in all proteins, of a signal peptide indicative of their targeting through the inner bacterial membrane. The cleavage sites of most of the released proteins are located after the AAXA consensus sequence,

5702

BOIGEGRAIN ET AL.

which is also present in the outer membrane Omp25/Omp31 protein family (35). However, for two released proteins, the identity of the cleavage site is restricted to AXA, while two others exhibited a very different APXXA sequence, further illustrating the difficulty of predicting the cleavage sites of signal peptides in Brucella proteins. As anticipated, the sequences of the periplasmic B. suis proteins present in the acidic culture supernatant were almost identical to their respective B. melitensis orthologues. However, one interesting difference was noted with protein 4. This potential arginine-ornithine binding protein, involved in the late virulence stage of the ␣-proteobacterium plant pathogen A. tumefaciens, is also produced in B. suis and B. canis. However, in B. melitensis, because of a large DNA deletion, and in B. abortus, there was no Nop protein production. This suggested that Nop protein (and its putative release in the phagosome) was crucial for Brucella pathogenicity. On the other hand, the protein Rib was also of interest, since the B. abortus and B. melitensis orthologues were recently reported to be responsible for the agglutinating properties of human, hamster, and rabbit erythrocytes (32). In order to seek a possible role of the Rib protein in virulence, we tried to knock out the gene encoding the B. suis Rib protein, but without success (M.-T. Alvarez-Martinez, unpublished data). One possible explanation for this failure is the existence, in the direction opposite that of the rib ORF, of another ORF containing the full rib gene, which might encode a vital protein. Independently, in a recent work on epithelial cells, Castaneda-Roldan et al. confirmed that Brucella cells attach to their host cells by binding to sialic acid (9), without mentioning involvement of the Rib protein. Thus, it seems that neither Nop nor Rib, the two major proteins released in acidic acetate medium, is crucial for Brucella pathogenicity. In fact, besides the numerous solute binding proteins, our study also provided evidence of the production of two types of protein which were previously described as putative (hypothetical) proteins in many genomes. The first, protein 12, found here in B. suis, was recently also characterized by a proteomic approach in B. melitensis (42). Genes encoding homologues of this protein were found in various organisms, and the highest identity within the putative bacterial proteins was concentrated in a stretch of ⬍30 amino acids. This suggests that the conserved sequence constituted a bioactive site of the protein, whose function remains to be elucidated. The production of the second protein (protein 13) demonstrated in this study was not previously detected, probably because of its very basic pI value. It is almost exclusively composed of nine repeats of a five-amino-acid sequence (DXMXK), characterized by the presence of a methionine surrounded by negatively and positively charged amino acids. Interestingly, the putative B. melitensis homologue possesses three additional repeats, and other potential orthologues were reported in bacterial genomes. Furthermore, genes encoding homologues of these Brucella proteins with several five-aminoacid repeats, such as DA(G/X)KK or DAMRC, were also found in Anopheles gambiae (gi 27609861/gi 27599415) and the mouse (AK028328), respectively. In humans, a prototype of this five-amino-acid repeat protein was also present in the Homo sapiens genomic sequence (AJ325699; gi 15870093). After a putative 24-amino-acid signal sequence, the human ho-

INFECT. IMMUN.

mologue of protein 13 is composed of 16 DAMXK repeats with 10 consecutive DAMSK repeats, followed by 5 DAMKK repeats. This first demonstration of the production in bacteria of a protein of unknown function but whose features are conserved throughout the human genome should provide an incentive for further studies to determine its preferred structure and to identify its function. In summary, the release of Brucella periplasmic proteins in acidic acetate medium led to the characterization of many of these proteins, including the previously unidentified protein 13. Together with the comparative analysis of B. suis and B. melitensis genomes, our data confirmed the high identities among proteins in the various Brucella species. However, an interesting difference was noted, i.e., the deletion of DNA encoding Nop and six other putative proteins, which occurred in B. melitensis and probably also in B. abortus. Further concrete evidence of the in vivo occurrence of Brucella periplasmic protein release into the phagosome remains to be obtained. Our data support the idea that Omp25, in addition to its previously reported inhibitory action on macrophage TNF-␣ release, is also involved in the outer membrane properties in acidic medium, thus providing another clue to its possible contribution to Brucella virulence. ACKNOWLEDGMENTS We thank V. Jubier-Maurin for providing us with B. suis 1330 mutants with the omp31 and omp25 genes, and also the complemented omp25 (omp25⫺/⫹), impaired. We also thank D. O’Callaghan for the virB mutants of B. suis, S. Ko ¨hler for the bvrS of B. suis, A. Cloeckaert for the monoclonal anti-Omp25/Omp31 and anti-p17 antibodies, and A. Kajava for discussions on the structure of proteins composed of repeated sequences. This work was supported by INSERM and the French Association pour la Recherche contre le Cancer (ARC; number 5566). We also thank the Institute of Molecular Biology and Medicine (University of Scranton, Scranton, Pa.) and TIGR for providing access to preliminary B. melitensis and B. suis sequence data, respectively. REFERENCES 1. Al Dahouk, S., H. Tomaso, K. Nockler, H. Neubauer, and D. Frangoulidis. 2003. Laboratory-based diagnosis of brucellosis—a review of the literature. Part I. Techniques for direct detection and identification of Brucella spp. Clin. Lab. 49:487–505. 2. Al Dahouk, S., H. Tomaso, K. Nockler, H. Neubauer, and D. Frangoulidis. 2003. Laboratory-based diagnosis of brucellosis—a review of the literature. Part II. Serological tests for brucellosis. Clin. Lab. 49:577–589. 3. Alvarez-Martinez, M.-T., J. Machold, C. Weise, H. Schmidt-Eisenlohr, C. Baron, and B. Rouot. 2001. The Brucella suis homologue of the Agrobacterium tumefaciens chromosomal virulence operon chvE is essential for sugar utilization but not for survival in macrophages. J. Bacteriol. 183:5343–5351. 4. Aragon, V., R. Diaz, E. Moreno, and I. Moriyon. 1996. Characterization of Brucella abortus and Brucella melitensis native haptens as outer membrane O-type polysaccharides independant from the smooth lipopolysaccharide. J. Bacteriol. 178:1070–1079. 5. Ausubel, S. F., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seichman, J. A. Smith, and K. Struhl. 1989. Current protocols in molecular biology. John Wiley & Sons, New York, N.Y. 6. Boschiroli, M. L., S. Ouahrani-Bettache, V. Foulongne, S. Michaux-Charachon, G. Bourg, A. Allardet-Servent, C. Cazevieille, J.-P. Liautard, M. Ramuz, and D. O’Callaghan. 2002. The Brucella suis virB operon is induced intracellularly in macrophages. Proc. Natl. Acad. Sci. USA 92:1544–1549. 7. Bowden, R. A., A. Cloeckaert, M. S. Zygmunt, S. Bernard, and G. Dubray. 1995. Surface exposure of outer membrane protein and lipopolysaccharide epitopes in Brucella species studied by enzyme-linked immunosorbent assay and flow cytometry. Infect. Immun. 63:3945–3952. 8. Bricker, B. J., L. B. Tabatabai, B. A. Judge, B. L. Deyoe, and J. E. Mayfield. 1990. Cloning, expression, and occurrence of the Brucella Cu-Zn superoxide dismutase. Infect. Immun. 58:2935–2939. 9. Castaneda-Roldan, E. I., F. Avelino-Flores, M. Dall’Agnol, E. Freer, L. Cedillo, J. Dornand, and J. A. Giron. 2004. Adherence of Brucella to human

VOL. 72, 2004

10. 11. 12.

13.

14.

15.

16.

17. 18. 19.

20.

21. 22.

23. 24. 25.

26. 27. 28.


epithelial cells and macrophages is mediated by sialic acid residues. Cell Microbiol. 6:435–445. Comerci, D. J., M. J. Martinez-Lorenzo, R. Sieira, J.-P. Gorvel, and R. A. Ugalde. 2001. Essential role of the VirB machinery in the maturation of the Brucella abortus-containing vacuole. Cell Microbiol. 3:159–168. Corbel, M. 1997. Brucellosis: an overview. Emerg. Infect. Dis. 3:213–221. Delrue, R., M. Martinez-Lorenzo, P. Lestrate, I. Danese, V. Bielarz, P. Mertens, X. De Bolle, A. Tibor, J. Gorvel, and J. Letesson. 2001. Identification of Brucella spp. genes involved in intracellular trafficking. Cell Microbiol. 7:487–497. Edmonds, M. D., A. Cloeckaert, N. Booth, W. T. Fulton, S. Hagius, J. V. Walker, and P. H. Elzer. 2001. Attenuation of a Brucella abortus mutant lacking a major 25 kDa outer membrane protein in cattle. Am. J. Vet. Res. 62:1461–1466. 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 omp25 deletion mutant. Res. Vet. Sci. 72:235–239. Edmonds, M. D., A. Cloekaert, and P. H. Elzer. 2002. Brucella species lacking the major outer membrane protein Omp25 are attenuated in mice and protect against Brucella melitensis and Brucella ovis. Vet. Microbiol. 2373:1– 17. Foulongne, V., G. Bourg, C. Cazevieille, S. Michaux-Charachon, and D. O’Callaghan. 2000. Identification of Brucella suis genes affecting intracellular survival in an in vitro human macrophage infection model by signature tagged mutagenesis. Infect. Immun. 68:1297–1303. Gamazo, C., and I. Moriyon. 1987. Release of outer membrane fragments by exponentially growing Brucella melitensis cells. Infect. Immun. 55:609–615. Gamazo, C., A. I. Vitas, I. Moriyon, I. Lopez-Goni, and R. Diaz. 1993. Brucella group 3 outer membrane proteins contain a heat-modifiable protein. FEMS Microbiol. Lett. 112:141–146. Gamazo, C., A. J. Winter, I. Moriyon, J. I. Riezu-Boj, J. M. Blasco, and R. Diaz. 1989. Comparative analyses of proteins extracted by hot saline or released spontaneously into outer membrane blebs from field strains of Brucella ovis and Brucella melitensis. Infect. Immun. 57:1419–1426. Guzman-Verri, C., L. Manterola, A. Sola-Landa, A. Parra, A. Cloeckaert, J. Garin, J. Gorvel, I. Moriyon, E. Moreno, and I. Lopez-Goni. 2002. The two-component system BvrR/BvrS essential for Brucella abortus virulence regulates the expression of outer membrane proteins with counterparts in members of the Rhizobiaceae. Proc. Natl. Acad. Sci. USA 99:12375–12380. Hong, P. C., R. M. Tsolis, and T. A. Ficht. 2000. Identification of genes required for chronic persistence of Brucella abortus in mice. Infect. Immun. 68:4102–4107. Jubier-Maurin, V., R.-A. Boigegrain, A. Cloeckaert, A. Gross, M.-T. AlvarezMartinez, A. Terraza, J. Liautard, S. Kohler, B. Rouot, J. Dornand, and J.-P. Liautard. 2001. Major outer membrane protein Omp25 of Brucella suis is involved in inhibition of tumor necrosis factor alpha production during infection of human macrophages. Infect. Immun. 69:4823–4830. Ka ¨stner, C. N., K. Scheinder, P. Dimroth, and K. M. Pos. 2002. Characterization of the citrate/acetate antiporter CitW of Klebsiella pneumoniae. Arch. Microbiol. 177:500–506. Keenan, J., T. Day, S. Neal, B. Cook, G. Perez-Perez, R. Allardyce, and P. Bagshaw. 2000. A role for the bacterial outer membrane in the pathogenesis of Helicobacter pylori infection. FEMS Microbiol. Lett. 182:259–264. Kulakov, Y. K., P. G. Guigue-Talet, M. R. Ramuz, and D. O’Callaghan. 1997. Response of Brucella suis 1330 and B. canis RM6/66 to growth at acid pH and induction of an adaptive acid tolerance response. Res. Microbiol. 148: 145–151. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685. Lopez-Goni, I., C. Guzman-Verri, L. Manterola, A. Sola-Landa, I. Moriyon, and E. Moreno. 2002. Regulation of Brucella virulence by the two-component system BvrR/BvrS. Vet. Microbiol. 90:329–339. O’Callaghan, D., C. Cazevieille, A. Allardet-Servent, M. L. Boschiroli, G. Bourg, V. Foulongne, P. Frutos, Y. Kulakov, and M. Ramuz. 1999. A homologue of the Agrobacterium tumefaciens VirB and Bordetella pertussis Ptl type

Editor: D. L. Burns

29. 30. 31.

32. 33.

34.

35.

36.

37.

38. 39.

40.

41.

42.

43. 44.

5703

IV secretion systems is essential for intracellular survival of Brucella suis. Mol. Microbiol. 33:1210–1220. Porte, F., J. Liautard, and S. Ko ¨hler. 1999. Early acidification of phagosomes containing Brucella suis is essential for intracellular survival in macrophages. Infect. Immun. 67:4041–4047. Rittig, M., M.-T. Alvarez-Martinez, F. Porte, J. P. Liautard, and B. Rouot. 2001. Intracellular survival of Brucella spp. in human monocytes involves conventional uptake but special phagosomes. Infect. Immun. 69:3995–4006. Rittig, M., 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. Rocha-Gracia, R. D. C., E. I. Castaneda-Roldan, S. Giono-Cerezo, and J. A. Giron. 2002. Brucella sp. bind to sialic acid residues on human and animal red blood cells. FEMS Microbiol Lett. 213:219–224. Rouot, B., M.-T. Alvarez-Martinez, C. Marius, P. Menanteau, L. Guilloteau, R. A. Boigegrain, R. Zumbihl, D. O’Callaghan, N. Domke, and C. Baron. 2003. Production of the type IV secretion system differs among Brucella species as revealed with VirB5- and VirB8-specific antisera. Infect. Immun. 71:1075–1082. Ruckdeschel, K., J. Machold, A. Roggenkamp, S. Schubert, J. Pierre, R. Zumbihl, J. P. Liautard, J. Heesemann, and B. Rouot. 1997. Yersinia enterocolitica promotes deactivation of macrophage mitogen-activated protein kinases extracellular signal-regulated kinase-1/2, p38, and c-Jun NH2-terminal kinase. Correlation with its inhibitory effect on tumor necrosis factor-alpha production. J. Biol. Chem. 272:15920–15927. Salhi, L., R. A. Boigegrain, J. Machold, C. Weise, A. Cloeckaert, and B. Rouot. 2003. Characterization of new members of the group 3 outer membrane protein family (Omp25/31) of Brucella spp. Infect. Immun. 71:4326– 4332. Sieira, R., D. J. Comerci, D. O. Sanchez, and R. A. Ugalde. 2000. A homologue of an operon required for DNA transfer in Agrobacterium is required in Brucella abortus for virulence and intracellular multiplication. J. Bacteriol. 182:4849–4855. Sola-Landa, A., J. Pizarro-Cerda, M.-J. Grillo, E. Moreno, I. Moriyon, J.-M. Blasco, J.-P. Gorvel, and I. Lopez-Goni. 1998. A two-component regulatory system playing a critical role in plant pathogens and endosymbionts is present in Brucella abortus and controls cell invasion and virulence. Mol. Microbiol. 29:125–138. Srisatjaluk, R., R. J. Doyle, and D. E. Justus. 1999. Outer membrane vesicles of Porphyromonas gingivalis inhibit IFN-gamma-mediated MHC class II expression by human vascular endothelial cells. Microb. Pathog. 27:81–91. Vesy, C. J., R. L. Kitchens, G. Wolfbauer, J. J. Albers, and R. S. Munford. 2000. Lipopolysaccharide-binding protein and phospholipid transfer protein release lipopolysaccharides from gram-negative bacterial membranes. Infect. Immun. 68:2410–2417. Vizcaino, N., A. Cloeckaert, M. S. Zygmunt, and L. Fernandez-Lago. 2001. Characterization of a Brucella species 25-kilobase DNA fragment deleted from Brucella abortus reveals a large gene cluster related to the synthesis of a polysaccharide. Infect. Immun. 69:6738–6748. Vizcaino, N., J. M. Verger, M. Grayon, M. S. Zygmunt, and A. Cloeckaert. 1997. DNA polymorphism at the omp-31 locus of Brucella spp.: evidence for a large deletion in Brucella abortus, and other species-specific markers. Microbiology 143:2913–2921. Wagner, M. A., M. Eschenbrenner, T. A. Horn, J. A. Kraycer, C. V. Mujer, S. Hagius, P. Elzer, and V. G. DelVecchio. 2002. Global analysis of the Brucella melitensis proteome: identification of proteins expressed in laboratory-grown culture. Proteomics 2:1047–1060. Watarai, M., S. Makino, Y. Fujii, K. Okamoto, and T. Shirahata. 2002. Modulation of Brucella-induced macropinocytosis by lipid rafts mediates intracellular replication. Cell Microbiol. 4:341–355. Watarai, M., S. Makino, and T. Shirahata. 2002. An essential virulence protein of Brucella abortus, VirB4, requires an intact nucleoside-triphosphate-binding domain. Microbiology 148:1439–1446.