Brucella abortus strain 2308 produces brucebactin, a highly efficient catecholic siderophore

Manuel I. González Carreró1, Félix J. Sangari1, Jesús Agüero1 and Juan M. García Lobo1

Departamento de Biología Molecular (Unidad Asociada al Centro de Investigaciones Biológicas, CSIC), Facultad de Medicina, Universidad de Cantabria, Cardenal Herrera Oria s/n, 39011-Santander, Spain1

Author for correspondence: Manuel I. González Carreró. Tel: +34 42 201943. Fax: +34 42 201945. e-mail: carrerom{at}unican.es


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Brucella abortus is known to produce 2,3-dihydroxybenzoate (2,3-DHBA) and to use this catechol as a siderophore to grow under iron-limited conditions. In this study a mutant (BAM41) is described that is deficient in siderophore production by insertion of Tn5 in the virulent B. abortus strain 2308. This mutant was unable to grow on iron-deprived medium and its growth could not be restored by addition of 2,3-DHBA. Production of catecholic compounds by both the Brucella mutant and parental strains under iron-deprivation conditions was assayed by TLC. Two catecholic substances were identified in the supernatant of the parental strain 2308. The faster migrating spot showed the same retention factor (Rf) as that of purified 2,3-DHBA. The mutant BAM41 overproduced 2,3-DHBA, but failed to form the slower migrating catechol. This defect could only be complemented by the addition of the slow-migrating catechol from strain 2308. The genomic region containing Tn5 in BAM41 was cloned and the position of the transposon was determined by nucleotide sequencing. The sequence revealed that the insertion had occurred at a gene with homology to Escherichia coli entF, a locus involved in the late steps of the biosynthesis of the complex catecholic siderophore enterobactin. Intracellular survival and growth rates of the B. abortus wild-type and entF mutant strains in mouse-derived J774 macrophages were similar, indicating that production of this siderophore was not essential in this model of infection. It is concluded that B. abortus synthesizes a previously unknown and highly efficient catecholic siderophore, different from 2,3-DHBA, for which the name brucebactin is proposed.

Keywords: iron acquisition, catechol, macrophage, 2,3-DHBA

Abbreviations: 2,3-DHBA, 2,3-dihydroxybenzoate; CAS, chrome-azurol-S; DIP, 2,2-dipyridyl; EDDA, diethylenediamine di(o-hydroxyphenylacetic acid); NTA Fe(III), ferric nitriletriacetate

The GenBank accession number for the sequence reported in this paper is AF361942.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Iron, the most abundant metal ion in biological systems, is involved in numerous metabolic processes and is essential for almost all living organisms examined to date. However, given the oxidative properties of free iron, the intracellular concentration of this metal needs to be tightly regulated, because otherwise it would react with reduced forms of oxygen, resulting in free radicals known to cause lipid peroxidation, protein modification, nucleic acid damage, etc. (Actis et al., 1986 ). Free iron in body fluids of animals is kept at very low levels by iron-sequestering proteins such as transferrin or lactoferrin; therefore, it is postulated that bacterial high-affinity iron acquisition systems are a prerequisite for most infections (Ratledge & Dover, 2000 ). Most iron inside host cells is in the cytoplasm, complexed with haem proteins or in the storage protein ferritin. However, little is known about iron content in the phagosomes where some intracellular parasites multiply. It is generally believed that phagosomal iron content is low and this limitation poses a restriction for bacterial growth. Stimulation of macrophages with interferon-{gamma} and/or lipopolysaccharide decreased iron uptake from transferrin caused by down-regulation of transferrin receptor expression (Mulero & Brock, 1999 ). Studies carried out with Legionella pneumophila, an intracellular parasite localized in phagosomes, indicate that IFN-{gamma}-activated monocytes inhibit intracellular multiplication of the parasite by limiting the availability of iron, showing that phagosomal iron is scarce and the pathogen needs it to survive (Byrd & Horwitz, 1989 ). Moreover, it has been shown that the iron content of host-cell vesicles containing facultative intracellular parasites is low enough to enhance the production of iron-regulated virulence factors (Garcia-del Portillo et al., 1992 ; Pope et al., 1996 ).

Brucella abortus is an intracellular parasite able to proliferate within macrophages, thereby successfully avoiding the bactericidal effects of phagocytes. Our knowledge about the mechanisms used by B. abortus for its intracellular survival has improved considerably in the last few years. Recent findings indicate that B. abortus is able to survive and replicate in phagosomes of the phagocytic cells by preventing the fusion between phagosomes and lysosomes (Frenchick et al., 1985 ; Pizarro-Cerda et al., 1998 ).

When grown under iron-limited conditions, B. abortus secretes 2,3-dihydroxybenzoate (2,3-DHBA), a simple catecholic compound. 2,3-DHBA was considered the only B. abortus siderophore responsible for Fe55 uptake and able to relieve the growth inhibition caused by the strong iron chelator EDDA (Lopez-Goñi et al., 1992 ). Production of other catecholic or hydroxamate siderophores in Brucella has not yet been reported. The role of 2,3-DHBA in protecting Brucella from the antibacterial activity of macrophages has been studied by the addition of exogenous 2,3-DHBA to B. abortus-infected murine macrophages in culture. In these experiments the number of intracellular brucellae recovered from 2,3-DHBA supplemented cells was higher than that of bacteria recovered from untreated macrophages (Leonard et al., 1997 ). This observation suggests that production of this catechol could provide intracellular brucellae with an efficient mechanism for acquiring iron in an iron-restricted environment that exists within the phagosome, thereby contributing to bacterial proliferation in the host. However, a recent report indicated that production of 2,3-DHBA was not required for replication of Brucella in murine macrophages or for the establishment of a chronic infection in the BALB/c mouse model (Bellaire et al., 1999 ).

In this work we report the isolation and characterization of a mutant B. abortus strain which, unlike the parental strain, was siderophore-negative in chrome-azurol-S (CAS) plates and unable to grow in iron-restricted medium, although it secreted 2,3-DHBA. The growth restriction observed under these conditions was overcome by the addition of iron in a soluble form or of a catecholic extract from culture supernatants of the parental strain grown in the presence of subinhibitory concentrations of EDDA. The mutant strain altered in siderophore production was deficient in iron assimilation but did multiply in macrophages.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids, culture media and chemical reagents.
Strains and plasmids used in this study are described in Table 1. Escherichia coli strains were grown either in LB broth or on LB agar plates (Pronadisa). Brucella strains were usually grown in Brucella Broth (BB), Brucella Agar (BA) or Tripticase Soy Broth (TSB) (Pronadisa). MMB, a semi-synthetic low-iron medium containing yeast extract (Lopez-Goñi, 1989 ), was used for the iron assimilation assays. 2,2-Dipyridyl (DIP; 150 mM; Sigma) or 10 mM diethylenediamine di(o-hydroxyphenylacetic acid) (EDDA; Fluka), were added to the medium as iron chelators. EDDA was previously deferrated with HCl and acetone as described by Rogers (1973) . EDDA stock solutions were stored at -20 °C in the dark. Agarose (0·7%; Pronadisa) was used instead of bacteriological agar in iron-deprived media. Every new batch of culture medium with chelator was tested for minimal inhibitory concentration for the strains tested. In some experiments iron bioavailability was increased by the addition of ferric nitriletriacetate [NTA Fe(III)] to the medium. Ampicillin (100 µg ml-1), nalidixic acid (15 µg ml-1) and kanamycin (50 µg ml-1) were added as appropriate.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids used in this study

 
Detection of siderophore in solid medium.
Siderophore production was detected using a modification of the CAS plating assay (Schwyn & Neilands, 1987 ). The modification consisted of using MMB/c medium instead of MM9 to prepare CAS plates. MMB/c medium has the same composition as MMB except that the K2HPO4 final concentration was 0·3 g l-1 and glycerol and lactic acid were omitted. The final pH was adjusted to 6·8 with PIPES (Sigma). The CAS assay is a universal assay for the detection of siderophores and operates independently of the siderophore structure. The presence of a siderophore is indicated by the decolorization of the blue-coloured ferric-CAS complex, resulting in a yellow-orange halo around colonies growing on CAS medium. Since B. abortus was unable to grow on this medium we centrifuged 1 ml of Brucella cultured in MMB to early stationary phase, washed the cell pellet and loaded it on top of the plate. Plates were incubated for 2–3 d at 37 °C and checked for colour change surrounding the zone where the samples had been deposited.

Isolation of mutants defective in siderophore production.
pSUP2021, an ampicillin-resistant plasmid containing Tn5, and unable to replicate in Brucella, was introduced into B. abortus 2308 by conjugation with E. coli S17-1(pSUP2021) as described previously (Sangari & Agüero, 1991 ). Transconjugants containing Tn5 were initially selected on BA containing nalidixic acid and kanamycin and purified on MMB plates with the same antibiotics. Isolated colonies from MMB plates were transferred to the blue CAS medium to screen for siderophore-deficient mutants. Colonies that were negative in this assay were tested for susceptibility to ampicillin to confirm the loss of the plasmid and were further tested for their growth on MMB plates containing different concentrations of the iron chelator EDDA.

Growth of B. abortus in iron-deprived medium.
B. abortus strains were cultured at 37 °C in liquid MMB medium containing different concentrations of EDDA. Samples were taken at different time intervals and growth was determined by measuring A590. In all these experiments the glass material was previously washed overnight with a 1 M HCl solution followed by three washes with ultrapure water (18 M{Omega}) to remove traces of iron.

Bioassays for siderophores.
The presence of siderophore was detected with a cross-feed assay, consisting of checking the growth of siderophore-negative mutants in iron-deprived solid medium around filter disks containing different substances. When the strain to be cross-fed was E. coli AN193 entA, half strength TSB plates containing 150 µM DIP were used, and in the case of BAM41 the medium used was MMB agarose with 150 µM EDDA. In both cases cells from 1 ml early-stationary-phase low-iron cultures were collected by centrifugation, resuspended in 10 µl sterile saline and seeded onto the indicator plates. Five microlitres of the supernatants were spotted onto a Millipore filter (0·22 µm) and allowed to dry before the disk was applied to the plate. The plates were then checked for the presence of a halo of growth after 48–72 h at 37 °C. To assay ethyl acetate extracts or pure dissolved substances, they were placed on sterile filter paper disks and dried for 10 min in the dark at room temperature to allow the solvent to evaporate. The size of the halo of growth represents the amount/strength of the siderophore present in the sample.

Preparation of ethyl acetate extracts and analysis by TLC.
Early-stationary-phase B. abortus cultures in MMB were centrifuged (4000 r.p.m. for 10 min at 4 °C) and the supernatants were filter-sterilized through Millipore 0·22 µm filters. Quantification of catechols in these supernatants was carried out by the colorimetric test of Arnow (1937) , using 2,3-DHBA as control. Catechol-containing supernatants were acidified to pH 2·0 by addition of 6 M HCl and catechols were extracted by addition of ethyl acetate (20 ml per 100 ml supernatant). The extracts were pooled and concentrated at low temperature in a vacuum system.

TLC of ethyl acetate extracts was carried out on silicagel G plates (Merck). Chromatograms were developed with benzene/acetic acid/water (125:72:3, by vol.). 2,3-DHBA and its derivatives were first detected by fluorescence under UV light. The presence of iron-reacting substances was revealed by spraying the plate with 0·12 M FeCl3 in 0·1 M HCl. When required, the compounds were recovered from the chromatograms by elution in aqueous acetic acid (15%, v/v), the resulting solutions were made up to 0·5 M HCl and the compounds were extracted with ethyl acetate.

Recombinant DNA techniques.
Chromosomal DNA from B. abortus strains was extracted by the guanidium thiocyanate method (Pitcher et al., 1989 ). Construction of plasmids, restriction enzyme analysis, agarose gel electrophoresis and Southern hybridizations were carried out by standard protocols (Sambrook et al., 1989 ). A 3·5 kb HindIII fragment from Tn5 was used as a hybridization probe and labelled with digoxigenin with the DIG-High Prime Kit (Roche Diagnostics). The DNA sequence flanking transposon mutants was cloned using arbitrary PCR (Caetano-Annoles, 1993 ), as described by O’Toole & Kolter (1998) , and sequenced. Database searches and alignments were performed at the National Center for Biotechnology Information (NCBI), using the BLAST network service.

Infection and intracellular viability assay of B. abortus in J774 cells.
J774 murine macrophages were cultured in RPMI medium with 2 mM L-glutamine, penicillin (100 U ml-1), streptomycin (0·1 mg ml-1) and 10% fetal calf serum (FCS) at 37 °C under 5% CO2 and 100% humidity. Confluent monolayers were trypsinized and 2x105 cells per well were incubated for 24 h before infection in 24-well tissue culture plates. B. abortus cells used for infection were opsonized with decomplemented antiserum for 30 min at 37 °C. Infection of macrophages was carried out for 45 min at 37 °C in RPMI without serum and antibiotics at a multiplicity of infection of 100. After infection the wells were washed five times with sterile phosphate-buffered saline (PBS) and further incubated in RPMI with 2 mM L-glutamine, 10% FCS and 40 µg gentamicin ml-1 to kill extracellular bacteria. The number of intracellular viable B. abortus was determined at different times post-infection. To this end, the monolayers were washed twice with sterile PBS and treated for 5 min with 1 ml 0·1% Triton X-100 (Sigma) in deionized water. Lysates were serially diluted and plated onto BA plates for determination of c.f.u.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Isolation of B. abortus mutants unable to secrete siderophore
A bank of B. abortus 2308 Tn5 insertion mutants was grown in MMB agar and tested for siderophore secretion. Isolated colonies were deposited on modified CAS plates. Although B. abortus was unable to grow on this medium, the siderophore present in the sample was able to decolorize CAS. The presence of a siderophore was considered positive when an orange halo was present under the mass of bacteria. Three hundred colonies were assayed by this method. The orange halo was absent in one isolate which was named BAM41 (Fig. 1).



View larger version (129K):
[in this window]
[in a new window]
 
Fig. 1. Detection of siderophore secretion on CAS plates. Bacterial samples recovered from early-stationary-phase culture in MMB were deposited on pieces of 0·22 µm Millipore filter on CAS plates and incubated for 72–96 h. Siderophore-positive strains produce an orange halo. A, E. coli AN193; B, B. abortus BAM41; C, 2,3-DHBA (10 µl of a 10 mM stock solution); D, B. abortus 2308; E, blank (MMB medium).

 
Growth of BAM41 in an iron-deprived medium
To assess the ability of the CAS-negative mutant BAM41 to grow in media with low availability of iron we tested the parental and the mutant strains in liquid MMB containing different concentrations of EDDA. The results showed that at EDDA concentrations higher than 15 µM the biomass of the mutant strain was considerably lower than that of the parental strain (Fig. 2). This result suggested that the mutant BAM41 was unable to assimilate iron when the metal was chelated by EDDA. Addition of Fe(III) in the form of NTA Fe(III) reverted the inhibitory effect of EDDA on strain BAM41, demonstrating that the inhibition was caused by the absence of bioavailable iron in the medium (Fig. 2)



View larger version (15K):
[in this window]
[in a new window]
 
Fig. 2. Effects of iron availability on the growth of mutant BAM41. (a) Growth of B. abortus 2308 (open bars) and BAM41 (filled bars) in MMB medium containing different concentrations of EDDA. The optical density of the cultures was measured in the stationary phase. (b) Effect of the addition of Fe(III) to a B. abortus BAM41 culture in the presence of an inhibitory concentration of EDDA. B. abortus BAM41 was inoculated in two different flasks containing MMB medium with 20 µM EDDA. NTA Fe(III) was added to one of the samples at the beginning of the experiment up to a final iron concentration of 15 µM. Samples were taken at different time intervals and growth was measured in a photometer. {bullet}, Iron-supplemented culture; {blacksquare}, non-supplemented culture.

 
Isolation and analysis of catecholic compounds present in B. abortus cultures
Supernatants from B. abortus cultures grown in MMB medium containing a subinhibitory concentration of EDDA were analysed by the Arnow test for catechol content at different time points. Maximal production of catechols was observed at the beginning of the stationary phase of growth. The amount of catechol detected for a particular strain under the same culture conditions varied considerably between different assays. However, when catechol was detected, the amount found was always proportional to the biomass of bacterial cells present in iron-limited cultures.

Catechols were extracted from supernatants with ethyl acetate, concentrated and analysed by TLC. This analysis indicated the presence of a catechol molecule (Arnow-positive) with an Rf identical to that of 2,3-DHBA both in the mutant BAM41 and in the parental strain, 2308. Strain 2308 presented, in addition, a second Arnow-positive spot that exhibited a lower Rf and was less abundant than 2,3-DHBA (Fig. 3).



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 3. Analysis of ethyl acetate extracts from culture supernatants by TLC. Culture supernatants from early-stationary-phase cultures in MMB containing subinhibitory concentrations of EDDA were extracted with ethyl acetate and analysed on TLC plates containing 2,3-DHBA as standard. The line at the top indicates the solvent front. A, 2,3-DHBA (3 µl of a 10 mM stock solution); B, B. abortus 2308; C, ethyl acetate extract of fresh MMB medium; D, B. abortus BAM41.

 
To confirm that the fast-migrating catechol described above was 2,3-DHBA we performed a cross-feed assay with the E. coli entA mutant AN193, unable to synthesize 2,3-DHBA, and hence enterochelin. This strain cannot grow in media with inhibitory concentrations of a strong iron chelator, such as DIP, unless 2,3-DHBA is present. The results showed that filtered supernatants from both B. abortus 2308 and mutant BAM41, were able to cross-feed the E. coli entA mutant, indicating that they contained 2,3-DHBA. This cross-feed assay was more efficient with strain BAM41 than with 2308, suggesting overproduction of 2,3-DHBA by the mutant. The fast-migrating catechol from BAM41 and 2308 supernatants was purified from the TLC plate, eluted and blotted on a filter. This purified substance restored the growth of AN193, confirming its identification as 2,3-DHBA.

Phenotypic complementation of mutant BAM41
To further characterize the Brucella siderophore-deficient mutant, we tried to restore its growth in EDDA-chelated MMB medium with different substances spotted on filters placed on the agar surface. 2,3-DHBA (3 µl, 10 mM) was unable to restore the growth of mutant BAM41. Ethyl acetate extracts from B. abortus 2308 culture supernatants, but not from BAM41, did complement the growth defect in the mutant. The two catechols in the 2308 supernatant were separately purified from a TLC plate and used in the complementation assay. Only the slow-migrating catechol was able to restore the growth of the mutant (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2. Phenotypic complementation of mutant BAM41

 
Survival of BAM41 in macrophages
The role of siderophore in the ability of B. abortus to survive in macrophages was tested by using the B. abortus BAM41 mutant in an in vitro macrophage survival assay. The results of the experiment are shown in Fig. 4. Both B. abortus 2308 and BAM41 survived and replicated in the macrophage without significant differences.



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 4. Intracellular survival of B. abortus BAM41 (open bars) and 2308 (filled bars) in J774 murine macrophages. Bacteria were recovered from the macrophages at different times after infection and intracellular live brucellae were determined by a c.f.u. count on BA plates.

 
Genetic characterization of mutants
Total DNA from BAM41 was digested with EcoRI and hybridized with an internal Tn5 probe as described in Methods. A single hybridization EcoRI band of 8 kb was observed. This band was excised from an agarose gel and cloned in the plasmid vector pUC18, resulting in pSUBA41. A physical map of this DNA fragment is shown in Fig. 5. The nucleotide sequence of both ends of this DNA fragment was determined by sequencing the recombinant plasmid pSUBA41 with universal primers. To analyse the transposon insertion point we used the arbitrary PCR method described by O’Toole & Kolter (1998) . Using this method we obtained and sequenced a PCR amplification product corresponding to the right flank of the transposon (Fig. 5). To determine the sequence of the left flank we used a synthetic primer derived from the previously determined sequence. These sequences were assembled in a contig of 1 kb corresponding to the B. abortus chromosome (GenBank accession no. AF361942). This sequence was compared with the non-redundant GenBank nucleotide sequence database using the TBLASTN program. The sequence showed strong homology with the gene vibH from Vibrio cholerae, which encodes an amidase involved in the condensation steps of vibriobactin synthesis (Keating et al., 2000a , b ). The sequence also showed homology with the amino-terminal part of E. coli entF, a gene encoding a subunit of the enterobactin biosynthesis multienzyme complex (Rusnak et al., 1991 ). According to these results we may conclude that the Tn5 insertion in BAM41 occurred at position 275, near the 5' end of a B. abortus gene homologous to E. coli entF. This insertion resulted in the inactivation of the gene, which could be involved in the biosynthesis of the new catecholic siderophore described above.



View larger version (6K):
[in this window]
[in a new window]
 
Fig. 5. Physical and genetic maps of the chromosomal EcoRI fragment carrying the Tn5 insertion in the mutant BAM41. The Tn5 insertion position is marked. The shadowed bars indicate the sequenced regions. The genes identified by sequence comparison are represented by arrows showing the direction of the transcription.

 
The sequence of the right end of plasmid pSUBA41 was also compared with sequences in the GenBank database. The sequence showed strong homology with the E. coli entC and entE genes, involved in the first stage of enterobactin biosynthesis. These data fit well in the physical map reported for the region encoding the 2,3-DHBA biosynthesis genes of B. abortus (Bellaire et al., 1999 ).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
2,3-DHBA has been reported to be the only catecholic siderophore present in the supernatant of Brucella cultures growing under iron-limiting conditions. 2,3-DHBA was able to supply iron to the bacterium and accordingly was considered as a true siderophore for Brucella (Lopez-Goñi et al., 1992 ).

The appearance of large amounts of 2,3-DHBA or salicylic acid in culture supernatants is a phenomenon commonly observed for bacteria producing phenolate siderophores in response to iron deprivation. However, those compounds are considered to be low-affinity siderophores (Actis et al., 1986 ), unable to compete with stronger iron-chelating compounds. Moreover, recent theoretical studies do not consider 2,3-DHBA to be capable of acting as a bacterial siderophore (Chipperfield & Ratledge, 2000 ). These facts question the role of 2,3-DHBA as a siderophore in B. abortus and suggest that Brucella either produces a stronger siderophore, different from 2,3-DHBA, or it uses a different iron assimilation system.

We have observed that B. abortus 2308 was able to decolorize CAS plates, indicating that this strain secretes a siderophore. To discover the role of this siderophore in intracellular survival we have isolated a Tn5 mutant of B. abortus 2308, named BAM41, which was negative in the CAS assay and thus considered to be siderophore-deficient. The minimal inhibitory concentration of EDDA for this mutant was lower than that of the parental strain, 2308. The correlation between absence of siderophore activity in the CAS assay and inability to grow in an iron-deprived medium indicated that the transposon insertion in this mutant was probably affecting genes involved in siderophore biosynthesis.

The growth of BAM41 in a low-iron medium was stimulated by filtered supernatants from a 2308 culture grown in an iron-deprived medium, and also by ethyl acetate extracts from the same culture, indicating that an iron-related growth factor was present in both samples. Analysis of ethyl acetate extracts of B. abortus 2308 low-iron culture supernatants by TLC showed the presence of two different catechol species, one with an Rf identical to that of 2,3-DHBA and the other with a lower Rf. The concentration of the low Rf catechol in culture supernatants was always considerably lower than that of 2,3-DHBA. When isolated from TLC plates, only the compound with the low Rf, but not 2,3-DHBA, was able to complement the growth of the mutant in an iron-deprived medium and to decolorize CAS plates. These results indicated that B. abortus 2308 produced a second catecholic substance with a higher affinity for iron than the formerly reported 2,3-DHBA. Since this new compound, capable of decolorizing CAS medium, could be an efficient siderophore for B. abortus, we propose to call it brucebactin. The presence of cathecol compounds in addition to 2,3-DHBA had been described by Lopez-Goñi et al. (1992) . Brucebactin could be identical to the cathecol with an Rf of 0·02 described in that report. However, this catechol was not further assayed and the possibility that it could be a siderophore was discarded by the authors.

An alternative explanation for the inability of BAM41 to grow in iron-deprived media is that the mutation in this strain affected siderophore transport instead of siderophore biosynthesis. However, siderophore transport mutants usually present a large halo in CAS plates (Schwyn & Neilands, 1987 ) and this was not the case for BAM41. Moreover, the fact that this strain could be cross-fed by B. abortus 2308 ethyl acetate supernatant extracts indicated that siderophore transport was not affected in this mutant.

The results obtained in a cross-feed assay with an E. coli entA mutant confirmed that the fast-moving catechol was 2,3-DHBA. Cross-feed experiments and TLC also demonstrated that BAM41 produced more 2,3-DHBA than the parental strain 2308. This would be the expected behaviour of a mutant blocked in the late steps of the biosynthesis of a complex catecholic siderophore structurally based on 2,3-DHBA. Genes encoding the enzymes for this process are usually organized in two clusters. The first encodes the enzymes for the production of 2,3-DHBA from chorismic acid. The second cluster encodes a multienzyme complex responsible for the condensation of 2,3-DHBA with amino acids or other molecules to produce the final catecholic siderophore (Ratledge & Dover, 2000 ). On these grounds, BAM41, which did not produce the final catechol, but produced 2,3-DHBA, might well be affected in the condensation step.

Genetic analysis revealed that BAM41 contained a single Tn5 copy inserted in a gene homologous to E. coli entF (Rusnak et al., 1991 ). The product of this gene (EntF) together with EntB and EntE constitute a multienzyme complex responsible for the biosynthesis of the catecholic siderophore enterobactin in E. coli (Gehring et al., 1998 ). The Brucella gene presumably should form part of an operon responsible for brucebactin biosynthesis.

Attempts to determine the conditions for optimal brucebactin production indicated that the better yields were always obtained in iron-limited media at the beginning of the stationary phase of growth. Attempts to isolate high amounts of brucebactin for structural and chemical analyses were unsuccessful because of the instability of the purified compound.

We conclude that 2,3-DHBA can be used by Brucella as a low-affinity siderophore when iron in the medium is readily soluble, but this substance cannot work as an efficient siderophore when iron is complexed by EDDA or high-affinity physiological chelators such as transferrin or lactoferrin, as theoretically predicted by Chipperfield & Ratledge (2000) . In this case Brucella must compete for iron by secreting another compound with similar or stronger affinity for iron (responsible for CAS decoloration). Moreover, concentrations of 2,3-DHBA higher that those secreted by BAM41 in iron-deprived medium were unable to decolorize CAS plates, indicating the lack of siderophore activity of this catechol.

Results presented here and previous evidence (Bellaire et al., 1999 ) have shown that siderophores appear not to play a relevant role in the intracellular survival of B. abortus. This finding has also been reported for other intracellular pathogens such as Salmonella (Benjamin et al., 1985 ), probably because iron availability in the phagosome is high enough for this pathogen or because other effective iron assimilation mechanisms are used by Salmonella. On the other hand, it has been shown that 2,3-DHBA improves the survival of B. abortus in the macrophage (Leonard et al., 1997 ). Considering the poor ability of 2,3-DHBA as a siderophore, this effect could be due to another circumstance different from iron assimilation. Brucella needs an acidic phagosome in the first hours of infection to survive (Porte et al., 1999 ). In this low pH compartment, iron is readily soluble and thus highly bioavailable. Apart from this, catecholic siderophores present less affinity for iron at acid than at neutral pH (Emery, 1978 ). Therefore, given the pH of Brucella-containing phagosomes, and taking into account the available results, we may conclude that the role of catecholic compounds in the phagosomal survival of B. abortus, if any, should be different from iron sequestration.


   ACKNOWLEDGEMENTS
 
This work has been partially supported by grants BIO96-1398 and 1FD1997-0593 from the Spanish Plan Nacional de Biotecnología and under European Contract QLK2-CT-1999-0014. The technical expertise of Beatriz Calvo is deeply acknowledged.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Actis, L. A., Fish, W., Crosa, J. H., Kellerman, K., Ellenberger, S. R., Hauser, F. M. & Sanders-Loehr, J. (1986). Characterization of anguibactin, a novel siderophore from Vibrio anguillarum 775(pJM1). J Bacteriol 167, 57-65.[Medline]

Arnow, L. (1937). Colorimetric determination of the compounds of 3,4-dihydroxyphenylalanine-tyroxine mixture. J Biol Chem 118, 531-537.

Bellaire, B. H., Elzer, P. H., Baldwin, C. L. & Roop, R. M.2nd (1999). The siderophore 2,3-dihydroxybenzoic acid is not required for virulence of Brucella abortus in BALB/c mice. Infect Immun 67, 2615-2618.[Abstract/Free Full Text]

Benjamin, W. H.Jr, Turnbough, C. L.Jr, Posey, B. S. & Briles, D. E. (1985). The ability of Salmonella typhimurium to produce the siderophore enterobactin is not a virulence factor in mouse typhoid. Infect Immun 50, 392-397.[Medline]

Byrd, T. F. & Horwitz, M. A. (1989). Interferon gamma-activated human monocytes downregulate transferrin receptors and inhibit the intracellular multiplication of Legionella pneumophila by limiting the availability of iron. J Clin Invest 83, 1457-1465.[Medline]

Caetano-Annoles, G. (1993). Amplifying DNA with arbitrary PCR primers. PCR Methods Appl 3, 85-92.[Medline]

Chipperfield, J. R. & Ratledge, C. (2000). Salicylic acid is not a bacterial siderophore: a theoretical study. Biometals 13, 165-168.[Medline]

Cox, G. B., Gibson, F., Luke, R. K., Newton, N. A., O’Brien, I. G. & Rosenberg, H. (1970). Mutations affecting iron transport in Escherichia coli. J Bacteriol 104, 219-226.[Medline]

Emery, T. (1978). Iron in model natural compounds. In Metal Ions in Biological Systems , pp. 77-121. Edited by H. Sigel. New York:Marcel Dekker.

Frenchick, P. J., Markham, R. J. & Cochrane, A. H. (1985). Inhibition of phagosome–lysosome fusion in macrophages by soluble extracts of virulent Brucella abortus. Am J Vet Res 46, 332-335.[Medline]

Garcia-del Portillo, F., Foster, J. W., Maguire, M. E. & Finlay, B. B. (1992). Characterization of the micro-environment of Salmonella typhimurium-containing vacuoles within MDCK epithelial cells. Mol Microbiol 6, 3289-3297.[Medline]

Gehring, A. M., Mori, I. & Walsh, C. T. (1998). Reconstitution and characterization of the Escherichia coli enterobactin synthetase from EntB, EntE, and EntF. Biochemistry 37, 2648-2659.[Medline]

Grant, S. G., Jessee, J., Bloom, F. R. & Hanahan, D. (1990). Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants. Proc Natl Acad Sci USA 87, 4645-4649.[Abstract]

Keating, T. A., Marshall, C. G. & Walsh, C. T. (2000a). Reconstitution and characterization of the Vibrio cholerae vibriobactin synthetase from VibB, VibE, VibF, and VibH. Biochemistry 39, 15522-15530.[Medline]

Keating, T. A., Marshall, C. G. & Walsh, C. T. (2000b). Vibriobactin biosynthesis in Vibrio cholerae: VibH is an amide synthase homologous to nonribosomal peptide synthetase condensation domains. Biochemistry 39, 15513-15521.[Medline]

Leonard, B. A., Lopez-Goñi, I. & Baldwin, C. L. (1997). Brucella abortus siderophore 2,3-dihydroxybenzoic acid protects brucellae from killing by macrophages. Vet Res 28, 87-92.[Medline]

Lopez-Goñi, I. (1989). Efecto de la carencia de hierro sobre el crecimiento, la producción de sideróforos y la composición de la membrana externa de Brucella abortus y Brucella melitensis. PhD thesis, University of Navarra, Spain.

Lopez-Goñi, I., Moriyón, I. & Neilands, J. B. (1992). Identification of 2,3-dihydroxybenzoic acid as a Brucella abortus siderophore. Infect Immun 60, 4496-4503.[Abstract]

Mulero, V. & Brock, J. H. (1999). Regulation of iron metabolism in murine J774 macrophages: role of nitric oxide-dependent and -independent pathways following activation with gamma interferon and lipopolysaccharide. Blood 94, 2383-2389.[Abstract/Free Full Text]

O’Toole, G. A. & Kolter, R. (1998). Initiation of biofilm formation in Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signalling pathways: a genetic analysis. Mol Microbiol 28, 449-461.[Medline]

Pitcher, D. G., Saunders, N. A. & Owen, R. J. (1989). Rapid extraction of genomic DNA with guanidinium thiocyanate. Lett Appl Microbiol 8, 151-156.

Pizarro-Cerda, J., Moreno, E., Sanguedolce, V., Mege, J. L. & Gorvel, J. P. (1998). Virulent Brucella abortus prevents lysosome fusion and is distributed within autophagosome-like compartments. Infect Immun 66, 2387-2392.[Abstract/Free Full Text]

Pope, C. D., O’Connell, W. & Cianciotto, N. P. (1996). Legionella pneumophila mutants that are defective for iron acquisition and assimilation and intracellular infection. Infect Immun 64, 629-636.[Abstract]

Porte, F., Liautard, J. P. & Kohler, S. (1999). Early acidification of phagosomes containing Brucella suis is essential for intracellular survival in murine macrophages. Infect Immun 67, 4041-4047.[Abstract/Free Full Text]

Ratledge, C. & Dover, L. G. (2000). Iron metabolism in pathogenic bacteria. Annu Rev Microbiol 54, 881-941.[Medline]

Rogers, H. J. (1973). Iron-binding catechols and virulence in Escherichia coli. Infect Immun 7, 438-444.[Medline]

Rusnak, F., Sakaitani, M., Drueckhammer, D., Reichert, J. & Walsh, C. T. (1991). Biosynthesis of the Escherichia coli siderophore enterobactin: sequence of the entF gene, expression and purification of EntF, and analysis of covalent phosphopantetheine. Biochemistry 30, 2916-2927.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sangari, F. & Agüero, J. (1991). Mutagenesis of Brucella abortus: comparative efficiency of three transposon delivery systems. Microb Pathog 11, 443-446.[Medline]

Schwyn, B. & Neilands, J. B. (1987). Universal chemical assay for the detection and determination of siderophores. Anal Biochem 160, 47-56.[Medline]

Simon, R., Priefer, U. & Pühler, A. (1983). A broad host range mobilization system for in vivo genetic engineering, transposon mutagenesis in Gram negative bacteria. Bio/Technology 1, 784-791.

Vieira, J. & Messing, J. (1982). The pUC plasmids, an M13mp7-derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19, 259-268.[Medline]

Received 5 September 2001; revised 26 October 2001; accepted 29 October 2001.