Identification of a Single-Nucleotide Insertion in the Promoter Region Affecting the sodC Promoter Activity in Brucella neotomae

Brucella neotomae is not known to be associated with clinical disease in any host species. Previous research suggested that B. neotomae might not express detectable levels of Cu/Zn superoxide dismutase (SOD), a periplasmic enzyme known to be involved in protecting Brucella from oxidative bactericidal effects of host phagocytes. This study was undertaken to investigate the genetic basis for the disparity in SOD expression in B. neotomae. Our Western blot and SOD enzyme assay analyses indicated that B. neotomae does express SOD, but at a substantially reduced level. Nucleotide sequence analysis of region upstream to the sodC gene identified a single-nucleotide insertion in the potential promoter region. The same single-nucleotide insertion was also detected in the sodC promoter of B. suis strain Thomsen, belonging to biovar 2 in which SOD expression was undetectable previously. Examination of the sodC promoter activities using translational fusion constructs with E. coli β-galactosidase demonstrated that the B. neotomae and B. suis biovar 2 promoters were very weak in driving gene expression. Site-directed mutation studies indicated that the insertion of A in the B. neotomae sodC promoter reduced the promoter activity. Increasing the level of SOD expression in B. neotomae through complementation with B. abortus sodC gene did not alter the bacterial survival in J774A.1 macrophage-like cells and in tissues of BALB/c and C57BL/6 mice. These results for the first time demonstrate the occurrence of a single-nucleotide polymorphism affecting promoter function and gene expression in Brucella.


Introduction
The genus Brucella consists of small, non-motile, non-spore forming, gram-negative, facultatively intracellular bacteria capable of infecting a variety of mammals. Infection with Brucella in animals leads to reproductive failure with abortions and infertility in females, and in some cases, orchitis in males [1,2]. Brucellosis is a zoonotic disease, and humans usually acquire the infection by consuming contaminated dairy or meat products or by coming in contact with the infected animal tissues and secretions [1]. There are six wellrecognized species of Brucella that show a marked host preference -B. abortus (cattle), B. melitensis (goat and sheep), B. suis (pig), B. canis (dog), B. ovis (sheep, especially ram), and B. neotomae (wood rat). In addition, isolates of B. abortus, B. melitensis, and B. suis are subdivided into biovars (or biotypes) based on the differences in their biological characteristics. Interestingly, B. suis biovar 2 is known to be incapable of causing significant clinical disease in pigs as well as in humans [3]. In the past few years, Brucella has been recovered from several marine mammals, including cetaceans (dolphin, whale and porpoise) and pinnipeds (seals and otters); these marine isolates belong to two potential new species, B. pinnipedialis and B.ceti [4]. Very recently, a new species of Brucella, B. microti, was isolated from wild common voles suffering from a systemic disease [5,6].
The virulence of Brucella can be attributed to the ability of these bacteria to escape the host defense mechanisms and survive and replicate within the host cells. Virulent Brucella organisms are capable of invading and replicating in professional phagocytes, such as macrophages, as well as several types of non-phagocytic cells [7]. The ability of Brucella to maintain long-term residence in macrophages is the basis for establishing and maintaining chronic infections, a hallmark of brucellosis. Adaptation to intracellular survival within macrophages necessitates Brucella to resist the bactericidal action of reactive oxygen intermediates generated by the host cells. One of the mechanisms some bacteria employ to resist the oxidative killing by host macrophages is the production of periplasmic Cu/Zn-cofactored superoxide dismutase (SOD) enzyme that catalyzes the dismutation of superoxide (O 2 2 ) to hydrogen peroxide [8], which can be subsequently detoxified by the enzymatic action of catalases and peroxidases. In Brucella spp., SOD is a periplasmic protein that is encoded by the sodC gene [9,10]. Studies with B. abortus indicated that SOD plays a significant role in setup and maintenance of persistent infections in murine brucellosis models [11,12].
B. neotomae was first isolated from pooled tissue suspensions of lung, spleen, liver and kidney of desert wood rats collected in the Great Salk Lake Desert in Utah [13]. In contrast to other Brucella spp., there is no documented evidence indicating the pathogenic potential of B. neotomae in any host species. Though not much further research has been done on B. neotomae, it is generally considered to be a less virulent Brucella species. Limited animal studies conducted immediately after the initial isolation indicated that B. neotomae can infect mice, pigs, guinea pigs, and wood rats but does not cause any apparent clinical disease [13][14][15]. Perhaps because of its restricted host range and/or low virulence characteristics, B. neotomae has never been isolated from any other animal species. No disease in human has been attributed to B. neotomae. Bricker et al. (1990) previously demonstrated that, of all the classical Brucella species and biovars examined, only B. neotomae and a B. suis biovar 2 isolate could not express detectable levels of SOD. The goal of this study was to determine the genetic basis for the lack of detectable SOD expression in B. neotomae. In this paper, we demonstrate that B. neotomae does express SOD, but at a substantially reduced level. Our studies discovered the presence of a single-nucleotide insertion in the promoter region of sodC gene as the cause for the reduced activity of this promoter in B. neotomae and B. suis biovar 2. Furthermore, we demonstrate that increasing the level of SOD expression in B. neotomae through complementation does not alter its persistence profile in mouse tissues.

Ethics statement
Some experiments described in this manuscript were performed in mice. The animal experiments were approved by Purdue Animal Care and Use Committee (PACUC No. 01-076-07).

Bacterial strains
B. neotomae strain 5K33 was purchased from American Type Culture Collection. B. abortus strain RB51 was available in our laboratory. B. abortus virulent strain 2308 was from culture collection at Virginia Tech. E. coli DH5a was purchased from Invitrogen. All of the bacteria were grown in tryptic soy broth (TSB) or tryptic soy agar (TSA) at 37uC. Bacteria containing plasmids were grown in presence of appropriate antibiotics at 30or 100-mg/ml concentrations of chloramphenicol and ampicillin, respectively. Colony forming units (CFU) of B. neotomae containing the recombinant plasmids were determined by plating the 10-fold serial dilutions of the cultures on TSA with and without appropriate antibiotics, to ascertain the stable maintenance of the plasmids in the recombinant bacteria.
Complementation of B. neotomae with functional sodC gene A 1.1 kb fragment containing the B. abortus sodC gene and its promoter sequence was excised from pBS/SOD [16] with ClaI restriction enzyme digestion and subcloned into pBBR4MCS to generate pBB4SOD. B. neotomae was transformed with pBB4SOD by electroporation following the previously described procedures [17]. B. neotomae containing the plasmid pBB4SOD (B. neotomae/ pBB4SOD) was selected by plating the transformed bacteria on TSA plates containing ampicillin. The expression of SOD in the complemented B. neotomae was determined by SDS-PAGE and Western blotting.

SDS-PAGE and Western blotting
For determining the expression of SOD in B. neotomae and in the complemented strain, SDS-PAGE and Western blot analyses were performed as described previously [18,19]. The antigen extract prepared from B. abortus strain RB51 was used for the positive control for SOD expression. The amount of antigen loaded in each lane of the gel corresponded to total antigens of 10 7 or 10 9 CFU of Brucella harvested during late log phase of growth curve. For Western blotting, goat anti-B. abortus SOD sera was used as the primary antibody [20]. The membranes were developed with rabbit anti-goat IgG conjugated with horseradish peroxidase and a colorimetric substrate (KPL Inc., Gaithersburg, Maryland).

Periplasmic extracts and SOD enzyme assay
Selective release of periplasmic contents of B. neotomae, B. neotomae/pBB4SOD, and B. abortus RB51 was performed as per the previously described procedure [21]. Briefly, bacteria harvested during different growth phases were pelleted down by centrifugation and resuspended in Tris-Hcl buffer, pH 7.5, at 100 ml per 10 mg of pellet wet weight. An equal volume of 0.2M Tris, pH7.5, containing 1M sucrose and 0.5% zwittergent 316 (Calbiochem) was added. Lysozyme was added to the suspension to achieve a final concentration of 100 mg/ml. Bacterial cells were exposed to a mild osmotic shock by adding an equal volume of distilled water. The suspensions were shaken for 2 hours at room temperature and then centrifuged at 8,0006 g for 30 min and the supernatants were used for measuring SOD enzyme activity. The protein concentration of the supernatant was determined using a Bio-Rad protein assay kit.
Quantitative determination of SOD activity was achieved by a colorimetric microtiter plate method using a commercially available SOD assay kit (Dojindo Molecular Technologies, Inc.). SOD activity was assayed at 450 nm as the inhibition of the reduction of the tetrazolium salt 2-(4-iodophenyl) 3-4-(nitrophenyl)-5(2,4-disulfophenyl)-2H tetrazolium (WST-1) by superoxide produced by xanthine oxidase. Commercially available E. coli SOD (Sigma-Aldrich) was used to construct standard curves. Enzyme activity, defined as percent inhibition of WST-1 reduction, was determined as for both samples and standards. SOD activity per well was determined from the standard curve. With each sample, the SOD assay was performed twice, each time in duplicates, and the data were reported as specific activity (SOD units/mg of protein).

PCR amplification and nucleotide sequence analysis
The coding sequences and the upstream region of the sodC gene from the genomic DNA of B. neotomae, B. suis biovars 2 (strain Thomsen) and 4 (strain 40) were amplified via PCR using specific primer-pairs -SOD-RBS-F (59-GGGAATGGCCTTACGGTT-39) and SOD-RBS-R (59-TTATTCGATCACGCCGCA-39) for coding region, and SOD-upstream-F (59-CGCAGCCACCCG-TTCATGTT-39) and SOD-upstream-R (59-GTCGGCAGCGC-CTCATAC-39) for upstream region. The genomic DNA of B. suis strains was previously obtained from National Animal Disease Center, Ames, Iowa [22]. The genomic DNA of B. neotomae and B. abortus RB51 was extracted using DNeasy Tissue kit (Qiagen) following the manufacturer's recommended procedure for DNA isolation from bacteria. Twenty ng of template DNA,10 pmol of each of the specific primer-pairs and reagents from NovaTaq PCR Kit PLUS (Novagen) were used to prepare PCR mixtures in a total volume of 50 ml. The PCR amplifications were performed using a thermocycler (iCyler, Bio-Rad Laboratories) with the following conditions: 95uC for 4 min followed by 30 cycles that each included 30 sec of denaturation at 95uC, 30 sec of annealing at 58uC, and 30 sec of extension at 72uC. The amplified PCR products were cloned into the pCR2.1 vector, using the TA cloning system (Invitrogen). The nucleotide sequences of both strands of the cloned fragments from 3 independent clones were determined by DNA sequencing using M13 forward and reverse primers at Purdue Genomics Core Facility. The nucleotide sequences were analyzed using LaserGene sequence analysis software (DNASTAR, Inc.). The B. neotomae nucleotide sequences were submitted to databases at GenBank (Accession number EU056817).

RNA extraction and RT-PCR
Total RNA was extracted from B. neotomae and B. abortus RB51 using RNeasy Mini Kit (Qiagen) according to the manufacturer's protocol for DNA-free RNA isolation from bacteria. Bacteria harvested during late log phase of growth curve were used for the RNA extraction. The extracted RNA concentration was determined using Quant-iT RiboGreen RNA reagent and kit (Molecular Probes). Using SuperScript TM One-step RT-PCR kit (Invitrogen), RT-PCR to detect sodC mRNA was performed with 40 ng of the extracted RNA and primers SOD-RBS-F and SODupstream-R. cDNA synthesis and PCR amplification were performed in the same tube with the following thermocycler conditions: cDNA synthesis at 45uC for 30 min, inactivation of reverse-transcriptase enzyme at 95uC for 10 min, amplification of cDNA sequences with 40 cycles of 95uC for 1 min, 58uC for 1 min, and 72uC for 1 min. As a control to rule out DNA contamination of the extracted RNA, a set of reaction tubes were subjected to direct PCR amplification by omitting the reverse transcription step (45uC for 30 min) of above mentioned thermocycler conditions. Following the PCR amplification, 12 ml of the reaction mixtures were separated on a 2% agarose gel electrophoresis and stained with ethidium bromide to visualize the amplified products.

RACE reaction
Amplification of the sodC cDNA 59 end was performed using the 59 RACE System Version 2.0 for rapid amplification of cDNA ends kit (Invitrogen) as per the manufacturer's suggested protocol. Briefly, the first strand cDNA was synthesized from the bacterial RNA using the gene specific primer 1 (GSP1) SOD-Pext-R (59-GGGCTTCGGAAATGACCACG-39) and SuperScript II under the following conditions: denaturation of the RNA at 70uC for 10 min, followed by cDNA synthesis at 42uC for 50 min. The template RNA was then removed by treating the reaction mixture with RNase mix. Unincorporated dNTPs and GSP1 were separated from the cDNA by SNAP column purification. A homopolymeric tail was added to the 39 end of the cDNA using terminal deoxynucleotidyl transferase (TdT) and dCTP at 37uC for 10 min. Following inactivation of TdT at 65uC for 10 min, the tailed cDNA was PCR amplified using gene specific primer 2 (GSP2), SOD-upstream R, and abridged-anchor primer with the following thermocycler conditions: initial denaturation at 94uC for 1 min, 35 cycles of 94uC for 1 min, 55uC for 1 min, and 72uC for 1 min, and a final extension step at 72uC for 7 min. In order to generate enough specific product, a nested PCR amplification was performed using GSP2 and abridged universal anchor primer using the previous thermocycler conditions. The PCR products were separated by electrophoresis on a 2% agarose gel, eluted using a QIAEX II gel extraction kit (Qiagen) and cloned into pGEM-T vector (Promega). The nucleotide sequences of the cloned fragments were determined by DNA sequencing using M13 reverse primer.

Construction of plasmids for determining the promoter strength
The strength of sodC promoters from B. neotomae, B. abortus, and B. suis biovars 2 and 4 was determined by generating translational fusion with E. coli b-galactosidase and then measuring the enzymatic activity of the expressed b-galactosidase in B. neotomae transformed with the plasmids. For this, a 138 bp fragment previously determined to contain the sodC promoter region, the ribosomal binding site and the start codon [16] was amplified from the genomic DNA of corresponding Brucella via PCR using a specifically designed primer-pair, SOD-pro-F (59-AAGTCGA-CATCGATAATTTCGGGGTGGAGA-39) and SOD-pro-R (59-AACTGCAGATCATCACTTCTCCTGAATATAGTTAGA-39). A restriction site was added at the 59 end of the each primer (SalI in SOD-pro-F and PstI in SOD-pro-R) to facilitate directional cloning into plasmid pNSGroE/lacZ [23]. The amplified DNA fragments were first cloned into pCR2.1 vector and sequenced to ascertain the integrity of the nucleotide sequences. The inserts were then excised from the pCR2.1 vector by digesting with SalI and PstI restriction enzymes and cloned into plasmid pNSGroE/lacZ digested with the same enzymes, resulting in the replacement of the groE promoter with the cloned sodC promoter. In the resulting recombinant plasmids, the start codon of the sodC gene was in-frame with the coding sequences of the lacZ gene, which was again confirmed by nucleotide sequencing using the SOD-pro-F primer. The resulting plasmids were designated pBnSODpro/lacZ, pBa-SODpro/lacZ, pBs2SODpro/lacZ, and pBs4SODpro/lacZ. B. neotomae was transformed with these plasmids and the expression of b-galactosidase in the recombinant bacteria was measured.

Site-directed mutagenesis
To alter specific nucleotide(s) in the predicted promoter region of B. neotomae sodC gene, PCR amplification was performed with a common reverse primer (SOD-pro-R) and one of the three forward primers, B.n-F-m1 (59-TTTAGCAATTGCTGTAAC-TATTTGTTTTAATACAGTA-39), B.n-F-m2 (59-TTTAGCAA-TTGCTGTAACTAATTTTTTTAATACAGTA-39), or B.n-F-2m1 (59-TTTAGCAATTGCTGTAACTATTTTTTTTAATA-CAGTA-39) containing the specific mutations. Similarly, two single-nucleotide mutations were generated in the potential promoter region of B. suis biovar 4 sodC gene using the forward primers B.s4-F-m1 (59-TTTAGCAATTGCTGTAACTAATGT-TTTAATACAGTA-39) and B.s4-F-m2 (59-TTTAGCAATTG-CTGTAACTATTGGTTTAATACAGTA-39). The 59 ends of the forward primers with the specific mutations encompass the sequence containing the site for MfeI restriction enzyme. The amplified 110 bp DNA product was digested with MfeI and PstI restriction enzymes and cloned into plasmid pBnSODpro/lacZ previously digested with the same enzymes, resulting in the replacement of the wild type sequences with that containing specific mutated nucleotide(s) which was confirmed by nucleotide sequencing. The resulting plasmids were designated pBnSODm1/ lacZ, pBnSODm2/lacZ, pBnSOD2m1/lacZ, pBs4SODm1/lacZ and pBs4SODm2/lacZ. B. neotomae was transformed with these plasmids and the expression of b-galactosidase in the recombinant bacteria was measured. B. abortus virulent strain 2308 and vaccine strain RB51 were also transformed with plasmids containing the B. neotomae sodC promoter variants and the b-galactosidase expression in the corresponding recombinant bacteria was measured.
b-Galactosidase enzyme assay b-Galactosidase was assayed in B. neotomae, B. abortus 2308 and B. abortus RB51 by the methods described previously [24]. The enzyme activity was calculated in Miller units with the following formula: (OD 420 61000/t6v6OD 600 ), where the OD 420 is the optical density at 420 nm, t is the incubation time in minutes, v is the volume of culture used in milliliters, and OD 600 is the optical density of the bacterial culture used for enzyme assay. The assays were performed in triplicate with bacteria grown from 3 separate colonies of each strain.
Survival and replication of B. neotomae and B. neotomae/ pBB4SOD in J744A.1 cells The intracellular growth characteristics of B. neotomae and B. neotomae complemented with the functional sodC gene (B. neotomae harboring pBB4SOD) was determined in J774A.1 macrophage-like cells using the method previously described [25]. Briefly, J774A.1 cells were cultured overnight in 6-well plates with antibiotic-free medium and infected with 10 8 CFU of B. neotomae or B. neotomae/ pBB4SOD as described above for intracellular b-galactosidase expression analysis. For each bacterial strain, the assay was performed in triplicate by infecting cells in 3 wells. At different time-points, the infected monolayers were washed three times with antibiotic free media, and the cells were lysed using 0.1% deoxychloate solution. The CFU of bacteria released from the cells were determined by plating 10-fold serial dilution of the lysates on TSA; for B. neotomae/pBB4SOD infected cells, the CFU were also determined by plating the diluted lysates on TSA with ampicillin.

Mice Experiments
The in vivo growth characteristics of B. neotomae and B. neotomae complemented with pBB4SOD were determined in mice. Fourweek old, female BALB/c and C57BL/6 mice were purchased from Harlan. Mice were given 1 week of rest before the experiments were started. Each strain of mice were divided into two groups, and mice in one group were injected intraperitoneally each with 3610 6 CFU of B. neotomae and mice in the other group were similarly injected with 1.8610 6 CFU of B. neotomae harboring pBB4SOD. The actual CFU injected was determined retrospectively by plating the inoculum on TSA plates. At specific timepoints after infection, three mice from each group were euthanized, their spleens and livers were removed and the numbers of CFU were determined by plating 10-fold serial dilutions of tissue homogenates on TSA, and also on TSA with ampicillin in case of B.neotomae/pBB4SOD inoculated mice.

Statistical analyses
One-way ANOVA with Tukey post test was performed using GraphPad Prism (GraphPad Software, San Diengo, CA, USA) to analyze the enzymatic activities of SOD and b-galactosidase. Student's t test was used to analyze the intracellular bacterial counts and the bacterial CFU in the spleens of mice.

Expression of SOD in B. neotomae
Low levels of SOD expression was detected in B. neotomae strain 5K33 by Western blot analysis using specific polyclonal antibodies. As shown in Fig. 1A and B, no SOD was detected in the total antigen extract from 10 7 CFU of B. neotomae, while considerable amount of SOD was present in the antigen extracts from the same number of B. neotomae containing plasmid pBB4SOD. When the Western blot analysis was performed using the total antigen extracts obtained from 10 9 CFU of B. neotomae, a thin protein band with size corresponding to the SOD consistently reacted with the SODspecific antibodies (Fig. 1C and D). As expected, SOD was present in the antigen extracts of both 10 7 and 10 9 CFU of B. abortus RB51.

SOD activity in periplasmic extracts
As shown in Fig. 2, SOD activity was detected in B. neotomae, B. neotomae/pBB4SOD and B. abortus RB51 at all time-points tested during the 48-hour culture period. The maximum enzyme activity was detected in B. neotomae during late log phase (24 hours) and stationary phase (48 hours). However, at all time-points tested, the SOD specific activity in B. neotomae was lower than that of B. abortus RB51 and B. neotomae/pBB4SOD.

Sequence analysis of B. neotomae sodC gene
In order to verify if the low level of SOD expression in B. neotomae is because of mutations within the coding or the upstream regions of sodC gene, we first performed PCR amplification using B. neotomae genomic DNA as template and a primer-pair designed based on the known sodC gene sequence from B. abortus. The PCR amplification resulted in a 575 bp product, the expected size in the presence of a complete sodC gene. Nucleotide sequence of the amplified product showed 99.7% identity with the corresponding region from B. suis and 99.5% identity with that from B. abortus or B. melitensis. The B. neotomae sequence differed in 2 and 3 nucleotides from that of B. suis and B. abortus or B. melitensis, respectively. All the nucleotide differences were located within the open reading frame of the sodC gene and only one of the differed nucleotides caused a change in the deduced amino acid sequence, a FRV change at amino acid position 52 of the precursor polypeptide (data not shown). We then amplified a 350 bp region flanking the 59 end of the start codon from the B. neotomae DNA. Nucleotide sequence analysis of this region revealed that B. neotomae differed in 2 nucleotides from that of the B. suis strain 1330 (biovar 1) sequence (GenBank accession no. NC_009504), one was a substitution of A with G and the other was an insertion of A (Fig. 3). Both the changes were within the 138 bp region upstream to the start codon that was previously shown to contain the sodC promoter element in B. abortus [16]. We also PCR amplified and sequenced the upstream region from strains belonging to B. suis biovars 2 (strain Thomsen) and 4 (strain 40). While the B. suis biovar 4 sequence was 100% identical with that of B. suis biovar 1, the sequence from B. suis biovar 2 differed in just 1 nucleotide, an insertion of A at the same location as that found in the B. neotomae sequence. Computer analysis of nucleotide sequences reported in the databases indicated 100% identity within the sodC upstream region between B. abortus and B. melitensis, and compared to B. suis biovars 1 and 4, these two bacteria contained 3 fewer nucleotides at the location where the nucleotide insertion was detected in B. neotomae and B. suis biovar 2 (Fig. 3). In addition, the nucleotide 48 upstream to the start codon was G in B. abortus and B. melitensis, but it was A in B. suis.
Determining the 59 end of sodC mRNA RT-PCR amplification resulted in specific amplification of 150 bp cDNA product from RNA extracted from both B. neotomae and B. abortus RB51 (Fig. 4A). No amplified products were detected when reverse-transcription step was omitted, ruling out the genomic DNA contamination as the source for the specific amplification obtained with RT-PCR (Fig. 4A). Nested PCR products of 59 RACE reaction from both B. neotomae and B. abortus RB51 contained a ,150 bp DNA fragment. The PCR products were cloned in pGEM-T vector and recombinant plasmids from 6 independent colonies for each Brucella species were used for sequencing the cloned products. For B. abortus RB51, two of the recombinant plasmids identified 'A' at 61 nucleotides upstream to the start codon as the 59 end of the cDNA (Fig. 4Bi). In all 6 of the B. neotomae and the remain 4 of the B. abortus RB51 recombinant plasmids, 'A' at 51 nucleotides upstream to the start codon was found to be the 59 end of the cDNA (Fig. 4Bii); however, because of the C-tailing of the RACE reaction, we cannot rule out the possibility of anyone of the five 'G'    enzyme by generating plasmids pBnSODpro/laZ, pBaSODpro/ lacZ, pBs2SODpro/lacZ, and pBs4SODpro/lacZ (Fig. 5A). The level of b-galactosidase expression in B. neotomae harboring these plasmids was determined by measuring the enzyme activity during late log phase (24 hour cultures). The expression of b-galactosidase under the B. neotomae and B. suis biovar 2 promoters was considerably lower than the expression under B. abortus and B. suis biovar 4 promoters (Fig. 5B).
To examine the effect of promoter activity on SodC expression, the sodC gene along with its own promoter was amplified from B. netomae and cloned into pBBR4MCS plasmid to obtain pBB4/ BnSOD. B. neotomae was transformed with pBB4/BnSOD and the resulting recombinant strain was designated B. neotomae/ pBB4BnSOD. SDS-PAGE analysis indicated that the increased expression of SOD in B. neotomae/pBB4BnSOD compared to that of B. neotome (Fig. 6), which could be attributed to the increased gene copies because of the number of plasmid copies (19). However, the level of SOD overexpression in B. neotomae/ pBB4BnSOD was clearly lower than that observed in B. neotomae/pBB4SOD (Fig. 6).

Determining the effect of single-nucleotide insertion on the promoter activity
The results of the above promoter strength assessment and 59 RACE experiments indicated that the region with the singlenucleotide insertion in B. neotomae and B. suis biovar 2 was part of the potential sodC promoter. To verify if nucleotide sequences in this region could affect the sodC promoter activity, we generated 3 separate site-directed mutation constructs with plasmid pBnSODpro/lacZ to change the B. neotomae sequences from TAATTG to TATTTG (pBnSODm1/lacZ), TAATTT (pBnSODm2/lacZ), or TATTTT (pBnSOD2m1/lacZ). Similarly, we also generated 2 separate site-directed mutations in the sodC upstream sequences of B. suis biovar 4 to change the sequences from TATTGT to TAATGT (pBs4SODm1) or TATTGG (pBs4SODm2). Plasmids containing the original and the altered sequences were electroporated into B. neotomae and the level of b-galactosidase expression was determined using 3 separate colonies for each plasmid construct. As shown in Fig. 7, with the B. neotomae protomer, both the single-nucleotide mutations did not alter the promoter activity while the double-nucleotide mutation increased the promoter activity significantly but not to the same levels seen with the wildtype sodC promoter of B. suis biovar 4. With the B. suis biovar 4 promoter, both single-nucleotide mutations decreased the promoter activity.
Activities of the B. neotomae wild-type and the mutated sodC promoters were also assessed in B. abortus virulent strain 2308 and vaccine strain RB51. The b-galactosidase expression analysis in B. abortus strains harboring the plasmid constructs showed that the original B. neotomae sodC promoter and its two single-nucleotide mutation variants were significantly less active than the variant with double-nucleotide mutation (Fig. 8).

cells
We determined the intracellular growth kinetics of B. neotomae and B. neotomae/pBB4SOD by infecting J774A.1 cells and then determining the number of viable intracellular bacteria at different time-points. As shown in Fig. 9, no obvious differences were detected between the two bacterial strains in their ability to enter and replicate within the cells up to 48 hours.
Infection and persistence profiles of B. neotomae and B. neotomae expressing SOD in mice BALB/c and C57BL/6 mice were infected with ,10 6 CFU of B. neotomae and B. neotomae/pBB4SOD, and the number of bacteria colonizing the spleens and livers was determined over time. In both strains of mice, no significant differences were observed in the bacterial counts between the B. neotomae and B. neotomae/ pBB4SOD infected groups at all time-points tested. Based on the bacterial CFU recovered from the spleens and livers of infected BALB/c mice at 1 and 7 days post-infection, both B. neotomae and B. neotomae/pBB4SOD were able to establish infection and replicate in the tissues. However, at 3 weeks post-infection, there was a 3 log decrease in the bacterial burden in the spleens and no bacteria were recovered from the livers, indicating the inability of these bacteria to maintain a chronic infection (Fig. 10A & B). In case of C57BL/6 mice, B. neotomae and B. neotomae/pBB4SOD were able to establish an infection in the spleens and livers, but in contrast to BALB/c mice, no increase in the bacterial numbers was detected at day 7 post-infection. While the livers were free of any bacteria by week 3 post-infection, bacterial counts in the spleens decreased significantly at that time-point and by week 5 post-infection, the spleens of the mice were also free any bacteria (Fig. 10C & D).

Discussion
Our Western blot and SOD enzyme assay analyses indicated that B. neotomae does express SOD, but at substantially low levels in comparison to B. abortus RB51. The reported preliminary data of Bricker et al. (1990) suggested no SOD expression in B. neotomae. Perhaps the amount of antigen and the concentration of the SODspecific antibodies used by the previous researchers in their immunoblot analyses were not sufficient to detect the low levels of SOD expression in B. neotomae. The time of harvesting bacteria for antigen preparation could also determine the amount of SOD present. Our analysis of enzymatic activity indicated that the maximum expression of SOD in B.neotomae occurred during the late log and stationary phases, which is in agreement with the previous reports of SOD expression in bacteria increasing in a growth-phase-dependant manner and reaching the maximal level during stationary phase [26]. Regulation of SodC expression in Brucella is not clearly understood. Studies in E. coli indicate that upregulation of expression of certain genes, including sodC, during the stationary phase depends on the promoter recognition by the alternate sigma factor s S (RpoS) encoded by rpoS gene [27]. Expression of RpoS is in turn controlled by host factor-I encoded by hfq gene [28]. In B. abortus also SodC expression is hfq dependent [11]. However, analysis of genome sequences does not indicate that Brucella possesses an rpoS-like gene [29]. How the hfq gene product regulates Brucella sodC expression remains to be unraveled.
The only nucleotide change with potential to affect the expression of SOD in B. neotomae was found within the promoter containing region. The same nucleotide change, insertion of A in comparison with B. suis biovars 1 & 4, was also found in B. suis  It should be noted that all the plasmid constructs tested in the present study were stably maintained in B. neotomae, as ascertained by the similar number of CFU obtained by plating the recombinant bacteria on TSA and TSA with appropriate antibiotics. Therefore, the difference in the b-galactosidase expression levels was not because of variation in plasmid stability.
Our 59 RACE analysis showed that the 59 end of sodC mRNA in B. neotomae was at 51-56 nucleotides upstream to the start codon. The B. abortus RB51 mRNA had two different 59 ends, but one of it was identical to that of B. neotomae. These mRNA 59 ends may represent the actual sodC transcription start sites, though the possibility that they could be products of post-transcriptional RNA processing or degradation cannot be ruled out. Identification of the 59 end of the sodC mRNA supported the previous observations about the potential location of the promoter region [16]. It should be emphasized that only analysis based on experiments such as DNase I footprinting can confirmatively identify the promoter sequences. Our objective in carrying out the site-directed mutagenesis studies of the promoter region was to determine if nucleotides present at the region where a nucleotide insertion was detected in B. neotomae can affect sodC promoter function. Not much information is available on the sequence features of Brucella promoters. Therefore, we examined the sodC upstream region for the consensus 210 sequence features known for E. coli; in E. coli s 70 -and s s -dependent promoters contain similar 210 core sequences [30]. We could identify two potential 210 sequences, TTTAAT and TATT(T/G)T, upstream to the identified mRNA 59 end(s). Our reasoning for mutating the targeted nucleotides was based on the assumption that the 'A' insertion in B. neotomae changed the potential 210 sequence from the optimal TATT(G/ T)T to weaker TAATTG. Though our experimental results with the mutated B. neotomae and B. suis biovar 4 promoters appear to support this hypothesis, we cannot rule out the possibility that the targeted nucleotides could be outside the 210 region but still be critical for the promoter function. Even if our assumption about the effect on the 210 region is not true, the b-galactosidase

html) and other
Brucella genome sequences present in the databases at GenBank indicates that all B. abortus and B. melitensis biovars/strains sequenced so far have identical nucleotide sequences in the sodC upstream region, whereas B. canis and B. suis biovars 1, 3 and 4 share 100% sequence identity in this region. However, B. ovis, B. suis biovar 5 (strain 513, isolated from a rodent), Brucella sp. 83/13 (isolated from a rodent) and all marine mammal isolates (B. pinnipedalis and B. ceti) also contain the same 'A' insertion in the sodC upstream region as reported here for that of B. neotomae and B. suis biovar 2, suggesting that the promoter activity in these Brucella could similarly be at a lower level.
To the best of our knowledge, there are no published studies documenting the growth characteristics of B. neotomae in J774A.1 cells and in tissues of BALB/c and C57BL/6 mice, some of the in vitro and in vivo model systems widely used in Brucella research.
Our studies show that B. neotomae is able to enter and replicate in J774A.1 cells similar to other Brucella spp. [25]. However, in BALB/c mice, B. neotomae exhibited an interesting growth and persistence kinetics, where it can establish an infection in spleen and liver, replicate until day 7 to increase the bacterial burden by up to 1.8 log, and then gradually decline in numbers. In C57BL/6 mice, B. neotomae did not replicate to a level to increase its numbers in livers and spleens, but maintained the similar bacterial load from day 1 to day 7 and then gradually declined. These findings suggest that B. neotomae can establish an acute infection in mice but it is unable to maintain a chronic infection like other known virulent Brucella spp. This characteristic feature makes B. neotomae an attractive candidate to be included in studies to understand the mechanisms employed by virulent Brucella spp. in causing chronic infections in mouse models. It should be mentioned that the pBB4SOD plasmid was stably maintained in B. neotomae, as ascertained by the similar number of CFU obtained by plating the tissue homogenates containing the recombinant bacteria on TSA and TSA with ampicillin.
Previous studies with B. abortus showed that SOD plays a role in establishing and maintaining a chronic infection in mice [11,12]. However, if reduction in the level of SOD expression by Brucella spp. affects their ability to cause chronic infection is not known. Increasing the level of SOD expression in B. neotomae did not alter the bacterial survival in BALB/c and C57BL/6 mice (Fig. 10). It is possible that the low level of SOD expression in B. neotomae is sufficient to inactivate the host-derived superoxide radicals and any additional increase in the amount of SOD has no functional effect. It is also possible that increasing the level of SOD expression per se is insufficient to compensate for other critical genetic deficiencies of B. neotomae affecting its ability to maintain a chronic infection.
In conclusion, we have identified the presence of a singlenucleotide insertion in the promoter region as the cause for the reduced activity of sodC promoter of B. neotomae. Though promoter nucleotide polymorphism has been shown to modulate gene expression in other bacteria [31,32], this is the first report demonstrating the occurrence of a single-nucleotide difference affecting promoter function and gene expression in Brucella spp. This finding highlights the possibility of single nucleotide polymorphisms in promoter regions contributing to the differences in expression of certain genes among Brucella species/biovars/ strains.