Characterization of Novel Brucella Strains Originating from Wild Native Rodent Species in North Queensland, Australia

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept. 2010, p. 5837 5845 Vol. 76, No. 17 0099-2240/10/$12.00 doi:10.1128/aem.00620-10 Copyright 2010, American Society for Microbiology. All Rights Reserved. Characterization of Novel Brucella Strains Originating from Wild Native Rodent Species in North Queensland, Australia Rebekah V. Tiller, 1 Jay E. Gee, 1 Michael A. Frace, 1 Trevor K. Taylor, 2 Joao C. Setubal, 3 Alex R. Hoffmaster, 1 and Barun K. De 1 * Centers for Disease Control and Prevention, Atlanta, Georgia 30333 1 ; Australian Animal Health Laboratory, Geelong, Victoria 3220, Australia 2 ; and Virginia Bioinformatics Institute, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 3 Received 10 March 2010/Accepted 4 July 2010 We report on the characterization of a group of seven novel Brucella strains isolated in 1964 from three native rodent species in North Queensland, Australia, during a survey of wild animals. The strains were initially reported to be Brucella suis biovar 3 on the basis of microbiological test results. Our results indicated that the rodent strains had microbiological traits distinct from those of B. suis biovar 3 and all other Brucella spp. To reinvestigate these rodent strains, we sequenced the 16S rrna, reca, and rpob genes and nine housekeeping genes and also performed multiple-locus variable-number tandem-repeat (VNTR) analysis (MLVA). The rodent strains have a unique 16S rrna gene sequence compared to the sequences of the classical Brucella spp. Sequence analysis of the reca, rpob, and nine housekeeping genes reveals that the rodent strains are genetically identical to each other at these loci and divergent from any of the currently described Brucella sequence types. However, all seven of the rodent strains do exhibit distinctive allelic MLVA profiles, although none demonstrated an amplicon for VNTR 07, whereas the other Brucella spp. did. Phylogenetic analysis of the MLVA data reveals that the rodent strains form a distinct clade separate from the classical Brucella spp. Furthermore, whole-genome sequence comparison using the maximal unique exact matches index (MUMi) demonstrated a high degree of relatedness of one of the seven rodent Brucella strains (strain NF 2653) to another Australian rodent Brucella strain (strain 83-13). Our findings strongly suggest that this group of Brucella strains isolated from wild Australian rodents defines a new species in the Brucella genus. Brucella species are facultative intracellular Gram-negative members of the Alphaproteobacteria class capable of causing brucellosis in a range of animal hosts, including domesticated livestock, wildlife, marine mammals, and humans (1, 5, 7, 29, 32, 33, 36, 47). Brucellosis is the most prevalent zoonotic disease worldwide, causing spontaneous abortion and fetal death in animals and severe flu-like symptoms, focal complications, and often, chronic disease in humans (7, 11, 22, 27, 40, 41, 49, 50). Brucella species are typically transmitted to humans through consumption of unpasteurized dairy products or exposure to fluids or tissues from infected animals (45, 49). Animals are primary hosts of all Brucella spp., which include Brucella abortus (cattle), B. canis (dogs), B. melitensis (goats, cows, and sheep), B. suis (swine), B. ovis (rams), and B. neotomae (desert rats) (3, 7, 8). Recently, three additional Brucella species have been recognized: B. pinnipedialis (seals), B. ceti (dolphins) (5), and B. microti. B. microti was initially isolated from the common vole in the Czech Republic (33, 35). In the mid-1980s, DNA-DNA hybridization studies demonstrated a very high level of genetic similarity (98.5%) among the Brucella spp., which led to the adoption of a monospecies concept for the Brucella genus, with all the species at that time renamed as biovars of B. melitensis (46). However, 20 years later, the use of a phylogenetic-evolutionary approach to Brucella taxonomy * Corresponding author. Mailing address: Centers for Disease Control and Prevention, Mail Stop G34, 1600 Clifton Road, Atlanta, GA 30333. Phone: (404) 639-5465. Fax: (404) 639-3023. E-mail: bkd1@cdc.gov. Published ahead of print on 16 July 2010. was accepted. By that approach, host preferences, virulence, and pathogenicity were considered important criteria in the delineation of Brucella species, and consequently, the multispecies taxonomy was restored to the Brucella genus (28). With the development of more advanced molecular typing methodologies such as multilocus sequence analysis (MLSA) (48), multiple-locus variable-number tandem-repeat (VNTR) analysis (MLVA) (23), and techniques interrogating single nucleotide polymorphisms (SNPs) (17, 47), Brucella spp. can be quickly genotyped and strains can be readily examined for their phylogenetic and evolutionary relationships (15). Recently, we reported on two unusual human brucellosis cases, one of which led to the identification of a novel species, Brucella inopinata, whose type strain is strain BO1 and which was associated with a breast implant infection in a patient in Oregon (12, 36). The second brucellosis case involved an atypical Brucella strain (strain BO2) isolated from the lung biopsy fluid of a patient with chronic destructive pneumonia in Australia (44). Because both patients denied common risk factors associated with human brucellosis, the primary hosts of these strains remain unknown. However, nucleotide sequence analysis of the outer membrane proteins (omp2a and omp2b) of both strain BO2 and strain BO1 T demonstrated close clustering to an atypical B. suis strain (strain 83-210) isolated from a rodent in Australia (30, 44). Further genetic analysis of the 16S rrna genes from strains BO1 T and BO2 and the 16S rrna gene from atypical Brucella strain 83-13 (available at http: //www.broadinstitute.org), which was isolated from a rodent in Australia and briefly described by Corbel and Brinley-Morgan in 1984 (9), showed that strain 83-13 yielded notable genetic 5837

5838 TILLER ET AL. APPL. ENVIRON. MICROBIOL. similarity to the novel human isolates, which led us to speculate that atypical human Brucella strains BO1 T and BO2 may have an animal reservoir in rodents from Australia. Rodent brucellosis is self-limiting and is mostly associated with wild rodents that cohabitate among domestic livestock presumably infected with classical Brucella spp. (27). Over 22 different wild rodent species worldwide have been reported to be susceptible to Brucella infection, as demonstrated by serology and/or culture (41). Earlier field studies in Argentina, Venezuela, and Denmark reported on the prevalence of B. suis in hares, opossums, and rats and B. abortus in ferrets and capybaras (10, 11, 27). However, two rodent-specific Brucella spp. have been identified, including B. neotomae, isolated from the desert wood rat (Neotomae lepida) in Utah (39), and most recently, B. microti, isolated from the common vole (Microtus arvalis) in the Czech Republic (35). In the early 1960s, Cook et al. reported on the isolation and biochemical identification of seven Brucella suis biovar 3 strains from three known species of wild native rats from Australia (6). We reinvestigated the microbiological characteristics of these seven rodent B. suis biovar 3 strains and performed genetic analyses with respect to the microbiological characteristics and genetics of the classical and atypical Brucella species. In this report, we describe this group of Brucella strains isolated from wild rodents in Australia, which confer unique microbiological and molecular characteristics distinct from those of any of the currently described species. MATERIALS AND METHODS Bacterial strains. The seven Brucella strains were isolated from native rat species Rattus assimilis (allied rat) (n 4), Melomys cervinipes (large climbing rat) (n 2), and Melomys lutillus (small climbing rat) (n 1) captured at Jordan Creek near Palmerston National Park, North Queensland, Australia. All of these Brucella strains were originally tested to determine their bacteriological and oxidative metabolic profiles and were identified as B. suis biovar 3, as described previously by Cook et al. (6). In 2009, these rodent strains were received by the Bacterial Special Pathogens Branch at the CDC, Atlanta, GA, from culture collections at the Australian Animal Health Laboratory, Geelong, Victoria, Australia. Upon receipt, these Brucella cultures were stored in defribrinated rabbit blood at 70 C until they were tested. All reference bacterial strains incorporated in this study are part of the CDC culture collection. Microbiological analysis. The rodent Brucella strains were maintained on Trypticase soy agar with 5% sheep blood (SBA) plates (BBL Microbiology Systems, Cockeysville, MD) incubated at 37 C under a 5% CO 2 atmosphere. The standard algorithm for phenotypic identification of Brucella at the species and biovar levels was performed as described previously (1, 4, 12). Freshly prepared cultures (48 h) were tested for CO 2 requirement, H 2 S production, tolerance to thionine and basic fuchsin dyes at three dilutions, gel formation, agglutination in monospecific antisera, and lysis by Tbilisi phage at two routine test dilutions (RTDs; 1 and 4 the RTDs) (1). Reagents for the microbiological tests were either prepared in-house or obtained from BBL Microbiology Systems (Cockeysville, MD). Tbilisi phage was purchased from the National Veterinary Services Laboratory (NVSL; Ames, IA). Rabbit monospecific anti-a and anti-m antisera from NVSL and the Veterinary Laboratories Agency (VLA; Addlestone, Surrey, United Kingdom) were used for the slide agglutination assays, as described by the providers. Real-time PCR. Genomic DNA from the rodent strains was purified and analyzed by a real-time TaqMan PCR assay capable of differentiating B. inopinata-like species from all other Brucella spp. using hybridization probes targeting base substitutions in the 16S rrna gene, as described previously (44). Positive results were expressed in log scale as the crossing threshold (C T ) value. 16S rrna, reca, and rpob gene sequencing. The nearly full-length 16S rrna genes from the seven rodent strains were amplified using eubacterial primers (primers F8 and R1492) and were sequenced using the BigDye Terminator (version 3.1) cycle sequencing kit (Applied Biosystems, Foster City, CA), as mentioned previously (18). The rodent Brucella 16S rrna gene consensus sequence was compared with the 16S rrna gene consensus sequence of the Brucella spp. (18), the Ochrobactrum intermedium type strain (strain CCUG 24694; GenBank accession no. AM114411), B. inopinata BO1 T (GenBank accession no. EU053207), and recently described strain BO2 using the GCG Wisconsin software package (version 10.2; Accelrys, San Diego, CA) and the MEGA (version 4.0) program (24, 44). Genomic DNA corresponding to the full-length reca gene (37) and rpob gene (26) of the rodent strains was amplified and sequenced using the BigDye Terminator (version 3.1) cycle sequencing kit (Applied Biosystems), as described previously (44). Contigs were assembled and edited before multiple-sequence alignments were constructed in the Lasergene (version 8) genetic analysis software suite (DNAStar Inc., Madison, WI). Pairwise analysis was performed, and neighbor-joining consensus trees inferred from 1,000 bootstrap replicates were constructed using the MEGA (version 4.0) program (24). MLSA. Genotyping of the rodent strains based on MLSA of nine housekeeping genes was performed as described previously (48). The full-length amplicons of the nine housekeeping genes from the rodent strains were sequenced, concatenated, and analyzed by comparison with those of the BO1 T, BO2, and 27 classical Brucella sequence types described by Whatmore et al. (48). All nine housekeeping gene sequences from Brucella sp. strain 83-13 were retrieved from http://www.broadinstitute.org for multiple-sequence alignments and comparison with the sequences of the seven rodent strains examined in our study. Sequence editing and gene concatenation were performed in the DNAStar Lasergene (version 8) software suite. Multiple sequences were aligned and neighbor-joining phylogenetic trees were constructed as described above (24). MLVA. Molecular typing based on MLVA was employed by examining 15 VNTR genetic markers (MLVA-1 to MLVA-15) of the seven rodent strains, as described previously (23, 43). Genomic DNA preparations from the rodent strains were used to amplify the 15 VNTR loci, and the amplicons underwent fragment analysis on an ABI Prism 3130 automated fluorescent capillary DNA sequencer (Applied Biosystems). Allele designations were assigned by internal binning capabilities in the GeneMapper (version 3.7) software package (Applied Biosystems). A distance tree was generated in BioNumerics (version 6.0) software (Applied Maths, Saint-Martens-Latem, Belgium) by clustering analysis using the unweighted-pair group method using average linkages (UPGMA) and was saved in the Newick format. Tree manipulations and labeling were done in the MEGA (version 4.0) program (24). Whole-genome sequencing and MUMi distance calculations. The whole genome of a representative rodent strain (strain NF 2653) was sequenced using Roche Applied Science/454 pyrosequencing, based on GS-titanium chemistry (25), at the CDC Biotechnology Core Facility. A random shotgun library of purified genomic DNA was produced using the Roche protocols for nebulization, end polishing, adaptor ligation, nick repair, and single-stranded library formation. Following emulsion PCR, labeled beads were isolated and sequenced using longread (LR) sequencing kits. After read trimming and refiltering to recover short reads of high quality, a draft de novo assembly was performed using a Newbler assembler (version 2.0), and the draft was forwarded to the Virginia Bioinformatics Institute, Blacksburg, VA, for annotation and comparative genomic analysis (finalized sequencing for submission of the sequences to GenBank are in progress). The genome of strain NF 2653 was compared to other Brucella genomes available at http://www.broadinstitute.org/and http://patric.vbi.vt.edu/. These include the genomes of B. neotomae 5K33, Brucella sp. strain 83-13, strain BO2, B. inopinata strain BO1 T, B. microti, and B. suis strain 686. The maximal unique exact matches index (MUMi) distance calculation was performed using the Mummer (version 3.0) program (13). The Mummer program was run on concatenated contigs (achieved by inserting a string of 100 nucleotides between contigs) of each incomplete genome. The only genome that was complete was that of B. microti, and its two chromosome sequences were concatenated. The distance calculations performed using the MUMi algorithm are based on the number of maximal unique matches of a given minimal length shared by two genomes being compared. The MUM index is used to rapidly estimate the distance between closely related bacterial genomes, and the distance calculations correlate with those obtained by the traditional average nucleotide identity estimation, which detects the level of DNA conservation of the core genome. MUMi values vary from 0 for very similar genomes to 1 for very distant genomes (13). RESULTS Phenotypic characterization. All seven rodent strains grew on SBA or Trypticase soy agar with 5% rabbit blood plates with or without 5% CO 2 at 35 to 37 C for 24 to 48 h under aerobic conditions; and the colonies were smooth, opaque con-

VOL. 76, 2010 NOVEL BRUCELLA STRAINS FROM WILD RODENTS IN AUSTRALIA 5839 Downloaded from http://aem.asm.org/ FIG. 1. Phylogenetic tree of reca gene sequences (948 bp) obtained using neighbor-joining analysis. Scale bar, 0.02 divergent nucleotide per site. vex, and mucoid in nature. The organisms were small Gramnegative coccoid rods; nonhemolytic; nonmotile; and positive for catalase, oxidase, and urease reactions ( 5 min); and they turned H 2 S paper black within 24 h at 37 C under aerobic conditions. Standard microbiological tests for biotyping of Brucella spp. were conducted under aerobic conditions. All seven strains grew in the presence of thionine and basic fuchsin dyes at all dilutions and were negative for gel formation in phenolized saline. All rodent Brucella strains showed variations in agglutination profiles using both live and heat-inactivated cells (1). Most of the rodent Brucella strains were highly susceptible to Tbilisi phage at two routine test dilutions; however, one strain, NF 2637, was resistant at both dilutions. All these rodent Brucella strains were negative for agglutination by the acriflavine test. Real-time PCR. Our initial real-time PCR assay revealed that the genomic DNAs from all seven rodent strains positively hybridized to the BI probe, designed to detect B. inopinata-like species, while they failed to hybridize to the BRU probe, which detects all classical Brucella species (44) (data not shown). These PCR results indicated that these rodent strains carry the same 4-bp nucleotide substitution in their 16S rrna genes as strains BO1 T and BO2 (44). 16S rrna, reca, and rpob gene sequence analysis. An analysis of the nearly full-length sequences of the 16S rrna genes (1,412 bp) of the Australian rodent Brucella strains verified the presence of the 4 base substitutions (GAAA) at positions 167 to 170, the single base substitution (C) at nucleotide position 234, and an additional single base difference at position 308, representing a unique 16S rrna gene sequence in the Brucella genus. In addition to the 16S rrna gene, the reca consensus sequence (875 bp) of the Australian rodent strains demonstrated 99.2% identity to the Brucella spp. consensus sequence and was 97.9% and 98.7% identical to the sequences of strains BO1 T and BO2, respectively. Thus, neighbor-joining analysis of the reca consensus sequence shows that the rodent strains form a distinct phylogenetic cluster nearest the classical Brucella spp. (Fig. 1). We observed several rodent strain-specific single nucleotide polymorphisms (n 5), all of which were silent mutations. We also performed SNP analysis with the rpob gene (4,093 bp) of the rodent strain consensus sequence and the rpob genes of 37 other Brucella spp. The consensus sequence of the rpob gene of the Australian rodent strain was 100% identical to the sequence of the rpob gene of Brucella sp. strain 83-13 (available at http://www.broadinstitute.org) and on June 16, 2018 by guest

5840 TILLER ET AL. APPL. ENVIRON. MICROBIOL. B. melitensis clone CT-18 (AY540343) 18 B. melitensis clone CT-14 (AY540338) B. melitensis CT-17 (AY54035) 95 B. melitensis Ether (DQ086122) 32 B. melitensis CT1p (AY540337) B. melitensis clone CT-11 (AY540339) 56 B. melitensis CT-15 (AY540340) 38 35 B. melitensis clone CT-9 (AY540342) 59 B. melitensis clone CT-19 (AY540341) 78 B. melitensis clone CT-1 (AY540344) B. melitensis B115 (DQ086120) 56 B. melitensis Rev1 (DQ086119) 46 B. melitensis 63/9 (DQ086121) B. suis Thomsen (DQ086123) B. suis 686 (DQ086124) 42 B. canis RM6/66 (DQ086128) 36 64 B. suis 40 (DQ086125) B. abortus clone Ba3R (AY562181) B. abortus C68 (DQ086139) 85 B. abortus B3196 (DQ086136) 68 B. abortus 870 (DQ086137) 56 B. abortus 63175 (DQ086138) B. abortus Tulya (AY562180) B. abortus RB51 (AY540346) 4538 B. abortus 99 (DQ086131) B. abortus 292 (DQ086135) B. abortus 19 (DQ086130) 61 B. abortus 45/20 (DQ086132) 100 B. abortus 2308 (AY562179) B. abortus 1119 (DQ086133) B. abortus 86/8/59 (DQ086134) B. neotame 5K33 (DQ086129) B. ovis 63/290 (DQ086127) B. suis 513 (DQ086126) BO2 B. inopinata BO1 Australia rodent Brucella spp. 0.001 FIG. 2. Phylogenetic tree of rpob gene sequences (4,093 bp) obtained using neighbor-joining analysis. Scale bar, 0.001 divergent nucleotide per site. 99% similar to the sequence of the rpob gene of BO1 T (40 SNPs). The rodent strain rpob consensus sequence displayed a lower level of identity ( 98.9%) to the rpob genes of BO2 and the classical Brucella sp. strains, differing by a range of from 49 to 76 base substitutions. There were a number of rodent strainspecific SNPs (n 20), with only one resulting in a protein change, at amino acid 436. Neighbor-joining analysis of the rpob sequences shows an evident subcluster of BO1 T and the rodent strains separate from the other Brucella spp. and BO2 (Fig. 2). MLSA. The use of MLSA has proven useful for describing the diversity among the known Brucella spp. and biovars and has aided with the characterization of new and emerging Brucella isolates, such as strains BO1 T and BO2 (12, 44). By

VOL. 76, 2010 NOVEL BRUCELLA STRAINS FROM WILD RODENTS IN AUSTRALIA 5841 FIG. 3. Unrooted phylogenetic profiles corresponding to the concatenated sequence of nine housekeeping genes (4,396 bp) representing 27 known STs of classical Brucella spp. and B. inopinata BO1 T and BO2, along with all seven rodent strains. MLSA, the sequences of the Australian rodent strain group showed 100% identity to the concatenated sequences from Brucella sp. strain 83-13. This genetically closely related group demonstrated the greatest divergence from the classical Brucella spp. evaluated thus far, having 1.7% sequence distance from Brucella sequence type 1 (ST1). Pairwise analysis revealed that the Australian rodent group has the greatest similarity to BO2 (98.7%) and the lowest similarity to B. melitensis STs 7 to 12 (98.2%). Clustering by the neighbor-joining method clearly demonstrated the unique diversity found in the rodent strain group when it was compared to the 27 previously described Brucella STs and the BO1 T and BO2 STs (Fig. 3). Sequence analysis of the rpob and reca genes and the nine MLSA genes (gap, aroa, glk, dnak, gyrb, trpe, cobq, omp25, and int-hyp) demonstrated the high degree of genetic relatedness of this group of rodent strains. Multiple-sequence alignments of these genes from all the seven strains yielded for each gene a representative consensus sequence that was then used for further statistical and phylogenetic analysis. MLVA. MLVA is a useful tool for evaluating strain diversity within Brucella species by exploiting highly polymorphic repeat regions in the genome. Each rodent Brucella strain had a different MLVA genotype, demonstrating the intraspecies diversity within this group of strains. All seven strains carried the same alleles at VNTRs 07 (null allele), 14, 20, 21, 25, and 31; and all alleles detected in these strains with the exception of three novel alleles at VNTR 28 have been observed in other Brucella species. UPGMA analysis of the MLVA data shows that the rodent Brucella strains form a distinct clade separate from the classical Brucella spp. (Fig. 4). MUMi distance calculations. The Brucella genome consists of two circular chromosomes of 2.1 Mb (chromosome I) and 1.16 to 1.2 Mb (chromosome II) (14, 20, 31). The core genomic DNA conservation based on MUMi analysis was used to specifically address the intra- and interspecies variability of the genome sequence of one representative strain (NF 2653) of this rodent group in comparison to the genome sequences of Brucella sp. strain 83-13, B. inopinata BO1 T and BO2, B. neotomae 5K33, B. microti, and B. suis 686 (WHO B. suis biovar 3). The MUMi distance values range from 0 (very similar) to 1 (most distant) and have been shown to correlate very well with DNA-DNA hybridization values, which is one criterion used by taxonomists to assign species designations (13). On the basis of the MUMi estimations presented in Table 1, the Brucella genomes show distance values ranging from 0.009 to 0.175. A remarkable similarity was observed between strain NF 2653 and Brucella sp. strain 83-13, for which the MUMi had a very low value of 0.009. These two rodent strains show an average MUMi value of 0.16 when their genome sequences are compared to those of the other Brucella spp. included in the comparison. The genomes of B. neotomae 5K33 and B. microti also show low MUMi values (0.02) when their sequences are compared to each other and to the sequence of strain B. suis 686. Finally, strains BO1 T and BO2 show greater similarity to each other (MUMi vaule, 0.107) than to the other Brucella spp. Analysis of these data revealed three clades: clade I (MUMi value, 0.02), made up of B. neotomae 5K33, B. suis 686, and B. microti; clade II (MUMi value, 0.009), made up of Brucella strain NF 2653 and Brucella sp. strain 83-13; and clade III (MUMi 0.107), made up of the two atypical human strains, B. inopinata BO1 T and BO2. DISCUSSION Earlier field studies have reported on the isolation and transmission of classical Brucella species in wild rodents around the areas of farms with infected livestock (2, 10, 11, 16, 27, 41, 42). Although wild rodents are self-limiting reservoirs, they are capable of transmitting common Brucella species within their populations and to comingling species (34, 41, 42). The epizootiology of Brucella infections in wild rodents has not been thoroughly explored and deserves special attention, par-

5842 TILLER ET AL. APPL. ENVIRON. MICROBIOL. B. inopinata BO1 BO2 NF2629 NF2640 NF2653 NF2668 Australian rodent Brucella strains NF2651 NF2637 NF2627 B. abortus (n=66) B. neotame 5K33 B. melitensis (n=215) B. canis (n=17) B. suis (n=60) 20 FIG. 4. UPGMA dendrogram of 15 MLVA markers of 7 rodent strains, strains BO1 T and BO2, and 359 characterized Brucella strains. ticularly with the continual identification of new and emerging Brucella species (12, 36, 44). The rodent Brucella strains (n 7) from North Queensland, Australia, described in this paper exhibit unique phenotypic traits and molecular diversity in comparison to those of the common Brucella spp. Cook et al., using standard microbiological tests, reported these seven strains to be B. suis biovar 3 (6). We found that all seven of the rodent strains shared the same colony morphology on SBA plates and demonstrated some biochemical characteristics similar to the those of classical B. suis biovar 3; e.g., it is a non-

VOL. 76, 2010 NOVEL BRUCELLA STRAINS FROM WILD RODENTS IN AUSTRALIA 5843 TABLE 1. MUMi results for the rodent strain NF 2653 genome along with six Brucella sp. genomes computed using the concatenated contigs of each incomplete genome a Strain MUMi value b 83-13 BO2 NF 2653 BO1 B. suis B. microti B. neotomae 5K33 0.145 0.168 0.146 0.168 0.017 0.022 Brucella sp. strain 83-13 0.172 0.009 0.175 0.147 0.150 Brucella sp. strain BO2 0.168 0.107 0.169 0.169 Brucella sp. strain NF 0.172 0.147 0.146 2653 B. inopinata BO1 0.169 0.167 B. suis biovar 3 686 0.020 a Only the sequence of B. microti was complete, and its two chromosome sequences were concatenated. Concatenation was done by inserting a string with 100 nucleotides between contigs. b The minimum MUMi value is 0, and the maximum MUMi value is 1. motile and Gram-negative coccobacillus, it has no CO 2 requirement for growth, it has a positive rapid urease test result, and there is no inhibition of growth in all dilutions of both the thionine and basic fuchsin dyes. However, these rodent strains also exhibited distinctive biochemical properties not characteristic of B. suis biovar 3, which include sensitivity to Tbilisi phage lysis at two dilutions (1 RTD and 4 RTD) and positivity for H 2 S production. Thus, our microbiological identification suggested that these rodent strains constitute a phenotypically unique Brucella sp. in comparison to all classical Brucella species, including B. neotomae and B. microti. Molecular characterization by standard PCR identified the Brucella-specific insertion sequence, IS711 (842 bp) (21), along with several large amplicons ( 1,000 bp), in all seven rodent Brucella strains. These rodent strains had IS711 profiles similar to those of B. inopinata BO1 T and the atypical strain Brucella BO2 (12, 44) (data not shown). These seven rodent strains also positively hybridized with the BI probe of the B. inopinataspecific real-time PCR assay targeting the 4 nucleotide substitutions in the 16S rrna gene (data not shown) (44). Sequence analysis of the 16S rrna, reca, and rpob genes and nine MLSA housekeeping genes revealed that all seven rodent Brucella strains are genetically identical at these loci, producing a consensus sequence for all the genes examined. The conserved nature of the 16S rrna gene among the brucellae has prevented discrimination at the species level but serves as a useful genetic locus for inclusion at the genus level. The novel B. inopinata BO1 T and BO2 strains were the first Brucella strains confirmed to have a 16S rrna gene sequence divergent from the 16S rrna gene sequences of the rest of the Brucella spp. (12, 44). Interestingly, the consensus 16S rrna gene sequence of the Australian rodent Brucella strains is nearly identical to the 16S rrna gene sequences of the B. inopinata BO1 T and BO2 strains, with the exception of one additional base substitution (T to C at position 308), representing another unique 16S rrna sequence in the Brucella genus. The phylogenetic similarity at the 16S rrna gene locus among the Australian rodent Brucella strains, including Brucella sp. strain 83-13 and B. inopinata BO1 T and BO2, led us to believe that these rodent strains may represent an ancestor close to the B. inopinata lineage. By reca gene sequence analysis, however, the rodent Brucella strains had a higher degree of genetic identity to the common Brucella sp. consensus sequence than to the sequence of either BO1 T or BO2, positioning these Australian rodent strains closer to the classical Brucella spp. than to the novel B. inopinata lineage. Clustering analysis at both the 16S rrna and the reca gene loci demonstrated that these rodent strains reside within the Brucella subclade, as opposed to grouping with the Ochrobactrum spp. included in the analysis. We also analyzed the rpob gene, the rifampin antibiotic target in prokaryotes, which has been shown by Marianelli et al. (26) to differentiate all the classical Brucella species and most biovars. Though the polymorphisms observed in the rpob gene do not confer rifampin resistance in the brucellae (26), the ropb gene is under different selective pressures, making this locus an interesting candidate for evaluating the phylogenetic relationships of BO1 T, BO2, and the novel rodent strains. The rpob gene in the rodent strains showed greater genetic identity to strain BO1 T than to strain BO2 or any of the other Brucella spp. Finally, by MLSA, the rodent strains, including Brucella sp. strain 83-13, displayed several rodent group-specific SNPs (n 25), causing 10 missense mutations, compared with the sequences of the other described Brucella STs and strains BO1 T and BO2. MLSA clustering analysis oriented the rodent strains in the B. inopinata BO1 and BO2 clade, distanced from the classical sequence types. Although multiple conserved genes have been used for phylogenetic analysis, it is known that horizontal gene transfer can generate conflicting dendrograms (38). Ultimately, whole-genome SNP analysis will provide the most accurate phylogenetic picture and give more insight into which genes best reflect the relationships within the brucellae (15). We performed whole-genome MUMi analysis with a limited set of genomes intended to represent those from species also isolated from rodents (B. neotomae, B. microti, and Brucella sp. strain 83-13) and B. suis biovar 3 (which was the original identification by Cook et al. [6]), and we included the genomes of BO1 T and BO2 because of their genuine diversity. MUMi estimations have been shown to significantly correlate with MLSA similarity matrices as well as with the average nucleotide index calculations; it has been considered an alternative tool for DNA-DNA hybridization studies (13, 19). In our calculations, the MUMi estimation results comparing a representative strain of the rodent group (strain NF 2653) shows significant genomic similarity to Brucella strain 83-13 and considerable distancing from the other Brucella spp. included in the comparison. These distancing calculations corroborate well with our MLSA results and provide very strong evidence that this group of rodent strains from Australia represents a new Brucella sp. In addition to the seven strains that we described in this work, we also received and characterized four additional Australian rodent strains (data not shown). These four additional isolates were genetically closely related to the group of seven strains that we describe in this paper, with each strain having a unique MLVA genotype and falling within the rodent strain subgroup. Though their origin in Queensland, Australia, is unknown, we suspect that these additional four strains correspond to the isolates referenced in the addendum of the paper of Cook et al. (6), which came from a retrapping in the Jordan Creek, Australia, area 1 year later, in 1965. Throughout our molecular analysis we have observed that Brucella strain 83-13, submitted to The Broad Institute by Adrian Whatmore, and B.

5844 TILLER ET AL. APPL. ENVIRON. MICROBIOL. suis strain 83-210 (26) are nearly identical to the strains from Jordan Creek that we have characterized. Through a personal communication with Adrian Whatmore of VLA (Surrey, United Kingdom) and Axel Cloeckaert of the Institut National de la Recherché Agronomique (INRA), Nouzilly, France, there is reason to speculate that these two strains (Brucella sp. strain 83-13 and B. suis 83-210) may belong to the Jordan Creek strains, although there are no formal records stating that this is the case. The publication of Cook et al. describing the animal hosts, the specific geographic origin, and the date and methods of isolation of the rodent strains that we describe in this work significantly improves our understanding of the epizootological role and ecological position of this novel group of Brucella strains found in the wild (6). Our phenotypic and molecular characterization of these seven Australian rodent Brucella strains strongly suggests that they are a new atypical Brucella species, representing yet another example of the expansion of both genetic and ecological diversity in the Brucella genus. Additional whole-genome SNP analysis and comparisons have begun to provide a better understanding of the relationships between these recently described species and how they relate to classical brucellae and their close relatives. Further work on the ecological persistence and distribution of these strains, their epizootological role in wild rodents, and their virulence and pathogenicity in other animal hosts is needed. REFERENCES 1. Alton, G. G., L. M. Jones, and D. E. Pietz. 1975. Laboratory techniques in brucellosis. Monogr. Ser. World Health Org., p. 1 163. 2. Boeer, W. J., R. P. Crawford, R. J. Hidalgo, and R. M. Robinson. 1980. Small mammals and white-tailed deer as possible reservoir hosts of Brucella abortus in Texas. J. Wildl. Dis. 16:19 24. 3. Boschiroli, M. L., V. Foulongne, and D. O Callaghan. 2001. Brucellosis: a worldwide zoonosis. Curr. Opin. Microbiol. 4:58 64. 4. Chu, M. C., and R. S. Weyant. 2003. Francisella and Brucella, p. 789 808. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of clinical microbiology, vol. 1, 8th ed. ASM Press, Washington, DC. 5. Cloeckaert, A., M. Grayon, O. Grepinet, and K. S. Boumedine. 2003. Classification of Brucella strains isolated from marine mammals by infrequent restriction site-pcr and development of specific PCR identification tests. Microbes Infect. 5:593 602. 6. Cook, I., R. W. Campbell, and G. Barrow. 1966. Brucellosis in North Queensland rodents. Aust. Vet. J. 42:5 8. 7. Corbel, M. J. 1997. Brucellosis: an overview. Emerg. Infect. Dis. 3:213 221. 8. Corbel, M. J. 1997. Recent advances in brucellosis. J. Med. Microbiol. 46:101 103. 9. Corbel, M. J., and W. J. Brinley-Morgan. 1984. Genus Brucella Meyer and Shaw 1920, 173 AL, p. 377 388. In N. R. Krieg et al. (ed.), Bergey s manual of systemic bacteriology, vol. 1. The Williams & Wilkins Co., Baltimore, MD. 10. Davis, C. E., and S. B. Troy. 2005. Brucellosis. N. Engl. J. Med. 353:1071 1072. 11. Davis, D. S. 1990. Brucellosis in wildlife, p. 321 334. In K. Nielsen and J. R. Duncan (ed.), Animal brucellosis. CRC Press, Boca Raton, FL. 12. De, B. K., L. Stauffer, M. S. Koylass, S. E. Sharp, J. E. Gee, L. O. Helsel, A. G. Steigerwalt, R. Vega, T. A. Clark, M. I. Daneshvar, P. P. Wilkins, and A. M. Whatmore. 2008. Novel Brucella strain (BO1) associated with a prosthetic breast implant infection. J. Clin. Microbiol. 46:43 49. 13. Deloger, M., M. El Karoui, and M. A. Petit. 2009. A genomic distance based on MUM indicates discontinuity between most bacterial species and genera. J. Bacteriol. 191:91 99. 14. DelVecchio, V. G., V. Kapatral, R. J. Redkar, G. Patra, C. Mujer, T. Los, N. Ivanova, I. Anderson, A. Bhattacharyya, A. Lykidis, G. Reznik, L. Jablonski, N. Larsen, M. D Souza, A. Bernal, M. Mazur, E. Goltsman, E. Selkov, P. H. Elzer, S. Hagius, D. O Callaghan, J. J. Letesson, R. Haselkorn, N. Kyrpides, and R. Overbeek. 2002. The genome sequence of the facultative intracellular pathogen Brucella melitensis. Proc. Natl. Acad. Sci. U. S. A. 99:443 448. 15. Ficht, T. 2010. Brucella taxonomy and evolution. Future Microbiol. 5:859 866. 16. Fitch, C. P., and L. M. Bishop. 1938. The wild rat as a host of Brucella abortus. Cornell Vet. 28:304. 17. Foster, J. T., R. T. Okinaka, R. Svensson, K. Shaw, B. K. De, R. A. Robison, W. S. Probert, L. J. Kenefic, W. D. Brown, and P. Keim. 2008. Real-time PCR assays of single-nucleotide polymorphisms defining the major Brucella clades. J. Clin. Microbiol. 46:296 301. 18. Gee, J. E., B. K. De, P. N. Levett, A. M. Whitney, R. T. Novak, and T. Popovic. 2004. Use of 16S rrna gene sequencing for rapid confirmatory identification of Brucella isolates. J. Clin. Microbiol. 42:3649 3654. 19. Goris, J., K. T. Konstantinidis, J. A. Klappenbach, T. Coenye, P. Vandamme, and J. M. Tiedje. 2007. DNA-DNA hybridization values and their relationship to whole-genome sequence similarities. Int. J. Syst. Evol. Microbiol. 57:81 91. 20. Halling, S. M., B. D. Peterson-Burch, B. J. Bricker, R. L. Zuerner, Z. Qing, L. L. Li, V. Kapur, D. P. Alt, and S. C. Olsen. 2005. Completion of the genome sequence of Brucella abortus and comparison to the highly similar genomes of Brucella melitensis and Brucella suis. J. Bacteriol. 187:2715 2726. 21. Halling, S. M., F. M. Tatum, and B. J. Bricker. 1993. Sequence and characterization of an insertion sequence, IS711, from Brucella ovis. Gene 133: 123 127. 22. Hatipoglu, C. A., G. Bilgin, N. Tulek, and U. Kosar. 2005. Pulmonary involvement in brucellosis. J. Infect. 51:116 119. 23. Huynh, L. Y., M. N. Van Ert, T. Hadfield, W. S. Probert, B. H. Bellaire, M. Dobson, R. J. Burgess, R. S. Weyant, T. Popovic, S. Zanecki, D. M. Wagner, and P. Keim. 2008. Multiple locus variable number tandem repeat (VNTR) analysis (MLVA) of Brucella spp. identifies species specific markers and insights into phylogenetic relationships, p. 47 54. In V. S. Georgiev, K. Western, and J. J. McGowan (ed.), National Institute of Allergy and Infectious Diseases, NIH: frontiers in research, vol. 1. Human Press, Totowa, NJ. 24. Kumar, S., M. Nei, J. Dudley, and K. Tamura. 2008. MEGA: a biologistcentric software for evolutionary analysis of DNA and protein sequences. Brief. Bioinform. 9:299 306. 25. Margulies, M., M. Egholm, W. E. Altman, S. Attiya, J. S. Bader, L. A. Bemben, J. Berka, M. S. Braverman, Y. J. Chen, Z. Chen, S. B. Dewell, L. Du, J. M. Fierro, X. V. Gomes, B. C. Godwin, W. He, S. Helgesen, C. H. Ho, G. P. Irzyk, S. C. Jando, M. L. Alenquer, T. P. Jarvie, K. B. Jirage, J. B. Kim, J. R. Knight, J. R. Lanza, J. H. Leamon, S. M. Lefkowitz, M. Lei, J. Li, K. L. Lohman, H. Lu, V. B. Makhijani, K. E. McDade, M. P. McKenna, E. W. Myers, E. Nickerson, J. R. Nobile, R. Plant, B. P. Puc, M. T. Ronan, G. T. Roth, G. J. Sarkis, J. F. Simons, J. W. Simpson, M. Srinivasan, K. R. Tartaro, A. Tomasz, K. A. Vogt, G. A. Volkmer, S. H. Wang, Y. Wang, M. P. Weiner, P. Yu, R. F. Begley, and J. M. Rothberg. 2005. Genome sequencing in microfabricated high-density picolitre reactors. Nature 437:376 380. 26. Marianelli, C., F. Ciuchini, M. Tarantino, P. Pasquali, and R. Adone. 2006. Molecular characterization of the rpob gene in Brucella species: new potential molecular markers for genotyping. Microbes Infect. 8:860 865. 27. Meyer, M. E. 1974. Advances in research on brucellosis, 1957 1972. Adv. Vet. Sci. Comp. Med. 18:231 250. 28. Osterman, B., and I. Moriyon. 2006. International Committee on Systematics of Prokaryotes; Subcommittee on the Taxonomy of Brucella: minutes of the meeting, 17 September 2003, Pamplona, Spain. Int. J. Syst. Evol. Microbiol. 56:1175. 29. Pappas, G., P. Papadimitriou, N. Akritidis, L. Christou, and E. V. Tsianos. 2006. The new global map of human brucellosis. Lancet Infect. Dis. 6:91 99. 30. Paquet, J. Y., M. A. Diaz, S. Genevrois, M. Grayon, J. M. Verger, X. de Bolle, J. H. Lakey, J. J. Letesson, and A. Cloeckaert. 2001. Molecular, antigenic, and functional analyses of Omp2b porin size variants of Brucella spp. J. Bacteriol. 183:4839 4847. 31. Paulsen, I. T., R. Seshadri, K. E. Nelson, J. A. Eisen, J. F. Heidelberg, T. D. Read, R. J. Dodson, L. Umayam, L. M. Brinkac, M. J. Beanan, S. C. Daugherty, R. T. Deboy, A. S. Durkin, J. F. Kolonay, R. Madupu, W. C. Nelson, B. Ayodeji, M. Kraul, J. Shetty, J. Malek, S. E. Van Aken, S. Riedmuller, H. Tettelin, S. R. Gill, O. White, S. L. Salzberg, D. L. Hoover, L. E. Lindler, S. M. Halling, S. M. Boyle, and C. M. Fraser. 2002. The Brucella suis genome reveals fundamental similarities between animal and plant pathogens and symbionts. Proc. Natl. Acad. Sci. U. S. A. 99:13148 13153. 32. Schlabritz-Loutsevitch, N. E., A. M. Whatmore, C. R. Quance, M. S. Koylass, L. B. Cummins, E. J. Dick, Jr., C. L. Snider, D. Cappelli, J. L. Ebersole, P. W. Nathanielsz, and G. B. Hubbard. 2009. A novel Brucella isolate in association with two cases of stillbirth in non-human primates first report. J. Med. Primatol. 38:70 73. 33. Scholz, H. C., E. Hofer, G. Vergnaud, P. Le Fleche, A. M. Whatmore, S. Al Dahouk, M. Pfeffer, M. Kruger, A. Cloeckaert, and H. Tomaso. 2009. Isolation of Brucella microti from mandibular lymph nodes of red foxes, Vulpes vulpes, in lower Austria. Vector Borne Zoonotic Dis. 9:153 156. 34. Scholz, H. C., Z. Hubalek, J. Nesvadbova, H. Tomaso, G. Vergnaud, P. Le Fleche, A. M. Whatmore, S. Al Dahouk, M. Kruger, C. Lodri, and M. Pfeffer. 2008. Isolation of Brucella microti from soil. Emerg. Infect. Dis. 14:1316 1317. 35. Scholz, H. C., Z. Hubalek, I. Sedlacek, G. Vergnaud, H. Tomaso, S. Al Dahouk, F. Melzer, P. Kampfer, H. Neubauer, A. Cloeckaert, M. Maquart, M. S. Zygmunt, A. M. Whatmore, E. Falsen, P. Bahn, C. Gollner, M. Pfeffer, B. Huber, H. J. Busse, and K. Nockler. 2008. Brucella microti sp. nov.,

VOL. 76, 2010 NOVEL BRUCELLA STRAINS FROM WILD RODENTS IN AUSTRALIA 5845 isolated from the common vole Microtus arvalis. Int. J. Syst. Evol. Microbiol. 58:375 382. 36. Scholz, H. C., K. Nockler, C. Gollner, P. Bahn, G. Vergnaud, H. Tomaso, S. Al-Dahouk, P. Kampfer, A. Cloeckaert, M. Maquart, M. S. Zygmunt, A. M. Whatmore, M. Pfeffer, B. Huber, H. J. Busse, and B. K. De. 2010. Brucella inopinata sp. nov., isolated from a breast implant infection. Int. J. Syst. Evol. Microbiol. 60:801 808. 37. Scholz, H. C., H. Tomaso, S. A. Dahouk, A. Witte, M. Schloter, P. Kampfer, E. Falsen, and H. Neubauer. 2006. Genotyping of Ochrobactrum anthropi by reca-based comparative sequence, PCR-RFLP, and 16S rrna gene analysis. FEMS Microbiol. Lett. 257:7 16. 38. Sicheritz-Ponten, T., and S. G. Andersson. 2001. A phylogenomic approach to microbial evolution. Nucleic Acids Res. 29:545 552. 39. Stoenner, H. G., and D. B. Lackman. 1957. A preliminary report on a Brucella isolated from the desert wood rat, Neotoma lepida Thomas. J. Am. Vet. Med. Assoc. 130:411 412. 40. Theegarten, D., S. Albrecht, M. Totsch, H. Teschler, H. Neubauer, and S. Al Dahouk. 2008. Brucellosis of the lung: case report and review of the literature. Virchows Arch. 452:97 101. 41. Thorne, E. T. 2001. Brucellosis, p. 372 395. In E. S. Williams and I. K. Barker (ed.), Infectious diseases of wild mammals, 3rd ed. Wiley-Blackwell, Ames, IA. 42. Thorpe, B. D., R. W. Sidwell, J. B. Bushman, K. L. Smart, and R. Moyes. 1965. Brucellosis in wildlife and livestock of west central Utah. J. Am. Vet. Med. Assoc. 146:225 232. 43. Tiller, R. V., B. K. De, M. Boshra, L. Y. Huynh, M. N. Van Ert, D. M. Wagner, J. Klena, T. S. Mohsen, S. S. El-Shafie, P. Keim, A. R. Hoffmaster, P. P. Wilkins, and G. Pimentel 2009. Comparison of two multiple locus variable number tandem repeat (VNTR) analysis (MLVA) methods for molecular strain typing human Brucella melitensis isolates from the Middle East. J. Clin. Microbiol. 47:2226 2231. 44. Tiller, R. V., J. E. Gee, D. R. Lonsway, S. Gribble, S. C. Bell, A. Jennison, J. Bates, C. Coulter, A. R. Hoffmaster, and B. K. De. 2010. Identification of an unusual Brucella strain (BO2) from a lung biopsy in a 52-year old patient with chronic destructive pneumonia. BMC Microbiol. 10:23. 45. Trout, D., T. M. Gomez, B. P. Bernard, C. A. Mueller, C. G. Smith, L. Hunter, and M. Kiefer. 1995. Outbreak of brucellosis at a United States pork packing plant. J. Occup. Environ. Med. 37:697 703. 46. Verger, J. M., F. Grimont, P. A. Grimont, and M. Grayon. 1987. Taxonomy of the genus Brucella. Ann. Inst. Pasteur Microbiol. 138:235 238. 47. Whatmore, A. M. 2009. Current understanding of the genetic diversity of Brucella, an expanding genus of zoonotic pathogens. Infect. Genet. Evol. 9:1168 1184. 48. Whatmore, A. M., L. L. Perrett, and A. P. MacMillan. 2007. Characterisation of the genetic diversity of Brucella by multilocus sequencing. BMC Microbiol. 7:34. 49. Young, E. J. (ed.). 2005. Brucella species. Elsevier, Philadelphia, PA. 50. Young, E. J., and M. J. Corbel. 1989. Brucellosis: clinical and laboratory aspects. CRC Press, Boca Raton, FL. Downloaded from http://aem.asm.org/ on June 16, 2018 by guest