Sylvia Valdezate,* Ana Navarro, Pilar Villalón, Gema Carrasco, and Juan A. Saéz-Nieto

JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 2010, p. 2734 2740 Vol. 48, No. 8 0095-1137/10/$12.00 doi:10.1128/jcm.00533-10 Copyright 2010, American Society for Microbiology. All Rights Reserved. Epidemiological and Phylogenetic Analysis of Spanish Human Brucella melitensis Strains by Multiple-Locus Variable-Number Tandem-Repeat Typing, Hypervariable Octameric Oligonucleotide Fingerprinting, and rpob Typing Sylvia Valdezate,* Ana Navarro, Pilar Villalón, Gema Carrasco, and Juan A. Saéz-Nieto Servicio de Bacteriología, Centro Nacional de Microbiología, Instituto de Salud Carlos III, Majadahonda, 28220 Madrid, Spain Received 12 March 2010/Returned for modification 14 May 2010/Accepted 8 June 2010 The severe morbidity of human brucellosis is one of the main reasons for using molecular typing in the epidemiological surveillance of this worldwide zoonosis. Multiple-locus variable-number repeat analysis (MLVA-16), hypervariable octameric oligonucleotide fingerprinting (HOOF-print), and the differences in the single nucleotide polymorphisms (SNPs) (codons 1249 and 1309) of the DNA-dependent RNA polymerase subunit (rpob) were used to type a human Brucella melitensis population (108 strains) collected from throughout Spain over 13 years. Eighty-six MLVA types (discriminatory index, 0.99) were detected, with a wide-ranging genetic similarity coefficient (37.2 to 93.7%). The population clustered into the following groups: American, with genotypes 47 (1 strain), 48 (13 strains), 53 (12 strains), 55 (2 strains), 80 (1 strain), and a new genotype (2 strains), Western Mediterranean, with genotype 51 (9 strains), and Eastern Mediterranean, with genotypes 42 (60 strains), 43 (4 strains), and 63 (4 strains). Two profession-related and two foodborne acquisitions were confirmed. Distributed throughout Spain, Eastern Mediterranean genotype 42 was the most common (55%). The low MLVA-16 allelic polymorphism (genetic similarity range, 75 to 94%) of the genotype 42 strains suggests that they recently evolved from a common ancestor. rpob typing grouped the strains as rpob type 1 (1249-ATG/1309-CTG; 28.7%), rpob type 2 (1249-ATG/1309-CTA; 62.9%), and rpob type 3 (1249-ATA/1309- CTG; 8.3%). According to the MLVA-16 results, the population clustered by rpob type. Given the correlation between B. melitensis MLVA groups and rpob types (American and rpob type 1, Eastern Mediterranean and rpob type 2, and Western Mediterranean and rpob type 3), the rpob type could be used as an initial marker for the epidemiological surveillance of brucellosis. Downloaded from http://jcm.asm.org/ * Corresponding author. Mailing address: Servicio de Bacteriología Centro Nacional de Microbiología, Instituto de Salud Carlos III, Majadahonda, Crta. Majadahonda-Pozuelo km 2, 28220 Madrid, Spain. Phone: 34 91 8223734 Fax: 34 91 509 7966. E-mail: svaldezate@isciii.es. Published ahead of print on 16 June 2010. Brucellosis is an anthropozoonosis with a worldwide distribution. Brucella spp. are responsible for considerable economic losses through their infection of livestock and cause severe morbidity in humans, involving multiple organs (7, 23, 24, 31). However, given the absence of specific signs and symptoms, the disease is commonly underdiagnosed (24, 31). Current global trends of brucellosis incidence show there to be nearly half a million new cases annually (31). Large differences are seen, however, between Spain and other European countries, with 3.17 cases/100,000 population in Greece and just 0.034/100,000 in France (http://www.oie.int/wahis/public.php) (33). Since 2005, the annual incidence of human brucellosis in Spain has decreased sharply (15, 16, 25), to 0.33/ 100,000 (151 estimated cases) in 2009 (Spanish Surveillance Net, Epidemiology National Centre). The reduction in human brucellosis in Europe has been attributed to effective preventive measures followed in livestock practices and veterinary medicine but also in food processing and public health monitoring (8). Certainly, typing techniques have been of great use in the epidemiological surveillance of human and animal disease. Based on tandem repeats, both hypervariable octameric oligonucleotide fingerprinting (HOOF-prints) (5) and multilocus variable-number tandem-repeat analysis (MLVA-16) (1, 18) have proven their discriminatory power in the study of Brucella species collected on a global scale (18, 19, 32). Recent studies have shown the latter technique to be useful in studies of different geographic and time sampling characteristics involving B. abortus (13) and B. melitensis (17, 21, 22, 28, 29). Spain is located in the Mediterranean basin, an acknowledged region of endemicity for brucellosis (24), where B. melitensis constitutes the major species responsible for human disease (30). This study examined a population of human B. melitensis strains from different areas of Spain collected over a period of 13 years (from 1997 to 2009). The aims of this work were the following: (i) to assess the usefulness of MLVA-16 in B. melitensis typing, (ii) to analyze the genotypes circulating in our country, (iii) to establish, via MLVA-16 and HOOF-print analysis, the occupational, laboratory, travel, and food origins of the disease in the patients, and (iv) to examine the potential of the DNA-dependent RNA polymerase subunit (rpob) for the identification and typing (21, 26) of Spanish B. melitensis strains. MATERIALS AND METHODS Bacterial strains. The studied population included 108 B. melitensis strains isolated from clinical samples (101 from blood, 4 from synovial fluid, 1 from a testicular abscess, 1 from a cerebral abscess, and 1 from a heart valve). These strains were collected from 1997 to 2009 in 29 Spanish provinces. This population on November 1, 2018 by guest 2734

VOL. 48, 2010 SPANISH CLINICAL BRUCELLA MELITENSIS GENOTYPES 2735 included strains suspected of being epidemiologically related. The different strains were isolated and identified by standard procedures (3). Genomic DNA for PCR assays was extracted from fresh cultures by boiling the crude lysate. Species-level identification was undertaken by B. abortus, B. melitensis, B. ovis, and B. suis PCR (AMOS-PCR) (4). rpob type assignment. The rpob types previously correlated with the three B. melitensis phenotypic biovars (20) were identified by rpob molecular targeting of the specific codon residues 985, 1249, and 1309 (in agreement with B. melitensis 16M numbering [EMBL accession no. AE009516]). Amplifications were performed using the previously described primers rbseq7 and 4143rB and by sequencing with rbseq7 and rbseq9 (20). MLVA-16 and HOOF-print typing. Sixteen tandem-repeat loci (TR) were identified in the 108 strains by previously described methods for MLVA (1, 18). TRs were grouped into 3 panels: panel 1, with 8 minisatellite loci (the bruce06, bruce08, bruce11, bruce12, bruce42, bruce43, bruce45, and bruce55 loci), panel 2A, with 3 microsatellite loci (the bruce18, bruce19, and bruce21 loci), and panel 2B, with 5 microsatellite loci (the bruce04, bruce07, bruce09, bruce16, and bruce30 loci). The bruce04, bruce09, and bruce30 markers of the last panel are equivalent to TR6, TR8, and TR2 in HOOF-print analysis. All strains were also typed by the HOOF-print method according to a previously described protocol (5, 30). The HOOF-prints of 70 strains collected in the period from 1999 to 2005 were known from earlier work (30). Sequence analysis. PCR products were purified using the GFX PCR DNA and gel band purification kit (Amersham Biosciences, Buckinghamshire, United Kingdom). Sequences were obtained using an ABI Prism 377 sequencer (Perkin- Elmer, Applied Biosystems Division, Foster City, CA) employing the BigDye Terminator protocol (Applied Biosystems). The sequences were assembled using the Lasergene Seqman II software program (DNAStar, Inc., Madison, WI) and aligned, and the amino acids were deduced using the CLUSTAL W routine of the MegAlign software program (v.6.1; DNAStar, Inc.). For MLVA-16 and HOOF-print typing, the number of repeats or alleles was estimated from their molecular sizes and sequences (1, 5, 18, 30). For each strain, the combination of the allelic profile of the MLVA-16 and HOOF-print markers was used to define genotypes. According to Al-Dahouk et al. (1), different weightings were given to each MLVA-16 panel (2 for panel 1, 1 for panel 2A, and 0.2 for panel 2B). The genetic relationships of the examined population were determined by clustering the corresponding character data sets for each strain using the InfoQuest program of the Bionumerics v. 4.5 software package (Applied Maths, Saint-Martens-Latem, Belgium) and employing different algorithms. As in previous MLVA-16 studies of Brucella sp. strains (1, 10, 17, 18, 28, 30, 32), the first clustering analysis was based on the use of the categorical coefficient and the unweighted pair group method with arithmetic averages (UPGMA). The second clustering was undertaken using the summed absolute distance employing the coefficient for calculating the minimum spanning tree (MST) (27). On the basis of the categorical coefficient, the MST was constructed with priority rules (to first link types with the largest number of single- and/or double-locus variants, the highest number of entries), allowing the creation of hypothetical types. Genetic diversity. The discriminatory capacity of individual or combined MLVA-16 markers was calculated using the Hunter-Gaston diversity index (HGDI) with 95% confidence intervals (12, 14) employing TR diversity and confidence extractor software (V-DICE), available at the Health Protection Agency s bioinformatic tools website (http://www.hpa-bioinformatics.org.uk/cgi -bin/dici/dici.pl). The HGDI is a measure of the variation of the TR copy number at each marker, ranging from 0 (no diversity) to 1 (highest diversity). The confidence interval indicates the precision of the HGDI by providing the upper and lower boundaries. RESULTS rpob types determined by SNPs 1249 1309. Three B. melitensis rpob types were assigned according to the combination of single nucleotide polymorphisms (SNPs) in codons 1249 and 1309: rpob type 1, 1249-Met (ATG) and 1309-Leu (CTG); rpob type 2, 1249-Met (ATG) and 1309-Leu (CTA); or rpob type 3, 1249-Ile (ATA) and 1309-Leu (CTG). The strain frequencies for these types, hereinafter referred to as rpob types 1, 2, and 3, were 28.7%, 62.9%, and 8.3%, respectively. No spatial grouping of these types was detected. Of the 68 rpob type 2 TABLE 1. Polymorphism index of individual or combined TR markers in a clinical population of B. melitensis a Panel and VNTR locus No. of alleles (allele s ) Diversity index b (HGDI) Confidence interval c MLVA-16 86 0.993 0.988 0.999 Allele mode (Max pi d ) Panel 1 10 0.694 0.613 0.775 bruce06 2 (1, 3) 0.471 0.424 0.518 1 (0.630) bruce08 2 (4, 5) 0.413 0.340 0.486 5 (0.713) bruce11 2 (2, 3) 0.037 0.000 0.086 3 (0.981) bruce12 1 (13) 0.000 0.000 0.065 13 (1.000) bruce42 6 (1 6) 0.597 0.516 0.678 2 (0.583) bruce43 3 (1 3) 0.428 0.331 0.525 2 (0.731) bruce45 1 (3) 0.000 0.000 0.065 3 (1.000) bruce55 2 (2 3) 0.471 0.424 0.518 2 (0.630) Panel 2A 23 0.703 0.617 0.789 bruce18 6 (3 7, 9) 0.526 0.432 0.621 5 (0.657) bruce19 6 (5, 18, 20 22, 25) 0.551 0.472 0.630 20 (0.611) bruce21 4 (5 8) 0.496 0.421 0.570 7 (0.648) Panel 2B 80 0.988 0.980 0.996 bruce04-tr6 6 (2 6, 9) 0.786 0.764 0.808 2 (0.287) bruce07 7 (2 8) 0.727 0.670 0.785 6 (0.444) bruce09-tr8 8 (3 9, 12,13) 0.443 0.328 0.557 3 (0.741) bruce16 8 (3 10) 0.708 0.636 0.780 3 (0.491) bruce30-tr2 6 (1, 3 7) 0.607 0.528 0.686 4 (0.574) a Markers were studied by MLVA-16 typing; n 108. b Hunter and Gaston index. c Precision of the diversity index, expressed as 95% upper and lower boundaries. d Fraction of samples that have the most frequent repeat number in this marker (range, 0.0 to 1.0). strains, three showed the missense mutation 985-Ala (GCC) instead of the corresponding Val (GTC) (20, 26). Typing and clustering of B. melitensis population by MLVA- 16. A total of 86 MLVA-16 types were identified among the 108 human strains examined. The corresponding diversity index (estimated from the HGDI) for panels 1, 2A, and 2B were 0.69, 0.70, and 0.98, respectively. The overall discriminatory index of MLVA-16 in this population was 0.99. Table 1 shows the number of alleles, the diversity coefficient, the confidence interval, the allele mode, and the frequency of mode (maximum [max] pi, or the fraction of samples harboring the modal value) for each marker and panel. Half of these markers had 5 different alleles. The most discriminatory markers were bruce04, bruce07, and bruce16 of panel 2B, with a diversity index of 0.71 and harboring 6, 7, and 8 alleles, respectively. The most homogeneous markers, in contrast, were bruce12, bruce11, and bruce45 of panel 1, with 1, 2, and 2 alleles, respectively. The identification of the strains, their geographical origins, the rpob type, the clinical specimen, and the panel 1 genotype are indicated in the dendrogram in Fig. 1. The UPGMA clustering analysis showed the genetic similarity to range from 37.2 to 100.0%. For the different MLVA types, the maximum similarity was 93.7% for single locus variants (SLVs). No link was identified among the MLVA-16 clusters and the geographical location or year of isolation. Using panel 1, the present population clustered into nine known genotypes and one new genotype. The nine known genotypes were included in the three previously named B. melitensis groups (1): the American group with genotypes 47 (1 strain), 48 (13 strains), 53 (12 strains), 55 (2 strains), and 80 (1 strain), the Western

2736

VOL. 48, 2010 SPANISH CLINICAL BRUCELLA MELITENSIS GENOTYPES 2737 FIG. 2. Minimum-spanning tree of the studied B. melitensis strains clustering according to their rpob types (MLVA-16 method). Each circle in the tree represents a different MLVA-16 type. Heavy, short lines connecting two MLVA-16 types denote types differing by a single MLVA-16 locus. Thinner, longer lines connect double-locus variants. Dotted lines indicate the most likely connection between two types differing by more than two MLVA-16 loci. The connections between types differing in more than a single locus are less likely to represent an evolutionary relationship but indicate the connection to the most similar type. The size of the circle indicates the number of strains with this particular MLVA-16 type. Downloaded from http://jcm.asm.org/ Mediterranean group, with genotype 51 (9 strains), and the Eastern Mediterranean group, with genotypes 42 (60 strains), 43 (4 strains), and 63 (4 strains). Two strains assigned to the American cluster were collected in two provinces of the Valencia Region (in 2003 and 2007) and showed a new panel 1 genotype (3 4 3 13 6 2 3 3), an SLV of genotype 55 of the bruce42 marker. Correlation among MLVA-16 panel 1 clusters and rpob types. The rpob type distribution was the following: rpob type 1, with 30 MLVA-16 types, rpob type 2, with 48 MLVA-16 types, and rpob-type 3, with 8 MLVA-16 types. By way of complementary analysis, the genetic relationships between the clinical B. melitensis strains were deduced by construction of an MST (Fig. 2). Three separate clusters were observed for each described B. melitensis group (the American, Eastern Mediterranean, and Western Mediterranean groups), each one gathering strains of identical rpob type. The rpob type 1 strains harbored the American genotypes 47, 48, 53, 55, and 80, the rpob type 2 strains harbored the Eastern Mediterranean genotypes 42, 43, and 63, and the rpob type 3 strains harbored the Western Mediterranean genotype 51. Correlation between MLVA-16 and HOOF-print typing results. A total of 104 different HOOF-prints were obtained for the studied population, showing a diversity index of 0.99 (confidence interval, 0.989 to 0.991; allele mode, 0.019). In the HOOF dendrogram constructed using the UPGMA method (Fig. 3), the different HOOF types showed a similarity range of 22.7 to 87.5%. In contrast to that seen for the MLVA-16 types, the HOOF-prints showed no gatherings with respect to rpob type or B. melitensis group in the MST model (data not shown). With respect to diversity, MLVA-16 typing showed 36 SLVs and double-locus variants (DLVs) in 48 strains of the largest group, i.e., Eastern Mediterranean/rpoB type 2. HOOF-typing increased the diversity recorded, with strains showing a smaller number of SLVs. Epidemiological tracing. Eleven MLVA-16 types were shared by 32 strains isolated in 12 Spanish provinces, with 9 out of those 11 belonging to genotype 42 (Fig. 1 and Table 2). Two sets of strains isolated in the same period in Girona (rpob type 3, genotype 51) and in A Coruña (rpob type 2, genotype 42) showed the same HOOF-prints (Fig. 3). For these strains, sets A and B in Table 2, both typing methods confirmed a work- on November 1, 2018 by guest FIG. 1. Dendrogram of MLVA-16 genotype clusters (UPGMA method) corresponding to the 108 Spanish human B. melitensis strains and their groups (American, Eastern Mediterranean, and Western Mediterranean). The columns show the identification numbers of the strains (year of isolation), their geographic origin, the rpob types, the human sources, and the panel 1 genotypes. Letter codes A to K indicate those MLVA-16 types shared by 11 sets of strains.

2738 VALDEZATE ET AL. J. CLIN. MICROBIOL. related infection (laboratory and livestock activities, respectively). The suspicion of laboratory acquisition was completely ruled out by the comparisons of the rpob, MLVA, and HOOF types of a Brucella strain isolated from a patient (rpob type 2, genotype 42) and another strain isolated from a microbiologist (rpob type 1, genotype 48) with a background of international travel. The respective HOOF-prints of the remaining strains from sets C to K (Table 2) harbored from 1 to 4 different alleles. Unpasteurized goat cheese was the source of infection for all patients whose infecting strains belonged to sets G and I (from Palma del Río, Province of Córdoba, 1999 to 2000, 100 estimated cases) or F and H (from Coria, Province of Cacéres, 2004-2005, 20 laboratory-confirmed cases) (30). Sets G and I were SLVs of bruce30 (repeats 4 and 5), and sets F and H were SLVs of bruce04 (repeats 3 and 5). It should be noted that the strains of sets G and H were also SLVs, differing only in one repeat of marker bruce04 (Fig. 1). DISCUSSION The epidemiological surveillance of animal and human brucellosis has clearly benefited from the appearance and improvement of molecular typing. The worldwide distribution of brucellosis has required laboratories of different nations to try to apply the same typing techniques, thus facilitating the exchange of information (see http://mlva.u -psud.fr/brucella). One of the aims of the present work was to examine the capacity of MLVA-16 typing (1, 18) to discriminate between B. melitensis strains collected over a period of more than 10 years in 29 Spanish provinces. In this collection (n 108), the complete MLVA-16 assay provided very good discriminatory power (close to 1), differentiating 86 genotypes grouped into three main clusters. Unlike other tandem repeat-based methods, MLVA-16 typing requires a large number of markers grouped into three complementary panels. The low diversities of panels 1 and 2A in the present work (Table 1) (HGDI of 0.7) restrict their role to the analysis of B. melitensis phylogeny (17), being useful for characterization purposes when strains come from different parts of the world. However, the high polymorphism of the panel 2B markers (HGDI 0.98) (a consequence of their high mutation rate) (11) allowed the surveillance of brucellosis in the present samples from Spain. Thus, the diversity of panel 1 was low but was useful for identifying genotypes 42, 43, 47, 48, 51, 53, 55, 63, and 80. The Eastern Mediterranean genotype 42 was the most common (55%), followed by the American genotypes 48 and 53 ( 11%) and the Western Mediterranean genotype 51 ( 8%). Genotype 42, widely distributed throughout Spain, has previously been reported to be predominant in Turkey and Portugal (1). This major genotype showed low allelic polymorphism for the MLVA-16 markers, with a genetic FIG. 3. Dendrogram of HOOF types (UPGMA method) corresponding to the 108 Spanish human B. melitensis strains. Columns show the identification numbers of the strains (year of isolation), their geographic origins, the rpob types, the human sources, and the MLVA-16 panel 1 genotypes.

VOL. 48, 2010 SPANISH CLINICAL BRUCELLA MELITENSIS GENOTYPES 2739 TABLE 2. Epidemiological and HOOF-print characteristics of clinical B. melitensis strains with identical MLVA types a Set b (n c ) Geographical location Yr of isolation similarity range of 75 to 94%. This suggests a recent evolution of its strains from a common ancestor in Spain or limitations in the usefulness of MLVA-16 typing. Genotype 48, here detected in six regions, was first reported in animal strains (1) and later identified in a Spanish human biovar 1 strain (11), while genotype 51 is one of the most common in Italy (1). In contrast, genotype 43, which is common in Turkey and Greece, was detected at a low frequency ( 4%) in Spain, and genotype 49, which is common in Italy, was not found (1). Consultations of the Brucella 2009 MLVA database confirmed that genotypes 51, 53, 63, and 80 were first detected in the present strains. Genotype 80 was initially reported in two Peruvian strains isolated in 2000 (28). To date it has not been found in any other European country. When the MLVA-16 types were analyzed using the UPGMA and MST models (Fig. 1 and 2), a clear correlation of the three B. melitensis groups was observed, in agreement with the assigned rpob types. Each of the three clusters brought together strains that harbored the same rpob type: American/rpoB type 1, Eastern Mediterranean/rpoB type 2, and Western Mediterranean/rpoB type 3. Molecular markers for characterizing Brucella strains as rpob types (20) could be used to complement the information obtained by serological methods, the traditional means of Brucella sp. identification (3). As for other bacteria (9), the study of the detected polymorphisms in rpob provides a new tool for inter- and intraspecies-level identification and typing (20, 26). The procedure is easy to perform, which should enable its rapid uptake by laboratories. For the present B. melitensis strains, it established three rpob types on the basis of two selected SNPs (codons 1249 and 1309). rpob type 2 was the most prevalent (63%) and showed the least MLVA-16 diversity, followed by rpob type 1 and rpob type 3 (the minority type). A known variant of rpob type 2 (26) was identified, but this was much less common than in human strains collected in Turkey (77%). Despite their being isolated in different provinces, the present rpob type 2 variants gathered together in the MLVA-16 dendrogram (genotype 43). With just half of the markers required for MLVA-16 typing, HOOF-printing showed a high degree of intraspecific diversity (6, 32); it is therefore a powerful tool for typing of B. melitensis strains collected from single countries (2, 30). In the present rpob type panel 1 genotype HOOF-print VNTR with differences population, the allelic richness was especially marked in TRs 1, 4, 5, and 7 (13 to 16 alleles). In addition, the HOOF-prints showed more diversity for the main genotype 42 strains than did MLVA-16 typing. It may be recommendable to undertake cluster analysis for this genotype using the MLVA-16 and HOOF-print markers, since for those strains involved in outbreaks with identical MLVA-16 profiles, the alleles of HOOFprint TRs 1, 4, 5, and 7 differed markedly. This suggests that apparently related strains may in fact be less strongly linked, a consequence of the hypermutability of HOOF-print loci being in continuous microevolution (6). In addition, the role of rpob as an initial typing marker should be noted; its study allowed the present strains to be assigned to their corresponding B. melitensis groups. The MLVA-16 analysis of the present B. melitensis population showed a high diversity of genotypes. Thus, it can be ruled out that few strains are present even when disease incidence is high and might be thought to be caused by a single strain. The Eastern Mediterranean genotype 42 was the most common, with a low diversity of MLVA-16 types, suggesting its strains to have a common ancestor. The correlation found among B. melitensis groups and rpob types suggests the study of the SNPs of rpob as an initial marker of brucellosis may be of use in epidemiological surveillance. ACKNOWLEDGMENTS Genetic similarity index (%)/variants d A (2) Girona 2007 3/51 None 100 B (2) A Coruña 2005-2006 2/42 None 100 C (2) León, Burgos 2005 2007 2/42 1, 5, 7 62.5 D (2) Madrid 2003 2/42 4, 5 75/DLVs E (2) Cáceres, Madrid 1999 2001 2/42 1, 4, 5, 7 50 F (2) Cáceres, Ciudad Real 2004, 2007 2/42 1, 4, 5, 7 49.7 G (6) Córdoba 1999 2/42 1, 4, 5, 7 46.8 87.5/SLVs, DLVs H (9) Cáceres, Guipúzcoa, León, Córdoba 1997, 2000, 2004-2005, 2008 2/42 1, 4, 5, 7 49.7 100/SLVs, DLVs I (2) Córdoba 1999-2000 2/42 1, 5 75.0/DLVs J (2) Málaga, Teruel 2000, 2004 2/42 1, 4, 5, 7 45.1 K (2) Valencia 2004 1/53 1, 4, 5, 7 50.0 a n 32. b Set referred to in MLVA-16 UPGMA dendrogram (see Fig. 1). c n, no. of strains. d Detection of SLVs and DLVs. This work was funded by a grant to A.N. from the Instituto de Salud Carlos III (MPY1116/07). We are grateful to the CNM Biopolymers Unit for assistance in sequencing, to Adrian Burton for linguistic assistance in preparation of the manuscript, and to the MLVA database Brucella 2009 for help in coding the MLVA alleles and profiles. REFERENCES 1. Al Dahouk, S., P. Le Flèche, K. Nöckler, I. Jacques, M. Grayon, H. C. Scholz, H. Tomaso, G. Vernaud, and H. Neubauer. 2007. Evaluation of Brucella MLVA typing for human brucellosis. J. Microbiol. Methods 69:137 145. 2. Almendra, C., T. K. L. Silva, A. Beja-Pereira, A. C. Ferreira, L. Ferrão-Beck, M. I. de Sá, B. J. Bricker, and G. Luikart. 2009. HOOF-Print genotyping and haplotype inference discriminates among Brucella spp. isolates from a small spatial scale. Infect. Genet. Evol. 9:104 107. 3. Alton, G. G., L. M. Jones, R. D. Angus, and J. M. Verger. 1988. Techniques for the brucellosis laboratory. Institut National de la Recherche Agronomique, Paris, France.

2740 VALDEZATE ET AL. J. CLIN. MICROBIOL. 4. Bricker, B. J., and S. M. Halling. 1994. Differentiation of Brucella abortus bv. 1, 2, and 4, Brucella melitensis, Brucella ovis, and Brucella suis bv. 1 by PCR. J. Clin. Microbiol. 32:2660 2666. 5. Bricker, B. J., D. R. Ewalt, and S. M. Halling. 2003. Brucella HOOF-Prints : strain typing by multi-locus analysis of variable number tandem repeats (VNTRs). BMC Microbiol. 3:15. 6. Bricker, B. J., and D. R. Ewalt. 2005. Evaluation of the HOOF-Print assay for typing Brucella abortus strains isolated from cattle in the United States: results with four performance criteria. BMC Microbiol. 5:37. 7. Corbel, M. J. 1997. Brucellosis: an overview. Emerg. Infect. Dis. 3:213 221. 8. European Commission. 2009. Commission decision of 5 August 2009. Off. J. Eur. Union 52:L/204/39 42. 9. Fusco da Costa, A. R., M. L. Lopes, S. C. Leão, M. P. Schneider, M. S. de Sousa, P. N. Suffys, T. C. Corvelo, and K. V. Lima. 2009. Molecular identification of rapidly growing mycobacteria isolates from pulmonary specimens of patients in the State of Pará, Amazon region, Brazil. Diagn. Microbiol. Infect. Dis. 65:358 364. 10. García-Yoldi, D., P. Le Fleche, C. M. Marín, M. J. De Miguel, P. M. Muñoz, G. Vergnaud, and I. López-Goñi. 2007. Assessment of genetic stability of Brucella melitensis Rev. 1 vaccine strain by multiple-locus variable-number tandem repeat analysis. Vaccine 25:2858 2862. 11. García-Yoldi, D., P. Le Flèche, M. De Miguel, P. M. Muñoz, J. M. Blasco, Z. Cvetnik, C. M. Marín, G. Vergnaud, and I. López-Goñi. 2007. Comparison of multiple-locus variable-number tandem-repeat analysis with other PCRbased methods for typing Brucella suis isolates. J. Clin. Microbiol. 45:4070 4072. 12. Grundmann, H., S. Hori, and G. Tanner. 2001. Determining confidence intervals when measuring genetic diversity and the discriminatory abilities of typing methods for microorganisms. J. Clin. Microbiol. 39:4190 4192. 13. Her, M., S. I. Kang, D. H. Cho, Y. S. Cho, I. Y. Hwang, Y. R. Heo, S. C. Jung, and H. S. Yoo. 2009. Application and evaluation of the MLVA typing assay for the Brucella abortus strains isolated in Korea. BMC Microbiol. 9:230. 14. Hunter, P. R., and M. A. Gaston. 1988. Numerical index of the discriminatory ability of typing systems: an application of Simpson s index of diversity. J. Clin. Microbiol. 26:2465 2466. 15. Instituto de Salud Carlos III. 2008. Comentario epidemiológico de las Enfermedades de Declaración Obligatoria y Sistema de Información Microbiológica, España, Año 2007. Bol. Epidemiol. 16:85 96. http://www.isciii.es /htdocs/centros/epidemiologia/boletin_semanal/bes0814.pdf. 16. Instituto de Salud Carlos III. 2009. Comentario epidemiológico de las Enfermedades de Declaración Obligatoria y Sistema de Información Microbiológica, España, Año 2008. Bol. Epidemiol. 17:197. 17. Kattar, M. M., R. F. Jaafar, G. F. Araj, P. Le Flèche, G. M. Matar, R. Abi Rached, S. Khalife, and G. Vergnaud. 2008. Evaluation of a multilocus variable-number tandem-repeat analysis scheme for typing human Brucella isolates in a region of brucellosis endemicity. J. Clin. Microbiol. 46:3935 3940. 18. Le Flèche, P., I. Jacques, M. Grayon, S. Al Dahouk, P. Bouchon, F. Denoeud, K. Nöckler, H. Neubauer, L. A. Guilloteau, and G. Vergnaud. 2006. Evaluation and selection of tandem repeat assay for a Brucella MLVA typing assay. BMC Microbiol. 6:9. 19. Maquart, M., P. Le Flèche, G. Foster, M. Tryland, F. Ramisse, B. Djønne, S. Al Dahouk, I. Jacques, H. Neubauer, K. Walravens, J. Godfroid, A. Cloeckaert, and G. Vergnaud. 2009. MLVA-16 typing of 295 marine mammal Brucella isolates from different animal and geographic origins identifies 7 major groups within Brucella ceti and Brucella pinnipedialis. BMC Microbiol. 9:145. 20. 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. Microb. Infect. 8:860 865. 21. Marianelli, C., C. Graziani, C. Santangelo, M. T. Xibilia, A. Imbriani, R. Amato, D. Neri, M. Cuccia, S. Rinnone, V. Di Marco, and F. Ciuchini. 2007. Molecular epidemiologic and antimicrobial susceptibility characterization of Brucella isolates from humans in Sicily, Italy. J. Clin. Microbiol. 45:2923 2928. 22. Nöckler, K., R. Maves, D. Cepeda, A. Draeger, A. Mayer-Scholl, J. Chacaltana, M. Castañeda, B. Espinosa, R. Castillo, E. Hall, S. Al Dahouk, R. H. Gilman, F. Cabeza, and H. L. Smits. 2009. Molecular epidemiology of Brucella genotypes in patients at a major hospital in central Peru. J. Clin. Microbiol. 47:3147 3155. 23. Pappas, G., N. Akritidis, M. Bosilkovski, and E. Tsianos. 2005. Brucellosis. N. Engl. J. Med. 352:2325 2336. 24. 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. 25. Sanchéz-Serrano, L. P., P. Ordoñez, M. O. Díaz, and A. Torres. 2004. Vigilancia de la brucelosis. Bol. Epidemiol. Inst. Salud Carlos III 12:209 220. 26. Sayan, M., Z. Yumuk, O. Bilenoglu, S. Erdenling, and A. Willke. 2009. Genotyping of Brucella melitensis by rpob gene analysis and reevaluation of conventional serotyping method. Jpn. J. Infect. Dis. 62:160 163. 27. Schouls, L. M., H. G. van der Heide, L. Vauterin, P. Vauterin, and F. R. Mooi. 2004. Multiple-locus variable-number tandem repeat analysis of Dutch Bordetella pertussis strains reveals rapid genetic changes with clonal expansion during the late 1990s. J. Bacteriol. 186:5496 5505. 28. Smits, H., B. Espinosa, R. Castillo, E. Hall, A. Guillen, M. Zevaleta, R. Gilman, P. Melendez, C. Guerra, and A. Draeger. 2009. MLVA genotyping of human Brucella isolates from Peru. Trans. R. Soc. Trop. Med. Hyg. 103:399 402. 29. 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 analysis methods for molecular strain typing of human Brucella melitensis isolates from the Middle East. J. Clin. Microbiol. 47:2226 2231. 30. Valdezate, S., I. Cervera, P. Hernández, A. Navarro, and J. A. Saéz-Nieto. 2007. Characterisation of outbreaks of human brucellosis and sporadic cases by the use of hyper-variable octameric oligonucleotide fingerprint (HOOF) variable number tandem repeats. Clin. Microbiol. Infect. 13:887 892. 31. Vassalos, C. M., E. Vangelis, E. Vassalou, and C. Papadopoulou. 2009. Brucellosis in humans: why is it so elusive? Rev. Med. Microbiol. 20:63 73. 32. Whatmore, A. M., S. J. Shankster, L. L. Perrett, T. J. Murphy, S. D. Brew, R. E. Thirlwall, S. J. Cutler, and A. P. MacMillan. 2006. Identification and characterization of variable-number tandem-repeat markers for typing Brucella spp. J. Clin. Microbiol. 44:1982 1993. 33. World Animal Health Information Database. 5 March 2010, accession date. Zoonoses in humans. http://www.oie.int/wahis/public.php. World Organization for Animal Health, Paris, France.