- Research article
- Open Access
The CO2-dependence of Brucella ovis and Brucella abortus biovars is caused by defective carbonic anhydrases
Veterinary Research volume 49, Article number: 85 (2018)
Brucella bacteria cause brucellosis, a major zoonosis whose control requires efficient diagnosis and vaccines. Identification of classical Brucella spp. has traditionally relied on phenotypic characterization, including surface antigens and 5–10% CO2 necessity for growth (CO2-dependence), a trait of Brucella ovis and most Brucella abortus biovars 1–4 strains. Although molecular tests are replacing phenotypic methods, CO2-dependence remains of interest as it conditions isolation and propagation and reflects Brucella metabolism, an area of active research. Here, we investigated the connection of CO2-dependence and carbonic anhydrases (CA), the enzymes catalyzing the hydration of CO2 to the bicarbonate used by anaplerotic and biosynthetic carboxylases. Based on the previous demonstration that B. suis carries two functional CAs (CAI and CAII), we analyzed the CA sequences of CO2-dependent and -independent brucellae and spontaneous mutants. The comparisons strongly suggested that CAII is not functional in CO2-dependent B. abortus and B. ovis, and that a modified CAII sequence explains the CO2-independent phenotype of spontaneous mutants. Then, by mutagenesis and heterologous plasmid complementation and chromosomal insertion we proved that CAI alone is enough to support CO2-independent growth of B. suis in rich media but not of B. abortus in rich media or B. suis in minimal media. Finally, we also found that insertion of a heterologous active CAII into B. ovis reverted the CO2-dependence but did not alter its virulence in the mouse model. These results allow a better understanding of central aspects of Brucella metabolism and, in the case of B. ovis, provide tools for large-scale production of diagnostic antigens and vaccines.
Brucella is a genus of gram-negative bacteria of the α-2 subdivision of the class Proteobacteria  that includes the causal agents of brucellosis, a zoonosis producing important economical loses and human suffering in many developing countries . Currently, the genus contains twelve nominal species often showing host preference. Those spp. that were identified early (frequently referred to as the classical Brucella spp.) are Brucella abortus, preferentially infecting cattle, B. melitensis, usually infecting sheep and goats, B. suis, infecting pigs, hares, reindeer ad several wild rodents, B. canis, found in dogs, B. neotomae, isolated from desert wood rats, and Brucella ovis, a non-zoonotic species that is restricted to sheep and causes a condition known as B. ovis ovine epididymitis . More recently, brucellae have been isolated from marine mammals, voles and other wild life vertebrates, and new species proposed [4,5,6,7,8].
Conventional identification of the classical Brucella spp. and biovars has traditionally relied on dye and phage sensitivity, H2S production, urease activity, requirement of 5–10% CO2 atmospheres (0.04% in normal air) for growth (CO2-dependence) and surface antigens . Even though these methods are being rapidly replaced by molecular tests, antigenic structure and CO2-dependence remain of immediate practical interest as these properties affect the conditions for primary isolation and propagation in vitro and the implementation of diagnostic tests. Antigenically, the classical Brucella spp. are divided in two groups: the rough (R) species (B. canis and B. ovis), which carry R-type lipopolysaccharides (LPS), and the smooth (S) brucellae, which carry S-LPS [3, 10]. Concerning CO2-dependence, this is a trait of B. ovis and most strains of B. abortus biovars 1–4 [3, 10]. In both cases, CO2-independent variants may appear with low frequency  and, for obvious practical reasons, some of these variants have been used for B. ovis and B. abortus antigen production [3, 10]. Similarly, B. abortus vaccines have been developed on CO2-independent backgrounds .
The classical Brucella spp. are facultative intracellular parasites able to circumvent early proinflammatory responses and endowed with a type IV secretion system involved in the control of intracellular trafficking [12,13,14,15]. Moreover, it is postulated that these bacteria have progressively adapted their metabolism to the nutrients encountered within cells as an essential part of their intracellular strategy [16,17,18,19,20,21]. In this regard, despite being a notorious phenotype of practical importance, Brucella CO2-dependence has deserved no attention since the demonstration over 60 years ago that CO2-rich atmospheres are not required to reduce oxygen tension and that CO2 is used as a nutrient per se (reviewed in ). Indeed, CO2 assimilation requires carbonic anhydrases (CAs), a group of critically important ubiquitous enzymes distributed into six evolutionary distinct classes named α to η, with the β class present in bacteria . In heterotrophs, CAs are involved in C acquisition via assimilatory and anaplerotic reactions linked to several biosynthetic processes , and CO2-dependence has been related to defects in CA function in several microorganisms [24,25,26,27,28,29,30]. However, to the best of our knowledge, the role of CAs in CO2-dependence has not been investigated in Brucella spp. where the information is limited to recent investigations in search for targets for new drugs [31,32,33,34]. These investigations have described that B. suis (thus CO2-independent) strain 1330 has two ORFs (BRA0788 and BR1829) that code for β CAs (henceforth Bs1330CAI and Bs1330CAII). Both CAs are predicted to contain all the amino acid residues involved in the catalytic site and, more important, their activity was verified upon purification and found to be better for Bs1330CAII [31,32,33]. The demonstration that these B. suis ORFs actually code for enzymes with the predicted activity, together with the availability of the genome sequences of both CO2-independent and -dependent Brucella spp. and biovars, open the way to investigate the mechanisms underlying CO2-dependence in Brucella. The aim of the work described here was twofold: to investigate the genetic background behind the Brucella CO2-independent and -dependent phenotypes and, for B. ovis, a species that shows constant CO2-dependence, to construct a CO2-independent strain suitable for vaccine and antigen production.
Materials and methods
Bacterial strains and growth conditions
The bacterial strains and plasmids used in this study are listed in Table 1 and Additional files 1 and 2. B. abortus and B. suis strains were grown in standard Peptone-Glucose (Tryptic soy broth, TSB) or in this medium supplemented with agar (TSA). B. ovis strains were grown in TSB supplemented with yeast extract (0.5%) and fetal bovine serum 5% (TYSB) or this medium supplemented with agar (TYSA). For the studies in mice, strains were growth in Blood Agar Base supplemented with fetal bovine serum 5% (BABS). In addition, two minimal media were used. The components for 1 L of the defined medium of Gerhardt (Glutamate–Lactate–Glycerol)  are: glycerol (30 g), lactic acid (5 g), glutamic acid (1.5 g), thiamine (0.2 mg), nicotinic acid (0.2 mg), pantothenic acid (0.04 mg), biotin (0.0001 mg), K2HPO4 (10 g), Na2S2O3∙5H2O (0.1 g), MgSO4 (10 mg), MnSO4 (0.1 mg), FeSO4 (0.1 mg) and NaCl (7.5 g). The pH was adjusted to 6.8–7. The second minimal medium was a modification of Plommet’s [17, 36] and 1 L of this medium is composed of thiamine (0.2 g), nicotinic acid (0.2 g), pantothenic acid (0.07 g), biotin (0.1 mg), K2HPO4 (2.3 g), KH2PO4 (3 g), Na2S2O3 (0.1 g), MgSO4 (0.01 g), MnSO4 (0.1 mg), FeSO4 (0.1 mg); NaCl (5 g), (NH4)2SO4 (0.5 g) and 1 g/L of glucose. Incubation was at 37 °C, with (5%) or without CO2. When needed, media were supplemented with 5% sucrose (Sigma), kanamycin (Km) at 50 μg/mL, nalidixic acid (Nal) at 25 μg/mL, polymyxin (Pmx) at 1.5 μg/mL, chloramphenicol (Cm) at 20 μg/mL, spectinomycin (Spc) at 100 μg/mL or ampicillin (Amp) at 100 μg/mL (all from Sigma). All strains were stored in skimmed milk or TYSB-DMSO at −80 °C.
Genomic sequences of B. suis 1330, B. suis 513 (not annotated), B. abortus 544 (not annotated) and B. abortus 292 were obtained from the databases at National Center for Biotechnology Information (NCBI), Kyoto Encyclopedia of Genes and Genomes (KEGG) or The Broad Institute. The genomic sequence of B. abortus 2308W was obtained from the European Nucleotide Archive (ENA) and compared with its sibling 2308 sequence in KEGG. When genomic sequences were not available (B. ovis PA, B. ovis REO198, B. abortus AB0339, AB0339-CO 2 mut , B. abortus AB0127, AB0127-CO 2 mut , B. abortus AB0130, AB0130-CO 2 mut and BoPA-CO 2 mut ) ORFs were PCR amplified and then sequenced. DNA sequencing was carried out by “Servicio de Secuenciación de CIMA (Centro de Investigación Médica Aplicada, Pamplona, Spain)”. Sequence alignments were performed with Clustal Omega.
Plasmid and chromosomal DNA were extracted with QIAprep Spin Miniprep (Qiagen) and Ultraclean Microbial DNA Isolation kit (Mo Bio Laboratories), respectively. When needed, DNA was purified from agarose gels using QIAquick Gel Extraction Kit (Qiagen). Primers (Additional file 3) were synthesized by Sigma (Haverhill, United Kingdom). Restriction modification enzymes were used under the conditions recommended by the manufacturer.
Construction of B. abortus 2308W and B. suis mutants by gene disruption
For the construction of the CAI mutants, an internal region of 323 bp was amplified with oligonucleotides CAI-F1-ins (5′-GAATTTCTATGGATCGGCTGTT-3′) and CAI-R2-ins (5′-CGGTCCTGCGTGTTTTCTAT-3′). The resulting fragment containing an internal region of the ORF was cloned into pCR2.1-TOPO® vector (Invitrogen) to generate plasmid pCR2.1Ba2308WCAI (Additional file 2) and then, sequenced to verify the insertion. After sequencing, this fragment was cloned into the BamHI and XbaI sites of the suicide vector pJQKm . The resulting plasmid pJQKmBa2308WCAI (Additional file 2) was transformed into competent E. coli S17 λpir [38, 39] and transferred into B. abortus 2308W, B. suis 1330 and B. suis 513 by conjugation, where a single crossover led to disruption of the wild type locus. Integrative mutants were selected on a medium containing kanamycin and nalidixic acid or polymyxin and called B. abortus 2308W::pJQKm-CAI, B. suis 513::pJQKm-CAI and B. suis 1330::pJQKm-CAI (Additional file 1). Since the orientation of the insert in the pJQKm vector was known after sequencing, gene disruption was confirmed by detecting PCR products with primers CAI-Fw and M13Fw and primers CAI-Rv and M13Rv.
The CAII mutants were constructed in a similar way. A 302 bp internal fragment of CAII was amplified with oligonucleotides CAII-F1-Ins (5′-CAATGTGGCCAATCTCATTC-3′) and CAII-R2-ins (5′-GCGAATAGCGGATCGAAATA-3′). The resulting fragment was cloned into pCR2.1-TOPO® vector (Invitrogen) to generate plasmid pCR2.1Ba2308W CAII (Additional file 2), sequenced to verify the insertion and subsequently cloned into the BamHI and XbaI sites of the suicide vector pJQKm . The mutants were named B. suis 513::pJQKm-CAII and B. suis 1330::pJQKm-CAII (Additional file 1). Since the orientation of the insert in the pJQKm vector was known after sequencing, the site of the insertion was confirmed by independent PCR rounds with primers CAII-Fw and M13Fw, and primers CAII-Rv and M13Rv. After several attempts, no mutant in B. abortus 2308W CAII was obtained either under normal or CO2-enriched conditions.
Selection of CO2-independent spontaneous mutants
To obtain CO2-independent spontaneous mutants from CO2-dependent bacteria, B. ovis PA and three B. abortus isolates (one biovar 1 and two biovar 3) were plated on TYSA or TSA and incubated at 37 °C without CO2. After 5 days, one colony was picked and the genes encoding CAI and CAII were PCR amplified using primers CAI-Fw and CAI-Rv, and CAII-Fw and CAII-Rv (see Additional file 3). DNA sequencing with these primers and CAI-F1-Sec and CAII-F1-Sec primers (Additional file 3) allowed identification of mutations by comparison with the nucleotide sequence of the parental CO2-dependent strains.
Construction of the plasmid carrying CAIIba2308W and introduction into B. abortus 292 and 544
For the construction of the expression plasmid encoding Ba2308WCAII oligonucleotides CAII-Fw-Gw (5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTCCGCTGCCGTGTTTGAAATCA-3′) and CAII-Rv-Gw (5′-GGGGACCACTTTGTACAAGAAAGCTGGGTCTCAAAGTTCAGGGCGTTTGAA-3′) that contain sequences attB (underlined) were used to amplify CAII and the promoter from B. abortus 2308W. The resulting PCR product was cloned into pDONR223 to generate plasmid pDONOR223Ba2308WCAII (Additional file 2). After sequence verification, the ORF encoding CAII was transferred from pDONOR223Ba2308WCAII to pRH001 . The resulting plasmid, pRH001Ba2308WCAII (Additional file 2) was transformed into competent E. coli S17 λpir and introduced into Brucella strains by conjugation. The clones that had acquired the plasmid were selected by kanamycin resistance and confirmed by PCR using primers CAII-Fw-Gw and CAII-Rv-Gw, and M13F-M13R. The strains were called B. abortus 292 pRH001Ba2308WCAII and B. abortus 544 pRH001Ba2308WCAII (Additional file 1).
Construction of miniTn7T-KmR plasmids carrying CAI or CAII and introduction into Brucella strains
Using DNA from B. abortus 2308W, oligonucleotides CAII-IF-F1 (5′-CCGGGCTGCAGGAATTCGCTGCCGTGTTTGAAATCA-3′) and CAII-IF-R2 (5′-AGCTTCTCGAGGAATTTCAAAGTTCAGGGCGTTTGAA-3′) amplified a 966 bp region containing CAII and the promoter region. This fragment was cloned into the linearized vector (EcoRI) pUC18R6KT-miniTn7T-KmR  using the In-Fusion HD Enzyme Premix (Clontech). The resulting plasmid was called pUC18R6KT-miniTn7T-KmRBa2308WCAII (Additional file 2) and transformed into E. coli PIR1 and subsequently to E. coli S17 λpir. Then, it was transferred into Brucella by a tetraparental conjugation . The resulting constructs (B. abortus 292 Tn7Ba2308WCAII, B. abortus 544 Tn7Ba2308WCAII and B. ovis PA Tn7Ba2308WCAII; Additional file 1) were confirmed by PCR for the correct insertion and orientation of the mini-Tn7 between genes glmS and recG. Primers GlmS_B (5′-GTCCTTATGGGAACGGACGT-3′) and Ptn7-R (5′-CACAGCATAACTGGACTGATT-3′) were used to confirm insertion downstream glmS; Ptn7-L (5′-ATTAGCTTACGACGCTACACCC-3′) and RecG (5′-TATATTCTGGCGAGCGATCC-3′) insertion upstream recG and GlmS_B and RecG presence of transposon.
Oligonucleotides CAI-IF-F1 (5′-CCGGGCTGCAGGAATTTGTGGAATTGCACCGACAC-3′) and CAI-IF-R2 (5′-AGCTTCTCGAGGAATTCAATTATTCTGCCGGTTGG-3′) amplified a 987 bp fragment from B. suis 513 DNA containing CAI and the promoter. This fragment was subsequently cloned into the linearized vector (EcoRI) pUC18R6KT-miniTn7T-KmR using the In-Fusion HD Enzyme Premix (Clontech). The resulting plasmid was called pUC18R6KT-miniTn7T-KmRBs513CAI (Additional file 2) and was transformed into E. coli PIR1 and then to E. coli S17 λpir. After, the plasmid was introduced into the different Brucella strains by a tetraparental conjugation . The strains were called B. abortus 2308W Tn7Bs513CAI, B. abortus 292 Tn7Bs513CAI and B. abortus 544 Tn7 Bs513CAI. The insertion of the transposon was confirmed by PCR (see above and Additional file 3).
When necessary, constructs without kanamycin resistance cassette were obtained following the protocol set up by Martínez-Gómez et al. .
The strains were inoculated into 10 mL of TSB or TYSB in a 50 mL flask and incubated at 37 °C for 18 h with or without orbital shaking, in an atmosphere with 5% CO2 in the case of CO2-dependent strains. Then, these bacteria were harvested by centrifugation, resuspended in 10 mL of the test medium at an optical density at 600 nm (OD600nm) of 0.1, and incubated under the same conditions for 18 h. These exponentially growing bacteria were harvested by centrifugation, resuspended at an OD600nm of 0.1 (equivalent to 0.05 readings in the Bioscreen apparatus) in the test medium in appropriate multiwell plates (200 μL/well) and cultivated in a Bioscreen C (Lab Systems) apparatus with continuous shaking at 37 °C. Absorbance values at 420–580 nm were automatically recorded at 30 min-intervals. All experiments were performed in triplicate. Controls with medium and no bacteria were included in all experiments.
Studies in mice
Seven-week-old female BALB/c mice (Harlan Laboratories; Bicester, United Kingdom) were accommodated in the facilities of “Centro de Investigación y Tecnología Agroalimentaria de Aragón” (CITA; Registration code ES502970012025) for 2 weeks before and during the experiments, with water and food ad libitum under P3 biosafety containment conditions. The animal handling and other procedures were in accordance with the current European (directive 86/609/EEC) and Spanish (RD 53/2013) legislations, supervised by the Animal Welfare Committee of the CITA (2014-20).
To prepare inocula, BABS-grown bacteria were harvested, adjusted spectrophotometrically (OD600nm = 0.170) in sterile buffered saline (BSS; 0.85% NaCl, 0.1% KH2PO4, 0.2% K2HPO4; pH 6.85) and diluted in the same diluent up to approximately 5 × 107 CFU/mL. For each bacterial strain, five mice were intraperitoneally inoculated with 0.1 mL/mouse, the exact doses assessed retrospectively by plating dilutions of the inocula. The number of CFU in spleen was determined at 3 and 8 weeks post-inoculation. For this, the spleens were aseptically removed and individually weighed and homogenized in 9 volumes of BSS. Serial tenfold dilutions of each homogenate were performed, and each dilution was plated by triplicate. Plates were incubated at 37 °C, without CO2, for 5 days. The identity of the spleen isolates was confirmed by PCR. The individual number of CFU/spleen was normalized by logarithmic transformation, and the mean log CFU/spleen values and the standard deviations (n = 5) were calculated. Statistical comparisons were performed by Student’s t-test.
ORF sequences suggest a critical role of CAII in CO2-independence
We first analyzed whether the sequences of Bs1330CAI and Bs1330CAII, respectively encoded by BRA0788 and BR1829 of B. suis 1330 and with proved CA activity, had orthologues in reference and collection strains representative of the CO2-independent and -dependent Brucella phenotypes (Table 1). This analysis showed that all these brucellae carry Bs1330CAI and Bs1330CAII orthologues, with the peculiarities summarized below (for further details, see Additional files 4 and 5).
B. suis 513 Bs1330CAI orthologue differed from Bs1330CAI only at position 40 (serine instead of leucine) and carried a CAII identical to Bs1330CAII. The B. abortus 2308W Bs1330CAI orthologue differed from Bs1330CAI in that the serine at position 40 and valine in position 76 were both substituted by glycine. Similarly, the Bs1330CAII orthologue had an extra amino acid (alanine) at position 114. B. abortus 292 and 544, both CO2-dependent, contained a Bs1330CAI orthologue with the same serine and valine substitutions as strain 2308W, and a cytosine insertion at position 338 of the Bs1330CAII orthologue leading to a frameshift affecting almost 50% of the protein. In the B. ovis PA Bs1330CAI orthologue, a deletion of 24 nucleotides at positions 217–240 results in a protein lacking amino acids 74–81, and the insertion of a guanine at the Bs1330CAII orthologue originates a frameshift and a protein defective in the last 40 amino acids. B. ovis REO198, which is CO2-independent, is identical to B. ovis PA with respect to the Bs1330CAI orthologue. However, the lack of a guanine in the Bs1330CAII orthologue three positions after the 521 position guanine of its B. ovis PA counterpart restores the reading frame and should allow synthesis of a protein identical to that of B. suis 1330 (Additional files 4 and 5). Altogether, the observations strongly suggest that the CO2-dependence of B. ovis PA, B. abortus 292 and 544 is caused by the lack of an active CAII (activity defined empirically as that allowing growth in a normal atmosphere) and, conversely, that mutations in B. ovis CAII could account for the spontaneous emergence of CO2-independent strains in at least B. ovis. On the other hand, these analyses did not allow inferring the relevance of CAI, which was apparently complete in B. abortus and B. suis.
CAII is mutated in spontaneous CO2-independent mutants
We examined first the validity of the hypothesis on the relevance of CAII for growth in normal air by comparing the putative CA genes of several spontaneous CO2-independent mutants that appeared during routine laboratory manipulations with their parental counterparts (Table 1). For the B. ovis PA CO2-independent mutant (BoPA-CO 2 mut ), we observed that while the guanine in the CAII gene causing the above-described frameshift was absent, the CAI gene had not undergone any changes. The CO2-independent mutants of three recent B. abortus isolates (one biovar 1 and two biovar 3)  lacked a guanine at position 340 of the CAII gene that was however present in the CO2-dependent parental isolates. Altogether, these results support the starting hypothesis that CAII mutations are involved in the emergence of CO2-independent mutants and indirectly suggest that CAI is less relevant in the uptake of CO2.
An active CAI is enough by itself to support CO2-independent growth of B. suis but not of B. abortus
To compare the physiological importance of CAI and CAII we first carried out mutagenesis in different backgrounds and tested the mutants for CO2-independence. Using B. abortus 2308W, we found that its CAI mutant (B. abortus 2308W::pJQKm-CAI) kept the CO2-independent phenotype, proving that CAII by itself can sustain growth under normal atmospheric conditions (Figure 1). In contrast, and despite repeated attempts, we failed to obtain a similar mutant in CAII, suggesting that the B. abortus 2308W CAI cannot supply bicarbonate at a rate high enough for growth under normal atmospheric conditions. This is in keeping with the identity of CAI sequence between B. abortus 2308W on one hand and B. abortus 292 and 544 on the other, and supports the idea that CAI is not active in these three B. abortus strains.
Köhler et al.  reported recently the failure to obtain B. suis 1330 (biovar 1) double CAI–CAII mutants, which together with the analysis of the purified CAI [31,32,33] strongly suggests that CAI is active in this strain. Thus, we hypothesized that CAI could be inactive in some Brucella strains but active in others. Indeed, this possibility was consistent with the observation that the valine in position 76 in B. suis 1330 Bs1330CAI was substituted by glycine in B. abortus 292 and 544 (both CO2-dependent) as well as in strain 2308W (see above and Additional file 4) where CAII seemed essential for CO2-independence. To validate our hypothesis, we first constructed pJQKm insertion mutants in the Bs1330CAI and Bs1330CAII genes of B. suis 1330 (biovar 1) and found that both were CO2-independent (Figure 2). Then, we obtained similar pJQKm insertion mutants in B. suis 513 (biovar 5). Again, both mutants kept the CO2-independent phenotype of the parental strain, proving that CAI was also active in B. suis 513 despite the difference in position 40 (serine instead of leucine) with respect to Bs1330CAI (see above and Additional files 4 and 5).
Once we knew that CAII was essential for CO2-independent growth of B. abortus 2308 W and that CAI was active in B. suis 1330 and 513, we tested whether a functional CAII or CAI could accomplish the same role in CO2-dependent B. abortus strains. For this, we first constructed a low-copy plasmid (pRH001Ba2308WCAII) carrying the gene encoding B. abortus 2308W CAII (which we had proven to be active) under the control of its own promoter. When we introduced this plasmid into B. abortus 292 and 544 (both CO2-dependent), the pRH001Ba2308WCAII constructs were able to grow in a normal atmosphere (Figure 3). Then, to circumvent any gene dosage artifacts associated with plasmid constructs, we introduced a miniTn7 carrying Ba2308WCAII (Tn7Ba2308WCAII)  into a neutral site of the genomes of B. abortus 292 and 544. We found that, like the strain origin of the CAII gene, the two constructs grew in a normal atmosphere (Figure 3). Then, we did similar experiments with a miniTn7 carrying Bs513CAI (which we had proven to be active) and its promoter (Tn7Bs513CAI). In this case, however, we found that the B. abortus 292 and 544 Tn7Bs513CAI constructs failed to grow without CO2 enrichment (data not shown) leading to the conclusion that an active CAI was not enough by itself to support CO2-independent growth of B. abortus.
The CO2-independence mediated by CAI is conditioned by nutrient availability
While the above-described experiments show that an active CAII but not an active CAI was enough to bypass CO2-dependence in B. abortus, it was not immediately obvious why CAI by itself was enough to support growth of B. suis 1330 and B. suis 513. However, the B. suis CAII (and CAI) mutants were tested for CO2-independence in a medium rich in peptones and glucose, conditions that are likely to downplay the role of the anabolic pathways where CA activity is important. Therefore, we reasoned that, depending upon the metabolic abilities of Brucella spp. and biovars, experiments in complex media could be not stringent enough to reveal differences between CAI and CAII activities. To analyze this, we took advantage of the almost prototrophic characteristics of B. suis 513, a strain that only requires a few vitamins and grows efficiently with limited C supplies . When we tested B. suis 513::pJQKm-CAII insertion mutant on Glutamate–Lactate–Glycerol (a gluconeogenic medium [21, 45]) or Glucose as the only C sources, we found that the mutant failed to grow under a normal atmosphere (Figure 4). This result, which shows that B. suis 513 CAI cannot meet the biosynthetic demands of this strain in simple media, strongly suggest that CAI is not active enough in less prototrophic species such as B. abortus even in complex media and, therefore, that it adds little to the role of CAII. In keeping with this, we found that a B. abortus 2308 W construct carrying Bs513CAI and its parental strain did not differ in growth rates (Figure 5).
An active CAII reverts the CO2-dependence of B. ovis PA and does not alter its multiplication in the mouse model
Among the brucellae, B. ovis is notorious for its CO2-dependence and fastidious nutritional requirements caused in all likelihood by its comparatively genome degradation . Since the above-described experiments not only proved the chief role of CAII in the B. abortus and B. suis biovars tested but also provided molecular tools for relieving CO2-dependence in B. abortus, we introduced Tn7Ba2308WCAII into the B. ovis PA chromosome and examined the construct for CO2-dependence. As can be seen in Figure 6A, B. ovis PA Tn7Ba2308WCAII grew under normal atmospheric conditions. Then, we used the Tn7Ba2308WCAII construct to test whether the CO2-dependence and/or this genetic manipulation would alter the virulence of B. ovis in the standard mouse model. We found that the introduction of a functional CAII into B. ovis PA did not affect the multiplication (acute phase) and permanence (chronicity) of the bacteria in the spleens of BALB/c mouse (Figure 6B).
The metalloenzymes generically designated as CAs catalyze the reversible hydration of CO2 into bicarbonate, the substrate of key anaplerotic and biosynthetic enzymes . Many aerobic microorganisms can obtain enough bicarbonate from ambient air (about 0.04% CO2) and not surprisingly CA mutants of at least Ralstonia eutropha, Streptococcus pneumoniae, Escherichia coli, Pseudomonas aeruginosa and Corynebacterium glutamicum are unable to grow under these conditions [25,26,27,28,29,30]. All these observations made very likely the existence of a relationship between CA deficiencies and the CO2-dependence that is characteristic of some Brucella biovars and species, a hypothesis proved in this work. If we define CA inactivity as that not high enough to make the bacteria able to grow in a normal atmosphere, we have demonstrated that while B. suis 1330 and 513 genomes encode two active CAs (CAI and CAII) only CAII is active in B. abortus 2308W and none in representative strains of the CO2-dependent B. abortus biovars or in B. ovis PA. Indeed, the fact that the gene encoding CAII is conserved in B. abortus 2308W is in agreement with previous in vitro enzymatic analysis that show that B. suis 1330 CAII is a better catalyst for the conversion of CO2 to bicarbonate (with an enzymatic activity 1.85 times higher) than Bs1330CAI .
It has to be stressed that the functional definition of activity used here pertains to the particular physiology of each strain. Although the number of strains tested was necessarily limited, we found evidence supporting the hypothesis that CAI is insufficient to support growth of Brucella when the medium is limited to simple C substrates, or even in rich media for those brucellae that display comparatively reduced biosynthetic abilities. Indeed, whereas growth on the minimal media used here requires bicarbonate being incorporated by the reactions catalyzed by enzymes such as phosphoenolpyruvate carboxylase, pyruvate carboxylase, carbamoyl phosphate synthetase, 5-aminoimidazole ribotide carboxylase and enoyl-CoA carboxylases/reductases, growth on rich media does not entail an intense biosynthesis of amino acids and nucleic acid precursors and, therefore, most if not all of the linked pathways should pose no stringent demands for bicarbonate. In the context of this hypothesis, B. suis, the fast-growing B. suis biovar 5 (strain 513) in particular, on one hand, and B. ovis, on the other, would respectively represent two opposite situations. The existence of Brucella strains carrying inactivated CAs strongly suggests that this enzymatic activity is not necessary for the persistency in nature of at least B. abortus and B. ovis. Indeed, the presence of mutations inactivating the metabolic genes may result from the absence of a positive selective pressure, reflecting an adaptation during which the cognate functions become dispensable because of the nutritional environment. Such a Brucella adaptation would not be a novelty in intracellular parasites because, while a majority of the genome-sequenced Proteobacteria retain a CA gene, intracellular genera such as Buchnera and Rickettsia contain CA-defective representatives . It remains to be investigated whether such a CA dispensability represents a high CO2 tension in their niche, as described for Symbiobacterium thermophilum , the exploitation of host CAs or the presence of nutrients bypassing metabolic steps connected to CA activity.
It is important to highlight the practical implications of this work. Unraveling the genetic background of Brucella CO2-dependence allowed us to construct a B. ovis CO2-independent mutant with practical implications on the diagnosis and control of B. ovis infection. Because of the non-zoonotic nature of B. ovis, this disease may not always deserve the attention of official programs and it is often overlooked. Control and eventual eradication of B. melitensis brucellosis of small ruminants is based on the use of diagnostic tests detecting antibodies to the S-LPS  and vaccination with B. melitensis Rev 1, a vaccine that also protects against B. ovis. However, Rev 1 may interfere in serological diagnosis  and it is virulent for humans  and resistant to streptomycin (an antibiotic of choice to treat human brucellosis). Owing to these drawbacks, Rev 1 vaccination is discontinued and finally banned in those regions or countries where B. melitensis prevalence is considered low enough to implement an exclusively test and slaughter strategy. Withdrawal of Rev 1 vaccination leaves animals unprotected against B. ovis, thus favoring the emergence of the disease in areas where B. melitensis is almost or totally eradicated. Moreover, B. ovis has remained endemic in many areas where B. melitensis is not present and Rev 1 vaccination was never implemented [51, 52]. Accordingly, research on B. ovis-specific vaccines is an area of increasing interest as these vaccines would neither pose risk of zoonotic infection nor interfere in those B. melitensis serological tests detecting S-LPS O-polysaccharide antibodies (i.e., rose bengal and complement fixation tests) [53,54,55,56]. The CO2 requirement represents a significant obstacle in the development of a B. ovis live attenuated vaccine for large-scale production. We have demonstrated that the B. ovis PA Tn7Ba2308WCAII described here not only grows under normal atmospheric conditions but also retains the virulence in at least the accepted laboratory model, thus representing an appropriate tool for the development of such CO2-independent attenuated vaccines. For instance, B. ovis PA Tn7Ba2308WCAII can be used as the background to apply the strategy proposed by Conde-Álvarez et al.  based on the deletion of LPS core glycosyltransferases that results in a truncated structure that by uncovering innate immunity targets triggers a potent protective Th1 response. In fact, Soler-Lloréns et al.  recently demonstrated that deletion of two of such glycosyltransferases in B. ovis PA results in attenuation and suitable vaccine properties in the mouse model. Similarly, the B. ovis PA Tn7Ba2308WCAII construct could be used to produce the R-specific antigen currently used in B. ovis serological tests. This antigen is made of vesicles rich in outer membrane proteins and R-LPS, and both types of components have been shown to be important for optimal sensitivity [58, 59]. Currently, this R antigen is obtained from B. ovis REO198 taking advantage of the unusual CO2-independence of this strain (Table 1). Yet, B. ovis REO198 LPS carries a core oligosaccharide defect that damages the diagnostic epitopes of the R LPS and it is thus likely to yield suboptimal results in serodiagnosis . If this is confirmed, it could be advantageously replaced by B. ovis PA Tn7Ba2308WCAII. Research is in progress to evaluate the attenuation and protection against B. ovis of B. ovis PA Tn7Ba2308WCAII core glycosyltransferase mutants as well as the diagnostic properties of R antigens obtained from this strain. In summary, the evidence presented in this work not only clarifies the biochemical basis of an important Brucella phenotype but also provides a tool for large-scale production of B. ovis diagnostic antigens and vaccines.
Moreno E, Stackebrandt E, Dorsch M, Wolters J, Busch M, Mayer H (1990) Brucella abortus 16S rRNA and lipid A reveal a phylogenetic relationship with members of the alpha-2 subdivision of the class Proteobacteria. J Bacteriol 172:3569–3576
McDermott JJ, Grace D, Zinsstag J (2013) Economics of brucellosis impact and control in low-income countries. Rev Sci Tech Off 32:249–261
OIE (2015) Ovine epididymitis (Brucella ovis). In: Manual of diagnostic tests and vaccines for terrestrial animals. Paris, pp 1–14
Whatmore AM (2009) Current understanding of the genetic diversity of Brucella, an expanding genus of zoonotic pathogens. Infect Genet Evol 9:1168–1184
Whatmore AM, Davison N, Cloeckaert A, Al Dahouk S, Zygmunt MS, Brew SD, Perrett LL, Koylass MS, Vergnaud G, Quance C, Scholz HC, Dick EJ, Hubbard G, Schlabritz-Loutsevitch NE (2014) Brucella papionis sp. nov., isolated from baboons (Papio spp.). Int J Syst Evol Microbiol 64:4120–4128
Al Dahouk S, Köhler S, Occhialini A, Jiménez de Bagüés MP, Hammerl JA, Eisenberg T, Vergnaud G, Cloeckaert A, Zygmunt MS, Whatmore AM, Melzer F, Drees KP, Foster JT, Wattam AR, Scholz HC (2017) Brucella spp. of amphibians comprise genomically diverse motile strains competent for replication in macrophages and survival in mammalian hosts. Sci Rep. 7:44420
Scholz HC, Nockler K, Gollner C, Bahn P, Vergnaud G, Tomaso H, Al-Dahouk S, Kampfer P, Cloeckaert A, Marquart M, Zygmunt MS, Whatmore AM, Pfeffer M, Huber B, Busse HJ, De BK (2010) Brucella inopinata sp. nov., isolated from a breast implant infection. Int J Syst Evol Microbiol 60:801–808
Scholz HC, Revilla-Fernández S, Al Dahouk S, Hammerl JA, Zygmunt MS, Cloeckaert A, Koylass M, Whatmore AM, Blom J, Vergnaud G, Witte A, Aistleitner K, Hofer E (2016) Brucella vulpis sp. Nov., isolated from mandibular lymph nodes of red foxes (vulpes vulpes). Int J Syst Evol Microbiol 66:2090–2098
Alton GG, Jones LM, Angus RD, Verger J-M (1988) Techniques for the brucellosis laboratory. INRA, Paris
OIE (2016) Brucellosis (Brucella abortus, B. melitensis and B. suis) (Infection with B. abortus, B. melitensis and B. suis). In: Manual of diagnostic tests and vaccines for terrestrial animals. Paris, pp 1–44
Gerhardt P (1958) The nutrition of brucellae. Bacteriol Rev 22:81–98
Barquero-Calvo E, Chaves-Olarte E, Weiss DS, Guzmán-Verri C, Chacón-Díaz C, Rucavado A, Moriyón I, Moreno E (2007) Brucella abortus uses a stealthy strategy to avoid activation of the innate immune system during the onset of infection. PLoS One 2:e631
Conde-Álvarez R, Arce-Gorvel V, Iriarte M, Manček-Keber M, Barquero-Calvo E, Palacios-Chaves L, Chacón-Díaz C, Chaves-Olarte E, Martirosyan A, von Bargen K, Grilló M-J, Jerala R, Brandenburg K, Llobet E, Bengoechea JA, Moreno E, Moriyón I, Gorvel J-P (2012) The lipopolysaccharide core of Brucella abortus acts as a shield against innate immunity recognition. PLoS Pathog 8:e1002675
Palacios-Chaves L, Conde-Álvarez R, Gil-Ramírez Y, Zúñiga-Ripa A, Barquero-Calvo E, Chacón-Díaz C, Chaves-Olarte E, Arce-Gorvel V, Gorvel JPP, Moreno E, de Miguel MJ, Grilló MJ, Moriyón I, Iriarte M (2011) Brucella abortus ornithine lipids are dispensable outer membrane components devoid of a marked pathogen-associated molecular pattern. PLoS One 6:e16030
Gorvel J-P, Moreno E (2002) Brucella intracellular life: from invasion to intracellular replication. Vet Microbiol 90:281–297
Barbier T, Nicolas C, Letesson JJ (2011) Brucella adaptation and survival at the crossroad of metabolism and virulence. FEBS Lett 585:2929–2934
Barbier T, Collard F, Zúñiga-Ripa A, Moriyón I, Becker J, Wittmann C, Van Schaftingen E, Letesson J-J (2014) Erythritol feeds the pentose phosphate pathway via three new isomerases leading to d-erythrose-4-phosphate in Brucella. Proc Natl Acad Sci U S A 111:17815–17820
Barbier T, Machelart A, Zúñiga-Ripa A, Plovier H, Hougardy C, Lobet E, Willemart K, Muraille E, De Bolle X, Van Schaftingen E, Moriyón I, Letesson J-J (2017) Erythritol availability in bovine, murine and human models highlights a potential role for the host aldose reductase during Brucella infection. Front Microbiol 8:1088
Letesson J-J, Barbier T, Zúñiga-Ripa A, Godfroid J, De Bolle X, Moriyón I (2017) Brucella genital tropism: what’s on the menu. Front Microbiol 8:506
Ronneau S, Moussa S, Barbier T, Conde-Álvarez R, Zúñiga-Ripa A, Moriyón I, Letesson J-J, Zuniga-ripa A, Moriyón I, Letesson J-J, Ronneau S, Moussa S, Barbier T, Conde-Álvarez R, Zúñiga-Ripa A (2014) Brucella, nitrogen and virulence. Crit Rev Microbiol 7828:1–19
Zúñiga-Ripa A, Barbier T, Conde-Álvarez R, Martínez-Gómez E, Palacios-Chaves L, Gil-Ramírez Y, Grilló MJ, Letesson J-J, Iriarte M, Moriyon I (2014) Brucella abortus depends on pyruvate phosphate dikinase and malic enzyme but not on Fbp and GlpX fructose-1,6-bisphosphatases for full virulence in laboratory models. J Bacteriol 196:3045–3057
Supuran CT (2016) Structure and function of carbonic anhydrases. Biochem J 473:2023–2032
Erb TJ (2011) Carboxylases in natural and synthetic microbial pathways. Appl Environ Microbiol 77:8466–8477
Aguilera J, Van Dijken JP, De Winde JH, Pronk JT (2005) Carbonic anhydrase (Nce103p): an essential biosynthetic enzyme for growth of Saccharomyces cerevisiae at atmospheric carbon dioxide pressure. Biochem J 391:311–316
Hashimoto M, Kato J (2003) Indispensability of the Escherichia coli carbonic anhydrases YadF and CynT in cell proliferation at a low CO2 partial pressure. Biosci Biotechnol Biochem 67:919–922
Kusian B, Su D, Bowien B, Sültemeyer D (2002) Carbonic anhydrase is essential for growth of Ralstonia eutropha at ambient CO2 concentrations. J Bacteriol 184:5018–5026
Merlin C, Masters M, Mcateer S, Coulson A (2003) Why is carbonic anhydrase essential to Escherichia coli? J Bacteriol 185:6415–6424
Mitsuhashi S, Ohnishi J, Hayashi M, Ikeda M (2004) A gene homologous to β-type carbonic anhydrase is essential for the growth of Corynebacterium glutamicum under atmospheric conditions. Appl Microbiol Biotechnol 63:592–601
Burghout P, Cron LE, Gradstedt H, Quintero B, Simonetti E, Bijlsma JJE, Bootsma HJ, Hermans PWM (2010) Carbonic anhydrase is essential for Streptococcus pneumoniae growth in environmental ambient air. J Bacteriol 192:4054–4062
Lotlikar SR, Hnatusko S, Dickenson NE, Choudhari SP, Picking WL, Patrauchan MA (2013) Three functional β-carbonic anhydrases in Pseudomonas aeruginosa PAO1: role in survival in ambient air. Microbiology 159:1748–1759
Joseph P, Turtaut F, Ouahrani-Bettache S, Montero J-L, Nishimori I, Minakuchi T, Vullo D, Scozzafava A, Köhler S, Winum J-Y, Supuran CT (2010) Cloning, characterization, and inhibition studies of a β-carbonic anhydrase from Brucella suis. J Med Chem 53:2277–2285
Joseph P, Ouahrani-Bettache S, Montero JL, Nishimori I, Minakuchi T, Vullo D, Scozzafava A, Winum JY, Köhler S, Supuran CT (2011) A new β-carbonic anhydrase from Brucella suis, its cloning, characterization, and inhibition with sulfonamides and sulfamates, leading to impaired pathogen growth. Bioorg Med Chem 19:1172–1178
Winum J-Y, Köhler S, Supuran CT (2010) Brucella carbonic anhydrases: new targets for designing anti-infective agents. Curr Pharm Des 16:3310–3316
Köhler S, Ouahrani-Bettache S, Winum J (2017) Brucella suis carbonic anhydrases and their inhibitors: towards alternative antibiotics? J Enzyme Inhib Med Chem 32:683–687
Gerhardt P, Wilson JB (1948) The nutrition of brucellae: growth in simple chemically defined media. J Bacteriol 56:17–24
Plommet M (1991) Minimal requirements for growth of Brucella suis and other Brucella species. Zentralbl Bakteriol 275:436–450
Scupham AJ, Triplett EW (1997) Isolation and characterization of the UDP-glucose 4′-epimerase-encoding gene, galE, from Brucella abortus 2308. Gene 202:53–59
Simon R, Priefer U, Pühler A (1983) A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram negative bacteria. Nat Biotechnol 1:784–791
Miller VL, Mekalanos JJ (1988) A novel suicide vector and its use in construction of insertion mutations: osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR. J Bacteriol 170:2575–2583
Hallez R, Letesson J-J, Vandenhaute J, de Bolle X (2007) Gateway-based destination vectors for functional analyses of bacterial ORFeomes: application to the min system in Brucella abortus. Appl Environ Microbiol 73:1375–1379
Llobet E, March C, Giménez P, Bengoechea JA (2009) Klebsiella pneumoniae OmpA confers resistance to antimicrobial peptides. Antimicrob Agents Chemother 53:298–302
Choi KH, Gaynor JB, White KG, Lopez C, Bosio CM, Karkhoff-Schweizer RR, Schweizer HP (2005) A Tn7-based broad-range bacterial cloning and expression system. Nat Methods 2:443–448
Martínez-Gómez E, Ståhle J, Gil-Ramírez Y, Zúñiga-Ripa A, Zaccheus M, Moriyón I, Iriarte M, Widmalm G, Conde-Alvarez R (2018) Genomic insertion of a heterologous acetyltransferase generates a new lipopolysaccharide antigenic structure in Brucella abortus and Brucella melitensis. Front Microbiol 9:1092
Khames M, Mick V, de Miguel MJ, Girault G, Conde-Álvarez R, Khelef D, Oumouna M, Moriyón I, Muñoz PM, Zúñiga-Ripa A (2017) The characterization of Brucella strains isolated from cattle in Algeria reveals the existence of a B. abortus lineage distinct from European and Sub-Saharan Africa strains. Vet Microbiol 211:124–128
Zúñiga-Ripa A, Barbier T, Lázaro-Antón L, de Miguel MJ, Conde-Álvarez R, Muñoz PM, Letesson JJ, Iriarte M, Moriyón I (2018) The fast-growing Brucella suis biovar 5 depends on phosphoenolpyruvate carboxykinase and pyruvate phosphate dikinase but not on Fbp and GlpX fructose-1,6-bisphosphatases or isocitrate lyase for full virulence in laboratory models. Front Microbiol 9:641
Tsolis RM, Seshadri R, Santos RL, Sangari FJ, Lobo JM, de Jong MF, Ren Q, Myers G, Brinkac LM, Nelson WC, Deboy RT, Angiuoli S, Khouri H, Dimitrov G, Robinson JR, Mulligan S, Walker RL, Elzer PEH, Hassan KA, Paulsen IT (2009) Genome degradation in Brucella ovis corresponds with narrowing of its host range and tissue tropism. PLoS One 4:e5519
Ueda K, Nishida H, Beppu T (2012) Dispensabilities of carbonic anhydrase in proteobacteria. Int J Evol Biol 2012:324549
Nishida H, Beppu T, Ueda K (2009) Symbiobacterium lost carbonic anhydrase in the course of evolution. J Mol Evol 68:90–96
Ducrotoy MJ, Conde-Álvarez R, Blasco JM, Moriyón I (2016) A review of the basis of the immunological diagnosis of ruminant brucellosis. Vet Immunol Immunopathol 17:81–102
Blasco JM, Díaz R (1993) Brucella melitensis Rev-1 vaccine as a cause of human brucellosis. Lancet 342:805
Blasco JM (1997) A review of the use of B. melitensis Rev 1 vaccine in adult sheep and goats. Prev Vet Med 31:275–283
Agence française de sécurité sanitaire des aliments (2008) Avis de l’Agence française de sécurité sanitaire des aliments sur un protocole de lutte contre l’épididymite contagieuse ovine (Brucella ovis) dans les Pyrénées-Atlantiques
Da Costa Martins R, Irache JM, Gamazo C (2012) Acellular vaccines for ovine brucellosis: a safer alternative against a worldwide disease. Expert Rev Vaccines 11:87–95
Moriyón I, Grilló MJ, Monreal D, González D, Marín CM, López-Goñi I, Mainar-Jaime RC, Moreno E, Blasco JM (2004) Rough vaccines in animal brucellosis: structural and genetic basis and present status. Vet Res 35:1–38
Silva APC, Macêdo AA, Costa LF, Turchetti AP, Bull V, Pessoa MS, Araújo MSS, Nascimento EF, Martins-Filho OA, Paixão TA, Santos RL (2013) Brucella ovis lacking a species-specific putative ATP-binding cassette transporter is attenuated but immunogenic in rams. Vet Microbiol 167:546–553
Soler-Lloréns P, Gil-Ramírez Y, Zabalza-Baranguá A, Iriarte M, Conde-Álvarez R, Zúñiga-Ripa A, San Román B, Zygmunt MS, Vizcaíno N, Cloeckaert A, Grilló M-J, Moriyón I, López-Goñi I (2014) Mutants in the lipopolysaccharide of Brucella ovis are attenuated and protect against B. ovis infection in mice. Vet Res 45:72
Conde-Álvarez R, Arce-Gorvel V, Gil-Ramírez Y, Iriarte M, Grilló MJ, Gorvel JP, Moriyón I (2013) Lipopolysaccharide as a target for brucellosis vaccine design. Microb Pathog 58:29–34
Riezu-Boj JI, Moriyón I, Blasco JM, Marín CM, Díaz R (1986) Comparison of lipopolysaccharide and outer membrane protein-lipopolysaccharide extracts in an enzyme-linked immunosorbent assay for the diagnosis of Brucella ovis infection. J Clin Microbiol 23:938–942
Riezu-Boj JI, Moriyón I, Blasco JM, Gamazo C, Díaz R (1990) Antibody response to Brucella ovis outer membrane proteins in ovine brucellosis. Infect Immun 58:489–494
Soler-Lloréns P (2014) Desarrollo de nuevas vacunas frente a Brucella ovis: Estudio de genes implicados en la síntesis del núcleo del lipopolisacárido. University of Navarra, Navarra
Vershilova PA, Lyamkin GI, Malikov VE, Dranovskaia EA (1983) Brucella strains from mouse like rodents isolated in the USRR. Int J Syst Bacteriol 33:399–400
Suárez-Esquivel M, Ruiz-Villalobos N, Castillo-Zeledón A, Jiménez-Rojas C, Roop RM II, Comerci DJ, Barquero-Calvo E, Chacón-Díaz C, Caswell CC, Baker KS, Chaves-Olarte E, Thomson NR, Moreno E, Letesson JJ, De Bolle X, Guzmán-Verri C (2016) Brucella abortus strain 2308 wisconsin genome: importance of the definition of reference strains. Front Microbiol 7:1557
Blasco JM, Marín CM, Barberán M, Moriyón I, Díaz R (1987) Immunization with Brucella melitensis Rev 1 against Brucella ovis infection of rams. Vet Microbiol 14:381–392
Fensterbank R, Pardon P, Marly J (1982) Efficacy of Brucella melitensis Rev. 1 vaccine against Brucella ovis infection in rams. Ann Rech Vet 13:185–190
Blasco JM, Gamazo C, Winter AJ, Jiménez de Bagüés MP, Marín CM, Barberán M, Moriyón I, Alonso-Urmeneta B, Díaz R (1993) Evaluation of whole cell and subcellular vaccines against Brucella ovis in rams. Vet Immunol Immunopathol 37:257–270
Díaz R, Jones LM, Wilson JB (1968) Antigenic relationship of the gram-negative organism causing canine abortion to smooth and rough brucellae. J Bacteriol 95:618–624
The authors declare that they have no competing interests.
IM, MI and AZ-R conceived the study. LP-E, AZ-R, RC-Á, MK, MJM and PMM carried out the experimental work. IM, LP-E and AZ-R wrote the paper. All authors read and approved the final manuscript.
The authors are grateful to F. J. Sangari and J. M. Blasco for helpful discussions and suggestions. The authors thank Sara Serrano and María Uriarte for excellent technical assistance.
This research was supported by MINECO grant AGL2014-58795-C4-1-R (UNAV) and AGL2014-58795-C4-3-R (CITA), as well as by the Institute for Tropical Health funders (Obra Social La Caixa, Fundaciones Caja Navarra and Roviralta, PROFAND, Ubesol, ACUNSA and Artai). LP-E was supported by a grant from the Institute for Tropical Health. Work at CITA was sustained also by Aragon Government (Grupo de Investigación en Desarrollo A13-17D).
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.