Brucellosis is a bacterial disease, which, although affecting cattle primarily, has been associated with human infections, making its detection an important challenge. The existing gold standard diagnosis relies on the culture of bacteria which is a lengthy and costly process, taking up to 45 days. New technologies based on molecular diagnosis have been proposed, either through dip-stick, immunological assays, which have limited specificity, or using nucleic acid tests, which enable to identify the pathogen, but are impractical for use in the field, where most of the reservoir cases are located. Here we demonstrate a new test based on hybridization assays with metal nanoparticles, which, upon detection of a specific pathogen-derived DNA sequence, yield a visual colour change. We characterise the components used in the assay with a range of analytical techniques and show sensitivities down to 1000 cfu/ml for the detection of Brucella. Finally, we demonstrate that the assay works in a range of bovine samples including semen, milk and urine, opening up the potential for its use in the field, in low-resource settings.
Citation: Pal D, Boby N, Kumar S, Kaur G, Ali SA, Reboud J, et al. (2017) Visual detection of Brucella in bovine biological samples using DNA-activated gold nanoparticles. PLoS ONE 12(7): e0180919. https://doi.org/10.1371/journal.pone.0180919
Editor: Aristides Docoslis, Queen's University at Kingston, CANADA
Received: May 21, 2017; Accepted: June 25, 2017; Published: July 18, 2017
Copyright: © 2017 Pal et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data files are available from the University of Glasgow repository under DOI 10.5525/gla.researchdata.406.
Funding: This study was funded by the Department of Biotechnology, Ministry of Science & Technology, India (PC - BT/IN/Indo-UK/FADH/43/PC/2013) and the Biotechnology and Biological Sciences Research Council, UK (JC - BB/L00464X/1). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Brucellosis is a bacterial disease caused by Brucella species (facultatively intracellular coccobacilli) and remains the world’s most common zoonoses. Ten species have so far been identified and out of these, four classical species (Brucella melitensis, Brucella suis, Brucella abortus and Brucella canis) have been reported to be responsible for human infection.[2,3] Because of its non-specific clinical features and the lack of efficient diagnostic methods, the number of undetected cases remains high. Brucellosis is also linked to a number of serious challenges in the economy of developing countries, including India. The control of brucellosis in humans first requires its control in animals, leading to large programmes of vaccinations. However, this strategy cannot succeed without the ability to detect the reservoir animals in a robust and efficient way, so that they can be removed from the herd. Given the local infrastructure around many farms in rural India, there is a need to be able to provide a test that is sensitive and which can be detected visually. For example, rural veterinary clinics often do not have access to essential lab equipment necessary to run current gold standard tests, while access to expertise (for the preparation of reagents, assay processing and result interpretation) can also be lacking.
Currently, the isolation of the bacteria from clinical samples and the detection of anti-Brucella antibodies from suspected animals are the two primary methods employed for the diagnosis of brucellosis. Culture and isolation are time-consuming and labour intensive processes: the organism takes a minimum of 4–5 days to grow and the culture should be kept at least 45 days to rule out false negatives. There is also the need for culture facilities, often associated with a reference laboratory (certainly not always available in Low and Medium Income Countries, LMICs).
Alternatively, serological tests (e.g. Rose Bengal Plate Agglutination Test (RBPT), Standard Tube Agglutination Test (STAT)). are based on the antigenic properties of bacterial lipopolysaccharides (LPS). These not only require blood sampling and processing but also have been shown to lead to high cross reactivity with other organisms (e.g. Yersinia enterocolitica O:9,Vibrio cholerae, Escherichia coli O: 157, Francisella tularensis, Salmonella urbana group N and Stenotrophomonas maltophilia). The test therefore generally has a low specificity.
Nucleic acid tests (NATs) offer high performance by bridging the ability to identify the pathogen (similarly to culture) with the speed of molecular-based assays. As a consequence Brucella spp have been identified using multiplexed PCR  as well as isothermal LAMP assays. Both methods have limited application in the field as they require well-equipped laboratory and skilled personnel, not available in many rural settings in LMICs.
In contrast, hybridization tests based on metal nanoparticles are simple to implement and can be read out visually through a colour change, arising from the plasmonic interactions between the probes, making them an attractive system for diagnostic applications. Gold in particular offers the possibility of using a colour shift in the visible range, from red to purple, when aggregated, usually through the addition of simple reagents such as salts, and has thus been preferred in the implementation of these tests for the molecular detection of pathogens.
In this study, we designed a gold nanoparticle (AuNP) based hybridization assay to detect Brucella (Fig 1). This broad strategy has been successfully demonstrated previously, using a probe specific to IS711, an insertion sequence with multiple copies. Here we designed a thiol modified probe specific to BCSP31 a genus specific gene which codes for outer membrane protein.16 The direct visual detection of Brucella was optimized and the analytical sensitivity and specificity of the test was assessed. Importantly we show that the AuNP-oligo probe can be used for the simple, visual detection of Brucella from a broad range of bovine samples, including semen, milk and urine.
a) positive reaction; and b) negative reaction. After hybridisation of the AuNP-probe with the specific DNA (positive sample), the negative charge surrounding the AuNPs is increased. After adding salts (here HCl), the negative charges surrounding the unhybridised AuNP are neutralised resulting in the aggregation of AuNP (change to purple colour), whereas in the case of a positive sample (with more negative charge), the negative charges are not neutralised completely resulting in the non-aggregation of AuNP (the colour remains red).
Materials and methods
Gold (III) chloride hydrate, HAuCl4 (Sigma, USA); Sodium citrate tribasic (Sigma); Tris (2-carboxyethyl)- phosphine, TCEP (Sigma); Sodium Chloride, NaCl (SRL); Sodium dodecyl sulphate, SDS (Sigma); disodium hydrogen phosphate, Na2HPO4 (Merck, Germany); sodium dihydrogen phosphate, NaH2PO4 (Merck); sodium hydroxide, NaOH (SRL, India); potassium dihydrogen phosphate, KH2PO4 (Merck). All the chemical solutions and buffers were prepared using sterile double distilled water (ddH2O).
Sequences for thiol-modified oligonucleotide probe and complimentary sequence were as follows and custom synthesized:
Brucella abortus S19 was used as the source of Brucella DNA. The organism was cultured in Brucella broth (Difco) incubated at 37°C, 180 rpm for 3–4 days. Non-Brucella species were obtained from the divisional laboratory of Indian Veterinary Research Institute, UP, India. Pure bacterial cultures were enumerated by the serial dilution method.
Preparation of AuNP
A colloidal solution of AuNP was prepared by citrate reduction of gold (III) chloride hydrate (HAuCl4) as described in the literature. All the glassware was first washed with aqua regia (HCl:HNO3 = 3:1) and rinsed with double distilled water (ddH2O). 250 ml of aqueous solution of HAuCl4 (1 mM) was allowed to boil on hot plate magnetic stirrer. While boiling, 25 ml of 38.8 mM pre boiled sodium citrate tribasic was poured quickly, which led to a change in colour from pale yellow to ruby red. The solution was refluxed for another 20 min. The solution was cooled overnight at room temperature and stored at 4°C in an opaque coloured bottle until used. The AuNP preparation was characterized spectrophotometrically (at 520 nm). The size of AuNP was also determined by transmission electron microscopy (Jem-1011 electron microscope, Jeol, USA). A drop of sample solution was loaded over a carbon coated copper grid. The sample was dried at 37°C and observed at an accelerating voltage at 80 KV.
Functionalization of AuNP
AuNPs were functionalized by conjugation with thiol modified oligonucleotide probes following previously published protocols, with minor modifications. Before conjugation, 8 μl of 2.5 μmol.l-1 thiol modified oligo probe was first reduced with 30 μl of 0.1 M TCEP following incubation for 1 h at room temperature (RT). After incubation, 5 μl of 3 M sodium acetate (pH 5) was added.
The DNA probe was allowed to precipitate by adding 20 μl of ethanol at -20°C for 20 min. The solution was centrifuged at 12,000 rpm for 10 min. The pellet was air dried and resuspended in 100 μl of nuclease free water. 1 ml of 1 mM AuNP was added and incubated at 25°C for 1 day in a shaker incubator. Using 100 mM sodium phosphate buffer (pH 7), the final AuNP-probe solution was diluted to 9 mM phosphate buffer. SDS (final concentration of 0.1% (w/v)) was added and the solution incubated for 30 min room temperature. The volume of buffer (2 M NaCl in 10 mM phosphate buffer, pH 7) to make a final concentration of 0.3 M NaCl was added over 2 days in 6 intervals.
The AuNP-probe was allowed to equilibrate overnight at room temperature and was centrifuged at 13,000 rpm for 15 min. The soft pellet was washed with 500 μl of washing buffer (10 mM PBS, pH 7.4, 150 mM NaCl). Finally, the pellet was resuspended in 500 μl of the assay buffer (10 mM PBS, pH 7.4, 150 mM NaCl, 0.1% SDS) and stored at RT until used.
Probe specific to BCSP31 gene of Brucella was designed in Genetool using gene information available in NCBI (accession no. KF56031). Sequence homologies with other organisms were ruled out by using NCBI blastn program. The possibility of dimer and hairpin loop formation of oligonucleotide were minimised by analysing the sequence in Integrated DNA technology (IDT) online Oligoanalyser software.
1 μl of DNA was mixed with 9 μl of 10 mM sodium phosphate buffer (pH 5.0). The mixture was heated to 95°C for 5 min in a water bath to denature the DNA. After denaturation, 15 μl of AuNP-oligonucleotide probe was added for hybridization. Hybridization reaction was carried out in water bath at 60°C for 15 min. Following hybridization, 0.1 M HCl was added. The colour change of AuNP-probe mixture due to aggregation was observed visually and by spectrophotometry.
Agarose gel electrophoresis
The visual detection of the hybridized DNA was optimized by using 24-mers oligo-nucleotide complementary to probe sequences. The probe was allowed to hybridize with complementary sequences as described. These reaction mixtures were analysed on 2% agarose gel (100V for 1 h)–S1 Fig.
Preparation of biological spiked samples
Institute animal ethics committee (IAEC) approval was obtained for all experiments and procedures conducted on animals. Bovine biological samples used in this study were semen, milk and urine. Bovine semen samples were obtained from the Germplasm centre (GPC) of the Indian Veterinary Research Institute (IVRI). Semen from pedigree bulls were collected and processed for distribution as frozen semen straw by GPC following standard guidelines, as approved by the IAEC. Bovine milk was purchased from the Livestock Product & Management (LPM) section of IVRI which sells cattle and buffalo milk. Bovine urine samples were collected during normal urination at dairy farm maintained by the LPM section of IVRI. Animals were not subjected to any experiment in the present study. The obtained biological samples were artificially infected with Brucella and non-Brucella microorganisms. Samples were first spiked with counted number of bacteria. Urine samples (1 ml) spiked with bacteria were centrifuged at 10,000 rpm for 10 min. The cell pellet was washed in ddH2O and harvested again by centrifugation.
Semen samples (100 μl) were spiked with bacteria and were processed by dilution with 1 ml of sterile distilled water. Milk samples were spiked with bacteria and were processed as described previously with minor modification. Briefly, 0.5 ml of spiked milk sample was mixed with 100 μl of NET buffer (50 mM NaCl, 125 mM EDTA, 50 mM Tris-HCl, pH 7.6). The mixture was incubated at 80°C for 10 min and then cooled on ice. Proteinase K (20 μg/ml) was added and incubated for 90 min at 50°C. DNA was finally extracted with phenol-chloroform-isoamyl alcohol and precipitated with isopropanol.
All experiments were performed at least three times independently and by at least two different operators.
Results and discussion
Characterization of conjugated Au-NPs
We characterized our newly synthesised AuNP and probe conjugated AuNP (Fig 2) using visible and IR spectroscopies, particle sizing and TEM. We first investigated their optical properties with UV-vis spectroscopy (Cary 60 UV- Vis, Agilent). The unconjugated colloidal AuNP showed a peak absorbance at 520 nm. Upon conjugation of the probe onto the AuNP, this peak shifted to 524 nm (Fig 2a), indicative of successful conjugation. This shift was similar to that obtained for a similar nanoparticle system, characterised in the same manner for exploring antibody conjugation.
(a) UV-Vis spectra showing a red shift of the absorbance peak from 520nm for the bare AuNP (black trace), to 524nm for AuNP conjugated with the BSCP31 probe (grey). (b) Transmission Electron Microscopy image of the colloidal AuNP before conjugation (left—scale 50 nm), AuNP conjugated with BCSP-31 probe (middle—scale 50 nm) and zoomed in (scale 20 nm). Conjugated AuNP was characterised by the presence of a halo surrounding the AuNP. (c) FTIR spectra of unconjugated (red) and conjugated (black) AuNP showing peaks at: 1000–1200 cm-1 that represent the vibrational spectra due to phosphate and sugar phosphate bonding; 1630–1640 cm-1 represent the double bond in the bases ring. Also noted is the absence of peaks at 2500–2600 cm-1 due to S-H vibration, indicating that the probes were conjugated onto the AuNP surface by forming Au-S bond. The full spectra are available in S2 Fig.
The probe conjugation was also validated through the measurement of the Zeta potential of the solutions. The citrate ions provide the negative charge surrounding the AuNP (potential of -9mV). The negative charge of oligonucleotides further increased the zeta potential of the particles after conjugation to -25mV. All measurements were carried out at identical pH.
The hydrodynamic diameter of our bare AuNP increased from 29.9 nm to 54.76 for AuNP-BCSP31, as measured by a Zeta sizer (Malvern Zetasizer Nano ZS). In TEM, the conjugated AuNP showed a circular hollow configuration, which was not observed in the case of bare AuNP at the same magnification (Fig 2b).
We confirmed that these changes in the AuNP is due to the binding of the DNA sequence using Fourier transform infrared spectroscopy (FTIR). The spectra shown in Fig 2c have the characteristic peaks of oligonucleotides: C = O stretch (peaks within 1665–1760 cm-1, and around 1630–1640 cm-1 in the bases ring), phosphate and sugar phosphate vibrations from the DNA backbone 1000–1200 cm-1). We also note the absence of the characteristic S-H vibration band (2500–2600 cm-1), indicating that this bond was broken to form the S-Au link during conjugation.
Hybridization was also verified using gel electrophoresis (S1 Fig). All these measurements concur to ensure the effective conjugation of our AuNP along with the validity of the assay approach.
In this study, our probe-conjugated AuNP (BCSP-AuNP) was allowed to react with specific and non-specific target DNA to validate the concept of the detection and investigate the assay performance (sensitivity and specificity). Genomic DNA (gDNA) from Brucella, E.coli and other non-Brucella organisms were extracted by the NaOH digestion method. After injection of the target DNA into the AuNP solutions, the colour change upon the addition of HCl was monitored visually as well as by using absorbance. Due to the negative charges of the DNA surrounding the AuNP, the addition of HCl in the test reaction (where the complementary DNA hybridizes with the probe DNA on the AuNP) resulted in limited aggregation and the solution stayed red.
In contrast, when the target DNA was non-complementary (e.g. E.coli DNA), the AuNP aggregated readily, giving the color purple to the reaction (Fig 3a). This direct visual observation was confirmed spectroscopically, showing that positive samples exhibited the characteristic absorbance peak at around 520 nm. The negative controls generated a spectral shift with peaks above 560nm (Fig 3b).
(a) Photographs of the assay results (after addition of HCl), for the detection of genomic DNA from Brucella at different concentrations (negative control is E.coli genomic DNA at 60 ng/μl) performed in tubes. Total volumes of the reaction mix in each tube is 25 μl. Quantities are indicated for 1 μl of sample before dilution in the reaction mix. (b) UV-Vis spectra of gDNA sensitivity showing the absorbance peak at ~520nm for positive samples and red shifting with absorbance peak at higher wavelength in both negative and control samples, with a limit of detection of 5ng/μl in the sample before dilution in the reaction mixture. S3 Fig. Enhanced photographs of all figures (3–6) are available as supplementary information to provide more apparent color differentiation.
Sensitivity and specificity of AuNP-oligonucleotide probes
Fig 3a shows the visual sensitivity of the assay. Brucella gDNA at different concentration was used for hybridization assay. E.coli DNA was used as negative control. Addition of HCl resulted in a change of colour from red to purple in the negative control (unhybridised), whereas in the positive reactions (hybridized, e.g. at 50ng/μl), the colour of AuNP-probe remained red. The sensitivity of the BCSP31 probe was found to be 5ng/μl. When challenged with DNA isolated from pure bacteria cultures, the detection limit was 102 cfu/ml (Fig 4).
Quantities indicated are in one ml (25 μl total volume of reaction). The assay could detect up to 102 cfu/ml. Enhanced in S4 Fig.
To determine the specificity of the assay, non-Brucella microorganisms (E.coli, L. Icterohaemorrhagiae, P. multocida, S. typhimurium, S. aureus, M. leachii, S. suis and M. bovis) and a virus (BoHV1) were subjected to hybridization with AuNP-probes after DNA extraction. All non-Brucella samples showed a colour change to purple (Fig 5). This indicates a high specificity for the designed probe towards the Brucella spp.
Reactions for Brucella spp and nine non-Brucella spp (E.coli, L. icterohaemorrhagiae, BoHV1, P. multocida, S. typhimurium, S. aureus, M. leachii, S. suis and M. bovis) with BCSP-AuNP. Changes in colour were observed in all DNAs samples except in Brucella DNA (tube on the left).
Brucella infected animals routinely shed organisms in the natural secretion and excretion liquids such as milk, urine and uterine discharges. The detection of Brucella in semen has a specific interest as a consequence of recent changes in animal husbandry driven by the need to improve agricultural efficiencies to feed a rapidly expanding population, particularly in India. For example, in order to achieve a step change in the production of milk, artificial insemination of cattle is now common. In such facilities, it is widely known that Brucella infections amongst the donor can lead to reduced fertility and in some circumstances spontaneous abortion.
To study the feasibility of detection of Brucella using the designed AuNP-probes, we spiked bovine semen, milk and urine samples with bacterial count ranging from 105 cfu/ml to 102 cfu/ml. Artificially spiked biological samples were processed for DNA extraction and the hybridization assay was performed. The assay could detect organism copy numbers down to 103 cfu/ml of Brucella organisms in semen and urine samples. In the case of milk, the detection limit was 104 cfu/ml (Fig 6). This lower performance could be due to the presence of fat in which some organisms may get trapped. This sensitivity is on par with the bacterial loading commonly seen in contaminated samples, which ranges from 100 to 50,000 organisms/ml .
Current techniques for the determination of Brucella are not wholly appropriate for the needs of rural counties, particularly in LMICs, where there may be reduced access to infrastructure. In this study, we have demonstrated a simple and sensitive AuNP based detection of Brucella spp. with high specificity. Our test relies upon the plasmonic colour change when specific DNA hybridization events occur at a metal nanoparticle, thus needing only visual inspection of the results. The hybridization reaction does not require more than a heater (no thermal cycling, no extensive control required) or technical expertise beyond the addition of the sample. In terms of cost and convenience, the assay can be implemented readily in the field. We show that this assay can be used for detection of bacteria from milk, semen and urine which are considered the most relevant samples for the diagnosis of reproductive diseases, including brucellosis.
This assay could be performed within 30 min. The visual detection limit of 103 cfu/ml Brucella organism in biological samples is appropriate for the translation of the assay at the “point-of-care”. Importantly, in the future, this test could either be used for selection of bulls for semen collection at germplasm centres, for the examination of frozen semen before artificial insemination or for the examination of bulk milk before packaging, to ensure the safety of consumers.
S1 Fig. Agarose gel-electrophoresis.
(2% agarose W/V) of AuNP conjugated with DNA sequence: Lane 1: only BCSP31-AuNP; Lane 2: BCSP31-AuNP with hybridized with non-complementary sequence; Lane 3: BCSP31-AuNP hybridised with target sequence. Upon hybridisation, a retardation is observed.
The authors are thankful to the Director, IVRI for providing necessary facilities to carry out this study, as well as to staff from the GPC of IVRI for providing semen samples and from the LPM section of IVRI for providing milk and urine samples. They also acknowledge help of Dr. Zhugen Yang (University of Glasgow) in preparing Fig 1.
- 1. Christopher S, Umapathy B, Ravikumar K. Brucellosis: Review on the recent trends in pathogenicity and laboratory diagnosis. J Lab Physicians. 2010;2: 55. pmid:21346896
- 2. Corbel M. Brucellosis: an Overview. Emerg Infect Dis. 1997;3: 213–221. pmid:9204307
- 3. Smirnova EA, Vasin AV, Sandybaev NT, Klotchenko SA, Plotnikova MA, Chervyakova OV, et al. Current Methods of Human and Animal Brucellosis Diagnostics. Adv Infect Dis. 2013;3: 177–184.
- 4. Mohan Hari, Kharb Subhash. Human brucellosis: A silent but dreadful disease | Journal of Innovative Biology. J Innov Biol. 2014;1: 163–167.
- 5. Singh BB, Dhand NK, Gill JPS. Economic losses occurring due to brucellosis in Indian livestock populations. Prev Vet Med. 2015;119: 211–215. pmid:25835775
- 6. Smits HL, Kadri SM. Brucellosis in India: a deceptive infectious disease. Indian J Med Res. 2005;122: 375–384. pmid:16456249
- 7. Godfroid J, Nielsen K, Saegerman C. Diagnosis of Brucellosis in Livestock and Wildlife. Croat Med J. 2010;51: 296–305. pmid:20718082
- 8. Al Dahouk S, Tomaso H, Nöckler K, Neubauer H, Frangoulidis D. Laboratory-based diagnosis of brucellosis—a review of the literature. Part II: serological tests for brucellosis. Clin Lab. 2003;49: 577–589. pmid:14651329
- 9. Mayer-Scholl A, Draeger A, Göllner C, Scholz HC, Nöckler K. Advancement of a multiplex PCR for the differentiation of all currently described Brucella species. J Microbiol Methods. 2010;80: 112–114. pmid:19887090
- 10. Soleimani M, Shams S, Majidzadeh-A K. Developing a real-time quantitative loop-mediated isothermal amplification assay as a rapid and accurate method for detection of Brucellosis. J Appl Microbiol. 2013;115: 828–834. pmid:23795718
- 11. Zhao W, Brook MA, Li Y. Design of Gold Nanoparticle-Based Colorimetric Biosensing Assays. ChemBioChem. 2008;9: 2363–2371. pmid:18821551
- 12. Kelly KL, Coronado E, Zhao LL, Schatz GC. The Optical Properties of Metal Nanoparticles: The Influence of Size, Shape, and Dielectric Environment. J Phys Chem B. 2003;107: 668–677.
- 13. Bakthavathsalam P, Rajendran VK, Mohammed JAB. A direct detection of Escherichia coli genomic DNA using gold nanoprobes. J Nanobiotechnology. 2012;10: 8. pmid:22309695
- 14. Sattarahmady N, Tondro GH, Gholchin M, Heli H. Gold nanoparticles biosensor of Brucella spp. genomic DNA: Visual and spectrophotometric detections. Biochem Eng J. 2015;97: 1–7.
- 15. Bounaadja L, Albert D, Chénais B, Hénault S, Zygmunt MS, Poliak S, et al. Real-time PCR for identification of Brucella spp.: A comparative study of IS711, bcsp31 and per target genes. Vet Microbiol. 2009;137: 156–164. pmid:19200666
- 16. Fu Z, Zhou X, Xing D. Sensitive colorimetric detection of Listeria monocytogenes based on isothermal gene amplification and unmodified gold nanoparticles. Methods San Diego Calif. 2013;64: 260–266. pmid:23948710
- 17. Hill HD, Mirkin CA. The bio-barcode assay for the detection of protein and nucleic acid targets using DTT-induced ligand exchange : Article : Nature Protocols. Nat Protoc. 2006;1: 324–336. pmid:17406253
- 18. Romero C, Lopez-Goñi I. Improved Method for Purification of Bacterial DNA from Bovine Milk for Detection of Brucella spp. by PCR. Appl Environ Microbiol. 1999;65: 3735–3737. pmid:10427076
- 19. Pal N, Sharma S, Gupta S. Sensitive and rapid detection of pathogenic bacteria in small volumes using impedance spectroscopy technique. Biosens Bioelectron. 2016;77: 270–276. pmid:26414023
- 20. Dibbern AG, Botaro BG, Viziack MP, Silva LFP, Santos MV. Evaluation of methods of DNA extraction from Staphylococcus aureus in milk for use in real-time PCR. Genet Mol Res. 2015;14: 227–233. pmid:25729954
- 21. Amin AS, Hamdy MER, Ibrahim AK. Detection of Brucella melitensis in semen using the polymerase chain reaction assay. Vet Microbiol. 2001;83: 37–44. pmid:11524164