Veterinary Research Communications, 31 (2007) 951–965 DOI: 10.1007/s11259-006-0109-6
C Springer 2007
Diagnosis of Canine Brucellosis: Comparison between Serological and Microbiological Tests and a PCR Based on Primers to 16S-23S rDNA Interspacer L.B. Keid1 , R.M. Soares1 , N.R. Vieira1 , J. Megid2 , V.R. Salgado2 , S.A. Vasconcellos1 , M. da Costa3 , F. Gregori4 and L.J. Richtzenhain1,∗ 1 Department of Preventive Veterinary Medicine and Animal Health, School of Veterinary Medicine, University of S˜ao Paulo, S˜ao Paulo, SP, Brazil; 2 Department of Veterinary Hygiene and Public Health, School of Veterinary Medicine, S˜ao Paulo State University, Botucatu, S˜ao Paulo, Brazil; 3 Departament of Microbiology, Health Sciences Institute, Federal University of Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil; 4 Research and Developmental Center of Animal Health, Biological Institute, S˜ao Paulo, S˜ao Paulo, Brazil ∗ Correspondence: E-mail:
[email protected]
ABSTRACT A pair of primers directed to 16S-23S rDNA interspacer (ITS) was designed directed to Brucella genetic sequences in order to develop a polymerase chain reaction (PCR) putatively capable of amplifying DNA from any Brucella species. Nucleic acid extracts from whole-blood from naive dogs were spiked with decreasing amounts of Brucella canis RM6/66 DNA and the resulting solutions were tested by PCR. In addition, the ability of PCR to amplify Brucella spp. genetic sequences from naturally infected dogs was evaluated using 210 whole-blood samples of dogs from 19 kennels. The whole-blood samples collected were subjected to blood culture and PCR. Serodiagnosis was performed using the rapid slide agglutination test with and without 2-mercaptoethanol. The DNA from whole blood was extracted using proteinase-K, sodium dodecyl sulphate and cetyl trimethyl ammonium bromide followed by phenol–chloroform purification. The PCR was capable of detecting as little as 3.8 fg of Brucella DNA mixed with 450 ng of host DNA. Theoretically, 3.8 fg of Brucella DNA represents the total genomic mass of fewer than two bacterial cells. The PCR diagnostic sensitivity and specificity were 100%. From the results observed in the present study, we conclude that PCR could be used as confirmatory test for diagnosis of B. canis infection. Keywords: Brucella canis, canine brucellosis, PCR, internal transcribed spacer, blood culture, rapid slide agglutination test Abbreviations: CTAB, cetyl trimethyl ammonium bromide; ITS, interspacer; PCR, polymerase chain reaction; RBPT, Rose Bengal Plate test; R-LPS, rough lipopolysaccharide; RSAT, rapid slide agglutination test; SDS, sodium docdecyl sulphate
INTRODUCTION Brucella canis, the aetiological agent of canine brucellosis, is an important cause of abortions and infertility in dogs. Canine brucellosis is currently diagnosed by serodiagnosis and the most widely used serological tests are the rapid slide agglutination test (RSAT) (George and Carmichael, 1978; Carmichael and Joubert, 1987), rapid slide agglutination test with 2mercaptoethanol (2ME-RSAT), and agar gel immunodiffusion test (Zoha and Carmichael, 1982). The 2-mercaptoethanol is used to improve specificity of the agglutination tests as
951
952
a result of the denaturation of the less specifically reacting IgM antibodies (Carmichael, 1976; Badakhsh et al., 1982). All these tests detect mainly antibodies directed to surface antigens of Brucella, of which the rough lipopolysaccharide (R-LPS) is the most important. It has been established that these tests are highly accurate in identifying non-infected animals (Carmichael, 1976; Wanke et al., 2004), but they are subject to considerable interpretative error because LPS antigens of several bacterial species cross-react with B. canis antigens (Carmichael et al., 1984). Therefore, false-positive reactions are very common, indicating that any positive result should be confirmed by a specific test (Carmichael, 1976; Carmichael et al., 1984). B. canis infection should be confirmed using a direct method of diagnosis, such as bacterial isolation. Blood is the sample of choice for microrganism isolation because of the prolonged period of bacteraemia caused by B. canis infection (Carmichael and Kenney, 1970; Johnson and Walker, 1992; Carmichael and Shin, 1996). Although bacterial isolation is considered the gold standard for Brucella diagnosis, the polymerase chain reaction (PCR) is a good alternative to overcome some major drawbacks of bacteriological methods (Bricker, 2002; Al Dahouk et al., 2003; Navarro et al., 2004). Bacterial isolation has the disadvantage of being time-consuming since it takes about 10 days or longer for the identification of the causative agent. In addition, microbiological methods depend on the bacterial viability and, as a consequence, may pose a threat to laboratory personnel (Wallach et al., 2004). Detection of Brucella spp. DNA in whole blood by PCR has already been reported in cattle, goats, bubaline and humans (Leal-Klevezas et al., 1995, 2000; Matar et al., 1996; Queipo-Ortu˜no et al., 1997, 1999; Morata et al., 1998, 1999; Navarro et al., 1999, 2002; Guarino et al., 2000; Zerva et al., 2001; Al-Nakkas et al., 2002, 2005; Nimri, 2003; Efalki et al., 2005a). Nevertheless, the performance of PCR for the detection of Brucella spp in dogs is barely known (Baek et al., 2003). The primer pair B4 and B5 directed to the gene encoding a 31 kDa protein of Brucella has been widely and successfully used for PCR detection of Brucella infection in blood samples (Matar et al., 1996; Queipo-Ortuno et al., 1997; Morata et al., 1998, 1999, 2002; Navarro et al., 1999, 2002; Zerva et al., 2001; Efalki et al., 2005a,b). However, primers B4 and B5 also amplify DNA of Ochrobactrum spp. (Velasco et al., 1998; Casanas et al., 2001), which may cause bacteraemia in humans (Gransden and Eykyn, 1992). The rDNA array of the Brucella genome is present in three copies, whereas the gene coding for the 31 kDa protein, in contrast, is present only once. As reported elsewhere (Navarro et al., 2002; Al Dahouk et al., 2003), a PCR system based on multiple-copy genes may be more sensitive than other PCR systems based on amplification of singlecopy genes. In fact, comparison of different primers for detection of Brucella spp. in human blood samples has shown that primers directed to 16S rDNA are more sensitive than primers directed to 31 kDa protein coding gene (Navarro et al., 2002). Despite the high evolution rate of the spacer region (ITS) between 16S rDNA and 23S rDNA in closely related organisms (Hillis and Dixon, 1991), ITS sequences are essentially identical within Brucella genus, as one can infer by sequence analysis of ITS from B. melitensis, B. suis and B. abortus available in molecular databases. Therefore, the ITS region was chosen to design a primer pair potentially capable of amplifying genetic sequences from any one of the six recognized species of Brucella without cross-reaction with Ochrobactrum spp.
953
TABLE I Bacterial strains used in the study Organism
Strain
Source
Brucella abortus bv. 1 Brucella abortus bv. 1 Brucella abortus bv. 2 Brucella abortus bv. 3 Brucella abortus bv. 4 Brucella abortus bv. 5 Brucella abortus bv. 6 Brucella abortus bv. 9 Brucella melitensis bv. 1 Brucella melitensis bv. 2 Brucella melitensis bv. 3 Brucella suis bv. 1 Brucella suis bv. 2 Brucella suis bv. 3 Brucella suis bv. 4 Brucella ovis Brucella neotomae Brucella canis Ochrobactrum anthopi
544 1119-3 86/8/59 Tulya 292 B3196 870 C68 16M 63/9 Ether 1330 Thomsen 686 40 63/290 5K33 RM 6/66 LMG 3331
ATCC 23448; BCCN R4 BCCN V1 ATCC 23450; BCCN R5 ATCC 23450; BCCN R6 ATCC 23451; BCCN R7 ATCC 23452; BCCN R8 ATCC 23453; BCCN R9 ATCC 23455; BCCN R11 ATCC 23456; BCCN R1 ATCC 23457; BCCN R2 ATCC 23458; BCCN R3 ATCC 23444; BCCN R12 ATCC 23445; BCCN R13 ATCC 23446; BCCN R14 ATCC 23447; BCCN R15 ATCC 25840 ATCC 23459; BCCN R16 ATCC 23365; BCCN R18 ATCC 49188
In view of the very few reports of PCR describing the direct detection of Brucella in dogs, the present study aimed to standardize and evaluate a PCR for the detection of Brucella spp. in whole blood of naturally infected dogs using a primer pair designed specifically for the ITS of Brucella spp. The performance of the developed PCR was compared with that of bacterial isolation from blood, RSAT and 2ME-RSAT. MATERIALS AND METHODS Bacterial reference strains and growth conditions The bacterial strains used in this study are shown in Table I. Brucella strains were kept in tryptose phosphate broth (Oxoid, Basingstoke, Hampshire, UK)–25% glycerol, at −80◦ C. Subcultures were performed on tryptose agar (Oxoid) at 37◦ C for a further 72 h. Ochrobactrum anthropi strain was kept in nutrient broth (Oxoid)–25% glycerol, at −80◦ C. Subcultures were performed in nutrient broth incubated at 25◦ C under aerobic atmosphere for the first 24 h and then on nutrient agar (Oxoid). Blood samples Blood samples were collected from 210 dogs (174 females and 36 males, of several breeds) from 19 commercial breeding kennels located in S˜ao Paulo state, Brazil. A total of 3 ml
954
of blood was collected from each animal by jugular vein puncture, using sodium citrate as anticoagulant. Of this, 2 ml was immediately submitted to bacterial isolation and the 1 ml of remaining sample was stored at −20◦ C until used for PCR. For the serodiagnosis, 4 ml of blood was collected without anticoagulant from each dog by jugular vein puncture and the serum obtained was stored at –20◦ C until used. The samples were collected from March 2002 to January 2005. Clinical examination The dogs were clinically evaluated for the detection of the following clinical signs suggestive of brucellosis: lymphoid hyperplasia, scrotal dermatitis, enlarged testicles or epididymis, abortion, stillbirths, vaginal discharges and pain in the spinal cord. In addition, a questionnaire was used to discover previous history of clinical signs suggestive of brucellosis in each animal, such as abortion, conception failure, and birth of weak or dead puppies (Johnson and Walker, 1990). Bacterial isolation from blood samples Blood samples (2 ml) were cultured in tryptose phosphate broth at 37◦ C, under aerobic conditions, for 30 days. Subcultures were performed on tryptose agar every 5 days and the plates were incubated at 37◦ C, under aerobic atmosphere for 5 days. Brucella canis organisms were presumptively identified by morphological, cultural and biochemical characteristics such CO2 requirement, oxidase, urease, catalase, citrate utilization, sugar fermentation, H2 S production, nitrate reduction, growth in the presence of thionine and basic fuchsin (20 μg/ml), agglutination in the presence of acriflavin and crystal violet dye (Alton et al., 1976). Serological tests Brucella canis serum antibodies were detected in the 210 dogs by rapid slide agglutination test (RSAT) and by RSAT after treatment of sera with 2-mercaptoethanol (2ME-RSAT) using Brucella ovis antigen (D-tec CB, Symbiotics Corp., Kansas City, MO, USA) as described by Badakhsh and colleagues (1982). In order to discard smooth Brucella infections, all dog sera were tested by Rose Bengal Plate test (RBPT) using B. abortus 1119-3 strain antigen. DNA extraction from Brucella spp. and O. anthropi bacterial cultures Bacterial colonies grown on tryptose agar (Brucella spp.) and nutrient agar (O. anthropi) were incubated overnight at 37◦ C in 400 μl of lysis solution (1% sodium dodecyl sulphate, 100 mmol/L NaCl, 10 mmol/L Tris-HCl pH 8.0, 25 mmol/L disodium EDTA pH 8.0, 10 μg/ml proteinase K). Cell wall debris and polysaccharides were removed by precipitation with 5 mmol/L NaCl and CTAB-NaCl solution at 65◦ C for 10 min and a further chloroform
955
extraction (Wilson, 1990). The aqueous phase containing nucleic acids was extracted from lysates by standard protocol using phenol–chloroform–isoamyl alcohol and concentrated by alcohol precipitation (Sambrook et al., 1989). The final sediment was resuspended in 30 μl TE buffer (10 mmol/L Tris-HCl pH 8.0, 1 mmol/L disodium EDTA pH 8.0) and used as template in PCR. DNA extraction from whole-blood samples One ml of each blood sample was thawed and centrifuged at 13 000g for 5 min. Pellets were resuspended in 1000 μl of TE buffer, mixed, and centrifuged again at 13 000g for 5 min. This washing step was repeated (five times) until the sediment lost the reddish colour. The final pellet was resuspended in lysis solution and treated as earlier indicated. TE buffer instead of blood was used as negative control and blood spiked with DNA of B. canis RM6/66 was used as positive control in each set of DNA extracts. Primer design The 16S-23S rDNA ITS region of Brucella melitensis (X95890) was submitted to BLAST search and the most similar sequences were recovered. The recovered genetic sequences from Brucella melitensis, B. suis, B. abortus, Bartonella bacilliformis, Agrobacterium tumefaciens, Acetobacter aceti, Mesorhizobium mediterraneum and Ochrobactrum anthropi (GenBank accession numbers X95890, X95891, X95889, AJ422182, AF345274, AJ007831, AF345261 and X95892, respectively) were aligned and compared with each other using the software BioEdit (Hall, 1999), allowing the design of a pair of oligonucleotides specific for Brucella spp. (ITS66: ACA TAG ATC GCA GGC CAG TCA and ITS279: AGA TAC CGA CGC AAA CGC TAC). The expected size of the amplification product from Brucella is 214 bp. DNA amplification by PCR Amplification reaction mixture was prepared in a volume of 50 μl containing 200 μmol of each deoxynucleoside triphosphate, 50 mmol/L KCl, 10 mmol/L Tris-HCl (pH 9.0), 1.5 mmol/L MgCl2 , 0.5 μmol of each primer, 1.5 U platinum Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA) and 5 μl of template DNA. The reaction was performed in a DNA thermal cycler (MJ Research PTC 200 DNA engine, Watertown, MA, USA) without mineral oil. Ultrapure water was used as negative control and DNA extracted from blood spiked with DNA of B. canis RM6/66 was used as positive control in each set of samples. After an initial denaturation at 95◦ C for 2 min, the PCR profile was set as follows: 30 s of template denaturation at 95◦ C, 30 s of primer annealing at 62◦ C and 30 s of primer extension at 72◦ C, for a total of 40 cycles, with a final extension at 72◦ C for 5 min. The samples were analysed by electrophoresis in a 2% agarose gel and then stained with ethidium bromide (0.5 μg/ml). DNA bands were visualized under UV light. DNA extracts from all the reference strains and all the blood samples from the 210 dogs were tested.
956
Evaluation of analytical sensitivity and specificity of PCR For the evaluation of the analytical sensitivity (Saah and Hoover, 1997) of PCR based on primers ITS66 and ITS279, a concentrated suspension of B. canis RM6/66 was prepared and 10-fold diluted (from 1.0 × 10−1 to 1.0 × 10−10 ) in sterile saline. From the last five dilutions, 0.1 ml suspension was inoculated onto tryptose agar plates and incubated at 37◦ C for 5 days, and the colonies grown on the plates were counted. The original undiluted B. canis suspension was estimated to contain 3.0 × 109 cfu/ml. In order to calculate the analytical sensitivity of the PCR primers, decreasing amounts of bacterial cells (from 1.0 × 104 to 1.0 × 100 cfu) were added to 1 ml of non-infected canine blood. The mock-infected wholeblood samples were then subjected to DNA extraction and PCR amplification. In addition, DNA of B. canis RM6/66 was extracted and suspended in TE buffer. The DNA concentration of the resulting suspension was estimated spectrophotometrically (Beckman DU 640, Fullerton, CA, USA) by reading the optical densities A260 and A280 , as described elsewhere (Sambrook et al., 1989). Suspensions of Brucella DNA in TE buffer were prepared with the following concentrations: 38 pg/μl, 380 fg/μl, 38 fg/μl and 3.8 fg/μl. One microlitre of each dilution was mixed with 4 μl of ultrapure water (final volume of 5 μl) and the resulting four mixes were used as template in the PCR. To study the influence of canine DNA on PCR amplification of Brucella spp. genetic sequences, 450 ng of DNA from a non-infected dog was added to different amounts (38 pg, 380 fg, 38 fg and 3.8 fg) of DNA of B. canis RM6/66 and the resulting mixes were tested by PCR. The DNA of non-infected dogs was obtained from a pool of blood from animals free of Brucella infection according to clinical, serological and microbiological tests. The analytical specificity (Saah and Hoover, 1997) of the PCR based on primers ITS66 and ITS279 was evaluated by testing DNA of Ochrobactrum anthropi extracted from pure colonies diluted in TE buffer.
Statistical analyses The association between the presence of clinical signs suggestive of brucellosis and the positive results of each test was evaluated using chi-squared test. Statistical analyses were performed by using SPSS 9.0 for Windows. To calculate relative diagnostic sensitivity and specificity of PCR, the 210 dogs were divided in three groups, according to the results of blood culture, rapid slide agglutination test, clinical examination and disease-associated clinical signs in the kennel of origin. • Group 1 (Brucella canis infected dogs): characterized by dogs positive by blood culture (n = 64). • Group 2 (Brucella canis non-infected dogs): dogs negative by blood culture, RSAT and 2 ME-RSAT and coming from kennels where clinical, serological and bacteriological evidence of brucellosis was not observed (n = 50). • Group 3 (dogs suspected of brucellosis): dogs negative by blood culture but positive by RSAT or 2 ME-RSAT or dogs negative by blood culture, RSAT and 2 ME-RSAT but coming from kennels where B. canis was isolated (n = 96).
957
Groups 1 and 2 were used to calculate diagnostic sensitivity and specificity of PCR in blood samples.
RESULTS The PCR designed with primers ITS66 and ITS279 was able to detect as little as 3.8 fg of B. canis DNA diluted in TE buffer (Figure 1). The analytical sensitivity of PCR was not affected by the presence of 450 ng of canine DNA (Figure 1). In addition, the PCR amplified Brucella genetic sequences from the mock-infected whole-blood sample containing 1.0×100 cfu/ml. The DNA of all Brucella strains used in this study was successfully amplified, yielding fragments of the expected size. The PCR showed a good analytical specificity since no PCR product was observed when DNA of O. anthropi was used as template. None of the 210 animals was positive by RBPT using B. abortus antigen, suggesting that smooth Brucella infection did not occur. PCR presented a relative diagnostic sensitivity and specificity of 100% when animals of group 1 (positive) and group 2 (negative) were analysed. If group 3 (negative suspected dogs) is considered as negative animals, the relative diagnostic specificity decreases to 86.45% (Table II). In group 1 (blood culture positive dogs) 82.81% (53/64) of the dogs were positive by RSAT and 39.06% (25/64) by 2ME-RSAT. In group 2 (blood culture negative dogs) all the animals were negative by RSAT, 2ME-RSAT. In group 3 (blood culture negative but
Figure 1. Analytical sensitivity of the PCR. Left gel: lane 1, 100 bp DNA ladder (Invitrogen, Carlsbad, CA, USA); lanes 2 to 5, 38 pg, 380 fg, 38 fg and 3.8 fg of B. canis RM6/66 DNA mixed with 450 ng of genomic DNA of a non-infected dog; lane 6, 450 ng of genomic DNA of a non-infected dog without Brucella DNA; lane 7, TE buffer; lane 8, 38 pg of B. canis RM6/66 DNA mixed with TE buffer. Right gel: lanes 9 to 12, 38 pg, 380 fg, 38 fg and 3.8 fg of B. canis RM6/66 DNA mixed with TE buffer; lane 13, GeneRuler100 bp DNA ladder (Fermentas, Hanover, MD, USA)
958
TABLE II Diagnostic sensitivity and specificity of PCR in blood samples for Brucella canis diagnosis in dogs Dog condition Infecteda PCR status Group 1 53 RSAT+ 25 2ME-RSAT+
Non-infectedb Group 2
Suspectedc Group 3
0
13
Positive
64
Negative
0
50
Total
64
50
50 RSAT− 50 2ME-RSAT−
7 RSAT+ 0 2ME-RSAT+ 28 RSAT+ 83 2 2ME-RSAT+ 96
Total 77 133 210
a Dogs positive by blood culture b Dogs negative by blood culture, RSAT, 2ME-RSAT and from healthy kennels where clinical, serological
and bacteriological evidence of brucellosis was not observed c Dogs negative by blood culture but positive by RSAT or 2ME-RSAT or from kennels where B. canis was isolated PCR relative diagnostic sensitivity in Group 1 = 100% PCR relative diagnostic specificity in Group 2 = 100% PCR relative diagnostic specificity in Group 3 = 86.45%
suspected dogs) 36.45% (35/96) of the dogs were positive by RSAT and 2.08% (2/96) were positive by 2ME-RSAT (Table II). The proportion of positive dogs detected by each of the diagnostic method used (RSAT, ME-RSAT, blood culture and PCR) was higher in animals presenting clinical evidence of brucellosis than in animals not presenting clinical signs ( p < 0.05) (Table III).According to the chi-squared test, the proportions of positivity detected by RSAT, 2ME-RSAT, blood culture and PCR were equal between males and females ( p > 0.05).
DISCUSSION In this paper we discuss the performance of a PCR based on novel primers directed to 16S23S rDNA interspacer (ITS) for the detection of Brucella spp. DNA in naturally infected dogs. Comparing ITS of Brucella with homologous sequences searched by the BLAST tool, the closest organisms found were: Bartonella bacilliformis, Agrobacterium tumefaciens, Acetobacter aceti, Mesorhizobium mediterraneum and Ochrobactrum anthropi. In the present study we designed a primer pair (ITS66 and ITS279) potentially capable of amplifying genetic sequences from any one of the six recognized species of Brucella and none of the closely related organisms searched by BLAST. The anchorage site of primer ITS66 is identical within all Brucella species and O. anthropi but showed no similarity with sequences from any other bacteria closely related to Brucella.
6 6
4 1 0 3 3 0 1 124
16 5
4 1 1 2 0 1 0 32
Abortion n = 22 Enlargement of lymph nodes n = 11 Conception failure n = 8 Weak puppies n = 2 Dead puppies n = 1 Vaginal discharge n = 5 Enlarged testicles n = 3 Enlarged epididymis n=1 Diskospondylitis n = 1 Absence of clinical signs n = 156
Negative
Positive
Blood culture
disorder
Clinical
46
0
1
3
1 1 3
4
18 5
Positive
PCR
110
1
0
0
1 0 2
4
4 6
Negative
54
1
1
2
1 1 1
5
18 5
Positive
102
0
0
1
1 0 4
3
4 6
Negative
RSAT
9
1
1
0
0 0 1
1
10 5
Positive
147
0
0
3
2 1 4
7
12 6
Negative
2ME-RSAT
TABLE III Results of clinical examination, rapid slide agglutination test (RSAT), 2-mercaptoethanol rapid slide agglutination test (2ME-RSAT), blood culture and PCR for diagnosis of canine brucellosis caused by Brucella canis
959
960
Primer ITS279 is unique to Brucella spp. sequences and has no identity with O. anthropi. In fact, using the same protocol as that adopted for amplification of Brucella sequences, the primers ITS66 and ITS279 do not amplify genetic sequences of O. anthropi, rendering good analytical specificity of the developed PCR. It is reported that high DNA concentrations inhibit PCR amplification (Cogswell et al., 1996; Morata et al., 1998; Zerva et al., 2001, Navarro et al., 2002). Navarro and colleagues (2002) compared three different PCR methods for the detection of Brucella spp. and studied whether human genomic DNA affected the sensitivity of the three primer pairs for the detection of Brucella DNA in peripheral-blood PCR assay. The most sensitive of the tested primers were F4/R2 (directed to 16S rDNA sequence), which amplified 8 fg of purified B. melitensis DNA. Primers B4/B5 and JPF/JPR (directed to the 31 kDa protein and Omp2 coding genes, respectively) amplified, respectively, 5 and 20 pg of purified genomic DNA. The sensitivity of the amplification using primers F4/R2 and B4/B5 decreased from 8 fg to 800 fg and from 5 pg to 50 pg, respectively, in the presence of 200 ng of human genomic DNA. In order to evaluate the influence of the concentration of host DNA in the PCR, we tested different amounts of Brucella DNA and Brucella colony-forming units either in the presence or in the absence of a fixed amount of host DNA (450 ng). The results of this experiment showed that PCR was able to detect as little as 3.8 fg of Brucella DNA and 1 × 100 Brucella cfu irrespective of the presence of canine genomic DNA. As inferred from the molecular mass of the Brucella genome, 3.8 fg of DNA corresponds, theoretically, to the DNA of fewer than two bacterial cells (Baily et al., 1992). According to Morata and colleagues (1998), amounts of total DNA per PCR higher than 4 μg can lead to inhibition of reaction. This supports the hypothesis that, below certain limits, host DNA does not impair the amplification. Other compounds such the haem group, heparin and EDTA (Morata et al., 1998; QueipoOrtu˜no et al., 1999) are indicated as potential PCR inhibitors. In this study we used sodium citrate as anticoagulant to avoid the inhibitory effects of heparin or EDTA. We believe that the presence of the haem group was not responsible for inhibition, as we washed the leukocyte pellet five times to eliminate the haemoglobin, as described by Morata and colleagues (1998). The developed PCR showed a relative diagnostic sensitivity of 100%, since all dogs of group 1 were PCR positive. Of these 64 dogs, 53 were also positive by RSAT. However, 11 of these animals showed negative results by RSAT, which should indicate an early phase of infection, when bacteraemia is present but the agglutination titre of antibodies is not yet detectable (Carmichael and Kenney, 1968; Flores-Castro and Carmichael, 1978; George and Carmichael, 1978; Carmichael and Joubert, 1987; Carmichael and Greene, 1990) (Table II). Again considering group 1, only 39% (25/64) gave positive results in 2ME-RSAT. Most of the blood culture and PCR positive dogs (60.93%) were negative by 2ME-RSAT. Since 2ME-RSAT is frequently used to confirm positive results obtained by RSAT, these results suggest that 2ME-RSAT is highly specific in detecting anti-B. canis antibodies, but falsenegative results should be observed. The presence of dogs positive by blood culture or PCR but negative by 2ME-RSAT could indicate an early stage of infection, when antibodies are predominantly of the IgM class and IgG was not yet detectable, in addition to the presence of bacteraemia (Badakhsh et al., 1982). RSAT results are often negative during the first 4–8 weeks post infection, but when 2ME is used antibodies can be detected even later,
961
usually at 8–12 weeks post infection, which may lead to false-negative results (Carmichael and Kenney, 1968; Flores-Castro and Carmichael, 1978; George and Carmichael, 1978; Carmichael and Joubert, 1987; Carmichael and Greene, 1990). The relative diagnostic specificity of PCR was 100% as none of the dogs from group 2 (B. canis non-infected dogs) showed positive PCR results. All of them were RSAT negative and came from kennels with no clinical or epidemiological evidence of infection. However, when group 3 (blood culture negative but suspected dogs) was considered as negative animals, the relative diagnostic specificity of PCR decreased to 86.45%, since 13 animals showed positive results. This loss of relative diagnostic specificity of PCR must be interpreted carefully as this group was composed of animals positive by RSAT or 2ME-RSAT or coming from kennels where B. canis had been isolated from at least one dog (Table II). The presence of dogs positive by PCR but negative by blood culture in the suspected group of dogs could be a consequence of the presence of non-viable brucellae in these samples or the presence of low number of circulating organisms. Infection by other species of Brucella that require CO2 atmosphere for isolation (such as some biovars of B. abortus) could also be an explanation, since blood samples were incubated only under aerobic atmosphere (Taylor et al., 1975; Forbes, 1990). However, all the animals were negative by RBPT and came from kennels located in the urban area. Contact with livestock or ingestion of livestock products were not reported. Also regarding group 3, of the 83 dogs that were negative by PCR and blood culture, 28 were positive by RSAT, leading to at least two interpretations of serological results: (i) perhaps not all the serological reactions observed were specific reactions to B. canis (Carmichael, 1976; Zoha and Carmichael, 1982, 1984; Carmichael and Joubert, 1987); (ii) dogs with low titres that nevertheless are infected but abacteraemic (these dogs should be in intermittent periods of bacteraemia or in chronic phase of infection) (Carmichael, 1976; Carmichael and Joubert, 1987; Carmichael and Greene, 1990). Dogs in these conditions can harbour B. canis in organs such as spleen, lymph nodes, prostate gland and epididymis (Moore and Kakuk, 1969; Ueda et al., 1974). Ten of these dogs came from B. canisinfected kennels (kennels where B. canis was isolated from at least one dog) and two of them presented clinical signs of the disease. The presence of clinical signs of brucellosis or history of contact with infected dogs, in animals serologically positive but negative by blood culture and PCR, may support the interpretation that they were infected but abacteraemic. It may be concluded that a negative blood culture or PCR cannot always be relied upon to exclude a diagnosis of brucellosis, especially in chronically infected dogs. Varying values of diagnostic sensitivity and specificity of PCR have been obtained in different studies, which may be a consequence of different protocols for DNA extraction and amplification, different target genes and choice of the reference test (Table IV). Clinical signs are usually not adequate to diagnose canine brucellosis as the disease is insidious in its onset and variable in its clinical evolution. However, the proportions of positive results detected by each of the four diagnostic tests used (RSAT, 2ME-RSAT, blood culture and PCR) were higher in animals presenting clinical evidence of infection. Consequently, brucellosis should be considered in the differential diagnosis whenever there is a history of abortion in females or poor reproductive performance in both sexes (Carmichael, 1990). Considering group 1 (blood culture positive dogs), 38 of the 64 dogs did not show clinical history of brucellosis, but all of them were from kennels where B. canis was isolated from at
Caprine n = 43 Humans n = 50 Humans n = 110 Humans n = 15 Humans n = 76 Humans n = 84 Humans n = 70 Humans n = 333 Humans n = 20
Leal-Klevezas et al. (1996)a
Navarro et al. (1999)
Al-Nakkas et al. (2002)b
Elfaki et al. (2005)
a RBT, Rose Bengal test; CFT, b Nested PCR assay used
Al-Nakkas et al. (2005)b
Nimri (2003)
Zerva et al. (2001)
Blood culture AND Serology (RBT) Serology (SAT and ELISA) AND Blood culture Clinical signs AND Serology (RBT)
Blood culture OR Serology (RBT/CFT) Serology (SAT) AND Clinical signs Blood culture OR Serology (Wright’s tube agglutination and Coombs test) Blood culture OR Serology (SAT) Blood culture OR Serology (Wright’s tube agglutination test) Serology (SAT and ELISA)
Gold standarda
complement fixation test; SAT, standard agglutination test
Queipo-Ortu˜no et al. (1997)
Matar et al. (1996)b
Species
Author
B4 B5
O1/O2 I1/I2 F4 R2 O1/O2 I1/I2
B4 B5 B4 B5
JPR JPF B4 B5 B4 B5
Primers
31 kDa protein
IS711
16S rDNA
IS711
31 kDa protein
31 kDa protein
31 kDa protein
31 kDa protein
omp-2
Target gene
70%
100%
100%
100%
61%
50%
100%
100%
62.79%
PCR diagnostic sensitivity
100%
100%
100%
100%
100%
60%
98.3%
100%
100%
PCR diagnostic specificity
TABLE IV PCR relative diagnostic sensitivity and specificity for Brucella spp. detection in blood samples of different species, using different primers and gold standard tests
962
963
least one dog. It should be emphasized that clinical signs of brucellosis may not be observed in dogs during the period of reproductive quiescence, but these animals may be important source of infection. In addition, female dogs may abort and subsequently whelp normal litters (Carmichael and Greene, 1990). From the results, it may be inferred that RSAT should be used with care as a screening test for the detection of Brucella infection in dogs, as a significant number of infected dogs detected by blood culture were negative by RSAT. Only 25 of 64 dogs of group 1 (blood culture positive dogs) showed 2ME-RSAT-positive results, and all of the 50 dogs of group 2 showed 2ME-RSAT-negative results. Thus, 2MERSAT should be a good alternative for confirming positive results by RSAT, but in dogs showing RSAT-positive and 2ME-RSAT-negative results, B. canis infection cannot not be excluded. Blood culture may be a better alternative for confirming positive results by RSAT, when compared to 2ME-RSAT; however, blood culture is a time-consuming procedure, as it takes at least 10 days for Brucella isolation and identification. Here, we have presented the first evaluation of PCR for the detection of Brucella spp. infection in dogs using bacterial isolation and agglutination test as reference. Considering the analytical specificity of ITS primers and the high diagnostic sensitivity of the PCR assay, we have evidence that the PCR proved to be a good alternative to blood culture for direct detection of brucellosis in dogs. In fact, PCR and blood culture showed good concordance in detection of Brucella-infected dogs, but methods based on molecular biology allow faster diagnosis, which is particularly of interest when brucellosis examination is performed in confined canine populations, as it leads to a rapid confirmation of infection and allows segregation of potential sources of infection. We conclude that PCR should be used as a confirmatory test for positive RSAT results for a rapid brucellosis diagnosis. For diagnosis of Brucella canis infection, we propose careful analysis of the results obtained by serodiagnosis, bacteriological culture, PCR assays and clinical examination of the dogs. The presence of positive results in 2ME-RSAT, blood culture or PCR strongly suggests infection in a suspected animal. However, since bacteraemia is frequently absent in chronically infected dogs, blood culture or PCR negative results do not ensure the absence of infection. RSAT-positive dogs should never be condemned as infected until confirmatory tests are performed, even in the presence of clinical signs suggestive of brucellosis. Problems in the detection of infected kennels are minor, since several individuals are usually involved. However, individual cases pose a variety of problems, especially when complete clinical histories or knowledge of opportunities for infection are not available. Because of the seriousness attending a positive diagnosis, especially as regards those dogs used for breeding purposes, all diagnostic methods should be employed.
ACKNOWLEDGMENTS The authors thank FAPESP and CNPq for financial support through grants numbers 99/02662-1 and 140748/2002-4, respectively.
964
REFERENCES Al Dahouk, S., Tomaso, H., Nockler, K., Neubauer, H. and Frangoulidis, D., 2003. Laboratory-based diagnosis of brucellosis—a review of the literature. Part I: Techniques for direct detection and identification of Brucella spp. Clinical Laboratory, 49, 487–505. Al-Nakkas, A.F., Wright, S.G., Mustafa, A.S. and Wilson, S., 2002. Single-tube, nested PCR for the diagnosis of human brucellosis in Kuwait. Annals of Tropical Medicine and Parasitology, 96, 397–403. Al-Nakkas, A., Mustafa, A.S. and Wright, S.G., 2005. Large-scale evaluation of a single-tube nested PCR for the laboratory diagnosis of human brucellosis in Kuwait. Journal of Medical Microbiology, 54(Pt 8), 727–730. Alton, G.G., Jones, L.M. and Pietz, D.E., 1976. Metodos bacteriologicos. In: Las tecnicas de laboratorios en la brucelosis, 2nd edn, (WHO, Geneva), 11–67. Badakhsh, F.F., Carmichael, L.E. and Douglass, J.A., 1982. Improved rapid slide agglutination test for presumptive diagnosis of canine brucellosis. Clinical Microbiology, 15, 286–289. Baek, B.K., Lim, C.W., Rahman, M.S., Kim, C.H., Oluoch, A. and Kakoma, I., 2003. Brucella abortus infection in indigenous Korean dogs. Canadian Journal of Veterinary Research, 67, 312–314. Baily, G.G., Krahn, J.B., Drasar, B.S. and Stoker, N.G., 1992. Detection of Brucella melitensis and Brucella abortus by DNA amplification. Journal of Tropical Medicine and Hygiene, 95, 271–275. Bricker, B.J., 2002. PCR as a diagnostic tool for brucellosis. Veterinary Microbiology, 90, 435–446. Carmichael, L.E., 1976. Canine brucellosis: an annotated review with cautionary comment. Theriogenology, 6, 105–116. Carmichael, L.E., 1990. Brucella canis. In: K. Nielsen and J.R. Duncan (eds), Animal Brucellosis, (CRC Press, Boca Raton, FL), 335–350. Carmichael, L.E. and Greene, C.E., 1990. Canine brucellosis. In: C.E. Greene (ed.), Infectious Diseases of the Dog and Cat, (W.B. Sauders, Philadelphia), 573–584. Carmichael, L.E. and Joubert, J.C., 1987. A rapid slide agglutination test for the serodiagnosis of Brucella canis infection that employs a variant (M-) organism as antigen. Cornell Veterinarian, 77, 3–12. Carmichael, L.E. and Kenney, R.M., 1968. Canine abortion caused by Brucella canis. Journal of the American Veterinary Medical Association, 152, 605–616. Carmichael, L.E. and Kenney, R.M., 1970. Canine brucellosis: the clinical disease, pathogenesis, and immune response. Journal of the American Veterinary Medical Association, 156, 1726–1734. Carmichel, L.E. and Shin, S.J., 1996. Canine brucellosis: a diagnostician’s dilemma. Seminars in Veterinary Medicine and Surgery (Small Animal), 11, 161–165. Carmichael, L.E., Zoha, S.J. and Flores-Castro, R., 1984. Problems in the serodiagnosis of canine brucellosis: dog responses to cell wall and internal antigens of Brucella canis. Development in Biological Standardization, 56, 371–383. Casanas, M.C., Queipo-Ortuno, M.I., Rodriguez-Torres, A., Orduna, A., Colmenero, J.D. and Morata, P., 2001. Specificity of a polymerase chain reaction assay of a target sequence on the 31-kilodalton Brucella antigen DNA used to diagnose human brucellosis. European Journal of Clinical Microbiology and Infectious Disease, 20, 127–131. Cogswell, F.B., Bantar, C.E., Hughes, T.G., Gu, Y. and Philipp, M.T., 1996. Host DNA can interfere with detection of Borrelia burgdorferi in skin biopsy specimens by PCR. Journal Clinical Microbiology, 34, 980–982. Elfaki, M.G., Al-Hokail, A., Nakeeb, S.M. and Al-Rabiah, F.A., 2005a. Evaluation of culture, tube agglutination, and PCR methods for the diagnosis of brucellosis in humans. Medical Science Monitor, 27, MT69–74. Elfaki, M.G., Uz-Zaman, T., Al-Hokail, A.A. and Nakeeb, S.M., 2005b. Detection of Brucella DNA in sera from patients with brucellosis by polymerase chain reaction. Diagnostic Microbiology and Infectious Disease, 53, 1–7. Flores-Castro, R. and Carmichael, L.E., 1978. Canine brucellosis. Current status of methods for diagnosis. Cornell Veterinarian, 68, 76–88. Forbes, L.B., 1990. Brucella abortus infection in 14 farm dogs. Journal of American Veterinary Medical Association, 196, 911–916. George, L.W. and Carmichael, L.E., 1978. Development of a rose bengal stained plate-test antigen for the rapid diagnosis of Brucella canis infection. Cornell Veterinarian, 68, 530–543. Gransden, W.R. and Eykyn, S.J., 1992. Seven cases of bacteremia due to Ochrobactrum anthropi. Clinical and Infectious Diseases, 15, 1068–1069. Guarino, A., Serpe, L., Fusco, G., Scaramuzzo, A. and Gallo, P., 2000. Detection of Brucella species in buffalo whole blood by gene-specific PCR. Veterinary Record, 147, 634–636. Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, 95–98.
965
Hillis, D. M. and Dixon, M.T., 1991. Ribosomal DNA: molecular evolution and phylogenetic inference. Quarterly Review of Biology, 66, 411–452. Johnson, C.A. and Walker, R.D., 1992. Clinical signs and diagnosis of Brucella canis infection. The Compendium on Continuing Education for the Practicing Veterinarian. Small Animal, 14, 763–772. Leal-Klevezas, D.S., Martinez-Vazquez, I.O., Lopez-Merino, A. and Martinez-Soriano, J.P., 1995. Single-step PCR for detection of Brucella spp. from blood and milk of infected animals. Journal of Clinical Microbiology, 33, 3087–3090. Leal-Klevezas, D.S., Martinez-Vazquez, I.O., Garcia-Cantu, J., Lopez-Merino, A. and Martinez-Soriano, J.P., 2000. Use of polymerase chain reaction to detect Brucella abortus biovar 1 in infected goats. Veterinary Microbiology, 75, 91–97. Matar, G.M., Khneisser, I.A. and Abdelnoor, A.M., 1996. Rapid laboratory confirmation of human brucellosis by PCR analysis of a target sequence on the 31-kilodalton Brucella antigen DNA. Journal of Clinical Microbiology, 34, 477–478. Moore, J.A. and Kakuk, T.J., 1969. Male dogs naturally infected with Brucella canis. Journal of the American Veterinary Medical Association, 155, 1352–1358. Morata, P., Queipo-Ortuno, M.I. and de Dios Colmenero, J., 1998. Strategy for optimizing DNA amplification in a peripheral blood PCR assay used for diagnosis of human brucellosis. Journal of Clinical Microbiology, 36, 2443–2446. Morata, P., Queipo-Ortuno, M.I., Reguera, J.M., Garcia-Ordonez, M.A., Pichardo, C. and Colmenero, J.D., 1999. Posttreatment follow-up of brucellosis by PCR assay. Journal of Clinical Microbiology, 37, 4163–4166. Morata, P., Queipo-Ortuno, M.I., Reguera, J.M., Miralles, F., Lopez-Gonzalez, J.J. and Colmenero, J.D., 2001. Diagnostic yield of a PCR assay in focal complications of brucellosis. Journal of Clinical Microbiology, 39, 3743–3746. Navarro, E., Fernandez, J.A., Escribano, J. and Solera, J., 1999. PCR assay for diagnosis of human brucellosis. Journal of Clinical Microbiology, 37, 1654–1655. Navarro, E., Escribano, J., Fernandez, J. and Solera, J., 2002. Comparison of three different PCR methods for detection of Brucella spp in human blood samples. FEMS Immunology and Medical Microbiology, 34, 147–151. Navarro, E., Casao, M.A. and Solera, J., 2004. Diagnosis of human brucellosis using PCR. Expert Review of Molecular Diagnostics, 4, 115–123. Nimri, L.F., 2003. Diagnosis of recent and relapsed cases of human brucellosis by PCR assay. BMC Infectious Disease, 28, 1–7. Queipo-Ortuno, M.I., Morata, P., Ocon, P., Manchado, P. and Colmenero, J.D., 1997. Rapid diagnosis of human brucellosis by peripheral-blood PCR assay. Journal of Clinical Microbiology, 35, 2927–2930. Queipo-Ortuno, M.I., Garcia-Ordonez, M.A., Colmenero, J.D. and Morata, P., 1999. Hydrogen peroxide improves the efficiency of a peripheral blood PCR assay for diagnosis of human brucellosis. Biotechniques, 27, 248–250, 252. Saah, A.J. and Hoover, D.R., 1997. ‘Sensitivity’ and ‘specificity’ reconsidered. Annals of Internal Medicine, 126, 91–94. Sambrook, J., Fritsch, E.F. and Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual, 2nd edn (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY). Taylor, D.J., Renton, J.P. and McGregor, A.B., 1975. Brucella abortus biotype 1 as a cause of abortion in a bitch. Veterinary Record, 96, 428–429. Ueda, K., Magaribuchi, T., Saegusa. J., Urano. T. and Ito. K., 1974. Spontaneous Brucella canis infection in beagles: bacteriological and serological studies. Nippon Juigaku Zasshi, 36, 381–389. Velasco, J., Romero, C., Lopez-Goni, I., Leiva, J., Diaz, R. and Moriyon, I., 1998. Evaluation of the relatedness of Brucella spp. and Ochrobactrum anthropi and description of Ochrobactrum intermedium sp. nov., a new species with a closer relationship to Brucella spp. International Journal of Systematic Bacteriology, 48, 759–768. Wallach, J.C., Giambartolomei, G.H., Baldi, P.C. and Fossati, C.A., 2004. Human infection with M-strain of Brucella canis. Emerging Infectious Disease, 10, 146–148. Wanke, M.M., 2004. Canine brucellosis. Animal Reproduction Science, 82/83, 195–207. Wilson, K., 1990. Preparation of genomic DNA from bacteria. In: F.M. Ausubel, R. Brent, R.E. Kimgston, et al. (eds), Current Protocols in Molecular Biology, (Greene Publishing Associates/John Wiley and Sons, New York), 241–245. Zerva, L., Bourantas, K., Mitka, S., Kansouzidou, A. and Legakis, N.J., 2001. Serum is the preferred clinical specimen for diagnosis of human brucellosis by PCR. Journal of Clinical Microbiology, 39, 1661–1664. Zoha, S.J. and Carmichael, L.E., 1982. Serological responses of dogs to cell wall and internal antigens of Brucella canis (B. canis). Veterinary Microbiology, 7, 35–50. (Accepted: 28 August 2006; Published online: 16 February 2007)
