The current research on the epidemiology and diagnosis of Brucella, also stimulated by the zoonotic and bioterrorism aspect, is exploiting novel methods to develop more rapid, reliable and safe diagnostic tools. This paper evaluates the various diagnostic strategies to suggest the most reliable technique to diagnose Brucella spp. infection in bovines. The pros and cons of phenotypic identification, serological diagnosis as well as genotypic identification have been discussed. MALDI-TOF-MS technology has been found to have significant advantages over other conventional and molecular identification methods and can be used as a first-line screening tool for epidemiological studies and outbreak investigations of bovine brucellosis with minimal time, labor and cost.
Brucellosis is a highly contagious disease impacting the dairy sector in India as it causes reproductive impairment in the form of abortion storms, retained placentae and infertility (Priyanka et al., 2019a, b). A precise diagnosis of Brucella spp. infection is important for the control of the disease in animals including the bovines. Besides, brucellosis being a highly contagious zoonotic disease, may also be transmitted to human beings, thus the risk of exposure of human beings to brucellosis can be indirectly assessed by finding the status in animals (Priyanka et al., 2017, 2018a). A presumptive diagnosis of brucellosis can be made by the clinical history of reproductive failures in livestock, but it must be confirmed by laboratory methods. The factors determining the diagnosis of bovine brucellosis are: the absence of clinical signs other than abortion, the incubation period, the high proportion of in apparent infections, the degree of resistance, either natural or resulting from vaccination, and the presence of natural or nonspecific agglutinins (Morgan, 1982). The diagnosis should be based upon the disease history of the herd, demonstration of the causal organism epidemiological observations, serum antibody tests and the cell-mediated immunity. Despite the vigorous attempts for more than one century to come up with a definitive diagnostic technique for brucellosis, the diagnosis still relies on the combination of several tests to avoid false negative results (Poiester et al., 2010). The reviews of the various diagnostic strategies have been mentioned below.
The conventional method of bacterial isolation is still the unequivocal method of reliable and definitive diagnosis in livestock as well as human beings. Isolation from a single animal is sufficient evidence to establish the infection status of a herd. The bacteriological method includes culturing of samples such as aborted fetal stomach contents, milk, blood, lymph nodes and vaginal discharges from suspected cases for isolation and identification of the infecting Brucella organisms (Quinn et al., 1999). As per OIE (2009), brucellae can also be isolated from vaginal swabs, discharges from the uterus, stomach contents, spleen, and lung collected from aborted fetuses, placental bits, semen, and synovial fluids. Associated lymph nodes and spleen, udder tissues, testes and epididymis are useful samples for culture from animal carcasses (Liu, 2009).
The isolated organisms are further tested using molecular-based tests. This sequence of confirmatory procedures is referred to as the “gold standard” method of identifying Brucella (Keid et al., 2007). However, in spite of its high specificity, the culture of Brucella spp. is challenging. Brucella spp. is a fastidious bacterium, thus it requires rich media for primary cultures. Furthermore, its isolation requires a large number of viable bacteria in clinical samples, proper storage and quick delivery to the diagnostic laboratory and it requires BSL3 facilities which are not available in most developing countries (Refai, 2003; Seleem et al., 2010; Hadush and Pal, 2013). A wide range of selective media can be used for the cultivation of Brucella spp. The suitable media include Brucella agar medium (BAM) base, trypticase soy agar, modified Thayer-Martin medium, Farrell’s medium, serum dextrose agar (SDA), glycerol dextrose agar and Castaneda’s medium (OIE, 2004). The culture media are then incubated in an atmosphere of 5 to 10% CO2 at 37ºC for a period of 7 to 14 days till the colonies start to appear (Alton et al., 1988; Bridgewater, 1989; Meyer et al., 2008).
Vaginal secretions should be sampled after abortion or parturition, preferably using a swab with transporter medium, allowing isolation of the organism up to six weeks post-parturition or abortion; samples must be immediately sent to the laboratory, preferentially frozen at -20 ºC (+4 ºC for milk sample), and they must be identified as suspect of Brucella spp. infection (Poiester et al., 2010). Vaginal swabs, semen and seminal fluid have low numbers of viable organisms, therefore enrichment media containing selected antibiotics can improve the sensitivity in these cases (De Miguel et al., 2011). After 3-5 days of incubation on selective agar, growth appears as pinpoint, smooth, glistening, bluish, translucent colonies. As they age the colonies become opaque and measure about 2-3 mm in diameter. Smooth colonies in a clear growth medium such as serum-dextrose agar, are convex with entire-edge, and a smooth shiny surface and are pale yellowish-brown when viewed under transmitted light. Smooth forms are often markedly pathogenic whereas the rough variants are usually less pathogenic (Quinn et al., 1999). The smears from suspected colonies stained with Modified Ziehl Neelsen’s stain show small red-staining i.e. MZN positive coccobacilli. Brucella is Gram-negative, non-motile, catalase-positive, oxidase-positive, indole-negative, gives a rapid urease activity (except some B. melitensis strains) and reduce nitrates to nitrites.
For routine identification, a combination of growth characteristics, colonial and cellular morphology, staining properties, agglutinating antiserum and biochemical reactions allow an accurate identification (Quinn et al., 1999). Presently, the fully automated microbial identification systems like Vitek 2 Compact system are also available which perform bacterial identification by biochemical analysis using colorimetry. The Vitek 2 Compact has been used for the identification of isolates of different Brucella species by various researchers including Borriello et al. (2013); Xu et al. (2013); Maymona et al. (2014); Alsharabasi (2015), Deshmukh et al. (2015); Eisenberg et al. (2017); Jia et al. (2017) and Paul et al. (2017). It provides more precise and rapid identification with minimum handling of cultures by simultaneous application of a multitude of biochemical tests, followed by the precise prediction of Brucella upto species level.
Classical differentiation between the different Brucella species and biovars is based on phage typing, sensibility to dyes, oxidative metabolic profiles, CO2 requirements, H2S production and agglutination with monospecific antisera (Alton et al., 1988). These typing methods are time-consuming, often subjective and risk for the laboratory personnel; that is why methodological improvements are desirable. Because of the costs inducted, lack of sensitivity and difficulty in performing the culture procedures, there is an indirect method of diagnosis by way of serological tests. The current epidemiological and diagnostic Brucella research, also stimulated by the bioterrorism debate, is further exploiting novel molecular typing methods to develop more rapid, reliable and safe diagnostic tools.
Major Brucella species (B. abortus, B. melitensis and B. suis) contain O-polysaccharide on their cell surface, which is a part of the lipopolysaccharide. The O-polysaccharide is lacking in other species (B. ovis and B. canis) of Brucella. The major species which contain O-polysaccharide are diagnosed serologically using either whole cell antigen or smooth lipopolysaccharide. B. abortus antigens were the major antigens used in most of the serological tests. The serological tests rely on a reaction between the Brucella antigen and antibodies produced in the host in response to the infection. A number of classes and subclasses of antibody (isotypes) may occur in positive sera and the various serological tests vary in their ability to detect the different isotypes.
There are many serological tests for demonstrating that Brucella antibodies exist in serum, milk, whey, vaginal mucus, semen, and muscle juice. The commonly used tests are the Milk Ring Test (MRT), Serum Agglutination Test (SAT), Rose Bengal Plate Test (RBPT), Anti-globulin (Coombs) Test, 2- Mercaptoethanol, Rivanol, and the Enzyme-linked Immunosorbent Assay (ELISA) (Morgan, 1982). The reliability of serological tests to detect brucellosis depends on the antibodies present at the time of examination thus the infected animals may escape from detection. Serological tests may show cross-reactions with other Gram-negative organisms such as Salmonella group N, Eschericia coli O:157, E. coli O:116, and Pseudomonas maltophilia; however, the most notable cross-reaction is between smooth lipopolysaccharide (S-LPS) found in Brucella and Yersinia enterocolitica O:9 making diagnosis difficult due to the sharing of antigenic determinants in the O-polysaccharide (O-PS) molecule, which is the basis for most serological tests (Corbel et al., 1983; Muñoz et al., 2005; Nielsen et al., 2006). Several countries have reported such cross-reactions during serological screening (Weynants et al., 1996; Bercovich, 1998), and false positives in addition to false negatives have often limited accurate diagnosis and disease eradication programmes.
Historically, the Standard Tube Agglutination Test (STAT) for brucellosis described by Wright and Smith (1897), has been recognized as the principal serological test used for the diagnosis of brucellosis. The test may give a false positive reaction due to the cross-reacting antibody (IgM). Therefore, its discontinuation is recommended by the OIE (OIE, 2000). However, the test is still in use after a large number of modifications to inactivate IgM agglutination, such as the incorporation of the acidified antigen, Rivanol precipitation and 2-mercaptoethanol.
The Milk Ring Test (MRT) is a screening test for bulk milk used to detect infected animals on a herd basis or to monitor the clean herds. Herds of which the MRT is positive should be examined by serological tests to identify the infected animals (Alton et al., 1988). Although it is a relatively insensitive test subject to wrong interpretation caused by various milk conditions such as mastitis, colostrum and milk at the end of the lactating cycle, it is recommended by the OIE as a screening test for bovine brucellosis.
The Rose Bengal Plate Test (RBPT) is a spot agglutination test used for rapid herd screening. In RBPT, the antigen is used at a low pH of 3.65 which prevents some agglutination by IgM and encourages agglutination by IgG1 thereby reducing nonspecific interactions (Corbel, 1972). The drawbacks of RBPT include: low sensitivity, particularly in chronic cases, relatively low specificity in endemic areas and prozones make strongly positive sera appear negative in RBPT (Diaz et al., 2011). The present World Health Organization (WHO) guidelines recommend the confirmation of the RBPT by other assays such as serum agglutination tests (Ruiz Mesa et al., 2005; Diaz et al., 2011).
The Complement Fixation Test (CFT) is a very specific and sensitive test. According to Radostits et al. (1994a; b), the CFT rarely exhibits non-specific reaction and is useful in differentiating the titres of calfhood vaccination from those due to infection. However, this method has certain disadvantages like high cost, complexity for execution, and the requirement for special equipment and trained laboratory personnel.
The Indirect Enzyme Linked Immunosorbent Assay (I-ELISA) is a highly sensitive test, but sometimes it is not capable of differentiating between antibody resulting from S19 vaccination or other false-positive reactions from those induced by pathogenic Brucella strains. Therefore, it should be considered more as a screening test than a confirmatory test in the testing of vaccinated herds affected by false-positive results (Nielsen, 2002; OIE, 2004). Competitive ELISA (c-ELISA) can distinguish vaccine antibody, whereas both conventional serological tests and the indirect ELISA can’t (Nielsen, 2002). This test can be invariably employed in the diagnosis of brucellosis in livestock in the endemic areas where vaccination is routinely carried out.
Other serological tests include fluorescence polarization assay (FPA), agar gel immunodiffusion test (AGID), Coombs test, Dipstick assay, immunocapture agglutination for anti-Brucella (BCAP) assay, lateral flow assay and rapid slide agglutination assay test (RSAT).
The brucellae genome is encoded on two circular chromosomes with sizes close to 2.05 Mb and 1.15 Mb for each species (Michaux-Charachon et al., 1997). Only the small chromosomes of B. suis, B. canis and B. neotomae are 50 kb longer. The Guanine / Cytosine (G + C) contents in the DNA of various members of the genus Brucella are 55 to 58 per cent (Hoyer and McCullough, 1968 and Verger et al., 1987). Almost identical proportions of potential coding regions (1028 and 1035, respectively) are present in both chromosomes. Housekeeping genes are evenly distributed all over the genome, which makes a highly probable coexistence (Moreno and Moriyon, 2002). A high conservation of restricted sites and genes order has been revealed by chromosomal mapping. Variability is localized to certain regions, most often on the small chromosome. The nucleotide sequence similarity between all Brucella species is also high and DNA-DNA homology exceeds 90 per cent (Patel, 2007).
Despite the high genetic similarities, a range of different molecular techniques has been established for the differentiation of Brucella species and to some extent the biovars. DNA-based methods such as gene probes and PCR utilize primers derived from different polymorphic regions in the genomes of Brucella spp. These molecular approaches include the analysis of outer-membrane genes and PCR-RFLP (Ficht et al., 1989; 1996; Cloeckaert et al., 1995, 1996, 2001; Vizcaino et al., 1997; 2004; Garcia-Yoldi et al., 2005; Al Dahouk et al., 2005), locus specific conventional and real-time PCR assays targeting different polymorphic regions (Fekete et al., 1990; Bricker and Halling, 1994; Redkar et al., 2001; Ferrao-Beck et al., 2006; Garcia-Yoldi et al., 2006; Ratushna et al., 2006), arbitrary primed and rep-PCR (Mercier et al., 1996; Tcherneva et al., 1996; Fekete et al., 1992; Tcherneva et al., 2000), ‘infrequent restriction site’ (IRS)-PCR (Cloeckaert et al. 2003), multilocus sequence typing (Whatmore et al., 2007), multilocus variable number tandem-repeat analysis (Bricker et al., 2003; Le Flèche et al., 2006; Whatmore et al., 2006; Garcia-Yoldi et al., 2007; Al Dahouk et al., 2007), AFLP (Whatmore et al., 2005), and SNP-PCR (Marianelli et al., 2006; Scott et al., 2007; Fretin et al., 2008; Foster et al., 2008).
The identification of Brucella at genus, species and even biovar levels has improved with the application of molecular methods, especially PCR. Molecular detection of Brucella spp. can be done directly on the clinical samples without the previous isolation of the organism. In addition, these techniques can be used to complement the results obtained from phenotypic tests (Bricker, 2002). The technique is chosen as per the type of biological sample and the goal, i.e. diagnosis or molecular characterization or epidemiological survey. Each type of clinical samples has inherent and unique difficulties for adequate sample preparation. The most common difficulties arose from co-purification of PCR inhibitors with the DNA and from interference by excessive host DNA (Morata et al., 1998).
Since the routine identification and differentiation of brucellosis suspected specimens, based on culture isolation and phenotypic characterization, requires a biosafety level-3 (BSL3) protocols for the high risk of laboratory-acquired infections (Boschiroli et al., 2001), molecular methods have been explored in order to overcome these difficulties. A number of PCR methods developed for the detection of Brucella are more and more used in the diagnosis of brucellosis owing to their more sensitivity than conventional culture methods and more specificity than serological methods (Al Dahouk et al., 2013). These are currently being used for the diagnosis of several infectious diseases caused by fastidious or slowly growing bacteria. Thus, the speed and sensitivity of the PCR assay coupled with the reduced risk to the laboratory workers, made this technique a very useful tool for the diagnosis of brucellosis.
However, molecular methods are relatively expensive, with variable sensitivity (Al Dahouk and Nockler, 2011), and their efficiency is highly dependent on primers specificity (Wang et al., 2014). Thus, they are more appropriate for the differential diagnosis rather than for establishing prevalence. The DNA extraction protocol, type of clinical sample, and detection limits of each protocol, are other factors that can influence the efficiency of the technique (Mitka et al., 2007). Also, the presence of large amounts of bovine genomic DNA may have inhibitory effects on the PCR assay (Priyanka et al., 2018b)
The methods based on PCR are becoming useful and to date, considerable progress has been made in the development of more sensitive, specific, easier and cheaper PCR techniques for Brucella detection (Yu and Nielsen, 2010). PCR had been successfully used for the identification of Brucella in bovine blood and milk (Leal-Klevezas et al., 1995; Romero and López-Goni, 1999; Romero et al., 1995b), aborted foetuses and associated maternal tissues (Fekete et al., 1992; Gallien et al., 1998; Cetinkaya et al., 1999, Cortez et al., 2001 and O’Leary et al., 2006), nasal secretion (Sreevatsan et al., 2000) and goats or sheep milk and cheese (Serpe et al., 1999; Tantillo et al., 2001,2003). For identification of Brucella spp. at the genus- level, the primers for sequences encoding bcsp31 (B4/B5) (Baily et al., 1992), 16S rRNA (F4/R2) (Romero et al., 1995a), 16S-23S intergenic transcribed spacers (ITS) (Rijpens et al., 1996; Bricker, 2000), 16S-23S rDNA interspace region (ITS66/ITS279) (Keid et al., 2007), IS711 (IS313/IS639) (Hénault et al., 2000), per (bruc1/bruc5) (Bogdanovich et al., 2004), omp2 (JPF/ JPR) (Leal-Klevezas et al., 1995), outer membrane proteins (Imaoka et al., 2007), proteins of the omp25/omp31 family (Vizcaino et al., 2004) have been used.
By increasing the number of molecular markers, both sensitivity and specificity can be increased accordingly. Molecular assays targeting the IS711 insertion element, which is found in multiple copies within Brucella chromosomes, also improve analytical sensitivity (Bounaadja et al., 2009). Genus-specific PCR assays are generally adequate for the molecular diagnosis of human brucellosis (Al Dahouk and Nöckler, 2011). The bcsp31 gene, coding for a 31-kDa immunogenic outer membrane protein conserved among all Brucella spp. is the most common molecular target in clinical applications (Baily et al., 1992). Such a genus-specific PCR can help to avoid false-negative results in individuals infected with unusual species and biovars.
Several studies have described PCR assays that make use of the specific occurrence of the multiple insertion element IS711 which was described by Halling et al. (1993) and is stable in number and position in the Brucella chromosomes (Bricker and Halling 1994; 1995; Bricker et al., 2000; Cloeckaert et al., 2000; 2003; Ohishi et al., 2004; Ocampo-Sosa et al., 2005). There has been steady progress towards more sophisticated differential assays despite the high level of conservation among Brucella species and strains. Hence, in the present study, primers targeting IS711 gene were employed for Brucella abortus species-specific PCR (Doust et al., 2007). The 16S-23S genes, the IS711 insertion sequence and the bcsp31 gene of Brucella spp. are validated for the detection of Brucella (Ouahrani-Bettache et al., 1996). Comparison of sensitivity of 3 pairs of primers amplifying 3 different fragments including a gene encoding the bcsp31, a sequence of 16S rRNA of B. abortus, and a gene encoding omp2 revealed the sensitivities of the bcsp31, omp2 and 16S rRNA to be 98%, 88.4% and 53.1%, respectively (Baddour and Alkhalifa, 2008). Navarro et al. (2002) compared detection ability of three primer pairs specific for the bcsp31, 16SrRNA and omp2 genes of Brucella in human blood samples and variation in sensitivity for detecting purified Brucella DNA was reported with the bcsp31 gene to be the most sensitive for detecting Brucella DNA.
Various methods have been optimized for a number of Brucella spp. using tissues, and blood or milk samples. The first published PCR based diagnostic assay was developed by Fekete et al. (1990) which was specific for brucellae, applicable to all species and biovars and very sensitive. This assay targeted a 635 bp sequence from a gene encoding a 43 kDa outer membrane protein of B. abortus strain 19. Fekete et al. (1992) also processed 105 aborted bovine tissues comprising of fetal tissues viz., lung, liver, spleen and stomach contents and maternal tissues viz., uterine exudates and placenta and compared the PCR sensitivity and specificity with bacteriological examination which were found to be 98 per cent and 96 per cent, respectively. Brucella gene 16S rRNA published by Dorsch et al. (1989) was explored to yield 800bp amplicon from B. abortus and other species of Brucella (Herman and De Ridder, 1992). In their opinion, the sequence was highly conserved and the test could be extended to the entire genus. A new PCR assay based on bcsp-31 gene of Brucella organisms was published by Baily et al. (1992) which contained a single pair of oligonucleotide primers designed to amplify a 223bp product.
Another PCR assay with primer derived from the 16S rRNA sequence of B. abortus, was developed by Romero et al. (1995a, b). In the study, they included all the representative strains of the species and biovars of Brucella and other non-Brucella species, which are related to Brucella species either phylogenetically or serologically. A 905bp fragment amplified from all Brucella species but not non-Brucella species except Ochrobactrum anthropi biotype D. They concluded that the assay was having high specificity and sensitivity and might provide a valuable tool for the diagnosis of brucellosis. The assay was evaluated with milk ELISA for the diagnosis of brucellosis in dairy cattle and a proportion of 0.91 agreement was found between the two tests. Finally, they opined that though ELISA was a better screening test than PCR, the combined sensitivity of the two assays was 100 per cent and their simultaneous application was found to be more useful than a single test for rapid screening of brucellosis in dairy cattle.
The PCR assay based on a gene encoding omp2 of brucellae was designed by Leal- Klevezas et al. (1995) for the detection of Brucella species from body fluids of infected animals. The test was found to be sensitive to detect the brucellae in clinical samples. Da Costa et al. (1996) thoroughly investigated the assay in which all the Brucella species, biovars and non-Brucella organisms were tested and found that all six Brucella species and one strain of Ochrobactrum were positive for amplification. The assay was reported to be robust, sensitive and specific.
The PCR technique provides a promising option in diagnosing brucellosis with high sensitivity in detecting Brucella from pure cultures. Vaid et al. (2004) applied PCR for detection of brucellosis using primers derived from the 43 kDa outer membrane protein gene of B. abortus, the 16S rRNA gene, insertion sequence IS711, BCSP31 (Brucella Cell Surface Protein) gene. PCR for detection of B. abortus infection in blood, milk and lymph tissues by using different primers that amplify various regions of the Brucella genome, IS711 genetic element, 31 kDa outer membrane protein and 16S rRNA was used by O’Leary et al. (2006) and they found that there was no amplification when PCR assays was applied to the blood samples, but obtained amplicons in a proportion of the culture-positive milk (44%) and lymph tissue samples by the same methods. Scholz et al. (2008b) described an assay targeting the conserved gene recA that detects and differentiates O. anthropi, O. intermedium and Brucella by conventional PCR in a single reaction.
A differentiation between some of the classical Brucella species was achieved by the use of the relatively time-consuming ‘infrequent restriction site’ (IRS)-PCR and even characteristic IRS-PCR patterns for B. suis biovar 2 could be found (Cloeckaert et al., 2003). In another study, DNA polymorphisms in the genes for the different outer membrane proteins (OMPs; omp2, omp25, omp31) were examined and tested for differentiation purposes. So it was shown that strains of B. melitensis are characterized by the lack of the signature sequence for the restriction enzyme EcoRV in the omp25 PCR product and B. ovis strains by a 50 bp deletion in the same gene. Furthermore, in the restriction patterns for the omp2 PCR product, species- and also biovar-specific features were found (Cloeckaert et al., 1995). A high variability was also found for the restriction patterns of this PCR product in marine brucellae which does not only hint at the existence of different species but also of different biovars (Cloeckaert et al., 2001).
The target sequence of another PCR based diagnosis of human brucellosis is omp31, encoding the 31-kDa Brucella outer membrane protein (Casañas et al., 2001). In this assay, which has already been adapted to a real-time PCR (Queipo-Ortuño et al., 2005), Brucella species cannot be differentiated from all members of the closely related genus Ochrobactrum. A comprehensive microarray study by Rajashekara et al. (2004) revealed several differences in the genomes of the classical Brucella species most of which were insertions or deletions Marianelli et al. (2006) reported that the rpoB gene contains enough DNA polymorphisms for the identification of Brucella spp.
A diagnostic assay should have the properties of being sensitive, specific, fast and easy to perform. Though, a number of assays are available for the diagnosis of brucellosis, but none is 100% sensitive and specific. To overcome this weakness various authors have evaluated the efficacy of serological, cultural and molecular assays to diagnose brucellosis. Ferris et al. (1995) compared results of 6 serological tests viz Particle Concentration Fluorescence Immunoassay, ACF Assay, Card Test, Buffered Acidified Plate Antigen Assay, STAT and Rivanol Test and isolation for diagnosis of brucellosis in pigs. Leal-Klevezas et al. (2000) compared PCR, serological and bacteriological techniques to diagnose goat brucellosis. Amin et al. (2001) compared PCR and cultural isolation for detection of Brucella melitensis DNA in bovine and ovine semen. Leyla et al. (2003) evaluated detection of Brucella DNA directly from the stomach contents of aborted sheep fetuses with culture isolation.
While comparing PCR, RBPT and STAT for diagnosis of brucellosis in human beings Varasada (2003) found the highest number of positive results by PCR followed by RBPT and STAT. Lavaroni et al. (2004) comparatively evaluated isolation, blood PCR, i-ELISA and CFT for diagnosis of bovine brucellosis. Rahman (2005) evaluated serological and cultural methods for diagnosis of B. abortus biotype 1 infection in experimentally infected Sprague-Dawley rats. Gupta et al. (2006) evaluated sensitivity and specificity of the tissue PCR in comparison to STAT and dot-ELISA. O’Leary et al. (2006) assessed the viability of using conventional and real-time PCR assays as potential diagnostic tools for the detection of Brucella abortus in naturally infected cows.
A study comparing the culture and PCR methods for detection of Brucella was conducted by Kanani (2007) using semen of serologically positive 101 bulls; PCR assay detected a highest number of positive bulls (19) than cultural isolation (8). Also, the highest numbers of bulls were found positive by B4 / B5 primer pair based PCR assay (18.81 per cent) as compared to JPF / JPR (1.98 per cent) and F4 / R2 (4.95 per cent) published primer pairs, respectively. The same finding was reported by Ghodasara (2008) that the B4/B5 primer pair was more suitable than the other two pairs of primers. Out of the 10 cultured isolates from 248 samples of vaginal swabs and aborted materials from cows, buffaloes, goats and bitches, the desired product of 223 bp using B4/B5 primer pair was amplified in all the 10 isolates, 8 by JPF/JPR primer pair and 8 by F4/R2 primer pair. But, when PCR and ELISA were compared by Chothe et al. (2013) for the detection of Brucella in 200 serum samples, only three samples yielded positive results in PCR against the 75 ELISA positive samples.
The PCR assay for detection of Brucella DNA using BCSP 31 target gene and IS711 locus was conducted by Garshasbi et al. (2014) which showed that an amplicon of 223 bp was obtained in 73.8 % (133/180) of the tested sera using primers (B4/B5) and an amplicon of 498 bp was obtained in 63.8% (115/180) of the samples using Brucella abortus-specific primers derived from a locus adjacent to the 3′-end of IS711. But, in another study conducted by Patel et al. (2015), all the samples (7 out of 33) from aborted buffaloes which yielded Brucella in genus specific PCR were confirmed as Brucella abortus in species specific PCR based on IS711 as well. Similarly, when Karthik et al. (2014) used the same primers i.e. bcsp31 and IS711 for the detection and identification of B. abortus in blood samples (n=370) of cattle from three states viz. Uttar Pradesh, Uttarakhand and Tamil Nadu, a total of 56 samples (15.03 %) were detected as positive by both the PCRs.
Several multiplex PCRs have been developed using different primer combinations which identify Brucella at the species level and partly at the biovar level as well. The AMOS-PCR (abortus-melitensis-ovis-suis) was the first species-specific multiplex PCR which identified and differentiated Brucella biovars including B. abortus biovars 1, 2 and 4, B. melitensis, B. ovis and B. suis biovar 1. It is based on polymorphism due to species-specific localization of the insertion sequence IS711 in the Brucella chromosome (Bricker and Halling, 1994). This PCR technique has a disadvantage of not being able to identify some Brucella species like B. canis and B. neotomae. Furthermore, some biovars within a given species gave negative results. The assay was improved to identify the vaccine strains S19 and RB51 by including further strain specific oligonucleotides into the reaction mixture (Bricker and Halling, 1995). This assay has been employed by several other laboratories with great success (Adone et al., 2001). A multiplex PCR for the detection of Brucella spp. and Leptospira spp. in aborted bovine fetuses was described by Richtzenhain et al. (2002).
In a study based on AMOS PCR, Baek et al. (2003) collected serum and blood sample from three dogs reared on serologically Brucella positive dairy farm in Korea with a high incidence for Brucella abortus which were serologically positive for RBPT and PAT. The amplified product with 498 bp DNA band of B. abortus in AMOS PCR as well as the specific primers of B. canis confirmed the identity of the organism. There was 100 per cent homology of the canine isolate and the bovine pathogen isolated from the farm, thus the only possible source of infection was infected cattle on the same farm. The results suggested that dogs should be routinely included in brucellosis surveillance and eradication programmes.
The AMOS PCR assay was further modified as BaSS-PCR (Brucella abortus Strain Specific PCR assay) which helped to identify and differentiate field strains of biotypes 1, 2 and 4 of B. abortus, and vaccine strains and other Brucella species from cattle (Bricker et al., 2003). Alain et al. (2005) developed a new PCR assay to identify B. abortus biovar 5, 6 and 9 and the subgroup 3b of biovar 3. This was the modified form of the assay developed earlier by Bricker and Halling (1994) and was named as AMOS-ERY PCR. For identification of B. abortus biovars 3, 5, 6 and 9 a new primer was added (Ocampo-Sosa et al., 2005). For recognition and discrimination of Brucella species and vaccine strains in a single step, a multiplex PCR assay, Bruce-ladder was developed (García-Yoldi et al., 2006). The multiplex was further enhanced to identify the marine strains, B. microti and B. inopinata. However, it does not differentiate at the biovar level, or below (López-Goñi et al., 2008; Hubber et al., 2009; Mayer-Scholl et al., 2010). Lopez-Goni et al. (2008) developed a multiplex PCR assay (Bruce-ladder) which could differentiate in a single step all of the classical Brucella species, including those found in marine animals and the S19, RB51 and Rev.1 vaccine strains. A comparable multiplex approach was described having the ability to discriminate the six classical species based on species-specific differences (Hinic et al., 2008), however, the approach was not as inclusive as “Bruce-ladder”.
The primer pair identifying B. microti (Scholz et al., 2008a) was included in the multiplex PCR described by Garcia-Yoldi et al. (2006), and the assay was set up on the DNA of Brucella reference strains and field isolates. The assay allowed the identification of all currently known Brucella species, also distinguishing between the marine species B. ceti and B. pinnipedialis besides identifying the recently described species B. microti and B. inopinata. An advancement of the Garcia Yoldi protocol for the differentiation of all currently described Brucella species was published by Mayer-Scholl et al. (2010). A multiplex PCR assay (Suis ladder) was also developed recently to differentiate among biovars of B. suis (López-Goñi et al., 2011). A PCR multiplex for Brucella and Leptospira was used by Scarcelli et al. (2004) to analyze the samples of abomasal contents, organs and/or fetal annexes of 67 aborted bovine fetuses, besides the bacteriological methods. PCR-multiplex showed 50.7% (34/67) of the samples positive for Brucella, whereas, the Brucella could be isolated from 38.8% (26/67) of the samples, thus showing an 88% agreement rate between the two methods used. As per the results, PCR was found to be more sensitive than culture in bovine brucellosis cases. Another multiplex PCR (mPCR) for the detection of Brucella spp. and Leptospira spp. in aborted bovine fetuses was described by Richtzenhain et al. (2002) who applied the mPCR to 63 clinical samples from bovine aborted fetuses. The 31 kDa outer membrane protein gene for Brucella spp. (primer B4/B5) was employed in the mPCR which amplified the expected size of amplicons (223 bp).
The mPCR assays for the detection of Brucella spp. (PCR/Bruce) and Salmonella abortus ovis (PCR/SAO) was performed by Sharifzadeh et al. (2008) for analyzing 54 clinical samples from aborted fetal stomach contents collected from the ovine abortion. The isolates amplified amplicons of 243 bp for Brucella spp. and 172 bp for Salmonella abortus ovis. Out of the 54 samples, 14 samples were totally negative for both, 10 resulted positive for Brucella spp., 24 came positive for Salmonella abortus ovis and 6 resulted positive for both the bacteria. Lopez-Goni et al. (2008) also used Bruce-Ladder PCR for the confirmation of Brucella species from 625 Brucella strains from different animal and geographical origins, and they obtained the expected profile of Brucella abortus in field isolates compared to known strains. The multiplex PCR Bruce-ladder based typing of 153 Brucella strains isolated from different regions of Mexico was reported by Morales-Estrada et al. (2012). The three vaccine strains (B. abortus S19 and RB51, and B. melitensis Rev1) were also included in this study. The results of microbiological typing and multiplex Bruce ladder amplification were identical for all the Brucella isolates tested. Because of its speed, Bruce-Ladder PCR served as a useful method in the typing of Brucella species isolated from animal and humans. The same assay was used by Orzil et al. (2016) for the identification and differentiation of Brucella species viz., Brucella abortus, B. suis, B. ovis, B. melitensis in the field as well as slaughter house samples in Brazil.
Real Time PCR
So far, acceptance of molecular diagnostics has been slow. This is reasonable considering the consequences at stake. However, threats of biological warfare and agro-terrorism may accelerate the process. Already, assays specific to this circumstance are being considered (McDonald et al., 2001) and the real-time decision is being explored (Redkar et al., 2001). The real-time PCR constitutes a further technological improvement for the molecular identification and differentiation of Brucella species. Many studies have shown that the conventional method for detecting Brucella spp. are technically time-consuming and labour intensive than real-time PCR assay (Bogdanovich et al., 2004; Yang et al., 2007). A measurable fluorescence signal is obtained during the real-time PCR process by different approaches, relying on the cleavage of fluorogenic probes, e.g. by double-stranded DNA intercalating dye (SYBR Green I), by enzymatically released fluorophores (5’exonuclease assay) or by fluorescence resonance energy transfer (hybridization probes). Using SYBR Green I assay, designing of probes is not required because of the non-specific intercalation of the dye in double-stranded DNA followed by fluorescence emission. However melting curve analysis is mandatory and sometimes, the amplicon sequencing can be necessary. Real-time PCR methods based on SYBR green have been shown to be less expensive than TaqMan based methods and more sensitive than conventional PCR techniques. Due to their ability to amplify larger fragments than TaqMan based real-time PCRs, these methods enable phylogenetic studies by sequencing PCR products. Finally, they can be easily adapted to conventional PCR protocols, and therefore can be used in a larger number of laboratories (Sacristán et al., 2015).
The real-time PCRs were described by Redkar et al. (2001); Newby et al. (2003); Probert et al. (2004) and Bogdanovich et al. (2004) as rapid, sensitive and specific diagnostic tools with a low risk of cross-contamination and the potential of automation. Using real-time PCR, the nucleic acids can be quantified and data can be automated for the individual samples. There is no requirement of post-amplification handling of PCR products and it can be performed in a very short time without electrophoretic analysis, thereby reducing the risk of laboratory contamination as well as false-positive results. Recently, the real-time PCR assays have been described for testing the Brucella cells (Redkar et al., 2001), urine (Queipo-Ortuno et al., 2005), blood, and paraffin-embedded tissues (Kattar et al., 2007). The real-time PCR assays targeting 16S-23S internal transcribed spacer region (ITS) and the genes coding omp25 and omp31 (Kattar et al., 2007), bcsp 31 (Colmenero et al., 2005; Debeaumont et al., 2005; Queipo-Ortuño et al., 2008), and IS711 (Cerekci et al., 2011; Zhang et al., 2013) have been developed for the rapid detection and differentiation of Brucella species in clinical samples.
Three separate real-time PCRs were developed to specifically identify seven biovars of B. abortus, three biovars of B. melitensis and biovar one of B. suis using fluorescence resonance energy transfer; with the upstream primers derived from the insertion element, IS711 whereas, the reverse primer and FRET probes were selected from unique species or biovar-specific chromosomal loci. The sensitivity of B. abortus-specific assay was as low as 0.25 pg DNA corresponding to 16-25 genome copies and similar detection levels were also observed for B. melitensis and B. suis-specific assays (Redkar et al., 2001). In spite of the IS711 based real-time PCR assay being specific and highly sensitive, its different copy number according to each species can affect the quantification of bacterial load (Chen et al., 2007). Another study for evaluation of three assays (SYBR Green I, 5-exonuclease and hybridization probes) to detect Brucella abortus was done by Newby et al. (2003) and the greatest specificity was achieved with the hybridization probe assay for use in a real-time PCR assay to detect Brucella abortus. Whereas, in another study, the majority (>90%) of the tested B. abortus and B. melitensis strains could be correctly identified in a real-time multiplex assay, employing only one reverse primer (binding IS711 of B. abortus and B. melitensis) and two specific forward primers (targeting the neighbouring alkB and BMEI1162 gene, respectively) (Probert et al., 2004).
Meanwhile, other real-time PCR assays were also developed, like the LightCycler-based real-time PCR (LCPCR) assay developed by Queipo-Ortuno et al. (2005) for the detection of Brucella based on a 223-bp gene sequence encoding an immunogenetic membrane protein (BCSP31) specific for the Brucella genus; melting curve and DNA sequencing analysis was performed to verify the specificity of the PCR products. This LC-PCR assay was found to be 91.9% sensitive and 95.4% specific for active brucellosis. The viability of using conventional and real-time PCR assays as potential diagnostic tools for the detection of B. abortus in naturally infected cows was assessed by O’Leary et al. (2006). In this study, PCR assays that amplified various regions of the Brucella genome, IS711 genetic element, 31 kDa omp and 16S rRNA, were optimized using nine known Brucella strains. To evaluate the various previously published real-time PCR assays targeting bcsp31, per, IS711, alkB/IS711 and BMEI1162/ IS711, an in-house assay was developed by Al Dahouk et al. (2007) using 248 Brucella strains representing the biotypes of all species and a large panel of clinically relevant, phylogenetically related and serologically cross-reacting bacteria. It was concluded that assays targeting the bcsp31 gene can be recommended to screen for Brucella. A comprehensive approach was made by Hinic et al. (2008) who used unique genetic loci of the six classical species to develop seven individual reactions for detection of the Brucella genus and the differentiation between the six species to be used in conventional as well as real-time PCR assay based on the Brucella-specific insertion sequence IS711.
Also, the further studies done by Bounaadja et al. (2009) by using TaqMan probes for the detection of Brucella at genus level, revealed that the real-time PCR assay targeting IS711 presented an identical or a greater sensitivity than those targeting the bcsp31 and per genes. The TaqMan probe-based real-time PCR with the target sequence of IS711 was carried out with the probe of Brucella genus Hinic Probe (IS711) by Hinic et al. (2009). The real-time PCR assays using such TaqMan probes have been used to study the prevalence of Brucella by various investigators, like Doosti and Ghasemi (2011) who performed TaqMan analysis on 425 bovine blood samples in southwest Iran and found 9, 69 and 5 of these samples to be positive for B. melitensis, B. abortus, and both bacteria respectively. A similar real-time assay was carried out by Dehkordi et al. (2012) in a total of 3710 DNA of abomasal contents of bovine, ovine, caprine and camel aborted fetuses. In the bovine fetuses, 281/892 (31.5%) gave positive results for Brucella species by conventional PCR and the TaqMan analysis confirmed that 45/281 and 231/281 were positive for B. meltensis and B. abortus, respectively.
Besides the studies being conducted in the animals, the human studies also revealed the real-time PCR assay to be an efficient monitoring tool, such as the one conducted by Sohrabi et al. (2014) who reported efficient diagnosis and treatment follow-up of human brucellosis by a novel quantitative TaqMan real-time PCR assay based on bcsp31 gene; by monitoring the DNA load of the 37 brucellosis patients for four weeks which decreased significantly by the end of the treatment period.
Other Molecular Approaches
Genomic fingerprinting is another approach to differentiate between Brucella species, biovars and strains. REP- and ERIC-PCR (rep-PCR) were used to examine whether these assays could be used for a differential typing of brucellae on the basis of the resulting banding patterns (genomic fingerprints) (Mercier et al., 1996; Tcherneva et al., 1996). The two studies yielded inconsistent results and REP-PCR assays proved to be difficult to reproduce and not the method of choice for reliable differentiation of species and strains. However, both groups showed that by the one or the other technique a range of different banding patterns could be obtained and that even B. canis strains could be differentiated from the other species by their banding patterns after REP-PCR (Tcherneva et al., 1996). Complex banding patterns were also found after PCR analyses of Brucella when short (10mer) and randomly designed primers were used for amplification (RAPD PCR) (Fekete et al., 1992; Tcherneva et al., 2000).
A fingerprinting method with a DIG-labelled IS711 probe was successfully used to differentiate between terrestrial and marine Brucella strains (Ouahrani-Bettache et al., 1996; Bricker et al., 2000). A similarly high variability in genomic fingerprints was detected in the typing of Brucella strains after a ‘multi-locus variable number tandem repeats’ analysis. However, the results showed neither species- nor biovar specificity (Bricker et al., 2003). The high genomic similarities between Brucella species is a major reason for the difficulty to differentiate between them on the basis of molecular techniques. Hence, the newer approaches including the proteomics-based (for e.g. MALDI-TOF-MS), and genomics-based techniques like fingerprinting and other approaches as the key strategy towards the identification of regions of variability in the Brucella genomes, are being evaluated to cover the sensitivity and specificity related challenges in the diagnosis of brucellosis.
Matrix-Assisted Laser Desorption Ionization- Time of Flight- Mass Spectrometry (MALDI-TOF-MS)
Still undisputable detection of Brucella is the classic isolation followed by genus and species identification with either phenotypic or nucleic acid recognition (OIE, 2012). But the molecular diagnostic methods, mainly 16S ribosomal RNA sequencing or real-time PCR detection of selected genes remain complicated and costly, and are not suited for use on the vast majority of routine samples. Further, the high genetic and phenotypic homology of Brucella renders its genospeciation an easier-said-than-done task. PCR based on specific genome sequences still suffer some inter-lab standardization problems (Yu and Nielsen, 2010). On the other hand, phenotypic methods reveal quantitative rather than qualitative differences among brucellae.
Analysis of cellular proteome is a method which occupies an intermediary position with respect to the phenotypic–genotypic dichotomy, since the proteins analyzed reflect gene products and metabolic functions (Singhal et al., 2015). The automation in proteomics technology, in recent years, has increased its throughput and potential use for microbial identification. The MALDI-TOF MS technique, combined with reference peptide databases and advanced software, has revolutionized microbial characterization (Seng et al., 2009). It is consistent with 16S rRNA gene sequencing and is expected to substitute for classic biochemical tests (Van Belkum et al., 2012). Its quickness and reliability make it fit for counter-bioterrorism, epidemiological tracing of field strains and detection of food contamination (Sandrin et al., 2013). MALDI-TOF MS is approximately two-thirds less expensive than conventional bacteriological methods (Böhme et al., 2012).
MALDI-TOF-MS has existed for a long time but it was in 1996 when MALDI-TOF spectral fingerprints could be obtained from whole bacterial cells for the first time (Holland et al., 1996). The same year, Krishnamurthy et al. (1996) obtained spectral fingerprints of pathogenic species such as Bacillus anthracis, B. melitensis, Yersinia pestis, and Francisella tularensis using MALDI-TOF (Carbonnelle et al., 2011). Ever since, the number of publications concerning the bacterial as well as mold and yeast identification have increased exponentially. However, use of MALDI-TOF in clinical microbiology as a routine first-line identification method started just during the past five years (Kostrzewa et al., 2013). Databases that include the main pathogenic microorganisms have been developed, thus allowing the use of this method in routine bacterial identification from plate culture. Recently, to identify Brucella species a reference library was constructed using 12 Brucella strains. With this ‘Brucella library’ discrimination was not possible to the species level (Ferreira et al., 2010).
MALDI-TOF-MS technology has fundamentally altered well established diagnostic testing methods because of its significant advantages over other conventional and molecular identification methods (Murray, 2012). It is rapid, and reliable, it takes only few minutes for correct identification (Fenselau, 2012). In addition, the MALDI technique is simple, does not require highly skilled personnel and is cost-effective (Seng et al., 2009). MALDI-TOF-MS works well for many bacterial species hence has the potential to replace conventional phenotypic identification for most bacterial strains isolated in clinical microbiology laboratories (Biswas and Rolain, 2013). The intrinsic property of MALDI-TOF-MS is to detect the mass-to-charge ratio (m/z) of bacterial ribosomal proteins, providing a unique mass spectrum of the microorganism within minutes (Carbonnelle et al., 2011). Importantly, MALDI approach does not rely on actual identification of the biomarker ion peaks in an MS spectrum but on the characteristic mass profile generated by a set of ion peaks that constitute a bacterial “fingerprint” (Dieckmann and Malorny, 2011). In this method, the biopolymer molecules are converted into isolated ionized molecules in the gas phase which are then separated according to their molecular weight after migration in an electric field. Each molecule detected is characterized by the molecular mass, the charge, and the relative intensity of the signal (Carbonnelle et al., 2011). A mass spectrum unique to the organism is produced, get compared to a library of spectra obtained from known reference organisms, and the organism’s likely identification is provided based on the closest match (Dekker and Branda, 2011). The SARAMIS (Spectral Archive and Microbial Identification System) database version 4.09 (originally developed by AnagnosTec, contains ReferenceSpectra for 1161 bacteria, and 263 mycota and yeast, and SuperSpectra for 552 bacteria, and 139 mycota and yeast (Martiny et al., 2012).
MALDI-TOF MS may be used to analyze samples of many types, including solutions of organic molecules, nucleic acids, proteins, and whole microorganisms, with the last two being the most useful in present clinical Microbiology applications (Dekker and Branda, 2011). Fatty acids were evaluated early on as biomarkers for bacterial identification, but rejected as they are too dependent on growth and storage condition (Fenselau, 2012). However, the most reliable MS biomarkers for bacterial identification are considered to be the major proteins, mainly ribosomal proteins which are abundant, basic, and of medium hydrophobicity, all biochemical traits that favor efficient ionization (De Carolis et al., 2014). A sufficient number of stable mass signals of these proteins(between 2000 Da and 20000 Da) can be detected, yielding profile spectra consisting of a series of peaks that are conserved at genus, species and subspecies as well (Barbuddhe et al., 2008).
Recently, the MALDI-TOF-MS has been used as a tool for classification and subtyping of bacteria. While MALDI-TOF spectrometry-based identification of bacteria at genus and species levels has been shown clearly to be rapid and effective, the utility of this approach at the strain level has not been completely explored and lack approved guidelines for data interpretation. Identification to the more specific ‘‘strain’’ requires higher resolution approaches and tends to be more challenging, because strains within a single species are quite often extremely similar, genotypically and phenotypically (Sandrin et al., 2013). There has been considerable interest in using MALDI-TOF MS to identify fastidious organisms and potential agents of bioterrorism. Ferreira et al. (2010) reported the reliability of MALDI-TOF in the identification of Brucella species at genus level by studying 131 clinical isolates on the MALDI Biotyper 2.0 profiles created for type strains belonging to B. melitensis biotypes 1, 2 and 3; B. abortus biotypes 1, 2, 5 and 9; B. suis, B. canis, B. ceti and B. pinnipedialis. As stated by him, an important problem for the routine use of MALDI-TOF-MS for identification of Brucella species is that no reference library for Brucella has been incorporated to the main databases, because of the issues derived from their potential bio-terrorist use.
Later, the accurate identification of Brucella upto species level was achieved by Lista et al. (2011) using MALDI-TOF-MS by constructing a Brucella reference library based on multilocus variable-number tandem repeat analysis (MLVA) data. The comparison of MS-spectra from Brucella species against a custom-made MALDI-TOF-MS reference library could be utilized to use MALDI-TOF-MS as a rapid identification method for Brucella species. In this manner, identification of 99.3% of the 152 isolates at the species level, and B. suis biovar 1 and 2 at the biovar level was obtained. This result demonstrated that even minimal genomic differences between these serovars of Brucella translate to specific proteomic differences.
The results from other studies also indicated that MALDI-TOF-MS assay is a reliable approach to identify Brucella genus and species and an increasing number of different Brucella strains in the database could provide a higher discriminatory power. A similar inference was given by De Maio et al. (2015) who assessed a new protein extraction protocol and constructed a home-made reference database to improve the efficiency of the method. The reliability of this database was evaluated by testing blind-coded Brucella field isolates and reference strains. The identification results at the genus level were always correct whereas, at the species level, a total of 94% bacterial samples were correctly identified. On the other hand, incorrect biovar assignments resulted in 23 out of 39 B. abortus strains and in 4 out of 53 B. melitensis strains.
In further studies, the biovar delineation of Brucella was further assessed. Sayour and Sayour (2015) evaluated the MALDI-TOF-MS for biotyping of 124 Brucella isolates from raw milk and tissues of cattle, buffaloes, sheep and goats in 9 governorates and unknown areas in Egypt for way faster and reliable genospeciation based on protein profiles. It was concluded that despite the high intrageneric similarity of Brucella, the MALDI Biotyper had enough resolution for binomial identification with good matching scores but MALDI identification at the biovar level was accurate in only B. melitensis bv. 3. Due to the limitation of the library created, it was hard to judge for the other Brucella species tested.
Lastly, for the routine application of MALDI-TOF-MS to unidentified bacteria as well as the safety of laboratory workers, a simple and safe protocol was needed which could allow storage, and eventual shipping, of inactivated samples. Mesureur et al. (2016) described such protocol for preparing Brucella samples prior to their analysis by matrix-assisted laser desorption ionization–time of flight mass spectrometry, which was also effective for several other bacterial pathogens.
MALDI-TOF-MS is a rapid as well as a reliable technique for the identification of bovine brucellosis and has the potential to become a first-line screening tool for epidemiological studies and outbreak investigations with minimal time, labor and cost, making it an attractive alternative to the relatively high investment required for other conventional and molecular settings.