A novel real-time PCR assay for specific detection of Brucella melitensis

Background Brucellosis is a zoonosis that occurs worldwide. The disease has been completely eradicated in livestock in Sweden in 1994, and all cases of confirmed human brucellosis are imported into Sweden from other countries. However, due to an increase in the number of refugees and asylum seekers from the middle-east to Sweden, there is a need to improve the current diagnostic methodology for Brucella melitensis. Whilst culture of Brucella species can be used as a diagnostic tool, real-time PCR approaches provide a much faster result. The aim of this study was to set up a species-specific real-time PCR for the detection of all biovars of Brucella melitensis, which could be used routinely in diagnostic laboratories. Methods A Brucella melitensis real-time PCR assay was designed using all available genomes in the public database of Brucella (N = 96) including all complete genomes of Brucella melitensis (N = 17). The assay was validated with a collection of 37 Brucella species reference strains, 120 Brucella melitensis human clinical isolates, and 45 clinically relevant non-Brucella melitensis strains. Results In this study we developed a single real-time PCR for the specific detection of all biovars of Brucella melitensis. Conclusions This new real-time PCR method shows a high specificity (100%) and a high sensitivity (1.25 GE/μl) and has been implemented in the laboratories of four governmental authorities across Sweden.


Background
Despite ongoing eradication programs, brucellosis is a widespread zoonosis that infects mainly cattle, sheep, goats, and pigs. It also leads to considerable financial losses in animal husbandry due to abortion and fertility problems in cattle, sheep and goats [1,2]. Some Brucella species can also infect humans and more than 500,000 human cases are reported annually worldwide [3]. Brucellosis is a febrile illness, sometimes with localized bone and tissue infection, or multiorgan disease. Transmission to humans occurs through different routes: the ingestion of unpasteurized milk and dairy products; direct contact with infected animal tissues; or accidental ingestion, inhalation or injection of cultured Brucella [4]. Most human Brucellosis cases occur in Syria, Turkey, Mexico and Iran, but also around the Mediterranean basin (Portugal, Spain, Southern France, Italy, Greece, Turkey, North Africa) [3]. Brucella comprises six classical species (B. abortus, B. canis, B. melitensis, B. neotomae, B. ovis, and B. suis) and five novel species (B. ceti, B. microti, B. inopinata, B. papionis and B. pinnipedialis). B. melitensis is recognized as the main human pathogen associated with human outbreaks worldwide [3,5,6]. The species of the genus Brucella can be distinguished on the basis of phenotype, genotype and preferred host. Sweden was officially declared free of brucellosis in 1994 [7] and all human cases in Sweden originate from countries abroad. The incidence in Sweden is between 1 and 20 reported cases per year according to the Swedish National Surveillance System SmiNet-2 [8]. The gold standard method for its diagnosis is the isolation of the bacteria from clinical samples via blood cultures and identification by classical microbiological tube testing. Brucella grows slowly and visible cultures appear after 3-4 days, but it can take more than 2 weeks to obtain a definitive result [9]. Due to its pathogenicity, a biosafety level 3 laboratory (BSL-3) is mandatory when handling Brucella organisms. Laboratory-acquired infections are rarely diagnosed or reported, however they do occur [9,10].
The application of DNA-based methods for Brucella diagnosis is challenging, since all Brucella species have a very high degree of genetic homology (up till 99.9%), as shown by whole genome sequencing of B. abortus, B. melitensis, and B. suis [11][12][13][14]. However, several groups have recently developed PCR-based assays for the discrimination among species and biovars of Brucella. Three Brucella species B. abortus, B. melitensis, and B. suis have been sub-typed into biovars [15,16]. All species within the genus Brucella show an average similarity of 99% across the entire genome and a range between 93% and 99.9% based on analysis and nonparametric inference (ANI) analysis. Data from all public available complete genome sequences of all type strains and reference strains were included in the design of primers and probes for the real-time PCR assay of this study. One multiplex PCR (AMOS) developed by Bricker and Halling in 1994, is applicable to differentiate between B. abortus biovars 1, 2 and 4, B. melitensis, and B. ovis, B. suis biovar 1 by specific PCR products based on unique chromosomal loci of the mobile genetic element IS711 in their genome [17]. This PCR was later improved by another laboratory by adding specific primers for the identification of B. abortus biovars 5, 6, 9 and genotype 3b of biovar 3 [18]. In 2009, Huber et al. developed a random amplified polymorphic DNA PCR assay to differentiate all recognized Brucella species, including the marine mammals-infecting species B. ceti and B. pinnipedialis [19]. However, all published methods are developed to distinguish species and biovars mainly by gel-based bar patterns and the majority of described methods were tested with a few strains.
Kim et al. [20] developed a new real-time PCR for distinguishing B. abortus from other Brucella species, which is based on a single nucleotide polymorphism. However, there was a need for a reliable real-time PCR for the detection of B. melitensis at the species level [21]. The method has to be validated for clinical purposes with a large number of human isolates to fulfill the validation requirements of the certified laboratory of the public health institute. A Brucella genus specific realtime PCR assay is currently in use at the Public Health Agency in Sweden (FOHM). However, B. melitensis is the most prevalent species, and we need to be able to differentiate to the species level, due to the epidemiological significance of B. melitensis [21].

Methods
With the aim of developing a real-time PCR assay for the detection of all known species and biovars of B. melitensis, the genomes of B. ceti, B. inopinata, B. neotomae and B. suis biovar 4 were sequenced and analyzed. This as necessary due to the lack of necessary genome data in the publicly available databases at the start of this project [14]. The panel to validate the inclusivity and exclusivity of the real-time PCR assay contained all known biovars of Brucella, as well as 45 other non-Brucella DNA from American Type Culture Collection (ATCC), Culture Collection University of Gothenburg (CCUG) and National Collection of Type Cultures (NCTC) with clinical relevance. A number of environmental samples, as well as the closely related species Ochrobactrum anthropi, were also tested in this study (Tables 1 and 2).

Bacterial strains and growth conditions
A collection of 31 Brucella sp. reference strains (Table 1) and 120 B. melitensis human clinical isolates were isolated in the BSL-3 laboratory at the FOHM, by cultivation on 5% sheep blood agar plates in a 5%-10% CO 2 atmosphere at 37°C for 48 h. All non-Brucella strains were routinely cultivated on 5% sheep blood agar over night at 37°C.
The B. melitensis human clinical isolates were collected between 1994 and 2016 from Swedish patients, who had returned from Brucella-endemic countries, in a biorepository of the FOHM and used as stipulated in the regulations for diagnostic development and quality assessment. The FOHM performs all Brucella diagnostics on human samples in Sweden, and the B. melitensis isolates used in this study have previously been confirmed by cultivation, a general real-time PCR amplifying DNA of all Brucella strains, as well as by Matrix Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry (MALDI-TOF MS) [22,23]. Ethical review of research involving humans is not applicable for diagnostic development and quality assessment.

DNA extraction
Bacterial DNA was extracted using the commercially available EZ1® DNA Tissue Kit from (Qiagen, Stockholm, Sweden) according to the protocol from the manufacturer and stored at 4°C until use. A volume of 5 μl of seal herpes virus cell culture was used in each sample of the total volume of 200 μl in the extraction step as a process control. Each sample was eluted in 50 μl elution buffer.

Bioinformatic analyses, primer & probe design and real-time PCR
In order to identify a B. melitensis-specific target(s), all available genomes in the public database of Brucella (N = 96) including all complete genomes of B. melitensis (N = 17) were used in the design of primers and probes. A 2 basepair deletion which is highly specific for B. melitensis was found in the acetyl-CoA acetyltransferase gene. Primers were designed flanking this deletion and a short 12-mer MGB probe was placed over the area with the deletion. The forward nucleotide sequence 5′-GC TCGACACAAAGGGCCA-3′ (Biomers, Germany) and the reverse nucleotide sequence 5′-CAAGCGTGGTCT GGCGA-3′ (Biomers, Germany) were used with the FAMlabelled hydrolysis probe -CCGAGATACAAA-MGB (Applied Biosystems®).
The real-time PCR assays were carried out in 25 μL reaction mixtures containing 5 μL template DNA, in PerfeCta Multiplex qPCR SuperMix (Applied Biosys-tems®) diluted in UltraPure™ DNase/RNase-Free Distilled Water (Invitrogen™), 0.9 μM of each primer, and 0.2 μM probe. Amplification and detection were performed using two different PCR machines, an ABI-7500/96-well plates real-time FAST PCR platform, and a StepOne Plus real-time PCR system (Applied Biosystems®).
Thermocycling parameters were as follows: inactivation 95°C for 3 min, followed by 45 cycles 95°C for 3 s, and annealing at 60°C for 30 s. The baseline and threshold were set using the auto-baseline and threshold feature in StepOneplus Software v2.2.2 (Applied Biosystems®). Samples were considered positive if target amplification was detected within 40 cycles.

Results and discussion
Real-time PCR is a rapid and reliable method for the analysis of Brucella in clinical samples. However, the high genomic similarity between different Brucella species makes the design of a species-specific real-time PCR assay difficult. An alternative attempt to use a probe (FAM-CCGCCGAGATACAAA) with Tm 68-70°C (Primer Express v 3.0) resulted in a signal for all Brucella species, but also a signal for non-melitensis species. To increase the discriminatory effect, a shorter probe (FAM-CCGAGATACAAA-MGB) with Tm 57°C was used. By using this shorter probe we could eliminate all the non-specific amplification derived from other Brucella species than B. melitensis.
The method was validated according to Broeders et al. 2014 [25] and according to the validation standards of the Swedish National Veterinary Institute (SVA), the Public Health Agency of Sweden (FOHM), the Swedish National Food Agency (NFA) and the Swedish Defense Research Agency (FOI). The validation comprised applicability, practicability, specificity, linearity, and sensitivity.

Applicability
No false positive or false negative result is acceptable in clinical BSL3 pathogen diagnostics. To guarantee a correct identification of BSL3-pathogens, we strongly recommend isolation of the bacteria on selective Brucella agar plates. This enables evaluation of the phenotypic properties of the strains, such as colony morphology, and enriches the molecular target of the PCR, as well as reduces the concentration of potential PCR inhibitors. Even after the enrichment an internal PCR process control is recommended as described above. All strains in our validation were cultured under the same conditions as bacterial strains were isolated in the clinical diagnostics workflow. The real-time PCR method was therefore recognized as applicable in combination with the isolation of bacteria from clinical specimens.

Specificity
The assay was tested with 120 human clinical B. melitensis isolates, 31 other non B. melitensis Brucella strains, 45 non B. melitensis strains, and 6 B. melitensis reference strains. There was no amplification from any other Brucella species other than B. melitensis. The specificity of this method was 100% because all 126 B. melitensis samples were tested positive while all 76 non-B. melitensis samples gave no amplification in the real-time PCR.

Practicability
The assay was used for the identification of Brucella samples from the first External Quality Assurance Exercises (EQAEs) on highly infectious agents (BSL-3) from the EU financed project "Efficient response to highly dangerous and emerging pathogens" (EMERGE). The results obtained with this assay conformed to the results of other laboratories.

Sensitivity
The limit of detection was 6.25 genome equivalents of B. melitensis (ATCC 23456) per reaction of template DNA (1.25 GE/μl). At a DNA concentration of 3.125 genome equivalents per reaction, 80% of the reactions were still positive (Fig. 1).

Linearity
The linear dynamic range of this PCR assay was determined by testing serially diluted DNA of B. melitensis

Conclusions
This new real-time PCR method for the detection of all biovars of B. melitensis shows a high specificity (100%) and a high sensitivity (1.25 GE/μl) and has been successfully implemented in laboratories of four governmental authorities across Sweden (National Food Agency, National Veterinary Institute, Swedish Defence Research Agency and The Public Health Agency of Sweden).