BMC Microbiology BioMed Central Methodology article

Background A fast and simple two-step multiplex real-time PCR assay has been developed to replace the traditional, laborious Salmonella serotyping procedure. Molecular beacons were incorporated into the assay as probes for target DNA. Target sequences were regions of the invA, prot6E and fliC genes specific for Salmonella spp. Salmonella Enteritidis and Salmonella Typhimurium, respectively, the two most clinically relevant serotypes. An internal amplification positive control was included in the experiment to ensure the optimal functioning of the PCR and detect possible PCR inhibition. Three sets of primers were used for the amplification of the target sequences. The results were compared to those of the Kauffmann-White antigenic classification scheme. Results The assay was 100% sensitive and specific, correctly identifying all 44 Salmonella strains, all 21 samples of S. Enteritidis and all 17 samples of S. Typhimurium tested in this work. Therefore, the entire experiment had specificity and sensitivity of 100%. The detection limit was down to 10 copies of DNA target per 25 μl reaction. Conclusion The assay can amplify and analyse a large number of samples in approximately 8 hours, compared to the 4 to 5 days conventional identification takes, and is thus considered a very promising method for detecting the two major serotypes of Salmonella quickly and accurately from clinical and environmental samples.


Background
The genus Brucella comprises Gram-negative bacteria that are the etiological agents of brucellosis, a zoonosis that represents one of the most important worldwide disease affecting animals and humans, especially in the Mediterranean areas [1,2]. The disease can be transmitted to humans directly by contact with infected animals or indirectly by contaminated dairy products. Brucella infections can cause chronic debilitating diseases in humans with the involvement of multiple organs and low mortality while in the domesticated animals can lead to reproductive failure with subsequent economic loss. Six classically species are recognized within the genus Brucella: Brucella abortus (7 biovars), Brucella melitensis (3 biovars), Brucella suis (5 biovars), Brucella ovis, Brucella canis, and Brucella neotomae [3]. Isolation and characterization of novel different Brucella strains from a wide variety wildlife species from terrestrial and marine mammals, recently classified as three additional species, B. ceti (cetaceans) B. pinnipedialis and B. microti underline the importance of wildlife as reservoir for zoonotic brucellosis [4,5]. In addition, Brucella spp. represent potential biological warfare agents due to the high contagious rates for humans and animals [6,7]. Therefore it is very important to have a strain typing epidemiological tool for source trace back in outbreaks of infection. Characterization of Brucella at species and biovar level can be performed by differential microbiological approaches used for phenotyping [8]. However, these biotyping methods are time consuming and potentially hazardous for laboratory personnel. In addition, the limited variation among some species and biovars may result in conflicting data and complicate interpretation [9,1]. Biotyping methods are not therefore suitable for epidemiological tracking where a more accurate identification is necessary. Thus, genetic characterization using molecular DNA technology has been investigated and several molecular techniques for subtyping have been proposed. In the last years PCR-based assays have been employed in an attempt to find suitable DNA markers for the molecular typing of Brucella. Examples are repetitive intergenic palindromic sequence-PCR (REP-PCR) [10], random amplified polymorphic DNA-PCR (RAPD-PCR) [11], arbitrary primed-PCR (AP-PCR) [12], amplified fragment length polymorphism (AFLP) [13], single nucleotide polymorphism (SNP) [14,15], and polymerase chain reactionrestriction fragment length polymorphism (PCR-RFLP) [16]. Recently, a highly discriminatory typing system called MLVA, based on multilocus characterization of variable number of tandem repeats (VNTRs), has been proposed for several pathogens such as Haemophilus influenzae [17], Mycobacterium tuberculosis [18], Bacillus anthracis and Yersinia pestis [19][20][21]. The multilocus VNTR analysis methods (MLVA-15 and -16) were also described for Brucella genotyping [22,23,6]. The MLVA-15 is based on 15 polymorphic markers subdivided in two panels: the panel 1, consisting of 8 minisatellite markers with a good species identification capability, and the panel 2, consisting of 7 microsatellite markers with higher discriminatory power. MLVA-15 assay generates multiple band profiles or fingerprint patterns and represents a useful molecular approach for Brucella typing in spite of the high DNA homology among Brucella species (90%) [24]. The obtained MLVA band profiles may be resolved by different techniques ranging from low cost manual agarose gels to the more expensive capillary electrophoresis sequencing systems. Recently, a rapid and inexpensive method based on the Lab on a chip technology (Agilent Technologies) has been proposed. This miniaturized platform for electrophoresis applications is able to size and quantitate PCR fragments, and was previously used for studying the genetic variability of bcIA gene of Bacillus cereus [25] and for genotyping of Bacillus anthracis and Yersinia pestis [26]. In this paper we evaluate the possibility to use the Agilent 2100 Bioanalyzer equipment for MLVA typing of Brucella strains using the selected subset of 15 loci proposed by La Flèche [23]. On that account, we analyzed by Agilent technology the collection of seventeen human Brucella isolates whose MLVA fingerprinting profiles were previously resolved by capillary electrophoresis sequencing system [27] and results compared, while twelve DNA samples, provided in 2007 for a MLVA VNTR ring trial, were de novo genotyped.

Results
In order to set up this method, we analyzed, by MLVA-15 [23] on Agilent 2100 Bioanalyzer, seventeen Brucella strains previously identified as B. melitensis biovar 3 by direct sequencing [27]. The 15 VNTR markers amplified for bacterial genotyping were arranged into three duplex and nine singleplex PCR reactions. The arrangement of different loci in the same multiplex was such to avoid overlapping of VNTR markers size ranges; in this way it was possible to use a single 12-wells chip (DNA 1000 LabChip Kit) for strain analysis. After PCR amplification 1 μl of each reaction was loaded into the chip and the amplification product size estimates were obtained by the Agilent Software. These values showed discrepancies compared with those expected [27], making difficult the allele assignment directly from rough data. In order to assign the correct alleles to the Agilent DNA fragment sizes, a conversion table containing for each locus the expected size, the range of observed sizes, including arithmetical average ± standard deviation, and the corresponding allele was produced (Additional File 1). We could establish experimentally the variability range for each allele. Even if we didn't conduct an extensive study on migration, these ranges were determined considering interchip/intrachip variability from the same amplification product or different amplification of the same strain allele or, for the same alleles, different strain amplification (data not shown). The data are considered valuable only if standard deviation is lower than the 50% of the repeat unit length.
In this way, we could measure a variation proportionated to the relative number of nucleotides in each repeat unit. All the allele measurements satisfied this criterion, allowing the unambiguous assignment of the correct allele to each observed value (Additional File 1). In order to validate this platform, we analyzed twelve unknown samples provided by Dr. Falk Melzer for MLVA Brucella 2007 ring trial [28]. The Agilent 2100 Bioanalyzer MLVA-15 products were separated and DNA fragment sizes were correlated to the alleles by the conversion table. The resulting fingerprint for each strain was matched against the MLVA database Brucella test [29], allowing identification of samples and their genetic relationship with the other database strains ( Table 1). The identified species were compared with the VNTR ring trial results [28], obtaining a full concordance.

Discussion
The renewed threat of biological weapons and the appearance of zoonotic infections caused by Brucella spp. have reinforced the need to provide a rapid and specific detection method for the characterization of Brucella isolates. In fact, the discrimination between natural outbreaks and/or intentional release of micro-organism agents may be of crucial importance in the context of the bioterrorism. Brucella strains differentiation in the last years was obtained by assays based on the analysis of the polymorphism of tandemly repeated DNA sequences. This strategy has proven to be appropriate for pathogenic bacterial species typing with a high genetic homogeneity. Recently a new multilocus VNTR analysis (MLVA-15) method, based on a subset of fifteen tandem repeat loci, was described demonstrating high discriminatory power [23]. The different alleles, amplified by standard PCR techniques, can be analyzed by several electrophoretic techniques that can be time-consuming as agarose gel, or require high costs for reagents and instruments as capillary electrophoresis sequencing. Our attention was addressed on the Agilent 2100 "Lab on a Chip" platform, in order to obtain a more rapid and inexpensive method of discrimination. The loading of the fifteen amplification products in a single chip provides an increased time-reduction (chip run time is approximately 30 minutes) as compared to assay run on agarose gels for the same analysed markers. After data collection and analysis, we observed that the size estimated by the Agilent 2100 Bioanalyzer were shifted by a variable value (offset) in respect to the real size. Indeed, the estimated sizes were shifted of a variable value (offset) respect to the real size estimated by sequencing and each offset showed a variable range. These differences could be due to the different nature of the gel matrix or to a slightly biased flanking sequence or repeat unit specific mobility pattern [21]. These discrepancies between observed and expected sizes have been overcome creating a correspondence table which allows for each observed value to assign the expected size and corresponding allele (Additional File 1). The comparison of the results obtained by the multilocus VNTR analysis (MLVA-15) method [23] on the Agilent 2100 "Lab on a Chip" platform showed a good size correlation with nucleotide sequence analysis, confirming the rapidity and efficiency of this platform respect to standard sequencing or ethidium bromide slab gel electrophoresis. Furthermore, the possibility to compare different chip data in different times made the system suitable for epidemiological purposes. We consider this system one the most promising platforms for MLVA-15 typing because offers a fair compromise among costs, speed and specificity compared to any of the conventional molecular typing techniques.

Conclusion
In this paper is described a rapid, accurate and reproducible system for genotyping of Brucella. The method is based on Multiple Locus Variable Number Tandem Repeats  [29] (VNTR) Analysis (MLVA) assay with 15 markers (MLVA-15), previously described, and the Agilent 2100 Bioanalyzer, for the separation of nucleic acid molecules. The validation of the system was performed reanalyzing seventeen DNA samples, previously genotyped by traditional techniques, and twelve DNA provided for a VNTR ring trials, obtaining a general concordance of the results. Therefore MLVA typing data produced by Agilent system represents an alternative to standard sequencing or ethidium bromide slab gel electrophoresis.

MLVA-15 analysis
The amplification products were loaded into chip wells prepared according to manufacturer recommendations (DNA 1000 LabChip Kit). Each chip contains 16 wells: 12 for the samples, 3 for gel mix. After gel preparation, each sample well was loaded with 1 μl of PCR reaction and 5 μl of internal marker (containing two MW size standards of 15 and 1500 bp). One microliter of DNA ladder was loaded in the ladder well. Finally, the chip was vortexed for 60 sec and inserted into Agilent 2100 Bioanalyzer. During the separation of the fragments, the samples were analyzed sequentially and electropherograms, virtual gel images and table data were shown. Amplification product size estimates were obtained by using the Agilent 2100 Expert Software version B.02.03.SI307 firmware C.01.055 (Agilent Technologies).