Two five-plex PCRs methods for identification of common Salmonella spp. serotypes

Abstract

In Tunisia, Salmonella is the most common bacterial agent responsible for childhood diarrhoea. Currently, isolation of the bacterium by microbiological and biochemical methods and confirmation of the serotype by serological method remain as the "gold standard". This study aimed to differentiate among the most common serotypes of Salmonella spp. via two rapid five-plex PCRs assay (MPCR) to evaluate the molecular serotyping method compared with the gold standard serotyping technique. The two five-plex PCRs assays were designed for the simultaneous detection of six genetic loci from Salmonella enterica serovar Typhimurium and four from S. enterica serovar Typhi. Sixty-one Tunisian strains (46 collected from patients and 15 from food) were isolated during the period 2002–2007. The STM and STY primers were able to discriminate all tested Salmonella serotypes that represent the most common clinical and food strains of S. enterica subsp. enterica in our laboratory. All strains belonged to 19 different serotypes: 15 serotypes gave unique amplification patterns compared each other and the other 4 serotypes were grouped into two pairs that gave the same molecular profile. We resolved this problem through the addition of a monoplex PCR. Salmonella typhimurium ATCC 14028 consistently produced the same molecular profile as S. typhimurium laboratory isolates. Interestingly, seven strains of Anatum serovar produced two different PCR profiles with these primers: five strains had the same amplification pattern STM 2,4,5 and STY 2; however, two strains had another molecular profile STM 2,3,5 and STY 2; so the reproducibility of this method was reduced to 93%. The MPCR system is a rapid, specific, and cost-effective molecular method that gas been proved to have efficient discrimination in serotyping of the most common isolates of S. enterica subsp. enterica.

Introduction

Salmonellosis is one of the most common causes of food-borne infections in humans all over the world. The genus Salmonella is extremely polymorphic. It has been classified into more than 2,400 different known serovars (Nair et al. 2002; Martinez-Urtaza et al. 2004) of which 1,478 are Salmonella enterica subsp. enterica (Bopp et al. 2003; Perch et al. 2003).

Many Salmonella serovars cause gastroenterites that are accompanied by diarrhoea, abdominal pain,s and fever. Among the Salmonella enterica serovars commonly implicated in most major outbreaks are serovars Enteritidis, Anatum, Newport, Cerro, Montevideo, Infantis, and Saint-Paul (Bopp et al. 2003; Perch et al. 2003).

Salmonella enterica subsp. enterica shows different disease syndromes and host specificities according to their antigenic profiles. Therefore, it is necessary and important to discriminate Salmonella serovars in order to ensure that each pathogen is correctly recognized. The differences within Salmonella serovars are based on the surface antigen differences of O and H antigens. The O antigens are derived from the polysaccharide domain of lipopolysaccharide (LPS) in the cell wall, while the H antigens are derived from flagellin protein in the flagella. O and H antigens are used for the identification of Salmonella serovars by agglutination tests using O and H antigen-specific anti-sera, then Salmonella serovars are determined by the combination of O and H antigen types using the Kauffmann-White scheme which is annually updated by the World Health Organization (Popoff et al. 1992; Bopp et al. 2003).

The serological method has been used to identify Salmonella serovars. However, this method is complex, time consuming, and 5–8% of isolates are partially typed or untyped. This can be due to capsular polysaccharides in mucoid strains and in “rough” strains that produce partially formed O antigens that can cross-react with different O antisera (Bopp et al. 2003). So these types of isolates can be partially typed by their H antigens. With H-antigen typing, both flagellar phases must be assayed. Non-motile isolates can be partially typed by their O antigens.

At present, alternative strategies for the identification of serotypes, such as ribotyping (Esteban et al. 1993), IS200 analysis (Ezquerra et al. 1993; Uzzau et al. 1999), real-time PCR, and DNA microarrays (Chan et al. 2003; Porwollik et al. 2004; O'Regan et al. 2008) are under development in many laboratories, but these methods require specialized equipment and there are problems with the reproducibility of results. For these reasons, an easier and simpler method is needed to identify Salmonella serovars.

In this study, we establish two simple five-plex PCRs methods (MPCR) to type the most common Salmonella enterica subsp. enterica serovars. This method is based on detection of genes present in specific serotypes. These genes were selected from analysis of previous work including whole-genome sequencing (Porwollik et al. 2004, 2005). Reproducibility of the assay was also evaluated. As our results show, the method described here is specific, fast, and cost-effective and can be applied in the clinical microbiology laboratory for the serotyping of Salmonella serovars.

Materials and methods

Bacterial strains and DNA extraction

The 19 most common serotypes of Salmonella spp. strains used in this study were collected from 61 Tunisian strains (46 collected from patients and 15 from food) isolated during the period 2002–2007. These serovars that represent more than 90% of all serotypes commonly identified by our clinical laboratory were selected according to their frequency and importance. Salmonella typhimurium ATCC 14028 was also tested as standard strain (Table 1).

Table 1 Molecular detection of common serotypes of Salmonella enterica subsp. enterica using STM and STY primers by multiplex PCR

Each isolate was previously serotyped according to the Kauffmann-White scheme based on slide agglutination to identify the somatic O antigens and flagellar H antigens using commercial antisera (Bio-Rad, Marnes-la-Coquette, France) in Centre de Référence des Salmonella, Shigella et Vibrions cholériques, Institut Pasteur Tunis, Tunisia. All strains were cultured on nutrient agar plates with overnight incubation at 37°C. DNA was prepared from each isolate by boiling (Agarwal et al. 2002).

Primer design and PCR amplification program

Genomic regions chosen for this serotyping study were derived from previous work that used microarray technology to compare the genomic complement of S. enterica serovar Typhimurium LT2 (STM) and S. enterica serovar Typhi CT18 (STY) to common serotypes of S. enterica subsp. enterica (Table 2; Porwollik et al. 2004). These loci were chosen for their ability to give unique results and so to differentiate the clinically most common serotypes of Salmonella. The biological functions and phylogenetic implications of these loci have been previously discussed (Shangkuan and Lin 1998; Chan et al. 2003; Porwollik et al. 2002, 2004).

Table 2 Chromosomal regions of Salmonella enterica serovars Typhimurium LT2 and Typhi CT18 used to create primers for multiplex PCR (Porwollik et al. 2004)

Suitable primers for two five-plex PCRs, each amplifying five different regions, were designed using Primer Express v2.0 (Applied Biosystems, Foster City, CA) and were synthesized by MWG-Biotech (High Point, NC). The other discriminatory monoplex PCR used primer set STM7, which was derived from serovar Typhimurium LT2 (Porwollik et al. 2005).

For all PCRs, primer sequences (5′ → 3′) are provided in Table 2 (Porwollik et al. 2004). The first five-plex PCR assay (named STM) incorporated loci from serovar Typhimurium. The second assay (named STY) was based on four loci in serovar Typhi C18 and one from serovar Typhimurium LT2.

All reactions were performed in a final volume of 34 µl containing 5 µl of template DNA, 0.2 mM dNTPs, 2 mM MgCl₂, 5.0 units of Taq polymerase, 50 ng of each primer and deionised water to make up the volume. All assays used the same cycling parameters under the following conditions: enzyme activation at 94°C for 5 min and then an additional 40 cycles with heat denaturation at 94°C for 30 s, primer annealing at 62°C for 30 s, and DNA extension at 72°C for 1 min. After the last cycle, samples were maintained at 72°C for 5 min to complete the synthesis of all strands.

Each PCR products (10 µl) were separated by electrophoresis and analyzed on 2% Tris-acetate EDTA agarose gel stained with ethidium bromide, visualized with UV induced fluorescence, and photographed.

Results

Optimization strategies

The optimization of MPCR is more tedious and difficult to achieve than monoplexes so the correct size of each primer is first established in a monoplex PCR rather than in multiplex PCR.

A monoplex PCR for each primer set was initially carried out based on a previous mixture protocol and PCR amplification program. Initial attempts are to amplify equally all the five STM and STY genes using a monoplex rather than a multiplex PCR reaction. The monoplex PCRs were successful and all STM and STY primer sets gave a single and a specific amplification product of a required gene except for STY 1 that cannot be detected in the tested Salmonella serovars (Fig. 1).

Fig. 1

Ethidium bromide-stained agarose gel of PCR products. PM 1,000-bp DNA ladder size standard; lanes 1, 3, 4, and 5 Salmonella Newport DNA amplified with STM 1, STM 2, STM 3, and STM 5 primers, respectively; lane 2 Salmonella Anatum DNA amplified with STM 4 primer; lanes 6 and 9 Salmonella Zanzibar DNA amplified STY 2 and STY 4 primers, respectively; lane 7 Salmonella Altona DNA amplified with STM 6 primer; lane 8 Salmonella Enteritidis DNA amplified with STY 3 primer; lane 10 a MPCR product obtained with Salmonella Newport DNA amplified with STM 1, STM 2, STM3, and STM 5 primers, respectively; lane 11a MPCR product obtained with Salmonella Zanzibar DNA amplified with STY 2 and STY 4 primers, respectively

Reproducibility

Reproducibility for the detection of STM and STY genes assayed in the MPCR, to confirm our PCR results, was 100% if we except Salmonella serovar Anatum (Table 1). In this way, all Salmonella-specific tested serotypes gave the same amplification profiles with the same Salmonella serotypes (Table 3) but Salmonella serovar Anatum produced two different STM amplicon patterns (Table 3).

Table 3 Antigenic formulas of Salmonella strains and results of numbers of tested isolates using the STM and STY assays to predict their serotype by PCR

Detection of amplified products with STM primers to discriminate between different serotypes by multiplex PCR

To identify Salmonella serovars using multiplex PCR, five primer sets were used in the same reaction mixture (STM 1, STM 2, STM 3, STM 4, and STM 5) (Table 2); the 19 most common serotypes of Salmonella spp. strains and the ATCC strain were tested (Table 1).

To confirm our results, we mostly tried to test more than one strain from each serotype (Table 3). For a total of 61 tested isolates, all Salmonella serotypes produced at least one amplicon. We can note that the amplified product obtained with STM 5 primer was produced with the most Salmonella serovars (80% of Salmonella serovars). Using these five STM primers with the 19 Salmonella serovars, we can identify four distinct groups: seven serotypes produced two amplicons; four serotypes produced one amplified product; five serotypes produced three amplified STM products, but the most common STM multiplex PCR amplicon pattern were 2,3,5; the last group was obtained with three serotypes that produced four amplified PCR products. Interestingly, Salmonella serovar Anatum produced either STM profile 2,4,5 or 2,3,5 (Table 1; Figs. 2 and 3).

Fig. 2

Ethidium bromide-stained agarose gel of Multiplex PCR products from extracted Salmonella serovars DNA amplified with STM 1, STM 2, STM 3, STM 4, and STM 5 primers. PM 1,000-bp DNA ladder size standard; lane 1 negative control; lanes 2 to 12 Salmonella serovars Corvallis, Mbandaka, Muenster, Bovis-morbificans, Braenderup, Infantis, Cerro, Typhimurium, Typhimurium ATCC 14028, Nikolaifleet, and Montevideo, respectively

Fig. 3

Ethidium bromide-stained agarose gel of Multiplex PCR products from extracted Salmonella serovars DNA amplified with STM 1, STM 2, STM 3, STM 4, and STM 5 primers. PM 1,000-bp DNA ladder size standard; lane 1 negative control; lanes 2 to 10 Salmonella serovars Anatum, Zanzibar, Enteritidis, Amsterdam, Carnac, Altona, Newport, Heidelberg, and Derby, respectively

Finally, we noted that Salmonella typhimurium ATCC 14028 consistently produced the same molecular profile compared with Salmonella typhimurium laboratory isolates.

Detection of amplified products with STY primers to discriminate between different serotypes by multiplex PCR

In a second approach, we validated the MPCR for Salmonella serovars detection by using STY primers: STY 1, STY 2, STY 3, and STY 4 genetic regions of S. enterica serovar Typhi; we also used STM 6 primer that designed the genomic region of S. enterica serovar Typhimurium (Table 2). The 19 different tested Salmonella serovars could be classified into three groups on the basis of scoring the presence or absence of appropriately size amplicons (Table 1): the first group contained the largest number of Salmonella serovars (58%) identified by the presence of only a single product, either STY 2, STY 3, STY 4, or STM 6 patterns, of which 45 and 36% of serotypes were obtained, respectively, with STY 2 and STM 6 primers; the second group contained six serotypes that produced two amplicons which were 100% STY 2 and STY 4 profiles; and the third group contained the two serovars Newport and Braenderup and was negative for all STY primers (Figs. 4, 5).

Fig. 4

Ethidium bromide-stained agarose gel of Multiplex PCR products from extracted Salmonella serovars DNA amplified with STY 1, STY 2, STY 3, STY 4, and STM6 primers. PM 1,000-bp DNA ladder size standard; lane 1 negative control; lanes 2 to 12 Salmonella serovars Corvallis, Muenster, Cerro, Mbandaka, Braenderup, Enteritidis, Bovis-morbificans, Infantis, Typhimurium, Typhimurium ATCC 14028, and Nikolaifleet, repectively

Fig. 5

Ethidium bromide-stained agarose gel of Multiplex PCR products from extracted Salmonella serovars DNA amplified with STY 1, STY 2, STY 3, STY 4, and STM6 primers. PM 1,000-bp DNA ladder size standard; lane 1 negative control; lanes 2 to 10 Salmonella serovars Zanzibar, Anatum, Amsterdam, Altona, Carnac, Newport, Montevideo, Heidelberg, and Derby, respectively

Combining the STM and STY primers to better discriminate between different Salmonella serotypes

To further evaluate the discriminatory method for Salmonella serotyping and to increase identified serovars, we combined both the STM and STY primers. With five STM primers (STM 1, STM 2, STM3, STM 4, STM 5), four STY primers (STY 1, STY 2, STY 3, STY 4) and STM 6 primer, it was possible to discriminate 15 different serotypes. However, four serotypes (Salmonella Anatum, Salmonella Bovis-morbificans, Salmonella Carnac and Salmonella Amsterdam) were grouped into two pairs that had the same amplification pattern (Table 1); for this reason, and to be able to discriminate between these four serotypes, we made a monoplex PCR using STM 7 primer set designed from S. enterica serotype Typhimurium (Table 4).

Table 4 Molecular detection of four serotypes of Salmonella enterica subsp. enterica using STM7 primer by a monoplex PCR

Using the STM 7 primer that creates an amplicon of 101 bp, we were able to discriminate between Anatum and Bovis-morbificans and also between Carnac and Amsterdam serotypes (Fig. 6). In this way, and using these panels of primers, we could type all of the tested Salmonella serotypes.

Fig. 6

Ethidium bromide-stained agarose gel of simple PCR products from extracted Salmonella serovars DNA amplified with STM 7 primer for discrimination of paired serotypes. PM 1,000-bp DNA ladder size standard; lanes 1 to 4 amplification products of Salmonella serovars Bovis-morbificans, Anatum, Amesterdam, and Carnac, respectively

Comparison between the two approaches: serotyping standard method and MPCR serotyping for identification of Salmonella serotypes

The results of the two five-plex PCRs were converted to amplification products and then compared to traditional serotyping. We noted that all tested Salmonella serovars gave the same PCR profile results with each specific tested serotype (Tables 1 and 3). In this way, there was 100% correlation between traditional serotyping and molecular serotyping. However, one exception was found with Salmonella Anatum serotype: two isolates gave the molecular combination named STM 2,4,5 STY 2 and the others gave the molecular combination STM 2,3,5 STY 2 (Table 1).

Discussion

Salmonella serotypes were determined on the basis of antigenic variability at lipopolysaccharide moieties (O antigen), the flagellar protein (H antigen), and the capsular polysaccharide (Vi antigen).

Salmonella isolates are diphasic with respect to the flagellar antigens, with several monophasic exceptions. In this way, Salmonella serovars were determined by the combination of O and H antigen types (phase-1 and phase-2) using commercial antisera (Bio-Rad). This serological method has been used to identify Salmonella serovars and specifically more than 1,478 different serovars of Salmonella enterica subsp. enterica (Bopp et al. 2003; Perch et al. 2003).

However, when using the serological method, antibodies must be produced for each serovar, and this is extremely complex and time consuming. So easier and simpler methods were needed for identification of Salmonella serovars. Recently, different molecular techniques were developed for detection of Salmonella serovars such as pulsed-field gel electrophoresis (PFGE) (Uzzau et al. 1999; Torpdahl and Ahrens 2004), ribotyping (Altwegg et al. 1989; Pang et al. 1992), and random amplification of polymorphic DNA (RAPD) (Shangkuan and Lin 1998), as well as the recent development of amplified fragment length polymorphism (AFLP) (Nair et al. 2000) and the real-time multiplex PCR (O'Regan et al. 2008).

In this context, different PCR protocols for specific detection of S. enterica subsp. enterica were developed, but they were limited to the number of strains that can be typed (Echeita et al. 2002; Herrera-Leon et al. 2004).

In this study, using suitable primers for the two five-plex PCRs and the monoplex PCR methods for molecular Salmonella serotyping, we could easily discriminate all the tested Salmonella serotypes that represented 100% of all Salmonella isolates in our regional Tunisian laboratory during the period 2002–2007. These results have been found elsewhere (Perch et al. 2003). The STM 7 primer was used in a monoplex PCR to differentiate between the four serotypes (Salmonella Anatum, Salmonella Bovis-morbificans, Salmonella Carnac, and Salmonella Amsterdam) that gave the same STM and STY profiles (Table 4). The same molecular serotyping profiles were also reported with other serovars (Kim et al. 2006). The reasons for this resemblance in molecular amplicon code can be explained by the presence of a very similar region in these serovars. It can also be explained by deletion problems that can concern a specific region and so the absence of appropriately sized amplicons with specific primers (Garaizar et al. 2002). A secondary discrimination problem that was interesting to note was that the serovar Anatum gave more than one amplicon code with STM primers which may reflect intraserovar variation.

To further discriminate each serovar, we can associate to this multiplex PCR serotyping the PFGE analysis, the enterobacterial repetitive intergenic sequence PCR (ERIC2-PCR), or the 16 S\23 S r RNA ribotyping. These methods provided a high degree of intraserovar discrimination. In this way, we describe the MPCR as a rapid, specific, and cost-effective molecular method that has demonstrated its efficient discrimination in serotyping of the most common clinical and food isolates of S. enterica subsp. enterica in our region. This method can be used as an alternative method of standard serotyping in many clinical laboratories.