Evaluation of a multiplex PCR method to serotype Salmonella in animal feeds pre-enrichment broth cultures

Graphical abstract Screening, detection and serotyping of Salmonella enterica from 24 h pre-enrichment of animal feed.

Molecular serotyping of Salmonella

DNA preparation
One ml of samples after 24 h pre-enrichment. The samples were collected in a 1.5 ml sterile Eppendorf microcentrifudge tube. The tubes were centrifuged at 10,000 g for 10 min DNA was purified from the 24 h pellets using Roche MagNA Pure Compact (Roche, Indianapolis IN) according to the manufacturer's instructions. DNA extraction quality and the PCR serotyping outcome are amplified by collecting samples at 24 h after inoculation. (Note: An important factor in the success of this assay is the ability to extract the DNA from the different matrices). The broth used for the pre-enrichment (Lactose or mBPW) showed no significant difference in recovery of Salmonella. Although the mBPW seems to be a better growth media for Salmonella in various matrices [11], in the animal feed it did not make a major difference.
PCR analysis. The PCR method consisted of two five-plex PCR reactions and one two-plex PCR reaction; the primers for each of the reactions are listed in Tables 1 and 2.
PCR amplification of each sample was conducted using 2.0 ml of DNA template, and 48 ml of a master mix containing primers, for a total volume of 50 ml.
The amplification parameters were as follows: a 5 min step at 94 C, followed by 40 amplification cycles that consisted of a 30 s denaturing step at 94 C, a 30 s annealing step at 56 C, and a 1 min elongation step at 72 C, and a final extension step of 5 min [11]. Inhibitory factors may interfere with PCR results. Lack of Salmonella growth in this particular feed may also minimize detection by PCR Table 3 show that each feed type is different and the growth obtained from each feed also differed.
Results were visualized as a gel image depicting the amplification of the PCR products (serotypespecific banding patterns) and the corresponding molecular weight for each PCR product. Table 1 Primers used in the PCR-Based Methods for serotyping Salmonella enterica.
Codes for the interpretation of the results.
Serotypic banding patterns are represented by absence or presence of specific genes labeled A-L as shown in Table 1 [10].
The multiplex PCR serotyping assay was performed concurrently with the Bacteriological Analytical Method (BAM). The optimal limit of detection in all the feeds was 50 CFU/25 g and lowest detection level was 10 CFU/25 g (Table 3 and Fig. 2a-f). Table 3 demonstrates the lowest limit of detection for Salmonella enterica serovar Typhimurium. S. Typhimurium was not detected in the DBP feed. Fig. 2b and d, which represent the Agilent gel image of the multiplex PCR, demonstrated that the detection level of the PCR for OT and WB was 10 CFU/25 g. Conversely, Salmonella identification and serotyping was not achieved in the DBP (Fig. 2f).
The PCR assay was further evaluated in Horse Feed, Whole Oats, and Calf Milk Replacer spiked with two concentrations of 10 CFU/25 g and 2.5 CFU/25 g of Salmonella enterica serovars      concentration (Table 4), but it was not a significant difference, equaling to less than 5% difference in recovery. However for the OT feed there were more positive cultures recovered using the lactose (13/20 and 9/20) as compared to mBPW (6/20 and 9/20) as demonstrated in Table 2. At the 2.5 CFU slightly more than 10% difference in recovery using the lactose broth (Fig. 4). There was not much difference observed with mBPW as compared to the lactose broth at 10 CFU per 25 g concentration ( Table 4). The 12 remaining samples were the positive controls of each serovar grown in each enrichment broth and un-inoculated enrichment broths for negative controls the results were as expected. The expected PCR amplicons or serotype banding patterns were obtained for serovars Typhimurium (ABCDEI), Agona (BCJ) and Hadar (BC) [10] at the lowest concentrations of 2.5 CFU/25 g ( Table 4, Fig. 3). The microbiological limit of detection in food matrices was 100 CFU/25 g. These results are significant in demonstrating the ability of the multiplex assay to detect Salmonella from a pre-enrichment broth with an initial inoculum of 2.5 CFU/25 g. These results also show that the assay is effective for screening in multiple animal feeds, and two different broths. In this study the pre-enrichment broth did not have a negative impact on the PCR assay results. However, the correct broth per feed must be used for optimal Salmonella enrichment. The assay is effective to screen for Salmonella after 24 h, an earlier time point than the current Salmonella method which takes 5-10 days to identify Salmonella and the serovar. In summary, since animal feed is a recognized source of S. enterica for farm livestock [15], it is also a potential Salmonella contamination source. Salmonella infections in animals can be due to contaminated feed. Outbreaks of human salmonellosis are sometimes linked to contact with infected animals and animal feed. Therefore, it is important to rapidly and efficiently detect Salmonella in feed. In this study, the multiplex PCR for molecular serotyping method was tested in the enrichment step and the results show that this method could prove to be a very useful and effective tool for rapid screening of animal feed for the presence of Salmonella. This multiplex assay can identify Salmonella directly from the pre-enrichment broth after 24 h, instead of waiting for a pure culture to serotype Salmonella, which typically may take 5-10 days for the entire process. This approach was verified in a study reported by Benhamed et al. (2017) for the identification of Salmonella enterica serovar Cubana in a naturally incurred chick feed [4,5]. Further testing in naturally contamined feeds would evaluate the limit of this method. The multiplex PCR serotyping assay may be used for rapid screening and serotyping of Salmonella contaminating animal feed, thereby decreasing the time it takes to detect Salmonella in animal feed and thus helping to prevent human disease.

Additional information
Background Salmonella enterica is a leading cause of food-borne illness and is a serious public health concern. Outbreaks of human salmonellosis are often linked to contact with infected animals and feed, as well as contact while preparing and eating contaminated meat. Global outbreaks of human salmonellosis linked to contaminated feed ingredients have been reported [3]. Crump et. al., reported that in 1993, the FDA tested for the presence of S. enterica in samples from 78 rendering plants that produced animal protein based animal feed and in samples from 46 feed mills that produced vegetable protein based animal feed. S. enterica were detected in 56% of the 101 animal protein based samples and 36% of the 50 vegetable protein based samples. Furthermore, in 1994, the FDA tested 89 finished feed samples collected from feed mills and from farms where animal feed is mixed and found that 25% of the samples were contaminated with S. enterica [3].
Animal feed and feed ingredients are important sources of zoonotic Salmonella infections [12] and might also act as an indirect source of infection in humans consuming foods of animal origin [3,10]. Therefore, it is imperative to rapidly and efficiently detect Salmonella in feed to reduce or prevent consumption of contaminated food/feed. Identification and removal of contaminated feed will provide a prevention step that will reduce and prevent foodborne illnesses associated with the "farm to fork continuum." This will help ensure that feed used for food animals is introduced into commerce free of Salmonella, especially since livestock may be both direct and indirect sources of both zoonotic and human infections.
Hence, the approach in this investigation is to test the capability of a molecular serotyping scheme previously used to serotype S. enterica from food matrices spiked with S. enterica serovars Newport, Typhimurium, Javiana and Saintpaul [2,11] to determine its suitability for screening of Salmonella in animal feed. The PCR method is a modified version of the multiplex PCR method reported by Kim et al. [6][7][8]13], and it identifies specific serotypes based on PCR amplification of serotype-specific target genes. Amplification of these gene targets in each Salmonella serotype produces a serotype-specific banding pattern [11]. This method can serotype 30 of the most clinically relevant Salmonella serotypes. The PCR method consists of two five-plex PCR reactions and one two-plex PCR reaction as described by Jean-Gilles Beaubrun et al. [11]. The most common method of serotyping Salmonella isolates is based on the serological discrimination of O (surface polysaccharide), H (flagellar) and Vi (capsular) antigenic properties [6][7][8][9]. The conventional method of serotyping employs more than 150O and H antisera for the characterization of over 2600 Salmonella serotypes, of which 1478 belong to the species S. enterica [14,16] and take 5 to 10 days. The multiplex PCR serotyping method can generate a result for the 30 serotypes after 24 h of pre-enrichment in a non-selective broth.
The advantage of this approach is that one PCR reaction can screen, identify and serotype Salmonella enterica from various contaminated feed matrices. This approach is significant since S. enterica serotypes identification remains a highly important public health concern for the microbiological analysis of foods and animal feeds. The capability of this approach over multiple food and feed may increases the global food safety community's ability to screen for Salmonella, which may lead to the prevention of Salmonella infections in animals and outbreaks of human salmonellosis.