Development of a multiplex loop-mediated isothermal amplification method for the simultaneous detection of Salmonella spp. and Vibrio parahaemolyticus

Rapid detection of food-borne pathogens is important in the food industry, to monitor and prevent the spread of these pathogens through contaminated food products. We therefore established a multiplex real-time loop-mediated isothermal amplification (LAMP) assay to simultaneously detect and distinguish Salmonella spp. and Vibrio parahaemolyticus DNA in a single reaction. Two target sequences, one specific for Salmonella and the other specific for Vibrio parahaemolyticus, were amplified by specific LAMP primers in the same reaction tube. After amplification at 65 °C for 60 min, the amplified products were subjected to melting curve analysis and thus could be distinguished based on the different melting temperatures (Tm values) of the two specifically amplified products. The specificity of the multiplex LAMP assay was evaluated using 19 known bacterial strains, including one V. parahaemolyticus and seven Salmonella spp. strains. The multiplex LAMP showed 100% inclusivity and exclusivity, and a detection limit similar to that of multiplex PCR. In addition, we observed and corrected preferential amplification induced by what we call LAMP selection in the multiplex LAMP reaction. In conclusion, our assay was rapid, specific, and quantitative, making it a useful tool for the food industry.

4~6 specifically designed primers that target 6~8 distinct regions of the target gene, a large amount of DNA can be synthesized under a constant temperature in less than 60 min, and the amplified products can be detected by the naked eye. As a result, the LAMP assay has been widely applied in the detection of pathogenic bacteria, viruses, and parasites [10][11][12][13] . However, restriction enzyme analysis of amplified products and probe-based methods, which has been attempted, is laborious and complex [14][15][16] . Furthermore, the LAMP assay can only detect a gene in a single reaction, and multiplex LAMP assays that can detect and discriminate two or more target genes in a single reaction are limited. In this study, a multiplex LAMP assay was developed to simultaneously detect Salmonella spp. and V. parahaemolyticus based on different melting temperatures determined by melting curve analysis of amplified products.

Optimization of primer concentration in the multiplex LAMP. The simultaneous amplification of
Salmonella and V. parahaemolyticus DNA targets was evaluated. Figure 1A shows the results when equal concentrations of Salmonella and V. parahaemolyticus primers were used. For S. typhimurium DNA, the mean C q and T m values were 13.96 min and 86.25 ± 0.06 °C, respectively (Fig. 1A, well 1), whereas for V. parahaemolyticus DNA, the mean C q and T m values were 12.68 min and 84.05 ± 0.06 °C, respectively (Fig. 1A, well 2). For a mixture of S. typhimurium and V. parahaemolyticus DNAs, the mean C q value was 12.14 min and the mean T m values for S. typhimurium and V. parahaemolyticus were 85.63 ± 0.15 °C and 82.88 ± 0.21 °C, respectively (Fig. 1A, well 3). No amplification occurred in the negative controls. These results indicate that the multiplex LAMP reaction with two sets of primers for S. typhimurium and V. parahaemolyticus successfully amplified the two target genes, and the amplified products showed different T m values (86.25 ± 0.06 °C and 84.05 ± 0.06 °C). In addition, the simultaneous amplification of the target genes from a mixture of S. typhimurium and V. parahaemolyticus DNAs resulted in two detection peaks, which could be differentiated by their different T m values (85.63 ± 0.15 °C and 82.88 ± 0.21 °C). However, the fluorescence produced by V. parahaemolyticus DNA amplification was clearly weaker than that produced by S. typhimurium DNA amplification, suggesting that there was preferential amplification of the S. typhimurium target sequence over the V. parahaemolyticus target sequence in the multiplex LAMP reaction. Thus, an adjustment of the relative concentration of the primers was required. The concentrations of S. typhimurium primers were reduced to half (1.3 μ M) and concentrations of V. parahaemolyticus primers were kept unchanged, enabling similar amplification efficiency and more intuitive identification of the two target sequences Specificity of the multiplex LAMP assay. To evaluate the specificity of the multiplex LAMP assay, all 19 bacterial strains were subjected to the multiplex LAMP reaction. The assay successfully detected Salmonella spp. and V. parahaemolyticus. The T m value for V. parahaemolyticus was 83.8 °C and the T m values for seven Salmonella spp. consistently fell between 86.5 °C and 86.8 °C, with an average of 86.61 ± 0.11 °C. For non-Salmonella spp. and non-V. parahaemolyticus strains, no T m value was obtained, suggesting that no amplification occurred. We conclude that the two sets of primers are highly specific for the detection of Salmonella spp. and V. parahaemolyticus ( Table 1).
Sensitivity of the multiplex LAMP assay. To estimate the detection limit of the multiplex LAMP assay, 10-fold serial dilutions of S. typhimurium CICC 21483 and V. parahaemolyticus ATCC 17802 DNA templates with concentrations ranging from 100 ng/μ L to 0.1 pg/μ L were used in the amplification reaction. The assay sensitivity for both the simultaneous detection of S. typhimurium and V. parahaemolyticus and the multiplex PCR was found to be 10 pg/μ L (Fig. 2).
Quantitative capability of the multiplex LAMP assay. The quantitative capability of the multiplex LAMP assay was also evaluated. Figure 3 shows the standard curve generated when simultaneously detecting S. typhimurium and V. parahaemolyticus in three independent replicates. In the multiplex LAMP assay, the mean C q values decreased linearly with increasing target concentrations, and the quantification equation for the assay was validated to be Y = − 3.688x + 32.734, with a correlation coefficient (R 2 ) of 0.9314. Moreover, comparison of C q values for each concentration in all three independent experiments showed minimal variation, indicating that the multiplex LAMP amplification is highly reproducible.

Discussion
In this report, a simple and rapid protocol for simultaneous detection of Salmonella spp. and V. parahaemolyticus based on the multiplex real-time LAMP technique was developed. This protocol incorporates the melting curve analysis, which allows the detection and discrimination of amplified products from a mixture by their different T m values within 60 min. In addition, the phenomenon of amplification bias, defined as uneven or preferential amplification of one target gene over another in a multiplex reaction, which was induced by what we call LAMP selection and can be balanced by primer concentration adjustments, was also observed in the present study.
Considerable time and effort can be saved by simultaneously amplifying and detecting two or more target sequences in a single reaction. However, multiplex LAMP in which multiple sequences can be amplified is limited owing to the difficulty of distinguishing the origin or specificity of the amplified products from the mixture. Several methods have been tried to resolve this difficulty, such as species-specific identification after restriction enzyme digestion 14,15 , visual detection by addition of fluorescence-labeled probes or primers coupled with cationic polymers [16][17][18] , colorimetric distinction using an immunochromatographic strip 19 , and melting curve analysis of amplified products 20 . The last approach was chosen in this study because of its simplicity, rapidity, and specificity. The detection assay takes advantage of the different T m values of target sequences, which are relatively stable and determined by melting curve analysis. Unlike PCR, the T m values of LAMP products are closely related to GC content rather than the length of target sequences. As shown in Fig. 1, two target sequences were simultaneously amplified with mean T m values for S. typhimurium and V. parahaemolyticus in mixed products of 85.63 ± 0.15 °C and 82.88 ± 0.21 °C, respectively, and were thus clearly distinguished. The presence of two or more primer pairs in the multiplex LAMP increases the chance of obtaining false positive amplification, primarily because of 4~6 primers contained in a single LAMP reaction, which is another factor limiting the development of multiplex LAMP. Although there is no theoretical limit to the number of primers to be used in a multiplex LAMP reaction, spurious-prone amplification among primers limits the establishment of specific reactions. A multiplex LAMP assay combining 21 primers targeting four Candida species has been reported 21 . The high specificity of primers used in this study has been reported and validated previously 10,12 , and there was no spurious amplification products in our multiplex LAMP reaction, which showed 100% inclusivity and 100% exclusivity for the detection of 19 bacterial strains including Salmonella spp. and V. parahaemolyticus. The detection limit for simultaneous detection of S. typhimurium and V. parahaemolyticus was found to be 10 pg/μ L, which was identical to that of LAMP amplification only for S. typhimurium or V. parahaemolyticus. Additionally, the sensitivity of multiplex LAMP is similar to that of multiplex PCR. The feasibility of the multiplex LAMP was evaluated and validated by the artificially inoculating Salmonella and V. parahaemolyticus into pasteurizad milk sample. The result showed that the assay could detect and distinguish Salmonella and V. parahaemolyticus contamination of food samples. Further work to optimize the detection of these two pathogens from clinical samples will also be important. The quantitative capability of LAMP has been examined previously 12,22 . In the present study, a standard curve was generated when detecting S. typhimurium and V. parahaemolyticus DNAs. In our multiplex LAMP assay, the C q values decreased linearly with increasing initial concentrations of template DNAs from 10 pg/μ L to 100 ng/μ L. The standard curve had a linear relationship with R 2 value of 0.9314, indicating the quantitative capability of the multiplex LAMP assay.
Ideally, all target sequences in a multiplex detection assay should have similar amplification efficiencies so that greater accuracy of species identification can be obtained. To our knowledge, very few reports have focused on the amplification efficiencies in the multiple amplification system. Amplification bias of one target gene over another was observed in this paper when equimolar primer concentrations were used. The results showed the preferential amplification of the Salmonella target sequence over that of V. parahaemolyticus, which was induced by what we call LAMP selection 23 . Therefore, adjustment of primer concentration in the multiplex reaction was required, and we found that halving the concentrations of the Salmonella primers led to equal amplification efficiency of both target sequences. Generally, optimal concentration of the primers in a multiplex detection reaction may vary greatly between targets and is established empirically 24 .
For the past few years, diseases caused by food-borne pathogens have become a significant public health issue globally 1 . As a leading cause of food-borne disease, Salmonella spp. and V. parahaemolyticus contamination of food products has become a vital concern for food safety. The multiplex real-time LAMP assay described here can simultaneously detect and distinguish Salmonella spp. and V. parahaemolyticus based on their different melting temperatures. We also observed preferential amplification in the multiplex LAMP reaction, which was balanced by primer optimization. The assay showed sensitivity similar to that of PCR and was rapid, specific, and quantitative, making it a useful tool for the food industry.

Materials and Methods
Bacterial strains, culture conditions, and DNA preparation. In total, 19 known bacterial strains, including a V. parahaemolyticus reference strain (ATCC 17802) and seven Salmonella sp. strains, were analyzed in this study (Table 1). All strains were stored in 10% (w/v) glycerol broth at − 70 °C. Salmonella typhimurium CICC 21483 and V. parahaemolyticus ATCC 17802 strains were used for primer optimization and sensitivity testing. Vibrio strains were cultured in trypticase soy broth supplemented with 2% NaCl at 35 °C overnight. Non Vibrio strains were grown in Luria-Bertani broth at 37 °C.
Genomic DNA extraction was conducted using the boiling method. Briefly, 1 ml of overnight bacterial culture was centrifuged at 5,000 × g for 10 min and the pellet was suspended in 200 μ l of TE buffer [10 mM Tris, 0.1 mM EDTA (pH 8.0)]. The bacterial suspension was boiled for 10 min and centrifuged at 12,000 × g for 5 min at 4 °C. The supernatant was used as the DNA template for multiplex LAMP and multiplex PCR assays. The genomic DNA of S. typhimurium CICC 21483 and V. parahaemolyticus ATCC 17802 was extracted with the Genomic DNA Isolation Kit (Sangon Biotechnology, Shanghai, China) following the manufacturer's instructions.
LAMP primers and the multiplex LAMP reaction. We used previously reported Salmonella-and V. parahaemolyticus-specific primers 10,12 ( Table 2). The primers were synthesized by Sangon Biotech (Shanghai). Aliquots (50 μ M) of each primer were prepared and used for the multiplex LAMP reaction.
The multiplex LAMP reaction was carried out in a final volume of 25 μ L containing 12.