Qualitative and Quantitative Detection of Potentially Virulent Vibrio parahaemolyticus in Drinking Water and Commonly Consumed Aquatic Products by Loop-Mediated Isothermal Amplification

Vibrio parahaemolyticus can cause acute gastroenteritis, wound infection, and septicemia in humans. In this study, a simple, specific, and user-friendly diagnostic tool was developed for the first time for the qualitative and quantitative detection of toxins and infection process-associated genes opaR, vpadF, tlh, and ureC in V. parahaemolyticus using the loop-mediated isothermal amplification (LAMP) technique. Three pairs of specific inner, outer, and loop primers were designed for targeting each of these genes, and the results showed no cross-reaction with the other common Vibrios and non-Vibrios pathogenic bacteria. Positive results in the one-step LAMP reaction (at 65 °C for 45 min) were identified by a change to light green and the emission of bright green fluorescence under visible light and UV light (302 nm), respectively. The lowest limit of detection (LOD) for the target genes ranged from 1.46 × 10−5 to 1.85 × 10−3 ng/reaction (25 µL) for the genomic DNA, and from 1.03 × 10−2 to 1.73 × 100 CFU/reaction (25 µL) for the cell culture of V. parahaemolyticus. The usefulness of the developed method was demonstrated by the fact that the bacterium could be detected in water from various sources and commonly consumed aquatic product samples. The presence of opaR and tlh genes in the Parabramis pekinensis intestine indicated a risk of potentially virulent V. parahaemolyticus in the fish.


Introduction
V. parahaemolyticus is a Gram-negative bacterium that can cause acute gastroenteritis, wound infection, and septicemia in humans [1]. The bacterium inhabits estuarine and marine environments worldwide, and is also frequently detected in aquatic products [2]. Clinical V. parahaemolyticus isolates produce two major toxins, thermostable direct hemolysin (TDH) and TDH-related hemolysins (TRH), both of which cause hemolysis and cytotoxicity of the host cells [3]. Their encoding genes trh and tdh, sharing approximately 70% homology, are molecular markers for the diagnosis of virulent V. parahaemolyticus isolates [4].
Previous studies have also revealed very important virulence-associated genes in V. parahaemolyticus; for example, an ureC gene encodes urease subunit alpha, which is known to be associated with enterotoxicity, a reasonably good clinical diagnostic marker for trh-positive V. parahaemolyticus isolates [5]. A tlh gene encodes a thermolabile hemolysin (TLH) that is present in pathogenic and non-pathogenic V. parahaemolyticus isolates [6]. The TLH is one of the phospholipases that can hydrolyze glycerophospholipids, the major Similarly, for the target genes tlh, ureC, and vpadF in the exclusivity tests, only the LAMP reaction tubes containing the genomic DNA of V. parahaemolyticus ATCC17802 (vpadF + /tlh + /ureC + ) showed the color change and emitted bright green fluorescence, while These results were confirmed by standard agarose gel electrophoresis analyses, in which the LAMP products from the positive reaction tube formed characteristic ladder-like DNA patterns [22], while those from the negative reaction tubes showed no DNA bands. Taken together, the LAMP method was highly specific to target each of the opaR, tlh, ureC, and vpadF genes of V. parahaemolyticus (Table 1), and no cross-reaction was observed with the other 7 species of Vibrios and 20 species of non-Vibrios strains tested in this study.  Tables 2 to 4. For the target gene opaR, for example, serial dilutions of V. parahaemolyticus B4-13 cell culture (1.32 × 10 9 to 1.32 × 10 0 CFU/mL) were added into the LAMP reaction tubes. After being reacted at 65 • C for 45 min, the limit of detection (LOD) was observed in the tube containing 4.40 CFU/reaction (25 µL) of V. parahaemolyticus B4-13 cells, which changed color to light green and emitted bright green fluorescence under visible light and UV light, respectively (Figure 2A(r1 and r2)). The LOD tube also formed characteristic ladder-like DNA patterns in the agarose gel electrophoresis analyses (Figure 2A(r3)). Similarly, for the detection of the V. parahaemolyticus B11-3 cell culture (7.00 × 10 8 -7.00 × 10 0 CFU/mL), the observed LOD was 2.33 × 10 1 CFU/reaction, while for the V. parahaemolyticus N3-3 cell culture (9.10 × 10 7 -9.10 × 10 0 CFU/mL), the LOD was 3.03 × 10 −1 CFU/reaction ( Figure 2B,C). Additionally, the cell cultures of the other 47 V. parahaemolyticus strains (opaR + ) were all tested in the inclusivity tests. For the target gene opaR, the observed LOD values of the LAMP method ranged from 1.03 × 10 −2 to 7.13 × 10 3 CFU/reaction for the detection of the cell cultures of the 50 V. parahaemolyticus strains (Table 2).
For the target gene tlh, the cell culture of 50 V. parahaemolyticus strains (tlh + ) was tested in the inclusivity tests. Serial dilutions of their cell culture ranged from 2.14 × 10 9 to 1.02 × 10 0 CFU/mL. The results showed that the LODs of the LAMP method targeting the tlh gene ranged from 1.37 × 10 0 to 9.00 × 10 3 CFU/reaction (Table 4).
Additionally, the target genes amplified from representative V. parahaemolyticus strains were confirmed by PCR and DNA sequencing analyses. The resulting sequences were deposited in GenBank under the accession numbers listed in Table S2.
Taken together, approximately 41.2% of the V. parahaemolyticus strains (21 of the 51 strains) could be detected in less than 10 CFU/reaction (25 µL) by the LAMP method developed in this study, and the average detection time was 1.5 h, which highlighted the high sensitivity of the LAMP method for the detection of a cell culture of V. parahaemolyticus.

For the Detection of Genomic DNA of V. parahaemolyticus
The sensitivity of the LAMP method for the detection of the genomic DNA of the 50 V. parahaemolyticus strains was also determined (Tables 2 to 4). For the target gene opaR, genomic DNA samples extracted from each of the 50 V. parahaemolyticus strains (opaR + ) were serially diluted with concentrations ranging from 6.58 × 10 −6 to 4.95 × 10 2 ng/µL, and examined by the LAMP method. To conduct this sensitivity test, genomic DNA dilutions of V. parahaemolyticus L5-1 (2.10 × 10 2 to 2.10 × 10 −5 ng/µL) were added into LAMP reaction tubes. After being reacted at 65 • C for 45 min, eight tubes had positive reactions, showing a light green color, bright green fluorescence, and characteristic ladder-like DNA patterns ( Figure 3A). The LOD tube contained 4.21 × 10 −5 ng/reaction (25 µL) of genomic DNA. Similarly, genomic DNA dilutions of each of the other 49 V. parahaemolyticus strains were examined by the LAMP method. The results indicated that the LODs targeting the opaR gene ranged from 1.46 × 10 −5 to 1.85 × 10 0 ng/reaction of genomic DNA of V. parahaemolyticus using the LAMP method ( Table 2). For the target gene ureC, genomic DNA dilutions (9.26 × 10 1 to 9.26 × 10 −6 ng/μL) of V. parahaemolyticus ATCC17802 (ureC + ) were examined by the LAMP method. The LOD tube contained 1.85 × 10 −3 ng/reaction of genomic DNA for the ureC gene (Table 4, figures not shown).
Taken together, approximately 76.5% of the genomic DNA samples from all the V. parahaemolyticus strains (39 of the 51 strains) could be detected at less than 10 pg/reaction (25 µ L) by the LAMP method developed in this study, which demonstrated the high sensitivity of the LAMP method for the detection of the genomic DNA of V. parahaemolyticus. To compare the sensitivity of the LAMP method with the standard PCR assay, serial dilutions of each of the 50 V. parahaemolyticus cell cultures (2.14 × 10 9 to 1.02 × 10 0 CFU/mL) were examined by the PCR assay. For the target gene opaR, the observed LOD values of 6 3 Similarly, for the target gene vpadF, genomic DNA dilutions (6.58 × 10 −6 to 4.95 × 10 2 ng/µL) of each of the 39 V. parahaemolyticus strains (vpadF + ) were tested in the LAMP tubes. The results showed that the LODs targeting the vpadF gene ranged from 1.85 × 10 −4 to 4.30 × 10 −1 ng/reaction using the LAMP method (Table 3, Figure 3B).

Sensitivity Comparison of the LAMP Method with the Standard PCR Assay
For the target gene tlh, genomic DNA dilutions of each of the 50 V. parahaemolyticus strains (tlh + ) were tested in the LAMP tubes. The results showed that the LODs of the LAMP method for the vpadF gene ranged from 1.85 × 10 −4 to 3.35 × 10 1 ng/reaction (Table 4, Figure 3C).
Taken together, approximately 76.5% of the genomic DNA samples from all the V. parahaemolyticus strains (39 of the 51 strains) could be detected at less than 10 pg/reaction (25 µL) by the LAMP method developed in this study, which demonstrated the high sensitivity of the LAMP method for the detection of the genomic DNA of V. parahaemolyticus. To compare the sensitivity of the LAMP method with the standard PCR assay, serial dilutions of each of the 50 V. parahaemolyticus cell cultures (2.14 × 10 9 to 1.02 × 10 0 CFU/mL) were examined by the PCR assay. For the target gene opaR, the observed LOD values of the PCR assay ranged from 7.13 × 10 6 to 1.37 × 10 3 CFU/reaction via the routine agarose gel electrophoresis analysis ( Table 2); for example, when serial dilutions of V. parahaemolyticus B4-13 cell culture were tested, the observed LOD of the PCR assay was 4.40 × 10 4 CFU/reaction ( Figure 2D), which was 1.00 × 10 4 -fold lower than that of the LAMP method (4.40 CFU/reaction) (Figure 2A(r4)).

figures not shown).
For the target gene tlh, the cell culture of the 50 V. parahaemolyticus strains (tlh + ) was tested by the PCR assay, and the observed LODs were recorded to range from 7.90 × 10 1 to 9.20 × 10 6 CFU/reaction. The LAMP method was 1.00 × 10 1 -to 1.00 × 10 4 -fold more sensitive than the PCR assay (Table 4).
For the target gene ureC, serial dilutions of V. parahaemolyticus ATCC17802 cell culture (ureC + ) were examined by the PCR assay. The LOD of the PCR assay was 4.40 × 10 0 CFU/reaction, which was 10-fold less sensitive than the LAMP method (4.40 × 10 −1 CFU/reaction; Table 4, figures not shown).
These results demonstrated that the lowest LODs, obtained using the PCR assay, targeting the opaR, tlh, ureC, and vpadF genes in the cell culture of V. parahaemolyticus strains, were 1.00 × 10 1 -to 1.00 × 10 7 -fold lower than those obtained using the PCR assay.

For the Detection of Genomic DNA of V. parahaemolyticus
Genomic DNA dilutions of each of the 50 V. parahaemolyticus strains (6.58 × 10 −6 to 4.95 × 10 2 ng/µL) were also examined by the PCR assay, and the resulting data are presented in Tables 2 to 4. For the target gene opaR, the observed LODs for the detection of the genomic DNA of the 50 V. parahaemolyticus strains (opaR + ) ranged from 1.96 × 10 −2 to 3.79 × 10 2 ng/reaction using the PCR assay, which was 1.00 × 10 1 -to 1.00 × 10 6 -fold lower than those obtained using the LAMP method (Table 2, Figure 3).
For the target gene tlh, the LODs for the detection of the genomic DNA of the 50 V. parahaemolyticus strains (tlh + ) were 3.05 × 10 −2 to 6.04 × 10 2 ng/reaction using the PCR assay, which was 1.00 × 10 1 -to 1.00 × 10 4 -fold lower than those obtained using the LAMP method (Table 4, Figure 3).
For the target gene ureC, the observed LOD for the detection of the genomic DNA of V. parahaemolyticus ATCC17802 (ureC + ) was 1.85 × 10 −1 ng/reaction using the PCR assay, which was 100-fold lower than that obtained using the LAMP method (Table 4,

figures not shown).
These results demonstrated that the sensitivity of the LAMP method was 1.00 × 10 1to 1.00 × 10 6 -fold higher than that of the routine PCR assay for the detection of the genomic DNA of V. parahaemolyticus strains.

Sensitivity Comparison of the LAMP Method with the Standard PCR Assay for the Detection of Spiked Aquatic Product Samples
The sensitivity of the detection of the spiked aquatic product samples with a cell culture of V. parahaemolyticus ATCC17802 (opaR + /vpadF + /tlh + /ureC + ) and N7-19 (opaR + /vpadF + /tlh + /ureC − ) was also determined by the standard PCR assay (Table 5); for example, when V. parahaemolyticus N7-19 was spiked into the samples, for the target gene opaR, the observed LODs using the PCR assay were 9.87 × 10 1 to 9.87 × 10 4 CFU/reaction for the spiked A. nobilis, C. auratus, C. idella, and P. pekinensis; 9.87 × 10 3 CFU/reaction for the spiked L. vannamei; 9.87 × 10 3 CFU/reaction for the spiked M. edulis samples. Similarly, for the target gene vpadF, the observed LODs using the PCR assay were 9.87 × 10 2 to 9.87 × 10 4 CFU/reaction for the spiked fish; 9.87 × 10 3 CFU/reaction for the spiked shrimp; 9.87 × 10 4 CFU/reaction for the spiked shellfish samples. For the target gene tlh, the observed LODs using the PCR assay were 9.87 × 10 4 to 9.87 × 10 5 CFU/reaction for the spiked fish; 9.87 × 10 3 CFU/reaction for the spiked shrimp; 9.87 × 10 5 CFU/reaction for the spiked shellfish samples (Table 5, Figure 4).
These results demonstrated that the sensitivity of the PCR assay was 1.00 × 10 1 -to 1.00 × 10 5 -fold lower than that of the LAMP method for the detection of the opaR, tlh, ureC, and vpadF genes in the spiked aquatic product samples.

Reproducibility of the LAMP Method
For the target genes opaR, tlh, ureC, and vpadF of V. parahaemolyticus, all the positive results could be repeated in all the tests performed for the detection of cell culture and genomic DNA samples, as well as of the spiked aquatic product samples, indicating high reproductivity (100%) of the LAMP method developed in this study. Similarly, when the cell culture of V. parahaemolyticus ATCC17802 (2.75 × 10 9 to 2.75 CFU/mL) was spiked into the aquatic product samples, for the target gene ureC, the observed LODs by the LAMP method ranged from 9.17 × 10 3 to 9.17 × 10 2 CFU/reaction for the spiked A. nobilis, C. auratus, C. idella, and P. pekinensis; 9.17 × 10 2 CFU/reaction for the spiked L. vannamei; 9.17 × 10 1 CFU/reaction for the spiked M. edulis samples (Table 5, Figure 4).

Sensitivity Comparison of the LAMP Method with the Standard PCR Assay for the Detection of Spiked Aquatic Product Samples
The sensitivity of the detection of the spiked aquatic product samples with a cell culture of V. parahaemolyticus ATCC17802 (opaR + /vpadF + /tlh + /ureC + ) and N7-19 (opaR + /vpadF + /tlh + /ureC − ) was also determined by the standard PCR assay (Table 5); for example, when V. parahaemolyticus N7-19 was spiked into the samples, for the target gene opaR, the observed LODs using the PCR assay were 9.87 × 10 1 to 9.87 × 10 4 CFU/reaction for the spiked A. nobilis, C. auratus, C. idella, and P. pekinensis; 9.87 × 10 3 CFU/reaction for the spiked L. vannamei; 9.87 × 10 3 CFU/reaction for the spiked M. edulis samples. Similarly, for the target gene vpadF, the observed LODs using the PCR assay were 9.87 × 10 2 to 9.87 × 10 4 CFU/reaction for the spiked fish; 9.87 × 10 3 CFU/reaction for the spiked shrimp; 9.87 × 10 4 CFU/reaction for the spiked shellfish samples. For the target gene tlh, the observed LODs using the PCR assay were 9.87 × 10 4 to 9.87 × 10 5 CFU/reaction for the spiked fish; 9.87 × 10 3 CFU/reaction for the spiked shrimp; 9.87 × 10 5 CFU/reaction for the spiked shellfish samples (Table 5, Figure 4).
These results demonstrated that the sensitivity of the PCR assay was 1.00 × 10 1 -to 1.00 × 10 5 -fold lower than that of the LAMP method for the detection of the opaR, tlh, ureC, and vpadF genes in the spiked aquatic product samples.

Reproducibility of the LAMP Method
For the target genes opaR, tlh, ureC, and vpadF of V. parahaemolyticus, all the positive results could be repeated in all the tests performed for the detection of cell culture and genomic DNA samples, as well as of the spiked aquatic product samples, indicating high reproductivity (100%) of the LAMP method developed in this study.

Detection of Drinking Water and Aquatic Product Samples by the LAMP Method
Water samples from various sources were collected in Shanghai, China, in August 2021, including mineral water (n = 3), tap water (n = 3), river water (n = 3), lake water (n = 3), and estuarine water (n = 3). The samples were promptly screened by the LAMP method for the virulence-associated genes opaR, vpadF, tlh, and ureC of V. parahaemolyticus. As shown in Table 6, all the water samples tested negative for the target genes. Table 6. Detection of the virulence-related genes of V. parahaemolyticus in drinking water and aquatic product samples by the LAMP method.

Sample
No. of Sample Virulence-Related Gene No. of Sample lected from the fish market in Shanghai, China, in September 2021, and were examin by the LAMP method. The results showed that all the meat samples and most the intest samples were negative for the opaR, vpadF, tlh, and ureC genes of V. parahaemolyticus. Ho ever, the opaR and tlh genes were detected in the intestine sample of P. pekinensis (Tabl Figure 5), which were confirmed by routine microbial isolation and identification metho [35] (data not shown). These results suggested the risk of potentially virulent V. parah molyticus strains in the fish product.

Discussion
V. parahaemolyticus is the most prevalent gastroenteritis-causing pathogen in Asian countries [36,37]. Appropriate tools for the diagnosis of V. parahaemolyticus contamination in drinking water and aquatic products are the key to fight against outbreaks of the disease [38]. In this study, for the first time, we successfully developed a LAMP method for the detection of toxins and infection process-associated genes opaR, tlh, ureC, and vpadF of V. parahaemolyticus. Our data demonstrated high specificity of the inner, outer, and loop primers designed for each of the target genes in this study. No cross-reaction was observed with the other 7 species of Vibrios and 20 species of non-Vibrios strains, including common pathogenic bacteria, such as V. cholerae, V. alginolyticus, V. fluvialis, V. harveyi, and V. vulnificus, as well as L. monocytogenes, K. pneumoniae, K. oxytoca, A. hydrophila, and S. aureus.
Different sensitivity of the LAMP technique has been reported in the detection of foodborne pathogens; for example, Anupama et al. reported LODs of 1 pg/reaction and 1 CFU/reaction when targeting the tdh and trh genes of V. parahaemolyticus by LAMP [3]. The toxR-LAMP assay was able to detect 47-470 V. parahaemolyticus cells per reaction tube [32]. The LODs of the LAMP assays targeting the rpoD and toxR genes of V. parahaemolyticus were 3.7 and 450 CFU per test, respectively [31]. For the artificially contaminated seafood and seawater, the LODs of the LAMP assay were 120 and 150 fg per reaction for the groEL gene of V. parahaemolyticus [33]. In this study, inclusivity tests were conducted for each of the target genes with 50 V. parahaemolyticus strains. In a 25 µL LAMP system, the lowest observed LODs were 14.6 fg/reaction and 0.0103 CFU/reaction when targeting the opaR gene; 1.85 × 10 −4 ng/reaction and 1.73 CFU/reaction when targeting the vpadF gene; 1.85 × 10 −4 ng/reaction and 1.37 CFU/reaction when targeting the tlh gene; 1.85 pg/reaction and 0.44 CFU/reaction when targeting the ureC gene. The LAMP method developed in this study was more sensitive, with lower LODs, than the previous reports [3,[31][32][33], not only for the detection of genomic DNA, but also for bacterial cell samples in water.
The influence of different aquatic product matrixes on the sensitivity of the LAMP method was observed in this study; for example, when the cell culture of V. parahaemolyticus ATCC17802 (2.75 × 10 9 -2.75 CFU/mL) was spiked into the aquatic product samples, the observed LODs ranged from 9.17 × 10 3 to 9.17 × 10 1 CFU/reaction when targeting the ureC gene for the spiked fish, shrimp, and shellfish samples, which was 2.08 × 10 2 -to 2.08 × 10 4 -fold lower than those obtained for the detection of V. parahaemolyticus cells in water (4.40 × 10 −1 CFU/reaction). Moreover, our data revealed that the L. vannamei matrix appeared to interfere with the LAMP method more than those from the fish and shellfish. The L. vannamei matrix contained higher contents of proteins (23.3%) and crude fat (15.09%) [39] than the P. pekinensis (15.6% and 6.6%, respectively) and M. edulis (10.8% and 1.4%, respectively) matrixes [40], which may explain the observation. It will be interesting to investigate possible components in the L. vannamei matrix that contributed to the influence.
The main limitation of the LAMP-based method is the complexity of the primer design to achieve the specificity of the detection. Another possible limitation of this method is that it could generate false-positive results due to the carry-over from previous experiments (due to its high sensitivity), especially when upgraded to an automated platform [41]. However, compared with the routine PCR and RT-PCR assays, the LAMP method developed in this study can be performed with a simple dry or water bath, which is more suitable for laboratories with less equipment [28]. Moreover, unlike the former two assays, the LAMP method does not require the reaction tubes to be opened, so there is no probable cross-contamination, and this method supports the field screening of potentially virulent V. parahaemolyticus with a larger diagnostic capacity.

Bacterial Strains and Culture Conditions
Bacterial strains used in this study are listed in Supplementary Table S1. Culture media were purchased as described previously [28]. Vibrio strains were incubated in 3% NaCl, pH 8.5 media, while non-Vibrio strains were incubated in 1% NaCl, pH 7.0 media [28].

Genomic DNA Preparation
Bacterial genomic DNA was prepared using TIANamp Bacterial Genomic DNA Extraction kit DP302 (Tiangen Biotech (Beijing) Co., Ltd., Beijing, China), or extracted by a thermal lysis method [28] with minor modifications. Briefly, 100 µL of bacteria cell culture was added into 900 µL 1 x phosphate-buffered saline (PBS, pH 7.4-7.6; Shanghai Sangon Biological Engineering Technology and Services Co., Ltd., Shanghai, China), mixed well, and then serially diluted. Cell pellet of each dilution was collected by centrifugation, resuspended with 200 µL sterile ultrapure water. The cell suspension was heated at 95 • C for 10 min, and then transferred onto ice for cooling. After centrifugation at 12,000 rpm for 5 min, the resulting lysis solution was used as DNA template. Extracted DNA samples were analyzed, and DNA concentrations and purity (A260/A280) were determined as described previously [28].

Designing of LAMP Primers
Sequences of target genes (opaR, tlh, ureC, and vpadF) in V. parahaemolyticus were retrieved from National Center for Biotechnology Information (NCBI, https://www.ncbi. nlm.nih.gov/; accessed on 9 December 2020 to 23 May 2021) with GenBank accession numbers listed in Supplementary Table S1. The FIP and BIP, F3 and B3, and LF and LB primers targeting conserved sequences of each gene were designed using Primer Explorer Version 5 and SnapGene Viewer version 4.1.4 software (GSL Biotech LLC, Chicago, IL, USA) as described previously [28]. All primers (Table 1) were synthesized and DNA sequencing of PCR products was carried out by Sangon (Shanghai, China).

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/10 .3390/pathogens11010010/s1: Table S1: bacterial strains and media used in this study; Table S2: target gene sequences in some representative V. parahaemolyticus strains used in this study.