Introduction

Vibrio cholerae is a gram-negative foodborne bacterial pathogen, which is associated with high prevalence and mortality in people after they consume contaminated water and food (Blackstone et al. 2007; Charles and Ryan 2011; Vu et al. 2003). V. cholerae is present in the water environment such as lakes, rivers, and coastal areas. It is usually active during summer. Drinking water and aquatic products may be contaminated by V. cholerae that can infect humans. Aquaculture animals are susceptible to V. cholerae that can cause severe damage to related industries. Moreover, the contamination will be able to spread to surrounding water areas. A violent diarrheal disease called cholera mainly attributed to Vibrio cholerae O1 and O139 strains had several serious outbreaks in the world historically, inflicting tremendous medical and economic damages (Gubala 2006; Khemthongcharoen et al. 2014; Tuteja et al. 2007). Hence, cholera is recognized as a serious public health concern and one of category-A infectious diseases of China. Moreover, due to lack of adequate sanitation, people in underdeveloped areas continue to be threatened by cholera (Fedio et al. 2007; Madoroba and Momba 2010). In addition, V. cholerae strains belonging to non-O1 and non-O139 serotypes can also lead to sporadic enterogastritis and bloodstream infection (Mantri et al. 2006). As such, rapid and sensitive detection of V. cholerae is crucial for preventing the disease.

Most traditional methods of V. cholerae detection are based on culture and biochemical procedures. They are often time-consuming and laborious (Chen et al. 2011a). In general, about 2–7 days are required for the entire detection process for these methods. Molecular diagnosis techniques like the PCR assay need complicated instruments and individuals with professional training, while the immunoassay is time-consuming for preparing antigen and antibody. As a consequence, these approaches are not amenable for grassroots clinical application. The loop-mediated isothermal amplification (LAMP) is a new nucleic acid amplification method since 2000 (Notomi et al. 2000), which utilizes strand displacement DNA synthesis of Bst DNA polymerase and corresponding circulatory system for rapid amplification of the target gene between 60 and 65 °C (Li et al. 2013). Owing to unique primers identifying six regions on the target sequence specifically (Tanner et al. 2012), sometimes with a pair of loop primers (Yang et al. 2011), this method has many outstanding advantages including high specificity, high sensitivity, and rapidity. Importantly, the entire progress of LAMP does not need thermocyclers that may not be readily available in many low-level clinical institutions. Because of the special electrophoretic bands and an accumulation of a reaction by-product called magnesium pyrophosphate, the result of LAMP assay is revealed by electrophoresis band or turbidity (Almasi et al. 2013; D’Agostino et al. 2016; Liang et al. 2009). In order to further simplify and optimize the procedure of LAMP assay, several dyes, such as calcein, are also chosen for observing the result (Tomita et al. 2008). The visual LAMP can show the result of reaction directly, while distinguishing between positive and negative ones clearly through the change of colors. Combined with dyes, the LAMP assay will be more suitable for a rapid screening in the basic level and decrease the risk of aerosol pollution. Traditional measurements for determining the LAMP result are generally end point methods, but we may also monitor the entire reaction in real time by detecting the accumulation of turbidity or fluorescence if it is taking some improvements appropriately (Denschlag et al. 2013). Calcein has absorption/emission maxima very similar to the SYBR Green I dye with a maximum of 497 nm and an emission maximum of 520 nm. Due to the characteristic described above, the SYBR channel can be used for monitoring the reaction process accurately.

However, most methods lack the ability of differentiating viable cells from dead cells. Although culture methods only detect viable bacteria, there exist some viable but non-culture cells (VBNC) which easily cause false negative results leading to potential security venture (Elizaquível et al. 2012a, b; Kramer et al. 2009). Besides, V. cholerae is prone to a VBNC physiological state when being exposed to an adverse growth environment (Xu et al. 1982). Hence, conventional measures are inadequate for point-of-care detection unless certain means like a heat shock are introduced (Wai et al. 1996). In the present study, researchers used the messenger RNA (mRNA) based on the degradation rate of mRNA which is much faster than DNA after cells died (Chen et al. 2011a), which was also stated in a review (Fittipaldi et al. 2012). But the expression level of mRNA depends on the physiological state of cells, and its mechanism is still unknown. False positive results will arise when the level of mRNA in a viable cell is higher than detection limits originally. Hence, some DNA-binding dyes such as ethidium bromide monoazide (EMA) and propidium monoazide (PMA) are applied instead of the mRNA analytic method, which is able to permeate damaged cell membranes of dead cells and then bind with DNA irreversibly. Afterwards, the DNA conformation could be transformed by giving an appropriate illumination and, at the same time, the excess dyes will be inactivated sufficiently (Elizaquível et al. 2012a, b; Fittipaldi et al. 2011; Schnetzinger et al. 2013). Moreover, PMA is much less toxic to viable cells than EMA, meaning PMA has a higher specificity for dead/live cell differentiation (Chang et al. 2010). Obviously, PMA offers an acceptable route for solving the problem of LAMP mentioned above.

In this study, we optimized the working concentration of PMA and the maximum concentration of dead cells. Finally, we developed a PMA-LAMP assay aiming to detect the viable V. cholerae in seafood.

Materials and Methods

Bacteria Strains and DNA Template Preparation

For preliminary experiment, 24 bacterial strains were prepared, including V. cholerae and non-V. cholerae (data shown in Table 1). For sensitivity trials, V. cholerae (GIMCC1.449) was used and was streaked onto thiosulfate citrate bile salts sucrose agar culture medium (TCBS, Hopebio, Qingdao, China) at 37 °C overnight. A single colony was picked into alkaline peptone water (APW, Hopebio, Qingdao, China) and was cultured at 37 °C for 18 h. V. cholerae cells were typically grown to the mid-exponential growth phase and harvested by centrifugation (10,000×g for 5 min). The bacteria suspensions obtained were tenfold serial diluted with APW followed by DNA extractions. In addition, a traditional method known as Bacteriological Analytical Manual (BAM, formulated by FDA) was applied for enumerating the bacterial colonies counted on agar. The harvested cells were resuspended in APW, followed by a serial tenfold dilution with APW. The initial viable cell numbers were quantified by plate counting on plate count agar (PCA, Hopebio, Qingdao, China), and each sample consisted of 0.5-mL volumes. After incubating at 37 °C for 24 h, the colony-forming units (CFU) of each plate were determined.

Table 1 Information of bacterial strains used for specificity tests in the study
Full size table

We put bacteria suspensions at −20 °C for 10–20 min in order to assist lysis of cells and improve efficiency of extraction. After that, DNA templates were extracted by heating at 100 °C for 10 min in the metal heating bath. Finally, we centrifuged them at 12,000 rpm for 5 min (Thermo Scientific IEC MicroCL 17) and collected the supernatant for LAMP reaction. When making DNA from unknown samples, 10 mg/mL lysozyme was added firstly at 37 °C for 5 min and then 20 mg/mL proteinase K was added at 60 °C for 15 min to help digest proteins in the matrix and break the cell wall of bacteria. All experiments related to viable cells were performed in a class II biosafety cabinet.

The Preparation of Heat-Killed Cell Suspensions

Dead cells used for PMA trials were prepared by boiling at 100 °C for 15 min that was evaluated by a spread plate method to confirm the heating completely. A volume of 100 μL dead cells was evenly spread onto TCBS. Three triple plates were placed at 37 °C culturing for 48 h. For PMA-LAMP reaction, a mixture of dead cells and viable cells was subjected to genomic DNA extraction. The quantity of dead cells was guaranteed in a certain magnitude that could be handled by PMA, while viable cells were prepared referring to the sensitivity test.

PMA Treatments

The PMA stock solution (20 mM in 100 μL ddH2O; Biotium Inc., Hayward, CA) was added to 500 μL V. cholerae suspension in 1.5-mL eppendorf (EP) tubes, followed by incubation in the darkness at room temperature for 5 min. Then, the tubes were transferred to PMA-Lite LED Photolysis Device (Biotium Inc., Hayward, CA) for the illumination for 30 min.

The working concentration of PMA dye varies with bacterial species. To optimize the concentration of PMA treatment, PMA was added to 500 μL V. cholerae suspension to final concentrations of 5, 10, 15, 20, 50, 80, and 100 μM. Cells without PMA treatment were used as the control. The expected concentration was the one that could not only eliminate amplified signals of DNA from dead cells but also had no harmful impact on LAMP reaction of viable cells. The ability of PMA treatment depended on many factors, including the cell number and the turbidity of the sample matrix. Some researchers found that PMA did not perform satisfactorily when the cell concentration is higher than 106 CFU/mL. In the study, we tried to treat five magnitudes of dead cells ranging from 108 to 104 CFU/mL respectively by utilizing PMA with optimized working concentration.

LAMP Primers and Establishment of Real-Time Visual LAMP Assay

After aligning by using Clustal W software, the thyA gene (one of the housekeeping genes of V. cholerae, GenBank accession no.AY143429.1) was selected as the target fragment. Based upon detailed analysis and comparison, a set of specific primers of LAMP including a loop primer (Table 2) was designed with the online tool Primer Explorer V3.0 (http://primerexplorer.jp/e/), which was produced by Eiken Chemical (Tokyo, Japan). All primers were synthesized by Invitrogen Biotechnology Co., Ltd. Based on previous optimization, a total volume of 25 μL LAMP reaction mixture was finally confirmed as follows: 20 mM Tris-HCl (pH 8.8), 10 mM KCl, 8 mM MgSO4, 10 mM (NH4)2SO4, 0.1% Tween 20, 0.8 M betaine, 1.4 mM each of dNTPs, 0.4 μM F3 and B3, 3.2 μM FIP and BIP, 1.6 μM LB, 0.2 M calcein, 0.5 M MnCl2, 8 U Bst DNA polymerase (NEB Inc., UK), and 2.5 μL DNA templates. Normally, tubes containing all reagents were incubated at 65 °C for 60 min in a thermal environment (metal bath) during practical application. After incubation, the results could be judged by naked eyes due to the colors. The positive results were fluorescent green while negative ones were dark yellow.

Table 2 Sequences of Vibrio cholerae LAMP primers and qPCR primers/probe
Full size table

qPCR Primers and Establishment of qPCR Assay

A TaqMan-based qPCR assay was also established with targeting thyA gene. Primers and a probe sequence were designed by Beacon Designer 7.9 (Table 2) and were synthesized by Invitrogen Biotechnology Co. Ltd. The qPCR reaction mix (20 μL) consisted of 2× SensiFAST Probe Lo-ROX (Bioline Inc., London, U.K.), 0.4 μM each of primers, 0.1 μM probe, and 2.5 μL DNA templates. The program of qPCR was conducted as follows: initial denaturation at 95 °C for 2 min, 40 cycles consisted of denaturation at 95 °C for 10 s, and annealing at 60 °C for 26 s. All reactions were performed by CFX96™ (Bio-Rad Laboratories, Inc., California, USA).

Analytical Specificity and Sensitivity of Real-Time Visual LAMP and qPCR Assays

A total of 24 bacteria species including eight V. cholerae strains and 16 non-V. cholerae strains were recovered and enriched in lysogeny broth (LB, Sangon Biotech, Shanghai, China) liquid medium overnight. These bacteria were used for assessing the specificity after the genomic DNA extraction. To determine the sensitivity of the two methods, a tenfold serial dilution of pure V. cholerae cells mentioned above was prepared from 1.1 × 109 to 1.1 × 100 CFU/mL. Tests were repeated three times.

Sensitivity of Real-Time Visual PMA-LAMP and PMA-qPCR Assays

Viable cells ranging from 1.1 × 106 to 1.1 × 100 CFU/mL were added in 1.1 × 105 CFU/mL of dead cells. These suspensions were mixed gently and sufficiently and then treated with the optimal concentration of PMA. Genomic DNA templates extracted were amplified by real-time visual PMA-LAMP and PMA-qPCR assays.

Sample Processing and Artificial Contamination

Fresh cod fillets, shrimps, and sleeve-fish (see Table 3) were obtained from local supermarkets in Hangzhou, China. These matrices were processed as indicated in the BAM Chapter 9 (Kaysner & Jr 2004) . Samples (25.0 ± 0.1 g) were weighed under a sterile condition with instruments pre-sterilized. Each 25-g food sample was spiked with V. cholerae cells with 5, 10, and 20 CFU/25 g food homogenate as final concentrations, measured by standard surface spread plating count. Three different 25-g portions of each food were blended individually with 225 mL of APW at 10,000 to 12,000 rpm for 2 min (Waring Variable Speed Laboratory Blender; Torrington, CT). After being homogenized fully, aliquots of samples were incubated at 35 ± 2 °C for 6 to 8 h. Uninoculated samples served as negative controls, and DNA referred in sensitivity assay was used as positive control. These three portions were later incubated at 35 ± 2 °C for 24 ± 2 h. After 24 h, 500 μL was taken from each sample for DNA testing. The aliquots of samples were heated at 100 °C for 10 min and then centrifuged for 2 min at 12,000 rpm. The treatment with PMA was boiled before the samples. The supernatants were recovered and used for PMA-qPCR and PMA-LAMP assays. Taken together, these steps comprised the enrichment step.

Table 3 Detection of Vibrio cholerae PMA-LAMP, PMA-qPCR assays, and BAM in artificially contaminated produce commodities
Full size table

Practical Application of Real-Time Visual PMA-LAMP and PMA-qPCR Assays

Actual samples were categorized as two groups based on the resource that one group was purchased randomly from the local market, while the other one was afforded by Zhoushan Entry-Exit Inspection and Quarantine Bureau. All of the samples were detected by real-time visual PMA-LAMP and PMA-qPCR assays, and a traditional method of FDA was utilized for a parallel comparison.

Statistical Analysis

Data collected from LAMP/PMA-LAMP and qPCR/PMA-qPCR reactions including standard curves were analyzed by CFX Manager 3.0 software and Microsoft Excel software (Microsoft Inc., USA). Results of qPCR were judged as positive ones when the Cq value ≤35.

Results

PMA Concentration and the Proper Magnitude of Treated Dead V. cholerae Cells

To gain the optimal concentration of PMA, an amount of 1.1 × 105 CFU/mL V. cholerae dead/viable cells was used. Cell suspensions were treated under different concentrations of PMA from 10 to 100 μM, and the result is shown in Fig. 1. The amplification signal of dead cells was suppressed fully when treated with 15 to 50 μM of PMA, while viable cells could not bear the concentration of 80 μM and above. The dead cells treated with 10 μM PMA could not be detected by the qPCR equipment with an insignificant accumulation of fluorescent curve (Fig. 2). In order to balance between an effect of treatment and the prime cost of rapid tests, the usage of 15 μM PMA was accepted for next experiments.

Fig. 1
figure 1

The preliminary determination of PMA concentration. After exposing live and heat-killed, Vibrio cholerae cells (1.1 × 105 CFU/mL) were treated by different PMA concentrations and tested by visual LAMP assay (a, b) and qPCR assay (d). c The concentration of residual cells detected by PMA-qPCR. Error bars in diagrams represent standard deviations from three independent replicates. The heat-killed cells to the fresh cells = 1:1 (v:v). 1 10 μM; 2 15 μM, 3 20 μM, 4 50 μM, 5 80 μM, 6 100 μM, 7 0 μM

Full size image
Fig. 2
figure 2

Optimization of PMA concentration by using visual LAMP assay (a, b) and qPCR assay (c). 1 dead cells with 5 μM, 2 dead cells with 10 μM, 3 dead cells with 15 μM, 4 dead cells with 20 μM, 5 viable cells with 5 μM, 6 viable cells with 10 μM, 7 viable cells with 15 μM, 8 viable cells with 20 μM, 9 viable cells without PMA

Full size image

The capability of PMA was vulnerable to a high concentration of bacterial cells that DNA of dead cells failed to be eliminated from LAMP and qPCR reactions. As shown in Fig. 3, PMA lost its partial function if the magnitude of dead cells was as high as 106 to 108 CFU/mL. In contrast, the two assays could still detect amplification signals of dead cells, and obviously, PMA played a good role aiming at cells with 105 CFU/mL at most. In addition, except for the cell number and turbidity, there were some other factors of the food matrix that would interfere with the result of PMA treatment such as pH value, salty ions, chemical adsorption, and organic and inorganic ingredients (Fittipaldi et al. 2012). Therefore, facing a tested object with an inappropriate concentration, mixed bacteria, or complex compositions, the tenfold serial dilution using APW was a considerable method worth attempting if necessary.

Fig. 3
figure 3

The detection limits of the optimized PMA (15 μM) treatment. The photograph (a) and amplification curve (b) of visual LAMP were accordant. c The concentration of residual dead cells detected by PMA-qPCR. As control, the qPCR showed good consistency. 1–5 1.1 × 108, 1.1 × 107, 1.1 × 106, 1.1 × 105, and 1.1 × 104 CFU/mL

Full size image

Specificity and Sensitivity of Real-Time Visual LAMP and qPCR Assays

To analyze the specificity of real-time visual LAMP assay and qPCR, a total of 24 bacteria strains were identified and no amplification signal of non-V. cholerae was detected. The two developed methods were performed specifically that only eight V. cholerae strains were observed to show positive results (Fig. 4).

Fig. 4
figure 4

Specificity of the LAMP detection (a) and DNA extracted from 24 bacteria strains were amplified by visual LAMP assay (b) and qPCR assay (c). 1–8 Vibrio cholerae, 9–24 non-V. cholerae, 25 NC

Full size image

According to the result of BAM plate counting, the colony number of the original bacterial suspension was 1.1 × 109 CFU/mL. Therefore, the detection sensitivity of real-time visual LAMP assay was 1.1 × 102 CFU/mL that could be determined by colors and fluorescence amplification curves simultaneously (Fig. 5a, b), while the limit of qPCR was also 1.1 × 102 CFU/mL (Fig. 5c) with a good correlation coefficient (R 2 = 0.9987). The R 2 value of visual LAMP was 0.9932 which was in accordance with the qPCR assay (Fig. 5d, e).

Fig. 5
figure 5

Sensitivity of visual LAMP assay (b) and qPCR assay (c) for V. cholerae. The photograph (a) was corresponded with the LAMP amplification curve. The standard curves of the visual LAMP assay (d) and qPCR assay (e) were generated from the standard plasmid DNA, ranged from 102 to 109 CFU/mL, with the Bio-Rad CFX Manager 3.0 software. 1–11 1.1 × 109, 1.1 × 108, 1.1 × 107, 1.1 × 106, 1.1 × 105, 1.1 × 104, 1.1 × 103, 1.1 × 102, 1.1 × 101, and 1.1 × 100 CFU/mL, negative control

Full size image

Comparison of Sensitivity Between Real-Time Visual PMA-LAMP and PMA-qPCR Assays

According to the sensitivity trial (Fig. 6a), the detection limit of real-time visual PMA-LAMP assay was 1.1 × 102 CFU/mL. The PMA-LAMP assay also exhibited a high sensitivity as LAMP assay without PMA (Fig. 6b). There is a similar result that applies to PMA-qPCR (Fig. 6c). In the experiment, an amount of 1.1 × 105 CFU/mL V. cholerae dead/viable cells was used. The amplification of DNA templates from dead cells was suppressed thoroughly, and there was no negative impact on DNA of viable cells.

Fig. 6
figure 6

Sensitivity of visual PMA-LAMP and PMA-qPCR. a The photograph of the visual PMA-LAMP reaction. b The visual PMA-LAMP amplification curve. c The amplification curve of PMA-qPCR assay. 1–7 1.1 × 106, 1.1 × 105, 1.1 × 104, 1.1 × 103, 1.1 × 102, 1.1 × 101, and 1.1 × 100 CFU/mL, 8 pure viable cells with PMA, 9 pure dead cells with PMA, 10 NC

Full size image

Application of the Real-Time Visual PMA-LAMP and PMA-qPCR Assays to V. cholerae Detection in Different Produce Commodities

To analyze how different samples affected the PMA-LAMP and PMA-qPCR assays, the assays were used to detect DNA of the target pathogens extracted from fresh cod fillets, shrimps, and sleeve-fish. A total of 21 artificially inoculated samples were analyzed (Table 3). After enrichment for 24 h, the samples were analyzed for the presence of V. cholerae using the PMA-LAMP, PMA-qPCR, and BAM methods. Samples of each produced commodity that were not inoculated were used as negative controls. All samples artificially contaminated at a high level were positive for V. cholerae using the three methods. Only the lower level of inoculation was shown in order to compare the performance of the two molecular detection methods with the culture method (Table 3). The levels of V. cholerae were detected as low as 5 CFU/25 g after enrichment for 24 h.

Detection of Practical Products by Real-Time Visual PMA-LAMP and PMA-qPCR Assays

Nine different samples were detected by real-time visual PMA-LAMP, and results are displayed in Table 4. All samples were clearly detected by the microbiological method. The same conclusion was also drawn by real-time visual PMA-LAMP and PMA-qPCR assays. The qPCR assay showed a false positive result in the Meretrix sample, which was obviously excluded by the methods mentioned above. The real-time visual LAMP method played an excellent role with a support of PMA that removed the interference of false positive or dead cells.

Table 4 Detection of unknown samples from the market by visual PMA-LAMP assay compared with PMA-qPCR, regular visual LAMP assay, qPCR, and the culture method
Full size table

Discussion

Several reports indicated that the toxicity of PMA dyes was far less than EMA. Under the same circumstances, the working concentration of PMA was always higher than EMA (Nagamine et al. 2002). Except when treating naked DNA samples directly, the dosage of PMA was more than 10 μM (Nkuipou-Kenfack et al. 2013), according to different bacterial strains. Some reports even used 200 μM as a final concentration (Nkuipou-Kenfack et al. 2013). Wu et al. (2015) used 20 μM of PMA for examining V. cholerae cells, while Ruike et al. (2016) stated that 100 μM was the optimized concentration in their research. In this study, although denatured by the exposure, free PMA still had apparent influence on LAMP reaction (qPCR was regulated to exclude the possibility of impacting viable cells). The amplification time of the assay treated by 50 μM PMA was about 7 min later than the one without PMA. This undesirable influence would decrease the sensitivity of LAMP assay that needs some further researches. In addition, making DNA templates by specialized DNA extraction kits costs more time and procedures and had low recoveries. In this study, we found that the concentration of 15 μM PMA met our prospect and requirements after the optimization.

PMA treatment did not always lead to complete removal of the qPCR signal of dead bacteria (Fittipaldi et al. 2012; Nkuipou-Kenfack et al. 2013; Seinige et al. 2014). In this study, the capacity of PMA to eliminate the signal from the dead cell DNA of V. cholerae was shown to increase with increased PMA concentration. Although higher PMA concentrations could reduce the amount of detectable DNA from dead cells, lower PMA concentrations may lead to a false positive signal with dead bacteria. In our study, to obtain dead cells, the V. cholerae pure cells were killed and the bacteria were treated at 100 °C for 15 min, which was a harsh treatment that would release DNA to the supernatant, making it readily accessible to PMA. However, naturally contaminated samples would contain dead cells with compromised cell membranes but still containing their DNA, and PMA would have to penetrate into the cells to reach and block the DNA. This might lead to overestimation of the presence of viable bacteria in a sample and a false positive signal with dead bacteria. Therefore, to further investigate the relationship of the V. cholerae cell wall and membrane, the integrity of PMA-LAMP or PMA-qPCR results compared with plate counting and microscopic observation needs more researches.

The turbidity of the sample matrix (≥10 NTU) and the concentration of cells (OD600 ≥0.8) would weaken the efficiency of PMA (Zhu et al. 2012). Relatively less live cell DNA was amplified after the same PMA treatment when the dead cells’ concentration increased. In this study, as shown in Fig. 3, PMA lost its partial function if the magnitude of dead cells was as high as 106 CFU/mL and played a good role aiming at cells with 105 CFU/mL at most. This was similar to previous other studies. On the other hand, species-specific differences in PMA-qPCR had been revealed by various studies (Wu et al. 2015). The PMA treatment should be optimized for different species, especially bacteria. In addition to the characteristics of the PMA treatment itself, our results suggested that the initial concentration of targeted bacteria should be considered to be an important influence. Besides, some factors like chemical adsorption, organic and inorganic ingredients, pH, and salty components (competing with dyes) would also interfere with PMA (Fittipaldi et al. 2012; Barth et al. 2012). Whether these factors would or would not affect the detection of V. cholerae by PMA-LAMP assay needs some further researches.

In naturally contaminated produce samples, V. cholerae cells were unlikely to be in this active physiological state. Moreover, the environmental water samples usually contained few, if any, V. cholerae cells (Fykse et al. 2012). For that reason, it is difficult to make a high concentration of cell suspension for the PMA-LAMP procedure even after centrifugation. Therefore, a longer enrichment step might be necessary (Chen et al. 2011b). However, any dead cells present in the practical samples will be diluted out and only live cells will be detected after a longer enrichment. The validity of the PMA may be diminished in LAMP detection of viable cells of V. cholerae. In this study, no difference was observed between LAMP with and without PMA in the only one shrimp sample that was positive for V. cholerae after an enrichment step of 24 h (Table 4). We hypothesize that such a result might be caused by too long enrichment time, which selected and multiplied only viable cells, making the treatment with PMA redundant. PMA was able to permeate damaged cell membranes of dead cells, bind with DNA irreversibly, and inhibit its amplification subsequently (Fittipaldi et al. 2012). Therefore, the shortest possible time for enrichment procedure was necessary in the detection of actual samples to guarantee the validity of PMA-LAMP detection (Chen et al. 2011b). To further investigate, the optimal enrichment time of PMA-LAMP detection needs more researches.

In this study, PMA-based LAMP was applied to detect the positive viable cells of V. cholerae in seafood and the results were compared with culture isolation and LAMP without PMA treatment. The results showed that in 50 seafood samples, the positive detection rates were 2.0 and 2.0% by using the PMA-LAMP and PMA-qPCR, respectively, which were in agreement with the determination by culture isolation. Nevertheless, as discussed previously, there was no difference between LAMP with and without PMA because of the enrichment step of 24 h. If the sample contained live cells, as confirmed by culture, the LAMP result would be positive, with or without PMA. A suspicious sample (Cq >35) of Meretrix was negative, which was validated by culture method. It was only one sample, and the result is not clear. At this stage of current knowledge on foodborne pathogenic bacteria detection systems, the use of enrichment before PMA-LAMP or PMA-qPCR analysis followed by confirmation of positive results with culture-based methods continues to remain as the most appropriate approach.

In this study, the real-time visual PMA-LAMP assay offers a 32–37-h analysis, including the sample pre-treatment and the procedure of enrichment. The total testing time is shorter than the traditional methods which require 4 to 7 days to complete. And in the practical application, reaction tubes could be put in a thermal environment like a metal heating bath directly so that they would not affect the amplification completely. This method had higher specificity and sensitivity in contrast to quantitative PCR (Wu et al. 2015), but need not rely on sophisticated instrument and professionals which PMA-qPCR must be provided with. In conclusion, the visual PMA-LAMP provided a simple, low-cost, and rapid method in local and remote areas for screening samples.

Conclusion

V. cholerae is a notorious foodborne pathogen that severely threatens human health and life. The pathogen is found in seafood. Therefore, determining V. cholerae contamination in water and food is crucial. This study established a rapid, simple, and sensitive PMA-LAMP assay for detecting viable V. cholerae. Through optimization experiment, the working concentration of PMA dye was determined as 15 μM. On this basis, the upper limit of PMA was optimized so that only the magnitude of dead cells below 106 CFU/mL could be handled normally. The real-time visual PMA-LAMP assay we developed in this study not only had a detection limit as low as 5 CFU/25 g of viable bacteria after a short-time enrichment step but also could eliminate the interference of the residuary DNA from dead cells. Compared with the traditional culture method, PMA-LAMP assay could issue detection results within 2 days. On the other hand, PMA-LAMP assay ensured the high specificity and sensitivity in contrast to PMA-qPCR at the same time. Therefore, the real-time visual PMA-LAMP assay can serve as a useful and valuable method for diagnosing live V. cholerae strains in commercial and clinical areas.