Incidence and characterisation of aerobic spore‐forming bacteria originating from dairy milk in Tunisia

The purpose of this study was to evaluate the incidence and characterisation of aerobic spore‐forming bacteria originating from dairy milk in Tunisia. The distribution of Bacillus species in raw milk, pasteurised milk and UHT milk were 47.5%, 27.5% and 25%, respectively. Seven Bacillus species, including Bacillus pumilus (10%), Bacillus subtilis (12.5%), Brevibacillus brevis (10%), Bacillus cereus (22.5%), Bacillus sphaericus (7.5%), Bacillus licheniformis (12.5%) and Bacillus sporothermodurans (25%) were identified in different milk samples. Bacillus cereus was predominant in raw and pasteurised milk. Although B. sporothermodurans was the predominant sporogenous micro‐organism in UHT milk, B. cereus, B. sphaericus and B. licheniformis were also present. This study showed that there is a high degree of diversity, both phenotypic and genotypic, among Bacillus isolates from Tunisian milk and the persistence of spoilage risk in UHT milk.     

Optimization of nutrient-induced germination of Bacillus sporothermodurans spores using response surface methodology

Spores of Bacillus sporothermodurans are known to be contaminant of dairy products and to be extremely heat-resistant. The induction of endospore germination before a heat treatment could be an efficient method to inactivate these bacteria and ensure milk stability. In this study, the nutrient-induced germination of B. sporothermodurans LTIS27 spores was studied. Testing the effect of 23 nutrient elements to trigger an important germination rate of B. sporothermodurans spores, only d-glucose, l-alanine, and inosine were considered as strong independent germinants. Both inosine and l-alanine play major roles as co-germinants with several other amino acids. A central composite experimental design with three factors (l-alanine, d-glucose, and temperature) using response surface methodology was used to optimize the nutrient-induced germination. The optimal rate of nutrient-induced germination (100%) of B. sporothermodurans spores was obtained after incubation of spore for 60 min at 35 °C in presence of 9 and 60 mM of d-glucose and l-alanine, respectively. The results in this study can help to predict the effect of environmental factors and nutrients on spore germination, which will be beneficial for screening of B. sporothermodurans in milk after induction their germination. Moreover, the chosen method of optimization of the nutrient-induced germination was efficient in finding the optimum values of three factors.     

肉及肉制品中沙门氏菌活细胞的PCR检测研究

将荧光染料叠氮溴化丙锭(propidium monoazide,PMA)与普通聚合酶链式反应(polymerase chain reaction,PCR)结合,通过对PMA的曝光时间、浓度进行优化,确定PMA-PCR区别死活细胞的最佳条件,并制作活细胞定量标准曲线,建立肉及肉制品中沙门氏菌活细胞的PMA-PCR检测方法。结果表明：使插入死细胞DNA中的PMA活化并且光解溶液中游离PMA的最佳曝光时间为15min;不抑制沙门氏菌活细胞DNA扩增的最大PMA质量浓度为10μg/mL;完全抑制热致死细胞DNA扩增的最小PMA质量浓度为4μg/mL。经PMA处理,含有不同比例的沙门氏菌热致死细胞和活细胞的混合液中活的沙门氏菌能够通过PCR被选择性的检测,最小检测限为20CFU/PCR。而且,经研究发现在20～2×105CFU/PCR范围内,电泳条带相对荧光强度与活细胞数的对数具有线性关系。采集30份肉及肉制品样品,利用PMA-PCR方法检测出两份生肉样品中存在沙门氏菌,经过6h的富集培养后的活菌浓度分别为2.5×103CFU/mL和3.4×103CFU/mL。     

Simultaneous detection and enumeration of viable lactic acid bacteria and bifidobacteria in fermented milk by using propidium monoazide and real-time PCR

This study describes a procedure that allows specific detection and enumeration of viable bacteria in four species of lactic acid bacteria (Streptococcus thermophilus, Lactobacillus delbrueckii subsp. bulgaricus, Lactobacillus casei subsp. casei and Lactobacillus acidophilus) and of Bifidobacterium lactis, mixed in fermented milk products. The procedure is based on the combined use of propidium monoazide (PMA), able to distinguish between viable and irreversibly damaged cells, with species-specific quantitative real-time PCR (RTi-PCR). Loss of viability of the species in a fermented milk through storage at 4 °C was similarly (P < 0.05) detected by PMA–RTi-PCR and selective plate counts. Furthermore, comparison of results obtained by both methods showed a Pearson linear correlation of 0.995. The enumeration of viable bacteria by PMA–RTi-PCR could be performed in 3 h, whereas enumeration by selective plate counts required three days. The procedure developed is a fast method for the identification, enumeration and discrimination of viability of lactic acid bacteria and bifidobacteria mixed in fermented milk products.     

Magnetic nano-beads based separation combined with propidium monoazide treatment and multiplex PCR assay for simultaneous detection of viable Salmonella Typhimurium, Escherichia coli O157:H7 and Listeria monocytogenes in food products

We developed a rapid and reliable technique for simultaneous detection of Salmonella Typhimurium, Escherichia coli O157:H7 and Listeria monocytogenes that can be used in food products. Magnetic nano-beads (MNBs) based immunomagnetic separation (IMS) was used to separate the target bacterial cells while multiplex PCR (mPCR) was used to amplify the target genes. To detect only the viable bacteria, propidium monoazide (PMA) was applied to selectively suppress the DNA detection from dead cells. The results showed the detection limit of IMS-PMA-mPCR assay was about 10² CFU/ml (1.2 × 10² CFU/ml for S. Typhimurium, 4.0 × 10² CFU/ml for E. coli O157:H7 and 5.4 × 10² CFU/ml for L. monocytogenes) in pure culture and 10³ CFU/g (5.1 × 10³ CFU/g for S. Typhimurium, 7.5 × 10³ CFU/g for E. coli O157:H7 and 8.4 × 10³ CFU/g for L. monocytogenes) in spiking food products samples (lettuce, tomato and ground beef). This report has demonstrated for the first time, the effective use of rapid and reliable IMS combined with PMA treatment and mPCR assay for simultaneous detection of viable S. Typhimurium, E. coli O157:H7 and L. monocytogenes in spiked food samples. It is anticipated that the present approach will be applicable to simultaneous detection of the three target microorganisms for practical use.     

Multiplex PCR coupled with propidium monoazide for the detection of viable Cronobacter sakazakii,Bacillus cereus,and Salmonella spp. in milk and milk products

Cronobacter sakazakii, Bacillus cereus, and Salmonella spp. are common food-borne pathogens. The aim of this study was to develop a sensitive, specific, and rapid method for the simultaneous detection of these 3 pathogens in milk and milk products. Three specific primers were designed based on ompA, invA, and cesB of C. sakazakii, Salmonella spp. and B. cereus, respectively, for use in a multiplex PCR (mPCR). To eliminate false-positive results, cells were pretreated with propidium monoazide (PMA) for the selective elimination of the genomic DNA of dead cells. An internal amplification control was applied as an indicator of false-negative results from the interference of inhibitors in the food matrix. Results showed that, in pure culture, the limits of detection of the assay for C. sakazakii, Salmonella Enteritidis, and B. cereus were 9.5 × 10⁴, 7.4 × 10², and 7.5 × 10² cfu/mL, respectively. Moreover, 8 cfu/mL of viable B. cereus cells were detected after 5 h of enrichment, and 9 cfu/mL of viable C. sakazakii and 7 cfu/mL of Salmonella Enteritidis were detected after 7 h of enrichment in spiked pure milk, walnut peanut milk, and whole-wheat milk. To validate the PMA-mPCR assay, the PMA-mPCR assay and the traditional culture method were performed to detect the 3 bacterial strains in 1,165 milk product samples. The PMA-mPCR assay obtained the same results as the culture-based method. Results demonstrated that the PMA-mPCR assay has excellent sensitivity and specificity for the simultaneous detection of viable C. sakazakii, Salmonella Enteritidis, and B. cereus in milk and milk products.     

Detection of culturable and viable but non‐culturable cells of beer spoilage lactic acid bacteria by combined use of propidium monoazide and horA‐specific polymerase chain reaction

Current methods of detecting beer spoilage lactic acid bacteria (LAB) are time‐consuming and do not differentiate between viable and non‐viable bacteria. In this study, a combination of the conventional polymerase chain reaction (PCR) and propidium monoazide (PMA) pretreatment has been described to circumvent the disadvantages. The horA‐specific PMA‐PCR described here identifies beer spoilage LAB based not on their identity, but on the presence of a gene that is shown to be highly correlated with the ability of LAB to grow in beer. The results suggest that the use of 20 µg/mL or less of PMA did not inhibit the PCR amplification of DNA derived from viable, but putatively non‐culturable (VPNC) Lactobacillus acetotolerans. The minimum amount of PMA to completely inhibit the PCR amplification of DNA derived from dead L. acetotolerans cells was 1.5 µg/mL. The detection limit of established PMA‐PCR assays was found to be 100 VPNC cells/reaction for the horA gene. Furthermore, the horA‐specific PMA‐PCR assays were subjected to 18 reference strains, representing 100% specificity with no false positive amplification observed. In conclusion, the use of horA‐specific PMA‐PCR allows for a substantial reduction in the time required for the detection of potential beer spoilage LAB and efficiently discriminates between live and dead cells. Copyright © 2016 The Institute of Brewing & Distilling     

Development of an immunomagnetic separation–propidium monoazide–polymerase chain reaction assay with internal amplification control for rapid and sensitive detection of viable Escherichia coli O157:H7 in milk

A rapid, reproducible, sensitive and specific combination assay was developed for Escherichia coli O157:H7 comprising sequential immunomagnetic separation (IMS), followed by propidium monoazide (PMA), exclusion by viable cells, internal amplification control (IAC) and polymerase chain reaction (PCR). IMS was used to improve sensitivity, reduce detection time and eliminate false positives. PMA was used to detect viable cells, and IAC was incorporated into the system to eliminate the inhibitors of PCR from the food matrix. In the presence of inhibitors, IAC produced no signal, thereby eliminating false negative results. The results indicated that the IMS cell capture efficiency was about 90% at a concentration of 1 × 10⁶ cfu mL⁻¹ with a detection limit of 2.1 × 10² cfu mL⁻¹ for pure culture. In an unoptimised system, IMS–PMA–PCR with IAC showed a detection limit of 5 × 10³ cfu mL⁻¹ in spiked milk over a 240 min assay, faster than conventional methods of detection.     

Quantitative PCR coupled with sodium dodecyl sulfate and propidium monoazide for detection of viable Staphylococcus aureus in milk

Conventional quantitative PCR (qPCR) are unable to differentiate DNA of viable Staphylococcus aureus cells from dead ones. The aim of this study was to use sodium dodecyl sulfate (SDS) and propidium monoazide (PMA) coupled with lysostaphin to detect viable Staph. aureus. The cell suspensions were treated with SDS and PMA before DNA extraction. The SDS is an anionic surfactant, which can increase the permeability of dead cells to PMA without compromising the viability of live cells. The lysostaphin was applied to improve the effectiveness of DNA extraction. The reliability and specificity of this method were further determined by the detection of Staph. aureus in spiked milk. The results showed that there were significant differences between the SDS-PMA-qPCR and qPCR when a final concentration of 200 μg/mL of lysostaphin was added in DNA extraction. The viable Staph. aureus could be effectively detected when SDS and PMA concentrations were 100 µg/mL and 40 μM, respectively. Compared with conventional qPCR, the SDS-PMA-qPCR assay coupled with lysostaphin was more specific and sensitive. Therefore, this method could accurately detect the number of viable Staph. aureus cells.     

Rapid and simultaneous quantification of viable Escherichia coli O157:H7 and Salmonella spp. in milk through multiplex real-time PCR

Escherichia coli O157:H7 and Salmonella spp. in milk are 2 common pathogens that cause foodborne diseases. An accurate, rapid, specific method has been developed for the simultaneous detection of viable E. coli O157:H7 and Salmonella spp. in milk. Two specific genes, namely, fliC from E. coli O157:H7 and invA from Salmonella spp., were selected to design primers and probes. A combined treatment containing sodium deoxycholate (SDO) and propidium monoazide (PMA) was applied to detect viable E. coli O157:H7 and Salmonella spp. only. Traditional culture methods and SDO-PMA-multiplex real-time (mRT) PCR assay were applied to determine the number of viable E. coli O157:H7 and Salmonella spp. in cell suspensions with different proportions of dead cells. These methods revealed consistent findings regarding the detected viable cells. The detection limit of the SDO-PMA-mRT-PCR assay reached 10² cfu/mL for Salmonella spp. and 10² cfu/mL for E. coli O157:H7 in milk. The detection limit of SDO-PMA-mRT-PCR for E. coli O157:H7 and Salmonella spp. in milk was significantly similar even in the presence of 10⁶ cfu/mL of 2 nontarget bacteria. The proposed SDO-PMA-mRT-PCR assay is a potential approach for the accurate and sensitive detection of viable E. coli O157:H7 and Salmonella spp. in milk.     

High-throughput detection of spore contamination in food packages and food powders using tiered approach of ATP bioluminescence and real-time PCR

A rapid and quantitative method for detection of Bacillus spores in food/non-alcoholic beverage packages and food powders has been developed using filtration-based ATP bioluminescence and real-time PCR, targeting the sporulation gene (spo0A). In combination with heat activation, the presence and amount of viable bacterial spores (i.e., Bacillus amyloliquefaciens, Bacillus licheniformis, and Bacillus thuringiensis) was determined within 20 min through ATP signal amplifications. The detection limits of heat activation-ATP bioluminescence assay for B. amyloliquefaciens and B. licheniformis spores on food packages were 1.4 × 10² and 1.0 × 10³ CFU/cm², respectively. In contaminated food powders, B. thuringiensis spores could be detected by the ATP assay within the range of 7.9 × 10⁰ to 3.2 × 10⁴ CFU/mg powder while the PCR detection limit was 614 CFU/mg. Linear relationships between luminescent signal (RLU/mg) and plate count (CFU/mg) were found. The same sample after heat activation-ATP assay could be directly used for real-time PCR as a streamlined detection to confirm the identity of Bacillus spores in food packages and food powders even though some bacterial DNA loss was observed. This tiered approach, filtration-based one-tube ATP luminescence method as a rapid, viable screening and using real-time PCR as confirmation, could serve as a high-throughput tool for the detection of Bacillus spores in the food and beverage industry.     

基于叠氮溴化丙锭与恒温核酸扩增技术快速检测亚致死状态食源金黄色葡萄球菌

建立一种将荧光染料叠氮溴化丙锭（propidium monoazide,PMA）与环介导等温扩增（loop-mediatedisothermal amplification,LAMP）技术相结合的方法,用于快速高效检测金黄色葡萄球菌（Staphylococcusaureus）。同时,采用人工污染金黄色葡萄球菌的速冻水饺和奶粉作为食品样品,研究PMA-LAMP方法的检测灵敏度。结果表明,PMA溶液质量浓度3 μg/mL,650 W卤素灯下曝光5 min,PMA能够完全抑制1.2×107 copies/mL金黄色葡萄球菌死菌核酸扩增。PMA-LAMP方法能够在恒温65 ℃、60 min内完成对亚致死型金黄色葡萄球菌特异性nuc基因的特异性检测,其对亚致死状态金黄色葡萄球菌的检出限为34 CFU/mL,对食物样品速冻水饺和奶粉的检出限分别为17 CFU/mL和1.70×102 CFU/mL。建立的PMA-LAMP方法可以有效检测亚致死态金黄色葡萄球菌,提供了一种新的检测技术和解决方案。     

Detection and quantification of spoilage and pathogenic Bacillus cereus, Bacillus subtilis and Bacillus licheniformis by real-time PCR

A new primer-probe set for the detection and quantification of Bacillus cereus, Bacillus licheniformis and Bacillus subtilis by real-time PCR (Rti-PCR) was developed. For it, forty-eight strains belonging to these species were considered. The DNA of these strains was isolated and a fragment of the 16S rRNA gene amplified. The amplicons were sequenced and the obtained sequences were aligned with reference sequences from the GenBank. For the development of the Real-Time PCR (RTi-PCR) methodology based on TaqMan probes, a primer pair and probe, specific for the studied Bacillus spp., were designed. To establish the quantification method, two RTi-PCR standard curves were constructed; one with DNA extracted from a serially-diluted B. cereus culture and a second curve with DNA extracted from a sterilised food product inoculated with serial dilutions of B. cereus. The curves exhibited R² values of 0.9969 and 0.9958 respectively. Linear correlations between the log₁₀ input DNA concentration and the threshold cycle (Ct) values were observed with a magnitude of linearity in the range of 1.65 × 10¹ CFU/mL to 1.65 × 10⁶ CFU/mL for both standard curves. The specificity of the designed primers and probe was tested with DNA extracted from B. cereus, B. licheniformis and B. subtilis strains, which gave Ct values between 14 and 15, whereas non-specific amplifications of the DNA from other microbial species of food interest exhibited a Ct value above 28.5. To our knowledge, this method represents the first study about the quantification of spoilage and/or pathogenic B. cereus, B. licheniformis and B. subtilis in food products, with the aim to prevent the presence of these undesirable species in the food chain.     

Seasonal and regional occurrence of heat-resistant spore-forming bacteria in the course of ultra-high temperature milk production in Tunisia

Spore-forming bacteria, principally Bacillus species, are important contaminants of milk. Because of their high heat resistance, Bacillus species spores are capable of surviving the heat treatment process of milk and lead to spoilage of the final product. To determine the factors influencing the contamination of milk, spore-forming bacteria occurrence throughout the UHT milk production line during winter, spring, and summer was studied. The obtained results confirm that the total viable rate decreases rapidly throughout the production line of UHT milk showing the efficiency of thermal treatments used. However, the persistent high rate of spore-forming bacteria indicates their high heat resistance, especially in spring and summer. In addition, a significant variation of the quality of raw milk according to the location of the collecting centers was revealed. The molecular identification showed a high degree of diversity of heat-resistant Bacillus species, which are isolated from different milk samples. The distribution of Bacillus species in raw milk, stored milk, bactofuged milk, pasteurized milk, and UHT milk were 28, 10, 16, 13, and 33%, respectively. Six Bacillus spp. including Bacillus licheniformis (52.38%), Bacillus pumilus (9.52%), Bacillus sp. (4.76%), Bacillus sporothermodurans (4.76%), Terribacillus aidingensis (4.76%), and Paenibacillus sp. (4.76%) were identified in different milk samples.     

Optimization of pressure-induced germination of Bacillus sporothermodurans spores in water and milk

Bacillus sporothermodurans produces highly resistant endospores that can survive ultra-high-temperature treatment in milk. The induction of endospore germination before a heat treatment could be an efficient method to inactivate these bacteria and ensure milk sterility. In this work, the rate of spore germination of B. sporothermodurans LTIS27 was measured in distilled water after high-pressure treatments with varying pressure (50–600 MPa), treatment temperature (20–50 °C), pressure-holding time (5–30 min) and post-pressurization incubation time (30–120 min) at 37 °C or 4 °C. The results showed that pressure-induced germination was maximal (62%) after a treatment at 200 MPa and 20 °C and increased with pressure-holding time and post-pressurization incubation time. Treatment temperature had no significant effect on germination. A central composite experimental design with three factors (pressure, pressure-holding time, and post-pressurization incubation time) using response surface methodology was used to optimize the germination rate in distilled water and in skim milk. No factor interaction was observed. Germination was induced at lower pressure and was faster in milk than in distilled water, but complete germination was not reached. The optimum germination obtained with experimental data was 5.0 log cfu/mL in distilled water and 5.2 log cfu/mL in milk from 5.7 log cfu/mL of spores initially present in the suspension. This study shows the potential of using high hydrostatic pressure to induce the germination of B. sporothermodurans spores in milk before a heat treatment.     

Biotin exposure–based immunomagnetic separation coupled with sodium dodecyl sulfate,propidium monoazide,and multiplex real-time PCR for rapid detection of viable Salmonella Typhimurium,Staphylococcus aureus,and Listeria monocytogenes in milk

In this study, we established a rapid and sensitive method for the detection of viable Salmonella Typhimurium, Staphylococcus aureus, and Listeria monocytogenes in milk using biotin-exposure-based immunomagnetic separation (IMS) combined with sodium dodecyl sulfate (SDS), propidium monoazide (PMA), and multiplex real-time PCR (mRT-PCR). We used IMS to lessen the assay time for isolation of target bacteria. We then optimized the coupling conditions and immunomagnetic capture process. The immunoreaction and incubation times for 5 μg of mAb coupled with 500 μg of streptavidin-functionalized magnetic beads using a streptavidin-biotin system were 90 and 30 min, respectively. Treatment with SDS-PMA before mRT-PCR amplification eliminated false-positive outcomes from dead bacteria and identified viable target bacteria with good sensitivity and specificity. The limit of detection of IMS combined with the SDS-PMA-mRT-PCR assay for the detection of viable Salmonella Typhimurium, Staph. aureus, and L. monocytogenes in spiked milk matrix samples was 10 cfu/mL and remained significant even in the appearance of 10⁶ cfu/mL of nontarget bacteria. The entire detection process was able to identify viable bacteria within 9 h. The combination of biotin-exposure-mediated IMS and SDS-PMA-mRT-PCR has potential value for the rapid and sensitive detection of foodborne pathogens.     

Rapid propidium monoazide PCR assay for the exclusive detection of viable Enterobacteriaceae cells in pasteurized milk

Pasteurized milk is a complex food and contains numerous PCR inhibitors and can often contain high levels of dead Enterobacteriaceae cells, depending on the condition of food sanitation. Usually, propidium monoazide (PMA) or ethidium monoazide PCR techniques decrease the number of dead bacteria by up to 3.5 log to the associated dead bacteria with no treatment. However, this difference could be insufficient to completely inhibit DNA amplification in the PCR from 10(6) cells of dead Enterobacteriaceae bacteria/mL, potentially contaminated in pasteurized milk. Actually, such potentially high levels of dead Enterobacteriaceae cells in milk has prevented milk researchers from applying PMA- or ethidium monoazide PCR to the assay of viable Enterobacteriaceae cells in milk. We, therefore, developed a rapid PMA real-time PCR whose minimum levels of detection were 1.5 log cfu/PCR for Cronobacter muytjensii and Escherichia coli, and 2.5 log cfu/PCR for Salmonella enteritidis without DNA purification in milk matrices. The PMA real-time PCR allowed us to specifically detect viable Enterobacteriaceae cells (5-10 cfu/mL) in pasteurized milk (20 mL) within 7.5h of total testing time, following the hygienic guidelines for pasteurized milk in the United States and European Union. The long DNA amplification (mainly 2,451 bp) of the 16S-23S rRNA gene was completely suppressed in highly contaminated dead Enterobacteriaceae cells (7.5 log cfu of Cronobacter muytjensii) in 20 mL of pasteurized milk by 23-μM PMA treatment. Although the contamination of the PCR reaction with 5% milk usually causes great inhibition, our method led to the successful elongation of PCR from viable Enterobacteriaceae cells still in the pasteurized milk matrices finally corresponding to 2 to 4 mL of milk PCR inhibitors without a DNA purification step. To comply with current customer demands for chilled pasteurized milk at the most excellent possible quality, our new technique could enable laboratory persons in a factory to conduct rapid milk coliform testing before shipping from a factory.     

Use of propidium monoazide and quantitative PCR for differentiation of viable Escherichia coli from E. coli killed by mild or pasteurizing heat treatments

Suspensions of Escherichia coli in peptone water were heated at temperatures between 52 and 90 °C, inclusive. Samples withdrawn at suitable times were not or were treated with propidium monoazide (PMA) or deoxycholate then PMA before extraction of DNA. DNA was quantified by real-time PCR for estimation of the numbers of E. coli from which template DNA for the PCR was obtained. Numbers of viable E. coli in suspensions at the times of sampling were determined from plate counts. For samples from suspensions heated at temperatures ≥52 ≤ 72 °C, PCR cycle threshold (Ct) values were little or no different for DNA from corresponding samples that were or were not treated with PMA. PMA treatment of samples heated to ≥80 °C largely inactivated E. coli DNA for PCR. When samples heated to ≤72 °C were treated with deoxycholate before treatment with PMA, Ct values for treated samples were greater than the Ct values for the corresponding untreated samples. Similar results were obtained with E. coli suspended in milk or fluid from ground beef pummeled with diluent. The results indicate that cells killed by heating to ≥80 °C are permeable to PMA, but most cells killed by heating to ≤72 °C are not. However, treatment with deoxycholate renders a substantial fraction of the latter cells permeable to PMA. Numbers of viable or dead E. coli can then be estimated from Ct values for samples not treated or treated with deoxycholate and PMA, provided viable cells are ≥1% of the total.     

Detection of viable Salmonella in lettuce by propidium monoazide real-time PCR

Contamination of lettuce by Salmonella has caused serious public health problems. Polymerase chain reaction (PCR) allows rapid detection of pathogenic bacteria in food, but it is inaccurate as it might amplify DNA from dead target cells as well. This study aimed to investigate the stability of DNA of dead Salmonella cells in lettuce and to develop an approach to detecting viable Salmonella in lettuce. Salmonella-free lettuce was inoculated with heat-killed Salmonella Typhimurium cells and stored at 4 °C. Bacterial DNA extracted from the sample was amplified by real-time PCR targeting the invA gene. Our results indicate that DNA from the dead cells remained stable in lettuce for at least 8 d. To overcome this limitation, propidium monoazide (PMA), a dye that can selectively penetrate dead bacterial cells and cross-link their DNA upon light exposure, was combined with real-time PCR. Lettuce samples inoculated with different levels of dead or viable S. Typhimurium cells were treated or untreated with PMA before DNA extraction. Real-time PCR suggests that PMA treatment effectively prevented PCR amplification from as high as 10(8) CFU/g dead S. Typhimurium cells in lettuce. The PMA real-time PCR assay could detect viable Salmonella at as low as 10(2) CFU/mL in pure culture and 10(3) CFU/g in lettuce. With 12-h enrichment, S. Typhimurium of 10(1) CFU/g in lettuce was detectable. In conclusion, the PMA real-time PCR assay provides an alternative to real-time PCR assay for accurate detection of Salmonella in food.