Real-time PCR method combined with immunomagnetic separation for detecting healthy and heat-injured Salmonella Typhimurium on raw duck wings

International Journal of Food Microbiology 186 (2014) 6–13

Contents lists available at ScienceDirect

International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro

Real-time PCR method combined with immunomagnetic separation for detecting healthy and heat-injured Salmonella Typhimurium on raw duck wings Qianwang Zheng a, Marta Mikš-Krajnik a,c, Yishan Yang a, Wang Xu b, Hyun-Gyun Yuk a,d,⁎ a

Food Science & Technology Programme, Department of Chemistry, National University of Singapore, Science Drive 4, 117543, Singapore Department of Chemistry, National University of Singapore, Science Drive 4, 117543, Singapore c Chair of Industrial and Food Microbiology, Faculty of Food Science, University of Warmia and Mazury in Olsztyn, Plac Cieszyński 1, 10-726 Olsztyn, Poland d National University of Singapore (Suzhou) Research Institute, No. 377 Linquan Street, Suzhou Industrial Park, Suzhou, Jiangsu 215123, China b

a r t i c l e

i n f o

Article history: Received 28 January 2014 Received in revised form 23 May 2014 Accepted 8 June 2014 Available online 13 June 2014 Keywords: Salmonella Typhimurium Raw duck wings Real-time PCR Immunomagnetic separation Heat injury

a b s t r a c t Conventional culture detection methods are time consuming and labor-intensive. For this reason, an alternative rapid method combining real-time PCR and immunomagnetic separation (IMS) was investigated in this study to detect both healthy and heat-injured Salmonella Typhimurium on raw duck wings. Firstly, the IMS method was optimized by determining the capture efficiency of Dynabeads® on Salmonella cells on raw duck wings with different bead incubation (10, 30 and 60 min) and magnetic separation (3, 10 and 30 min) times. Secondly, three Taqman primer sets, Sal, invA and ttr, were evaluated to optimize the real-time PCR protocol by comparing five parameters: inclusivity, exclusivity, PCR efficiency, detection probability and limit of detection (LOD). Thirdly, the optimized real-time PCR, in combination with IMS (PCR–IMS) assay, was compared with a standard ISO and a real-time PCR (PCR) method by analyzing artificially inoculated raw duck wings with healthy and heatinjured Salmonella cells at 101 and 100 CFU/25 g. Finally, the optimized PCR–IMS assay was validated for Salmonella detection in naturally contaminated raw duck wing samples. Under optimal IMS conditions (30 min bead incubation and 3 min magnetic separation times), approximately 85 and 64% of S. Typhimurium cells were captured by Dynabeads® from pure culture and inoculated raw duck wings, respectively. Although Sal and ttr primers exhibited 100% inclusivity and exclusivity for 16 Salmonella spp. and 36 non-Salmonella strains, the Sal primer showed lower LOD (103 CFU/ml) and higher PCR efficiency (94.1%) than the invA and ttr primers. Moreover, for Sal and invA primers, 100% detection probability on raw duck wings suspension was observed at 103 and 104 CFU/ml with and without IMS, respectively. Thus, the Sal primer was chosen for further experiments. The optimized PCR–IMS method was significantly (P = 0.0011) better at detecting healthy Salmonella cells after 7-h enrichment than traditional PCR method. However there was no significant difference between the two methods with longer enrichment time (14 h). The diagnostic accuracy of PCR–IMS was shown to be 98.3% through the validation study. These results indicate that the optimized PCR–IMS method in this study could provide a sensitive, specific and rapid detection method for Salmonella on raw duck wings, enabling 10-h detection. However, a longer enrichment time could be needed for resuscitation and reliable detection of heat-injured cells. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Salmonella is a leading cause of foodborne illness worldwide, with approximately 100,000 cases reported in the European Union (EU) (EFSA, 2013) and 40,000 cases in the United States (US) (Almeida et al., 2013) each year. Poultry appears to be the main reservoir of Salmonella spp., with an increase of salmonellosis linked to poultry products from 2010 to 2012 (CDC, 2012a, 2012b). Most cases showed transmittal through direct contact or consumption of undercooked contaminated products

⁎ Corresponding author. Tel.: +65 6516 1136; fax: +65 6775 7895. E-mail address: [email protected] (H.-G. Yuk).

http://dx.doi.org/10.1016/j.ijfoodmicro.2014.06.005 0168-1605/© 2014 Elsevier B.V. All rights reserved.

(CDC, 2013). Salmonella enterica subspecies enterica serovar Typhimurium and Salmonella Enteritidis are the most commonly reported serotypes in the outbreaks both in the US and Asian countries (CDC, 2012b). Recently, a total of 356 persons from 39 US states were infected during the outbreak of S. Typhimurium linked to the contact with backyard chicks and ducklings (CDC, 2013). Global duck meat production is dominated by Asia, reaching 4.4 million tons in 2013 and expected to approach 4.6 million tons by 2015 according to a forecast made by the Food and Agriculture Organisation (FAO) (Anonymous, 2012). Raw duck meat from Asia has been shown to be a reservoir for Salmonella spp., with S. Typhimurium as the predominant serovar (Adzitey et al., 2011). In food products, Salmonella spp. appears in relatively low concentrations and in a physiologically weakened state due to the environmental stresses

Q. Zheng et al. / International Journal of Food Microbiology 186 (2014) 6–13

or food processing conditions (Jasson et al., 2011; Li et al., 2010). Therefore, to ensure the safety of duck meat consumption, the presence of Salmonella spp. in these products should be screened using an accurate and effective detection method. The currently used standard culture method (ISO, 2002) for Salmonella spp. detection is laborious and time consuming, involving several steps: non-selective pre-enrichment, selective enrichment, followed by plating on the selective agars and finally biochemical identification. For these reasons, numerous approaches have been examined to shorten the detection time and increase the sensitivity, most commonly utilizing PCR as the detection tool (Almeida et al., 2013; Malorny et al., 2007; Zhou et al., 2013). The use of sequence-specific probes and DNA binding dyes to monitor a target DNA amplification (i.e., real-time PCR) has been widely adapted for the detection of foodborne pathogens. The specific real-time PCR probe Taqman has been employed to detect Salmonella spp. in various matrices including chicken rinses, fish, raw milk (Malorny et al., 2004, 2007), potato salads, ground beef (Warren et al., 2007), eggs, mayonnaise (Almeida et al., 2013), biscuit, juice, pork and spinach (Li et al., 2010). However, the real-time PCR method can have low detection sensitivity due to the loss of target DNA template during sample preparation (Lee et al., 2009), or the inhibition of reaction which is caused by various compounds presented in food samples (Margot et al., 2013). Moreover, the detection and isolation of Salmonella spp. cells in food containing high background microflora, such as raw duck meat (Hu et al., 2011), may be challenging (Fedio et al., 2011). Immunomagnetic separation (IMS) technique has been shown to be an effective approach for the selective isolation and concentration of target pathogens with respect to the reduction of analysis time and sensitivity improvement (Duodu et al., 2009; Fedio et al., 2011; Shields et al., 2012). IMS uses magnetic beads (ca. 1 to 5 μm) coated with specific antibodies to capture and isolate target pathogen cells from the homogenized food suspensions through antibody–antigen interaction. IMS has been recently applied to isolate Escherichia coli O157:H7 (Fedio et al., 2011; Fu et al., 2005), Listeria monocytogenes (Duodu et al., 2009; Wang et al., 2007, 2011; Yang et al., 2007), S. Typhimurium (Almeida et al., 2013; Mercanoglu and Griffiths, 2005; Wang et al., 2011) and other bacteria (Shields et al., 2012) from various food matrices, including eggs, milk, mayonnaise, ground beef, salmon, fresh-cut lettuce, and chicken carcass. Even though many studies have shown that real-time PCR combined with IMS (PCR–IMS) is effective in detecting Salmonella spp. in various foods, no studies have been conducted to evaluate the detection method for sub-lethally injured Salmonella on duck meat. Thus, the objective of the study was to optimize the protocol for a molecular-based, rapid detection of low concentrations of S. Typhimurium on raw duck wings by PCR–IMS. Firstly, the experimental factors for IMS of S. Typhimurium from raw duck wings using commercially available microparticles (Dynabeads®) were optimized. Next, the comparisons of three Taqman primers for S. Typhimurium detection in pure cultures and on raw duck wings were performed. Thirdly, the optimized PCR–IMS method for detecting S. Typhimurium was compared against the standard ISO (ISO, 2002) and real-time PCR (PCR) methods utilizing artificially inoculated raw duck wing samples with both healthy and heat-injured cells at low inoculum levels. Finally, the PCR–IMS method was validated utilizing naturally contaminated raw duck wings in comparison with the standard ISO method. 2. Materials and methods

7

use. S. Typhimurium cells were adapted to 200 μg/l of nalidixic acid (Sigma-Aldrich, St. Louis, MO, USA) and 40 μg/l of novobiocin (Sigma-Aldrich) by successive culturing with increasing concentrations of nalidixic acid in TSB to develop antibiotic resistance that allows isolation of inoculated S. Typhimurium cells from background microflora in raw duck wings. Before inoculation, 1 ml of nalidixic acid and novobiocin adapted S. Typhimurium (ca. 108 CFU/ml) was centrifuged at 3500 ×g for 10 min at 4 °C, washed twice with 1 ml of 0.1% (w/v) peptone water (PW, Oxoid) and finally the pellet was re-suspended in 1 ml of 0.1% (w/v) PW. The preparation of heat-injured Salmonella cells was described elsewhere (Zheng et al., 2013). Briefly, 2 ml of stationary-phase culture (ca. 108 CFU/ml) was transferred into a sterile aluminum cylindrical container (3.0 cm diameter, 1.2 cm height) and heated in a water bath at 60 °C for 1.5 and 2 min to achieve the final percent of injury of 50% and 85%, respectively. The percentage of sub-lethal injury was determined by comparing the counts obtained on tryptone soy agar (TSA; Oxoid) as the non-selective agar to brilliant green agar (BGA; Oxoid) as the selective agar (Uyttendaele et al., 2008). 2.2. Inoculation of Salmonella cells on raw duck wings Fresh, raw duck wings were purchased from Singapore local grocery stores. For each trial, 25 g portions of raw duck wings were weighed under sterile conditions and inoculated with spotting technique. Ten spots (an aliquot of 10 μl per spot) of appropriate bacterial dilutions were spotted on the raw duck wing surface (Zheng et al., 2013). Inoculated raw duck wing samples were placed in sterile Stomacher® bags (Gosselin™, Hazebrouck, France) and stored at 8 °C overnight to simulated storage condition in the grocery stores. 2.3. Optimization of immunomagnetic separation (IMS) A 25 g sample was homogenized in 225 ml one broth (OB; Oxoid) as recommended by Zheng et al. (2013), for 2 min using a paddle blender (Silver Masticator, IUL Instruments GmbH, Königswinter, Germany) and incubated at 42 °C for 24 h. Prior to IMS, 100 μl of the pure culture and 24-h OB-enrichment was spread plated on the modified BGA containing 0.8 g/l sulfadiazine (Sigma-Aldrich), 200 ppm of nalidixic acid, and 40 ppm of novobiocin to confirm the concentration of Salmonella spp. (106 CFU/ml) (Zheng et al., 2013). Twenty microliters of Dynabeads® anti-Salmonella (2.5 μm diameter) (Invitrogen™, Life Technologies, Carlsbad, CA, USA) was incubated with 980 μl of pure culture suspension or enriched sample at room temperature (25 °C) using a rotating mixer (BioSan, Riga, Latvia) for 10, 30 or 60 min. After bead incubation, the bead–bacteria complexes were magnetically separated from the suspension using a magnetic particle concentrator (Invitrogen™) for 3, 10 or 30 min, and then washed twice with 1 ml of phosphate-buffered saline (PBS, pH 7.4, Sigma-Aldrich) containing 0.05% (v/v) Tween 80 (Thermo Scientific, Hampton, NH, USA) to remove non-specifically binding bacteria from the complex (Tatavarthy et al., 2009), followed by re-suspending with 1 ml of PBS. A 100 μl sample was spread plated on the modified BGA to enumerate Salmonella spp. and the capture efficiency (CE) of Dynabeads® was calculated using the following equation (Wang et al., 2011): Capture efficiency ð%Þ ¼ Nc =N0  100 where Nc is an average total number of Salmonella cells captured with Dynabeads® (CFU/ml) and N0 is an average total number of Salmonella cells present in 1 ml of original suspension (CFU/ml).

2.1. Preparation of healthy and heat-injured Salmonella cells 2.4. Optimization of real-time PCR S. enterica subspecies enterica serovar Typhimurium ATCC 14028 (Poultry isolate; American Type Culture Collection, Manassas, VA, USA), was cultured twice in 10 ml of tryptone soy broth (TSB; Oxoid, Basingstoke, Hampshire, UK) overnight at 37 °C before

2.4.1. DNA template preparation To prepare DNA template for real-time PCR, a 1 ml aliquot of the OBenrichment of inoculated raw duck wings or S. Typhimurium pure

8

Q. Zheng et al. / International Journal of Food Microbiology 186 (2014) 6–13

culture was centrifuged at 3500 ×g for 10 min at 4 °C. The supernatant was carefully discarded and the pellet was re-suspended in 200 μl 10xPBS. The suspension was boiled at 100 °C for 5 min and immediately chilled on ice for 3 min, followed by a centrifugation at 13,300 ×g for 10 min at 4 °C. The supernatant was collected and stored at − 20 °C for further real-time PCR analysis.

with the ORIGIN program (ver. 4.0; Micro Software, Northampton, MA, USA). For the real-time PCR efficiency, standard curves were constructed by plotting Ct values as a function of log CFU/ml of 1:10 serially diluted pure S. Typhimurium cultures. The real-time PCR efficiency was calculated by the following formula (Gorski and Liang, 2010):

2.4.2. Primer sets and real-time PCR condition The Salmonella-specific primers tested in this study are presented in Table 1. All primers were synthesized by Integrated DNA Technology (IDT, Coralville, IA, USA). The Salmonella spp. target probes were labeled at the 5′-end with 6-carboxyfluorescein (FAM™; 494–520 nm) dye as the reporter and at the 3′-end with BHQ-2 as the dark quencher. The internal positive control (IPC) (Applied Biosystems®, Life Technologies, Carlsbad, CA, USA) was labeled at the 5′-end with VIC® (530–549 nm) dyes as reporter and at the 3′-end with TAMRA as quencher. A typical 20 μl PCR mixture contained: 900 nM of each primer, 250 nM of probe, 10 μl of Taqman Fast Universal PCR Master Mix (Applied Biosystems®), 2 μl of IPC mix, 0.4 μl of IPC DNA (Applied Biosystems®) and 2 μl of the template DNA. Thermocycling was performed with a StepOne™ instrument (Applied Biosystems®) using the programmed parameters: 95 °C for 20 s (primary denaturation step) followed by 40 cycles of 95 °C for 1 s and final extension at 60 °C for 20 s. All runs included sample without target DNA as a negative control and also included sample with target S. Typhimurium template as positive control. Negative values or lack of amplification was considered for the cycle threshold (Ct) values of N 40. Baseline of PCR was automatically set by the system and the threshold was ranged between 0.034 and 0.103, therefore the threshold was manually set to 0.15 when analyzed.

h i ð−1=slopeÞ −1: PCR Efficiency ¼ 10

2.5. Standard ISO method and real-time PCR for Salmonella detection The optimized PCR–IMS , PCR methods and ISO 6579:2002 (ISO, 2002) are depicted in Fig. 1. For the ISO method, artificially inoculated raw duck wings (25 g) were homogenized with 225 ml of buffered peptone water (BPW, Oxoid) for 2 min using a paddle mixer and subsequently incubated for 20 h at 37 °C. After pre-enrichment, 0.1 and 1 ml of enrichment were transferred to 10 ml of Rappaport-Vassiliadis soya (RVS) broth (Oxoid) and Muller–Kauffmann tetrathionate–novobiocin (MKTTn) broth (Oxoid), and incubated for 24 h at 42 and 37 °C, respectively. After the selective enrichment, a loopful of each enriched sample was streaked on differential media: xylose lysine deoxycholate agar (XLD, Oxoid) and Hektoen enteric agar (HE, Oxoid). The presumptive Salmonella colonies from the selective agar were plated on nutrient agar (Oxoid), followed by biochemical confirmation using an API 20E strip kit (BioMérieux®, Inc., Marcy I'Etoile, France) and the results were interpreted by APIWEB software (BioMérieux®). For PCR–IMS and PCR, artificially inoculated raw duck wings were incubated in OB at 42 °C for 5, 7 or 14 h. Following the incubation period, 980 μl of enriched sample was incubated with immunomagnetic beads for 30 min for PCR–IMS, followed by DNA extraction and realtime PCR, while another 1-ml portion of enriched sample was directly used for the PCR method without IMS. Real-time PCR was performed as described previously. For all three methods, healthy and heat-injured cells of S. Typhimurium (50% and 85% injury) at 101 and 100 CFU/ml were inoculated on raw duck wings. Ten samples were analyzed for both healthy and heat-injured cells at each concentration. Meanwhile, the same numbers of naturally contaminated samples were prepared as controls (ISO, 2002).

2.4.3. Inclusivity and exclusivity of primers Target Salmonella strains (n = 16) and the non-Salmonella strains (n = 36) were tested for the inclusivity and exclusivity of primers. All bacterial strains were grown aerobically at 37 °C for 18 to 24 h in TSB and a 2 μl aliquot of DNA was added into the PCR tube and amplified as described above. 2.4.4. Detection probability and real-time PCR efficiency The probability of detecting Salmonella spp. in a suspension of known concentration with or without IMS was determined according to Malorny et al. (2003). After enrichment in OB as described above (ca. 108 CFU/ml), the enrichment broth was serially diluted in 0.1% (w/v) PW to obtain the final Salmonella cell concentration in the range from 102 to 108 CFU/ml. Before the extraction of DNA, the diluted suspension was spread plated on the modified BGA for enumeration. A 2 μl aliquot of DNA template of each dilution was added to separate PCR tube and run with IPC template as described above. The detection probability of PCR assays for different primers, with or without IMS, was obtained by plotting the relative number of positive PCRs observed against the cell concentrations. A sigmoidal curve fitting was performed

2.6. Validation To validate the optimized PCR–IMS method, a total of 60 naturally contaminated raw duck wing samples were purchased from the local markets in Singapore. PCR and PCR–IMS methods were performed and compared. Sensitivity, specificity, accuracy and Kappa index of both methods were calculated according to ISO (2003) and the results were compared with the standard ISO method (ISO, 2003).

Table 1 Taqman primers and probe pairs for the detection of Salmonella spp., sequences, expected amplicon sizes, target sites and references. Target genes Sal

invA

ttr

a b

Sequence (5′ → 3′) Primer Sal-F: GCGACTATCAGGTTACCGTGGA Primer Sal-R: AGTACGGCCTGCTTTTATCG Probe Sal-P: FAM-TAGCCAGCGAGGTGAAAACGACAAAGG-BHQ2 Primer invA-F: CAACGTTTCCTGCGGTACTGT Primer invA-R: CCCGAACGTGGCGATAATT Probe invA-P: FAM-CTCTTTCGTCTGGCATTATCGATCAGTACCA-BHQ2 Primer ttr-F: CTCACCAGGACATTACAACATGG Primer ttr-R: AGCTCAGACCAAAAGTGACCATC Probe ttr-P: FAM-CACCGACGGCGAGACCGACTTT-BHQ2

Data not available (NA). Salmonella Pathogenicity Islands (SPI1 and SPI).

Amplicon size

Primer target a

Reference

261

NA

Wang (2006)

116

invA invasion gene at SPI1b

González-Escalona et al. (2009)

ttrRSBCA locus near SPI2 at centisome 30.5

Malorny et al. (2004)

94

Q. Zheng et al. / International Journal of Food Microbiology 186 (2014) 6–13

ENRICHMENT

Standard ISO method

Rapid detection methods

BPW 37°C 24h

RVS 41.5°C 24h

OB 42°C

MKTTn 37°C 24h

XLD, HE 37°C 24h

DETECTION

5 – 24 h

9

in pure culture (about 85%) were significantly higher (P ≤ 0.05) than in inoculated raw duck wings (about 64%) under the same condition. Therefore, to reduce total time of IMS procedure, 30 min for bead incubation and 3 min for magnetic separation were applied for all further experiments. 3.2. Optimization of real-time PCR

Without IMS

With IMS

Nutrient Agar 37°C 24h

DNA template

API 20E Biochemical identification

Real-time PCR

Fig. 1. Flow diagrams of standard ISO, real-time PCR combined with immunomagnetic separation (IMS) and real-time PCR for detecting Salmonella Typhimurium on raw duck wings in this study.

2.7. Statistical analysis All mean values were obtained from independent triplicate replicates with duplicate samples (n = 6). Significant differences were statistically analyzed by the analysis of variance (ANOVA) test using IBM SPSS Statistical Software (IBM Corporation, Armonk, NY, USA) when comparing more than two values and ANOVA (unpaired t-test) using unpaired GraphPad software (GraphPad Software Inc., San Diego, CA, USA) when comparing between two values (Table 2). The difference was judged to be statistically significant when P ≤ 0.05. Fisher's exact probability tests (GraphPad Software Inc.) for 2 × 2 contingency tables were applied to compare PCR–IMS with PCR methods at P = 0.05 (Table 5). 3. Results 3.1. Optimization of IMS The results of capture efficiency (CE) of Dynabeads® for S. Typhimurium at a concentration of 106 CFU/ml in both pure culture and inoculated raw duck wings are shown in Table 2. There was no significant (P N 0.05) difference among CEs for different magnetic separation times (3, 10 and 30 min) and 3 min seemed to be adequate to fully separate the bead–bacteria complexes from broth. For bead incubation time, significantly (P ≤ 0.05) higher CEs were obtained for 30 min, however, further extension of bead incubation time did not show a positive influence on the proportion of separated cells from pure culture and raw duck wings. Additionally, the CEs for S. Typhimurium cells

3.2.1. Inclusivity and exclusivity The results of inclusivity and exclusivity tests of three primer sets (Sal, invA, and ttr) with 16 target Salmonella strains and 36 non-Salmonella strains were shown in Table 3. All primer sets gave positive response to 16 Salmonella strains (100% inclusivity), while invA produced detectable amplicon in Citrobacter freundii and Streptococcus thermophilus, showing a 94.4% exclusivity. Neither Sal nor ttr produced positive response to non-Salmonella bacteria strains, showing 100% exclusivity. 3.2.2. Real-time PCR efficiency The real-time PCR efficiencies were calculated based on standard curves of 10-fold serially diluted pure cultures of S. Typhimurium (Table 4). A strong linear relationship (R2 ≥ 0.99) between Ct values and bacterial concentrations (103–108 CFU/ml) were observed for Sal and invA and 104–108 CFU/ml for ttr. The highest real-time PCR efficiency was obtained for Sal, which was 94.1%, followed by ttr (86.3%) and invA (84.3%) primers, which correspond to the slopes of −3.47, −3.84 and −3.77, respectively. The calculated standard deviation values (sr) (Malorny et al., 2004) expressed in percentage, which describe assay precision, were in a range of 0.8–4.9%, 1.2–4.2% and 0.6–3.7% of measured mean Ct values, for Sal, invA and ttr primers, respectively, indicating a high precision of the assay. 3.2.3. Detection probability with and without IMS Subsequently, the probability of detecting S. Typhimurium in raw duck wings with Sal, invA and ttr primers was tested at different cell concentration levels (100–107 CFU/ml) with or without the IMS process (Fig. 2). Without IMS, the detection probability of S. Typhimurium was 100% in the reactions containing 104 CFU/ml using Sal and invA and at the level of 105 CFU/ml with ttr (Fig. 2a). The levels of detection at 100% probability were determined to be 103 CFU/ml for Sal and invA, and 104 CFU/ml for ttr when IMS was applied (Fig. 2b). The reactions containing 102 CFU/ml resulted in a detection probability of 33.3% in Sal and invA without IMS, while a more than two times increase in detection probabilities was observed both for Sal (83.3%) and invA (66.7%) when cells were captured by Dynabeads®. Therefore, these results indicated that the IMS step could improve the detection probability at the level range of 102–103 CFU/ml. 3.2.4. Limit of detection (LOD) In the standard curves constructed from pure Salmonella cultures, the slopes of the standard curves for Sal and invA were linear down to

Table 2 Comparison of capture efficienciesa (CE, %) of immunomagnetic separation on Salmonella Typhimurium (ca. 106 CFU/ml) from pure culture and artificially inoculated raw duck wings for different bead incubation and magnetic separation times. Matrix

CE (%) Magnetic separation time

Bead incubation time 10 min

Pure culture

Raw duck wings

3 min 10 min 30 min 3 min 10 min 30 min

39.3 39.5 38.7 11.7 11.7 12.2

± ± ± ± ± ±

30 min 2.27a,x 1.88a,x 1.90a,x 0.51a,y 0.49a,y 1.49a,y

85.0 84.0 85.4 64.5 63.1 64.8

± ± ± ± ± ±

60 min 2.55b,x 2.37b,x 1.79b,x 3.11b,y 4.10b,y 2.52b,y

86.0 88.6 86.6 65.2 65.1 64.2

± ± ± ± ± ±

1.19b,x 2.44b,x 2.08b,x 2.60b,y 3.91b,y 2.26b,y

a All values are expressed as mean CE ± standard deviation (n = 6). Different letters (a, b) and (x, y) within the same row and column indicated significant (P ≤ 0.05) differences, respectively.

10

Q. Zheng et al. / International Journal of Food Microbiology 186 (2014) 6–13

Table 3 Exclusivity and inclusivity of real-time PCR tests using invA, ttr and Sal primers for non-Salmonella and Salmonella strains.

Non-Salmonella strains

Salmonella strains

Bacterial strains

ATCC number/ID no.

Source

Sal

invA

ttr

IPC

Bacillus cereus Citrobacter freundii Cronobacter muytjensii C. sakazakii C. sakazakii EB7 Enterococcus faecalis Escherichia coli E. coli E. coli E. coli O26:H11 E. coli O45:H2 E. coli O103:H11 E. coli O111:H8 E. coli O121:H19 E. coli O127:H6 E. coli O157:H7 E. coli serotype O55:b5:hE. coli 16 serotype: O157:H7 E. coli 22 serotype: O157:H7 E. coli O157:H7 Enterococcus faecalis Klebsiella pneumoniae Listeria monocytogenes 1/2a L. monocytogenes 1/2b L. monocytogenes 1 L. monocytogenes 4b L. innocua Shigella boydii S. flexneri S. sonnei Staphylococcus aureus S. aureus S. aureus Streptococcus thermophilus Vibrio parahaemolyticus V. vulnificus Salmonella Agona S. Enteritidis S. Enteritidis 109 serotype S. Enteritidis 124 serotype: 8 S. Enteritidis 125 serotype: 13A S. Enteritidis 130 serotype: 2 S. Gaminara S. Heidelberg S. Montevideo S. Newport S. Saintpaul S. Tennessee S. Typhimurium S. Typhimurium S. Typhimurium S. Typhimurium

14579 43836 51329 29544 6962 29212 25922 35218 BL21/pFPV CDC 0303014 CDC 00-3039 CDC 06-3008 CDC 2010C-3114 CDC 02-3211 E2348/69 35150 12014 C7927 F12 EDL933 29212 10031 BAA-679 BAA-839 19111 13932 33090 9207 12022 29031 6538 33591 33592 19258 17802 27562 BAA-707 13076 E-4 SE-5 5575, 164 50036, 203 D1286, 271 BAA-711 8326 BAA-710 6962 9712 10722 51812 25241 29629 13311

– – – Child's throat Isolated from food Urine Clinical isolate Canine Mutant strain from Dr. Walczyk – – – – – – Feces, human – Apple cider Seattle CDC – – Tissue, animal (rabbit) Clinical specimen Poultry Spinal fluid of child with meningitis Cow brain – – Human feces Human lesion – Blood Pasteurized milk Shirasu food poisoning, Japan Human blood Plant – Peter Holt MD Department of Health USDA CDC, clinical isolate NY Orange juice – Clinical isolate from tomato outbreak Food poisoning fatality, England Cystitis, Panama – Human blood Methionine auxotroph derived from strain LT-2 – Feces, human

− − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − + + + + + + + + + + + + + + + +

− + − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − + − − + + + + + + + + + + + + + + + +

− − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − − + + + + + + + + + + + + + + + +

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

Table 4 Comparison of Sal, invA and ttr primers on 10-fold serially diluted Salmonella Typhimurium from pure culture using real-time PCR. Salmonella culture (CFU/ml) 108 107 106 105 104 103 102 Linear regression, equation R2 PCR efficiency (%) a b c

Mean Ct values ± standard deviation sra Sal

invA

ttr

16.7 ± 0.83 21.5 ± 0.58 24.2 ± 0.43 28.1 ± 0.64 31.1 ± 0.25 34.5 ± 1.04 34.4 ± 0.09b y = −3.47x + 41.6 0.9953 94.12

17.0 ± 0.38 21.0 ± 0.64 25.0 ± 0.31 28.9 ± 0.64 32.6 ± 0.56 35.6 ± 1.51 36.4 ± 0.02b y = −3.77x + 43.6 0.9978 84.26

19.1 ± 0.60 23.4 ± 0.34 27.1 ± 0.18 31.2 ± 0.70 34.5 ± 0.62 37.3 ± 0.33b NDc y = −3.84x + 46.2 0.9984 86.29

Results expressed as mean Ct and repeatability standard deviation sr (n = 6). Two out of six PCR reactions were negative (Ct N 40). These were not included in the calculation. ND: not detectable.

Q. Zheng et al. / International Journal of Food Microbiology 186 (2014) 6–13

11

Table 5 Comparisona of real-time PCR combined with immunomagnetic separation (PCR–IMS) and real-time PCR (PCR) methods on the detection of healthy and heat injured Salmonella Typhimurium on raw duck wings at low inoculum levels (100–101 CFU/25 g) after 5, 7 and 14 h enrichment in one broth at 42 °C. All inoculated duck wing samples were confirmed as Salmonella positive by ISO method. Inoculum level (CFU/25 g)

Healthy

100 101

50% injured

Subtotal 100 101

85% injured

Subtotal 100 101 Subtotal Total

Enrichment time 5h

7h

14 h

PCR

PCR–IMS

PCR

PCR–IMS

PCR

PCR–IMS

NDb (0%) 2/10 (20%) 2/20a ND (0%) ND (0%) NDa ND (0%) ND (0%) NDa 2/60a

6/10 (60%) 7/10 (70%) 13/20b ND (0%) 2/10 (20%) 2/20ab ND (0%) ND (0%) NDa 15/60b

4/10 (40%) 9/10 (90%) 13/20b 4/10 (40%) 4/10 (40%) 8/20bc ND (0%) ND (0%) NDa 21/60b

10/10 (100%) 10/10 (100%) 20/20c 5/10 (50%) 8/10 (80%) 13/20c ND (0%) ND (0%) NDa 33/60c

10/10 (100%) 10/10 (100%) 20/20c 10/10 (100%) 10/10 (100%) 20/20d 10/10 (100%) 10/10 (100%) 20/20b 60/60d

10/10 (100%) 10/10 (100%) 20/20c 10/10 (100%) 10/10 (100%) 20/20d 10/10 (100%) 10/10 (100%) 20/20b 60/60d

a Different letters (a–d) within the same row indicated significant (P ≤ 0.05) differences. b ND: not detectable.

3.4. Validation

Fig. 2. Detection probability of Salmonella Typhimurium on raw duck wings using Sal, invA and ttr primers by real-time PCR without (a) and with (b) immunomagnetic separation (IMS).

103 CFU/ml, with 60% (b95%) probability at 102 CFU/ml, indicating that the LOD was 103 CFU/ml in pure Salmonella culture (Table 4). Higher LOD (104 CFU/ml) was observed for ttr primer (Table 4). The LOD on raw duck wing samples using real-time PCR was 104 CFU/ml for Sal and invA and at the level of 105 CFU/ml with ttr (Fig. 2a), showing that the LOD of Sal and invA primers was l log unit lower than that of ttr primer. When the real-time PCR method was combined with IMS, the LODs decreased to 103 CFU/ml for Sal and invA, and 104 CFU/ml for ttr (Fig. 2b). 3.3. Comparison of the optimized real-time PCR with IMS and conventional real-time PCR Raw duck wings were inoculated with S. Typhimurium, with healthy and heat-injured cells at low inoculum levels (101 and 100 CFU/25 g). Regardless of the physical state of the cell and inoculum level, the sensitivity of real-time PCR for the detection of S. Typhimurium increased with IMS and the extended enrichment time (Table 5). In total, after 5 h of OB-enrichment, S. Typhimurium detection with PCR– IMS (15/60) was significantly (P = 0.0011) improved compared to PCR (2/60). The same trend was observed after 7-h of enrichment, where a significant difference (P = 0.0431) between the results of the PCR–IMS and PCR methods was noted. When healthy cells were inoculated, after 7-h enrichment, 100% sensitivity of real-time PCR was achieved when IMS was applied. For 50% heat-injured cells, significant (P ≤ 0.05) differences were observed between different enrichment times, whereas there was no significant (P N 0.05) difference between the two methods with the same enrichment time. Both healthy and heat-injured cells at low inoculum levels were fully detected by the PCR–IMS and PCR methods with 14-h enrichment.

To validate the optimized PCR–IMS method in this study, a total of 60 naturally contaminated raw duck wing samples were analyzed with the ISO method (ISO, 2002), PCR and PCR–IMS for the detection of Salmonella spp. after 12, 14 and 24 h enrichment (Table 6). A total of 6 positive samples were identified by the standard ISO method. For the real-time PCR methods, after 12-h enrichment, 3 and 5 positive samples were detected by PCR and PCR–IMS, respectively. After 14-h enrichment, comparable results for both methods were obtained and 6 positive results were observed showing 100% accuracy, however, after 24-h enrichment, 1 false-positive result was obtained by both real-time PCR methods, lowering accuracy, specificity and Kappa index. 4. Discussion In a previous study, OB has shown to be better at recovering both healthy and heat-injured S. Typhimurium cells compared to other commercial enrichment broths (Zheng et al., 2013). Based on these results, OB was selected in this study as the pre-enrichment broth for the optimized PCR–IMS and PCR methods. The present results showed that 30 min for bead incubation and 3 min for magnetic separation were sufficient to capture Salmonella cells from both the pure culture and the raw duck wings with the IMS process. Similarly, Steingroewer et al. (2001) and Tatavarthy et al. (2009) reported that 30 min bead incubation and 3 min magnetic separation times were the most effective in isolation of Salmonella cells from pure suspensions (CE ~80%). Higher CEs were also reported for pure S. Typhimurium cultures by magnetic nano-beads (Wang et al., 2011) after a 30 min bead incubation time. This study also showed that the CE for S. Typhimurium cells from the broth with raw duck wings was approximately 20% lower than from pure culture. This is consistent with the findings of Fu et al. (2005) which showed that over 40% difference in CE values was observed for E. coli O157:H7 detection with IMS from PBS suspension comparing to ground beef suspension at the inoculum level of 104 CFU/ml. It was suggested that fat particulates or proteins might interfere with the antibody or the bacterial surface, causing the blockage and the formation of bead–bacteria complexes (Fu et al., 2005). Another possibility is that high background microflora might decrease the efficiency of IMS by

12

Q. Zheng et al. / International Journal of Food Microbiology 186 (2014) 6–13

Table 6 Validation of real-time PCR combined with immunomagnetic separation (PCR–IMS) and real-time PCR (PCR) for the detection of Salmonella Typhimurium on raw duck wing samples in one broth at 42 °C for 12, 14 and 24 h. Detection method

PCR PCR–IMS PCR PCR–IMS PCR PCR–IMS

Enrichment time

12 h 14 h 24 h

No. of samples

60 60 60 60 60 60

ISO method

Rapid method

Positive

Negative

False positive

False negative

6 6 6 6 6 6

54 54 54 54 54 54

0 0 0 0 1 1

3 1 0 0 0 0

Sensitivity (%)

Specificity (%)

Accuracy (%)

Kappa indexa

50.0 83.3 100 100 100 100

100 100 100 100 98.1 98.1

95.0 98.3 100 100 98.3 98.3

0.67 0.90 1 1 0.91 0.91

a Kappa values of b0.01 indicate no concordance, those between 0.1 and 0.4 indicate weak concordance, those between 0.41 and 0.60 indicate clear concordance, those between 0.61 and 0.80 indicate strong concordance, and those between 0.81 and 1.00 indicate nearly complete concordance.

cross-reactivity and nonspecific binding (Duodu et al., 2009; Fu et al., 2005; Tatavarthy et al., 2009). Except for the sample preparation, the optimization of the downstream detection method is critical to conduct accurate and rapid detection. It was reported that in primers used for different specific target genes amplification might contribute to PCR detection sensitivity (Duodu et al., 2009). For this reason, three major primers (Sal, invA and ttr) were compared to select an ideal primer for the rapid detection of Salmonella cells on raw duck wings using a real-time PCR detection system. The inclusivity of the PCR assay is an ability to detect the target bacteria from the wider group of bacterial strains within the same species, while exclusivity is defined as a lack of interference with nontarget bacteria (Malorny et al., 2003). In this study, 100% inclusivity and exclusivity was achieved for all primers, which was in the agreement with the previous studies that a high inclusivity was obtained in Sal (Wang, 2006), invA (González-Escalona et al., 2009; Mercanoglu and Griffiths, 2005; Malorny et al., 2007; Zhang et al., 2011) and ttr (Malorny et al., 2004). However, Rahn et al. (1992) reported that DNA for S. Litchfield and S. Senftenberg was not amplified by invA, indicating the lack of inclusivity. Furthermore, the invA gene shares high homology with Shigella spp. and Citrobacter spp. (Hu et al., 2011), which are common contaminants of raw duck meat and it has been reported that the presence of Citrobacter spp. in the sample could give false positive results in Salmonella spp. detection using real-time PCR (Margot et al., 2013). In this study, stable signal was observed for C. freundii and S. thermophilus using invA. The PCR efficiency under the optimum condition should be 90– 100%, which corresponds to a slope of − 3.3 to − 3.6 in the Ct vs log CFU/ml standard curve (Duodu et al., 2009) and in this study, it was achieved only for Sal primer. Based on these observations, Sal primer was chosen for further experiment. The LOD of Sal primer in raw duck wing samples was 103 CFU/ml, which was higher compared to the LOD of 102 CFU/ml achieved by Malorny et al. (2004) for Salmonella detection in the presence of different concentrations of background flora (from 104 to 108 CFU/ml) and previously reported by Kawasaki et al. (2010) for pure Salmonella culture. Comparable LOD (103 CFU/ml) was obtained for Salmonella (Hyeon et al., 2010) and Cronobacter (Hyeon et al., 2010; Wang et al., 2012) in the artificially inoculated infant powder, whereas a higher LOD of ≥ 105 CFU/ml in various leafy green produce was obtained by González-Escalona et al. (2012). It is known that the application of the IMS technique can decrease LOD by concentrating target bacteria (Shields et al., 2012; Yang et al., 2007). These results also showed that the LODs of PCR–IMS were l log unit lower than PCR alone. Similar results were obtained by Wang (2006) who reported that the LOD for Salmonella detection in beef with Sal primer decreased from 105 to 102 CFU/g after IMS application. Although several investigators have evaluated a real-time PCR method in combination with IMS for the rapid detection of Salmonella (Mercanoglu and Griffiths, 2005; Warren et al., 2007; Weagant et al., 2011), there was no study to determine the optimal enrichment time for these methods. In this study, a 7-h OB-enrichment at 42 °C was sufficient to detect healthy S. Typhimurium cells at low inoculum levels

(100–101 CFU/ml) by the PCR–IMS method, which was significantly better than PCR alone. Similarly, Weagant et al. (2011) reported that healthy E. coli cells (0.1–0.3 CFU/g) were undetectable in artificially inoculated alfalfa sprouts but they gave a fluorescence signal after capture with Dynabeads® after a 5-h enrichment time. Another study showed that a 10-h enrichment was sufficient in recovering Salmonella cells at 1.5 and 25 CFU/25 g in alfalfa sprout and ground beef samples, respectively, to achieve a detectable level by the PCR–IMS method (Mercanoglu and Griffiths, 2005). Warren et al. (2007) suggested that a 5-h enrichment was needed to detect Salmonella cocktail inoculated in tomatoes, potato salad and ground beef by real-time PCR combined with flow-through immunocapture, which is a circulating IMS system. Unlike the detection of healthy cells, a longer enrichment time was needed to detect heat-injured cells on raw duck wings, revealing that there was no significant (P N 0.05) difference between PCR–IMS and PCR. In general, shorter enrichment times (6–8 h) and higher temperature of incubation (42 °C) were recommend previously for food matrices with high background microflora, while longer times (18–24 h) and lower temperature (37 °C) were reported as more sufficient for recovery of sub-lethally injured pathogens (Fedio et al., 2011). Similar to our results, Malorny et al. (2004) indicated that approximately 20-h was sufficient for the recovery of sub-lethally injured Salmonella cells to detectable levels using real-time PCR method. In the validation study, the optimized PCR–IMS method had higher sensitivity, specificity, accuracy and Kappa index than the PCR method after a 12-h enrichment time. However, no difference was observed between PCR–IMS and PCR after 14-h enrichment. Notably, one more positive reaction occurred in the two real-time PCR methods than in the ISO method after 24-h enrichment. Similarly, Mokhtari et al. (2013) showed that nine food samples (8.8%) were positive for Shigella spp. with realtime PCR, while all negative results were obtained from the standard culture methods. Another study reported that two pork samples were found positive for Cronobacter spp. by real-time PCR while only one was detected by the ISO standard method (Wang et al., 2012). The European Food Safety Authority also reported that monitoring by traditional serotyping methods could result in a high underestimation of the real-prevalence of S. Enteritidis in poultry and poultry products based on a study on the prevalence of Salmonella in egg-laying flocks (EFSA, 2006). These studies indicate that the real-time PCR method was more sensitive than the conventional culture method for detecting Salmonella in food, especially in poultry, meat, and poultry-related products (Almeida et al., 2013). 5. Conclusion This study was the first attempt to optimize the real-time PCR method in combination with IMS for the detection of healthy and heatinjured S. Typhimurium cells on raw duck wings with different enrichment times. Although no significant difference was observed between the PCR–IMS and PCR methods with a longer enrichment time, the PCR–IMS method did shorten the enrichment time (7 h) when healthy Salmonella cells were detected in raw duck wings and increase the

Q. Zheng et al. / International Journal of Food Microbiology 186 (2014) 6–13

sensitivity under the same time enrichment. Thus, this study demonstrates the potential for combining real-time PCR with IMS to develop a reliable alternative rapid method for detecting S. Typhimurium on raw duck wings with a short enrichment time, while a longer enrichment time should be needed for the recovery of injured cells to obtain reliable detection. Acknowledgments This research was funded by the Ministry of Education - Singapore Academic Research Fund (AcRF) Tier 1 (R-143-000-451-112). The authors thank Dr. Keith Schneider of the University of Florida for the thoughtful review of this manuscript and Dr. Kunho Seo of Konkuk University for kindly providing S. Enteritidis and E. coli O157:H7 strains. The authors also thank DuPont Company (Singapore) for kindly providing enrichment broths. References Adzitey, F., Rusul, G., Huda, N., 2011. Prevalence and antibiotic resistance of Salmonella serovars in ducks, duck rearing and processing environments in Penang, Malaysia. Food Res. Int. 45, 947–952. Almeida, C., Cerqueira, L., Azevedo, N.F., Vieira, M.J., 2013. Detection of Salmonella enterica serovar Enteritidis using real time PCR, immunocapture assay, PNA FISH and standard culture methods in different types of food samples. Int. J. Food Microbiol. 161, 16–22. Anonymous, 2012. Global Poultry Trends 2012 — Asia, China dominate global duck and goose meat production. The PoultrySite. England. http://www.thepoultrysite.com/articles/2644/global-poultry-trends-2012-asia-china-dominate-global-duck-andgoose-meat-production. CDC, 2012a. Salmonella. Centers for Disease Control and Prevention, Atlanta, GA, (http:// www.cdc.gov/salmonella/). CDC, 2012b. Salmonellosis. Centers for Disease Control and Prevention, Atlanta, GA, (http://www.cdc.gov/nczved/divisions/dfbmd/diseases/salmonellosis/#what). CDC, 2013. Multistate outbreak of human Salmonella Typhimurium infections linked to live poultry in backyard flocks. Centers for Disease Control and Prevention, Atlanta, GA, (http://www.cdc.gov/salmonella/typhimurium-live-poultry-04-13/index.html). Duodu, S., Mehmeti, I., Holst-Jensen, A., Loncarevic, S., 2009. Improved sample preparation for real-time PCR detection of Listeria monocytogenes in hot-smoked salmon using filtering and immunomagnetic separation techniques. Food Anal. Methods 2, 23–29. European Food Safety Authority (EFSA), 2006. Large variation in prevalence of Salmonella in laying hen flocks in EU: EFSA preliminary report. Euro Surveill. 11 (24) (pii = 2975. http://www.eurosurveillance.org/ViewArticle.aspx?ArticleId=2975). European Food Safety Authority (EFSA), 2013. The European union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2011. EFSA J.. http://dx.doi.org/10.2903/j.efsa.2013.3129. Fedio, W.M., Jinneman, K.C., Yoshitomi, K.J., Zapata, R., Wendakoon, C.N., Browning, P., Weagant, S.D., 2011. Detection of E. coli O157:H7 in raw ground beef by Pathatrix™ immunomagnetic-separation, real-time PCR and cultural methods. Int. J. Food Microbiol. 148, 87–92. Fu, Z., Rogelj, S., Kieft, T.L., 2005. Rapid detection of Escherichia coli O157:H7 by immunomagnetic separation and real-time PCR. Int. J. Food Microbiol. 99, 47–57. González-Escalona, N., Hammack, T.S., Russell, M., Jacobson, A.P., De Jesus, A.J., Brown, E. W., Lampel, K.A., 2009. Detection of live Salmonella spp. cells in produce by a TaqMan-based quantitative reverse transcriptase real-time PCR targeting invA mRNA. Appl. Environ. Microbiol. 75, 3714–3720. Gorski, L., Liang, A.S., 2010. Effect of enrichment medium on real-time detection of Salmonella enterica from lettuce and tomato enrichment cultures. J. Food Prot. 73, 1047–1056. Hu, Q., Tu, J., Han, X., Zhu, Y., Ding, C., Yu, S., 2011. Development of multiplex PCR assay for rapid detection of Riemerella anatipestifer, Escherichia coli, and Salmonella enterica simultaneously from ducks. J. Microbiol. Methods 87, 64–69. Hyeon, J.Y., Park, C., Choi, I.S., Holt, P.S., Seo, K.H., 2010. Development of multiplex realtime PCR with internal amplification control for simultaneous detection of Salmonella and Cronobacter in powdered infant formula. Int. J. Food Microbiol. 144, 177–181. ISO, 2002. International Organization for Standardization 6579:2002(E). Microbiology of Food and Animal Feeding Stuffs — Horizontal Method for the Detection of Salmonella spp.,. ISO, 2003. International Organization for Standardization 16140:2003. Microbiology of Food and Animal Feeding Stuffs — Protocol for the Validation of Alternative Methods,. Jasson, V., Baert, L., Uyttendaele, M., 2011. Detection of low numbers of healthy and sublethally injured Salmonella enterica in chocolate. Int. J. Food Microbiol. 145, 488–491.

13

Kawasaki, S., Fratamico, P.M., Horikoshi, N., Okada, Y., Takeshita, K., Sameshima, T., Kawamoto, S., 2010. Multiplex real-time polymerase chain reaction assay for simultaneous detection and quantification of Salmonella species, Listeria monocytogenes, and Escherichia coli O157:H7 in ground pork samples. Foodborne Pathog. Dis. 7, 549–554. Lee, H.J., Kim, B.C., Kim, K.W., Kim, Y.K., Kim, J., Oh, M.K., 2009. A sensitive method to detect Escherichia coli based on immunomagnetic separation and real-time PCR amplification of aptamers. Biosens. Bioelectron. 24, 3550–3555. Li, A., Zhang, H., Zhang, X., Wang, Q., Tian, J., Li, Y., Li, J., 2010. Rapid separation and immunoassay for low levels of Salmonella in foods using magnetosome–antibody complex and real-time fluorescence quantitative PCR. J. Sep. Sci. 33, 3437–3443. Malorny, B., Hoorfar, J., Bunge, C., Helmuth, R., 2003. Multicenter validation of the analytical accuracy of Salmonella PCR: towards an international standard. Appl. Environ. Microbiol. 69, 290–296. Malorny, B., Paccassoni, E., Fach, P., Bunge, C., Martin, A., Helmuth, R., 2004. Diagnostic real-time PCR for detection of Salmonella in food. Appl. Environ. Microbiol. 70, 7046–7052. Malorny, B., Bunge, C., Helmuth, R., 2007. A real-time PCR for the detection of Salmonella Enteritidis in poultry meat and consumption eggs. J. Microbiol. Methods 70, 245–251. Margot, H., Stephan, R., Guarino, S., Jagadeesan, B., Chilton, D., O'Mahony, E., Iversen, C., 2013. Inclusivity, exclusivity and limit of detection of commercially available realtime PCR assays for the detection of Salmonella. Int. J. Food Microbiol. 165, 221–226. Mercanoglu, B., Griffiths, M.W., 2005. Combination of immunomagnetic separation with real-time PCR for rapid detection of Salmonella in milk, ground beef, and alfalfa sprouts. J. Food Prot. 68, 557–561. Mokhtari, W., Nsaibia, S., Gharbi, A., Aouni, M., 2013. Real-time PCR using SYBR green for the detection of Shigella spp. in food and stool samples. Mol. Cell. Probes 27, 53–59. Rahn, K., De Grandis, S.A., Clarke, R.C., McEwen, S.A., Galan, J.E., Ginocchio, C., Curtiss III, R., Gyles, C.L., 1992. Amplification of an invA gene sequence of Salmonella Typhimurium by polymerase chain reaction as a specific method of detection of Salmonella. Mol. Cell. Probes 6, 271–279. Shields, M.J., Hahn, K.R., Janzen, T.W., Goji, N., Thomas, M.C., Kingombe, C.B., Paquet, C., Kell, A.J., Amoako, K.K., 2012. Immunomagnetic capture of Bacillus anthracis spores from food. J. Food Prot. 75, 1243–1248. Steingroewer, J., Knaus, H., Bley, T., Boschke, E., 2001. A rapid method for the preenrichment and detection of Salmonella Typhimurium by immunomagnetic separation and subsequent fluorescence microscopical techniques. Eng. Life Sci. 5, 267–272. Tatavarthy, A., Peak, K., Veguilla, W., Cutting, T., Harwood, V.J., Roberts, J., Amuso, P., Cattani, J., Cannons, A., 2009. An accelerated method for isolation of Salmonella enterica serotype Typhimurium from artificially contaminated foods, using a short preenrichment, immunomagnetic separation, and xylose–lysine–desoxycholate agar (6IX method). J. Food Prot. 72, 583–590. Uyttendaele, M., Rajkovic, A., Van Houteghem, N., Boon, N., Thas, O., Debevere, J., Devlieghere, F., 2008. Multi-method approach indicates no presence of sub-lethally injured Listeria monocytogenes cells after mild heat treatment. Int. J. Food Microbiol. 123, 262–268. Wang, L.X., 2006. Simultaneous Quantitation of Escherichia coli O157:H7, Salmonella and Shigella in Ground Beef by Multiplex Real-time PCR and Immunomagnetic Separation. University of Missouri-Columbia, USA, (Master thesis). Wang, L., Li, Y., Mustapha, A., 2007. Rapid and simultaneous quantitation of Escherichia coli O157:H7, Salmonella, and Shigella in ground beef by multiplex real-time PCR and immunomagnetic separation. J. Food Prot. 70, 1366–1372. Wang, H., Li, Y., Wang, A., Slavik, M., 2011. Rapid, sensitive and simultaneous detection of three foodborne pathogens using magnetic nanobead-based immunoseparation and quantum dot-based multiplex immunoassay. J. Food Prot. 74, 2039–2047. Wang, X., Zhu, C., Xu, X., Zhou, G., 2012. Real-time PCR with internal amplification control for the detection of Cronobacter spp. (Enterobacter sakazakii) in food samples. Food Control 25, 144–149. Warren, B.R., Yuk, H.G., Schneider, K.R., 2007. Detection of Salmonella by flow-through immunocapture real-time PCR in selected foods within 8 hours. J. Food Prot. 70, 1002–1006. Weagant, S.D., Jinneman, K.C., Yoshitomi, K.J., Zapata, R., Fedio, W.M., 2011. Optimization and evaluation of a modified enrichment procedure combined with immunomagnetic separation for detection of E. coli O157:H7 from artificially contaminated alfalfa sprouts. Int. J. Food Microbiol. 149, 209–217. Yang, H., Qu, L., Wimbrow, A.N., Jiang, X., Sun, Y., 2007. Rapid detection of Listeria monocytogenes by nanoparticle-based immunomagnetic separation and real-time PCR. Int. J. Food Microbiol. 118, 132–138. Zhang, G., Brown, E.W., González-Escalona, N., 2011. Comparison of real-time PCR, reverse transcriptase real-time PCR, loop-mediated isothermal amplification, and the FDA conventional microbiological method for the detection of Salmonella spp. in produce. Appl. Environ. Microbiol. 77, 6495–6501. Zheng, Q.W., Bustandi, C., Yang, Y.S., Schneider, K., Yuk, H.G., 2013. Comparison of enrichment broths for the recovery of healthy and heat-injured Salmonella Typhimurium on raw duck wings. J. Food Prot. 76, 1963–1968. Zhou, B., Xiao, J., Liu, S., Yang, J., Wang, Y., Nie, F., Zhou, Q., Li, Y., Zhao, G., 2013. Simultaneous detection of six food-borne pathogens by multiplex PCR with a GeXP analyzer. Food Control 32, 198–204.