FOOD-06425; No of Pages 5 International Journal of Food Microbiology xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro
European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat Elisabetta Delibato a,⁎, David Rodriguez Lazaro b,c, Monica Gianfranceschi a, Alessandra De Cesare d, Damiano Comin e, Antonietta Gattuso a, Marta Hernandez b, Michele Sonnessa a, Frédérique Pasquali d, Zuzsanna Sreter-Lancz f, María-José Saiz-Abajo g, Javier Pérez-De-Juan g, Javier Butrón g, Estella Prukner-Radovcic h, Danijela Horvatek Tomic h, Gro S. Johannessen i, Džiuginta Jakočiūnė j, John E. Olsen j, Marianne Chemaly k, Francoise Le Gall k, Patricia González-García b, Antonia Anna Lettini e, Maja Lukac h, Segolénè Quesne k, Claudia Zampieron e, Paola De Santis l, Sarah Lovari l, Barbara Bertasi m, Enrico Pavoni m, Yolande T.R. Proroga n, Federico Capuano n, Gerardo Manfreda d, Dario De Medici a a
Department of Veterinary Public Health and Food Safety, Istituto Superiore di Sanità Viale Regina Elena, 299 00161 Rome, Italy Instituto Tecnológico Agrario de Castilla y León (ITACyL), Valladolid, Spain c Microbiology Section, Faculty of Sciences, University of Burgos, Burgos, Spain d Alma Mater Studiorum — Università di Bologna, Dip. Scienze e Tecnologie Agro-Alimentari, Ozzano dell'Emilia, BO, Italy e Istituto Zooprofilattico Sperimentale delle Venezie, Dept. Food Safety, Legnaro (PD), Italy f National Food Chain Safety Office, Food and Feed Safety Directorate, Food Microbiological National Reference Laboratory, Mester, Budapest, Hungary g CNTA, Centro Nacional de Tecnología y Seguridad Alimentaria, San Adrian, Navarra, Spain h University of Zagreb, Faculty of Veterinary Medicine, Dept. of Poultry Diseases with Clinic, Zagreb, Croatia i Section for Bacteriology — Food and GMO Norwegian Veterinary Institute, Oslo, Norway j Department of Veterinary Disease Biology, University of Copenhagen, Denmark k Anses, Laboratoire de Ploufragan-Plouzané, Unité Hygiène et Qualité des Produits Avicoles et Porcins, Ploufragan, France l Direzione Operativa Controllo degli Alimenti, Istituto Zooprofilattico Sperimentale del Lazio e Toscana, Rome, Italy m Istituto Zooprofilattico Sperimentale della Lombardia e dell'Emilia Romagna, Brescia, Italy n Istituto Zooprofilattico del Sperimentale Mezzogiorno Portici, NA, Italy b
a r t i c l e
i n f o
Available online xxxx Keywords: Validation Salmonella Real Time PCR ISO 6579 Pork meat Artificial contamination
a b s t r a c t The classical microbiological method for detection of Salmonella spp. requires more than five days for final confirmation, and consequently there is a need for an alternative methodology for detection of this pathogen particularly in those food categories with a short shelf-life. This study presents an international (at European level) ISO 16140-based validation study of a non-proprietary Real-Time PCR-based method that can generate final results the day following sample analysis. It is based on an ISO compatible enrichment coupled to an easy and inexpensive DNA extraction and a consolidated Real-Time PCR assay. Thirteen laboratories from seven European Countries participated to this trial, and pork meat was selected as food model. The limit of detection observed was down to 10 CFU per 25 g of sample, showing excellent concordance and accordance values between samples and laboratories (100%). In addition, excellent values were obtained for relative accuracy, specificity and sensitivity (100%) when the results obtained for the Real-Time PCR-based methods were compared to those of the ISO 6579:2002 standard method. The results of this international trial demonstrate that the evaluated Real-Time PCR-based method represents an excellent alternative to the ISO standard. In fact, it shows an equal and solid performance as well as it reduces dramatically the extent of the analytical process, and can be easily implemented routinely by the Competent Authorities and Food Industry laboratories. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Pork meat and derived products are the main sources of Salmonella infection in humans in Europe; a recent study estimated that more ⁎ Corresponding author at: Department of Food Safety and Veterinary Public Health, Istituto Superiore di Sanità, Viale Regina Elena, 299-00161 Rome, Italy. E-mail address: [email protected] (E. Delibato).
than 56% of the human salmonellosis cases could be attributable to pigs, while the contributions of total reservoirs associated with laying hens (eggs), broilers and turkeys were 17.0, 10.6 and 2.6%, respectively (EFSA, 2013). As a consequence, some EU countries have established Salmonella surveillance and control programs for pork production (Arguello et al., 2013; Berriman et al., 2013). According to the European microbiological food safety criteria Salmonella spp. must be absent in an established amount depending on the kind of the product
0168-1605/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
2
E. Delibato et al. / International Journal of Food Microbiology xxx (2014) xxx–xxx
Table 1 Interlaboratory results (mean ± relative standard error) from pork meat samples artificially contaminated with 10 CFU Salmonella spp. (low). Laboratory
1 2 3 4 5 6 7 8 9 10 11 12 13
Sample A
Sample B
Sample C
Sample D
Sample E
Sample F
Sample G
Sample H
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
+ + + + + + + + + + + + +
16.14 ± 0.13 14.25 ± 1.57 18.01 ± 0.53 21.67 ± 0.58 18.47 ± 0.14 16.13 ± 0.84 23.33 ± 0.53 15.58 ± 0.23 20.70 ± 0.20 13.16 ± 0.51 13.96 ± 0.65 19.18 ± 0.54 15.77 ± 0.65
+ + + + + + + + + + + + +
19.68 ± 0.18 15.42 ± 0.54 23.31 ± 1.64 21.62 ± 0.58 16.22 ± 0.21 17.81 ± 1.26 23.82 ± 0.66 15.52 ± 0.43 18.40 ± 0.16 12.56 ± 0.69 14.49 ± 0.50 22.58 ± 0.40 17.91 ± 0.50
+ + + + + + + + + + + + +
17.61 ± 0.99 14.21 ± 1.33 22.46 ± 1.00 26.50 ± 1.43 18.55 ± 0.10 15.57 ± 0.52 21.86 ± 0.25 15.11 ± 0.74 14.49 ± 0.23 17.97 ± 0.53 14.13 ± 0.17 17.63 ± 0.81 14.98 ± 0.86
+ + + + + + + + + + + + +
15.65 ± 0.81 13.99 ± 0.85 23.49 ± 0.61 25.26 ± 1.76 16.10 ± 0.31 16.64 ± 0.91 28.86 ± 6.38 15.79 ± 0.10 14.50 ± 4.17 16.37 ± 1.72 15.23 ± 0.06 23.03 ± 0.79 17.80 ± 0.46
+ + + + + + + + + + + + +
15.38 ± 0.21 14.88 ± 0.17 24.46 ± 2.78 31.46 ± 2.79 17.64 ± 0.28 18.31 ± 0.08 18.12 ± 0.48 15.52 ± 0.33 15.44 ± 0.48 12.41 ± 0.77 14.39 ± 0.57 18.70 ± 0.88 16.91 ± 0.05
+ + + + + + + + + + + + +
16.18 ± 0.42 14.47 ± 0.96 21.49 ± 2.02 21.22 ± 1.13 17.19 ± 0.27 17.56 ± 0.48 18.53 ± 0.18 35.02 ± 0.96 18.25 ± 2.10 13.99 ± 1.46 16.19 ± 0.18 16.19 ± 0.24 16.06 ± 0.69
+ + + + + + + + + + + + +
15.35 ± 0.14 16.72 ± 0.81 21.11 ± 0.57 18.24 ± 0.62 17.86 ± 0.53 13.93 ± 1.29 17.10 ± 0.30 16.54 ± 0.54 18.24 ± 0.17 16.79 ± 0.38 19.88 ± 0.29 19.88 ± 0.22 17.31 ± 1.03
+ + + + + + + + + + + + +
16.63 ± 0.27 15.98 ± 0.67 19.05 ± 4.04 24.02 ± 0.33 15.91 ± 0.42 17.01 ± 1.00 17.69 ± 0.53 16.29 ± 0.44 36.60 ± 0.81 13.16 ± 1.71 19.16 ± 0.39 19.16 ± 0.05 14.86 ± 0.39
(usually 25 g of sample), and a standardized ISO cultural method has to be applied to evaluate product compliance (Commission Regulation No. 2073/2005). However, the standard culture method for detecting Salmonella spp. requires up to five days to produce final results; it involves stages of pre-enrichment, selective enrichment, isolation on selective agars, and biochemical characterization of presumptive colonies, and final serological confirmation (Anonymous, 2002). As a result, this methodology is tedious and laborious, and sometimes requires more time than the established shelf-life of the product under analysis. In this context the development of rapid, cost-effective, and automated methods for the determination of Salmonella, integrated with preventive strategies such as Hazard Analysis Critical Control Points (HACCP), could significantly improve safety throughout the food chain (Delibato et al., 2009, 2011, 2013). To respond to this need, several methods, mainly involving Real-Time PCR, have been developed to detect the presence of Salmonella in foods (Malorny et al., 2003; Krämer et al., 2011; Löfström et al., 2009, 2012). The main advantages of Real-Time PCR are high sensitivity, high specificity, excellent efficiency, reduced amplicon size and no post-PCR steps that reduce risks of crosscontamination (Delibato et al., 2011; Rodríguez-Lázaro et al., 2003, 2013; Rodríguez-Lázaro and Hernandez, 2013). However, one important prerequisite for an internationally standardized and recognized methodology is that a multicenter validation is performed to show the reproducibility and robustness of the method (Josefsen et al., 2007; Malorny et al., 2007). The aim of this study was to validate an openformula, non-patented, diagnostic Real-Time PCR-based assay (Josefsen et al., 2007; Malorny et al., 2004) for detection of Salmonella spp. in artificially contaminated pork meat. The validation involved thirteen laboratories from seven European Countries and was based on a collaborative and inter-laboratory study according to the ISO 16140:2003, the International Standard for the
validation of alternative microbiological methods and the use of the ISO 6579:2002/Corr. 1:2004, as reference method. The Real-Time PCR method was validated within the FP7 European research project named BASELINE (http://www.baselineeurope.eu/). To our knowledge, this is the first study, addressing an international validation of a nonproprietary Real-Time PCR method for the detection of Salmonella, involving more than ten laboratories from different countries.
2. Materials and methods 2.1. Participating laboratories The IstitutoSuperiore di Sanità (ISS — Italy) was the organizing laboratory and led the international investigation. Thirteen laboratories from seven different European Countries participated in the trial: the University of Bologna (Italy); the National Veterinary Institute (Norway); the Centro Nacional de Tecnologia y Segurdad Alimentaria (Spain); the National Food Chain Safety Office (Hungary); the University of Zagreb (Croatia); the IstitutoTecnologicoAgrario de Castilla y León (Spain); the University of Copenhagen (Denmark); the French Agency for food, environmental and occupational health safety, Anses (France), the Istituto Zooprofilattico Sperimentale (IZS) delleVenezie (Italy); the IZS del Lazio e Toscana (Italy); the IZS della Lombardia e dell'Emilia Romagna (Italy) and the IZS del Mezzogiorno (Italy). Nine out of thirteen laboratories are accredited according to ISO 17025 (ISO standard for general requirements for the competence of testing and calibration laboratories) and two laboratories, National Food Chain Safety Office in Hungary and IZS delle Venezie in Italy, are also National Reference Laboratory for Salmonella. Each participant was provided with an operating procedure (SOP) for performance of this trial.
Table 2 Interlaboratory results (mean ± relative standard error) from pork meat samples artificially contaminated with 100 CFU Salmonella spp. (medium). Laboratory
1 2 3 4 5 6 7 8 9 10 11 12 13 a
Sample A
Sample B
Sample C
Sample D
Sample E
Sample F
Sample G
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
Sample H ISO
qPCR
+ + + + + + + + + + + + +
15.77 ± 0.56 14.20 ± 0.66 25.67 ± 5.97 30.74 ± 5.74 17.73 ± 0.18 17.31 ± 1.30 18.58 ± 0.33 16.63 ± 0.21 18.90 ± 0.49 12.07 ± 0.87 16.49 ± 0.58 22.70 ± 1.76 17.99 ± 4.07
+ + + + + + + + + + + + +
15.76 ± 0.57 14.96 ± 1.29 28.79 ± 2.06 26.00 ± 0.84 16.80 ± 0.19 18.18 ± 0.88 20.69 ± 0.36 15.69 ± 0.35 13.53 ± 0.34 12.57 ± 2.92 15.09 ± 1.68 16.33 ± 0.15 15.37 ± 0.28
+ + + + + + + + + + + + +
15.79 ± 0.20 16.62 ± 0.40 24.98 ± 0.89 29.31 ± 0.66 19.48 ± 1.11 16.81 ± 0.43 20.10 ± 0.34 14.92 ± 2.52 20.66 ± 0.62 12.20 ± 0.62 17.10 ± 0.26 18.98 ± 0.36 19.90 ± 1.71
+ + + + + + + + + + + + +
15.96 ± 0.40 15.37 ± 1.02 28.33 ± n.a.a 27.28 ± 0.81 15.51 ± 0.16 15.49 ± 1.43 17.05 ± 0.24 15.35 ± 0.71 18.89 ± 0.38 12.05 ± 0.44 16.02 ± 0.17 20.30 ± 1.55 18.55 ± 0.63
+ + + + + + + + + + + + +
17.58 ± 0.23 14.86 ± 0.26 27.00 ± 0.99 26.07 ± 0.23 16.94 ± 0.22 17.07 ± 1.59 13.45 ± 0.64 15.30 ± 0.28 18.00 ± 0.20 12.09 ± 1.17 16.37 ± 0.47 27.74 ± 0.32 16.19 ± 0.61
+ + + + + + + + + + + + +
16.01 ± 0.32 16.61 ± 0.87 30.65 ± 1.54 27.42 ± 0.76 15.90 ± 0.47 14.81 ± 1.09 14.63 ± 0.88 14.86 ± 0.44 18.92 ± 0.30 11.87 ± 0.25 14.60 ± 0.48 21.34 ± 0.31 14.64 ± 1.35
+ + + + + + + + + + + + +
15.79 ± 0.42 14.80 ± 0.90 25.29 ± 0.30 30.16 ± 1.09 15.26 ± 0.69 18.59 ± 1.26 15.67 ± 0.72 15.82 ± 0.18 14.42 ± 0.69 11.88 ± 0.65 17.40 ± 0.11 16.28 ± 0.32 13.36 ± 0.79
+ + + + + + + + + + + + +
18.56 ± 0.03 18.10 ± 0.20 27.73 ± 1.07 28.59 ± 1.08 16.97 ± 0.41 20.54 ± 0.83 16.99 ± 0.95 16.00 ± 0.16 14.18 ± 0.16 13.88 ± 0.59 19.32 ± 1.22 15.33 ± 0.65 13.66 ± 0.32
n.a.: not applicable; only 1 out of the three PCR replicate was positive.
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
E. Delibato et al. / International Journal of Food Microbiology xxx (2014) xxx–xxx
15
3
each level of contamination (M, L, and B). One sample of pork meat not artificially contaminated and one sample of Buffered Peptone Water (BPW) without inoculated food were used as negative controls. After the experimental contamination of food samples, a ten-fold dilution of each sample in BPW was performed. Samples were homogenized for 90 s, and incubated at 37 °C ± 1 °C for 18 h ± 2 h. Samples were subsequently analyzed following the two methodological alternatives: traditional culture method (ISO6579:2002/ Corr. 1:2004) and alternative method (Real-Time PCR-based method).
10
ΔCq
5
0
5-
2.4. Real-Time PCR-based method protocol -10
-15 Fig. 1. Deviation of the data from the overall average obtained from the samples artificially inoculated with low (■) and medium (○) contamination levels of Salmonella spp.
2.2. Trial materials The organizing laboratory provided to the participating laboratories, materials (pork meat and reference material as Lenticule® discs) and all reagents to perform Real-Time PCR. Food samples were delivered by a company of northern Italy and reference materials at defined inoculum >were purchased from Public Health England (Culture Collections, Microbiology Services, Salisbury, UK). Food samples were prepared at the ISS by specific personnel not to be involved in the actual trial, and sized and weighted for preparing individual samples of 25 g. Prior to use, six samples were analyzed by culture method to check the absence of Salmonella; and all samples were found to be negative. Total aerobic counts at 30 °C, lactic acid bacteria and Enterobacteriaceae were found by standard methods to be an average of 100, 40 CFU/g and 10 CFU/g, respectively. Lenticules were prepared to include at the moment of the artificial inoculation 3 contamination levels: a medium level (M) containing around 100 CFU Salmonella enterica serovar Typhimurium (Salmonella typhimurium) NCTC 12023; a low level (L) containing around 10 CFU of S. typhimurium NCTC 12023; and a non-Salmonella contamination level (B) not containing any microorganisms. The organizing laboratory provided the foodstuffs and the lyophilized strains in ready to use, coded containers, as well as all reagents to perform the Real-Time PCR assay by a courier service. This means that the laboratories performing the analysis were blind to the actual content of Salmonella spp. in each sample. 2.3. Artificial contamination and pre-enrichment in buffered peptone water The laboratories performed the artificial contamination of 8 pork meat samples using 8 blind coded lyophilized bacterial strains for
Two mL of the pre-enrichment broth (BPW) was transferred into a clean microcentrifuge tube, and centrifuged for 10 min at 10,000 ×g at 4 °C. The supernatant was discarded carefully and the pellet was washed with 1 mL of Phosphate Buffered Solution (PBS), and centrifuged for 5 min at 10,000 ×g at 4 °C. Afterwards, the pellet was re-suspended in 200 μL of 6% Chelex 100 (Biorad, Hercules, CA, USA) by vortexing, and incubated for 20 min at 56 °C and then for 8 min at 100 °C. The suspension was immediately chilled on ice for 1 min, and centrifuged for 5 min at 10,000 ×g at 4 °C. Four μL of DNA extracted were used as template for the Salmonella-specific Real-Time PCR detection assay (Josefsen et al., 2007). The ring trial was performed using 96-well plates. PCR mix for 90 reactions was prepared using 1125 μL of Master mix (QuantiTecMultiplex PCR No Rox Master Mix — Qiagen, Hilden, Germany), 90 μL of 400 nM of each forward and reverse primers and 54 μL of 240 nM ttr5 and IAC probes (ttr5 and IAC probes were labelled with FAM and HEX, respectively). Two negative and positive PCR controls were prepared, adding 4 μL of molecular biology grade water or 4 μL of DNA standard to 21 μL of Master mix, respectively. Twenty-one μL of the internal amplification control (IAC) (about 1200 copies per reaction) were added to the remaining master mix tube. The PCR mix was mixed, centrifuged briefly and 21 μL were aliquoted in the PCR plate. For the IAC test reactions, 4 μL of water for molecular biology were added to each well. For the negative media control (NMC) 4 μL of extract of BPW in the absence of the test sample were added to each well. For the positive PCR control 4 μL of DNA standard were added to each well. PCR 96-well-plate was spinned-down, and transferred into the Real-Time PCR platform. Five different Real Time PCR platforms were used in the study: Applied BioSystems 7500 fast – Applied BioSystems, Foster City, USA – (two labs); Applied BioSystems 7900HT (one lab); Bio-Rad CFX96 – Bio-Rad – (three labs); Roche Light Cycler 96 – Roche Diagnostics, Mannheim, Germany – (one lab) and Stratagene Mx3005P – Agilent technologies, Santa Clara, USA – (six labs). The amplification was performed using an initial hot-start step at 95 °C for 15 min, followed by 40 cycles of a denaturation step at 95 °C for 30 s, an annealing step at 65 °C for 60 s and an extension step at 72 °C for 30 s. The fluorescence was recorded only at the end of annealing step. Three PCR replicates were used for each sample.
Table 3 Statistical analysis of the data obtained in the collaborative trial.
ISO
qPCR
Contamination level
Diagnostic specificity
Diagnostic sensitivity
Positive predictive value
Negative predictive value
Accordance (%)
Concordance (%)
Concordance odd ratio (COR)
Low Medium Low + medium None Low Medium Low + medium None
– – – 100.0 – – – 100.0
100.0 100.0 100.0 –
– – 100.0 – – – 100.0 –
– – – 100.00 – – – 100.00
100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
100.0 100.0 –
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
4
E. Delibato et al. / International Journal of Food Microbiology xxx (2014) xxx–xxx
Table 4 ISO 16140 evaluation parameters obtained for the Real-Time PCR-based method from the trial data. Relative accuracy (%)
Relative specificity (%)
Relative sensitivity (%)
False negative ratio (%)
False positive ratio (%)
100.0
100.0
100.0
0.0
0.0
2.5. Reporting and interpretation of data Raw data were reported by each participant to the organizing laboratory (ISS), who translated the codes and analyzed the data in collaboration with ITACyL. When an assay showed a quantification cycle (Cq) value ≤35 independently of the IAC Cq value, the result was interpreted as positive. When an assay showed a Cq value ≥ 35 with the IAC Cq value ≤ 40 was interpreted as negative. When an assay showed both the target and its IAC Cq values ≥ 37 the reaction was considered to have failed. When a participant reported that at least one of the three Salmonella PCR replicates was positive, they were considered to have identified the sample as being Salmonella contaminated. When a participant reported that the three Salmonella PCR replicates were negative, but at least one replicate IAC assay was positive, they were considered to have identified the sample as being Salmonella uncontaminated. When a participant reported that both Salmonella and IAC assays had failed, they were considered to have reported that the analysis of that sample had failed. 2.6. Statistical analysis The raw data sent by each laboratory were statistically analyzed according to the recommendations of Scotter et al. (2001) and by the methods of Langton et al. (2002). Accordance (repeatability of qualitative data), concordance (reproducibility of qualitative data) and the concordance odds ratio (COR) values were calculated. The first two values take into account different levels of replication in different laboratories by weighting results appropriately, and the latter evaluates the degree of inter-laboratory variation (Langton et al., 2002). Confidence intervals for accordance, concordance and COR were calculated by the method of Davison and Hinckley (1997); each laboratory was considered representative of all laboratories in the “population” of laboratories, not just those participating in this analysis. In addition, other important parameters included in the ISO 16140:2033 (Anonymous, 2003) were calculated: relative accuracy, diagnostic sensitivity, diagnostic specificity, positive and negative predictive values, false negative results and the relative specificity and sensitivity. 3. Results 3.1. Results of the laboratories in the collaborative trial All the laboratories did not show appreciable problems during the ring trial, and none of them was excluded. Table 1 shows the results of each participant for the analysis of pork meat samples contaminated with a low level load of Salmonella spp. (10 CFU per 25 g of sample). At this level of contamination, all the samples were considered as positive using the reference method (ISO 6579:2002/Corr. 1:2004) (Anonymous, 2002) and the RealTime PCR-based method. In the Real-Time PCR method, laboratories 4 and 7 showed higher Cq values with samples at low level of contamination than those detected in the other laboratories participating in the ring trial. Table 2 shows the results of the analysis of pork meat samples artificially contaminated with a medium level load of Salmonella spp. (100 CFU per 25 g of sample). Similarly to the results obtained for the low contamination level, all the samples
were correctly reported as contaminated when both the reference culture method and the Real-Time PCR based method were used. Finally, when the pork meat samples not contaminated with Salmonella spp., were tested, all were correctly reported as uncontaminated by both the reference and the alternative methods (the signal for IAC were in all the cases positive). As all the results obtained from artificially contaminated samples in this study showed a correct determination by both methodologies, the variation of the Cq values between samples, level of contamination and laboratories was studied. Interestingly, the average Cq values for each level of contamination were not statistically different (p N 0.5) (18.10 ± 0.42 vs 18.17 ± 0.47 for low and medium levels of contamination — mean ± relative standard error). Therefore to calculate how each single sample Cq value diverged from the mean, an overall average was calculated considering together the results of both levels (18.14 ± 0.31); then the Cq value of the sample was subtracted from that overall average. Fig. 1 shows how each single sample results varied. As expected from the average results for each level of contamination, the results point cloud did not differentiate between levels of contamination regardless to the labs involved. However, the deviation from the mean values of some values from the results observed can be only due to the differences between laboratories, regardless to the level of contamination. Interestingly, excluding the results of one laboratory, all the differences between the Cq values per each sample and the overall mean was in a range of 5 Cq units (i.e. 32-fold deviation) regardless to the laboratory and the level of contamination. This finding demonstrates that the 10-fold dilution between the two levels of contamination assayed and the different performances of the pre-enrichment and bacterial DNA extraction in the different laboratories can be corrected during the pre-enrichment step in BPW. 3.2. Statistical analysis Table 3 shows the diagnostic specificity, diagnostic sensitivity, positive and negative predictive values, accordance and concordance values and the concordance odds ratio for the collaborative trial of both analytical methods for the detection of Salmonella spp. on pork meat. Both methods showed excellent values for all these parameters. Similarly, the relevant parameters defined in ISO 16140 for validation of alternative methods in food microbiology (relative accuracy, relative specificity, relative sensitivity, and false negative and positive ratios) showed an excellent performance (Table 4). 4. Discussion Monitoring the presence of foodborne pathogens is a key issue in food safety (Rodríguez-Lázaro et al., 2007). In terms of microbiological criteria, food safety requires the reliable detection of pathogens such as Salmonella spp. along the entire food chain by appropriate analytical methods. This study presents an international effort to validate an alternative method based on an ISO-compatible pre-enrichment coupled to bacterial DNA extraction and Real-Time PCR detection. The Real-Time PCR assay used is based on the co-amplification of a specific region of the Salmonella spp. ttr gene (Malorny et al., 2004; Josefsen et al., 2007) and an internal amplification control (IAC). The choice of the ttr locus as a target specific for Salmonella spp. over other published targets might have the advantage in the identification of all Salmonella spp. strains as the ability to respire tetrathionate is significant for Salmonella survival and outgrowth in anaerobic competitive environments (Winter et al., 2010). In addition, the use of an IAC in food molecular microbiology diagnostics is becoming mandatory. An IAC indicates the presence of DNA polymerase inhibitors, errors caused by PCR components, or malfunction of the thermal cycler. The IAC used in this assay is a chimerical DNA sequence which is recognized by a specific probe. The simultaneous use in a single reaction of two differently labelled fluorescent probes makes it possible to detect the target, and if
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
E. Delibato et al. / International Journal of Food Microbiology xxx (2014) xxx–xxx
negative results are obtained, the positive IAC signal can confirm that the negative result is not due to an inhibition during the amplification (Hoorfar et al., 2004). Importantly, the Real-Time PCR-based method has previously been shown to detect Salmonella spp. robustly down to 1 CFU of Salmonella spp. in 25 g of different types of foods (pork meat, cheese and lettuce samples) and has also been successfully tested in naturally contaminated food samples (RodríguezLázaro et al., submitted for publication). The results obtained in this international trial corroborate those findings, and demonstrate that the internationally validated Real-Time PCR method represents a reliable method to verify microbiological criteria for Salmonella spp. in pork meat. The results obtained in this ring trial study shows that Real-Time PCR method not only has a robust detection limit of at least 10 CFU per 25 g of pork meat sample, but also has a superior capacity for correctly determine the nature of sample tested (values of 100% for diagnostic specificity and sensitivity and for positive and negative predictive values) regardless the analysis is performed in different food samples or laboratories (values of 100% for accordance, concordance and concordance odd ratio). On top of that, the performance of the Real-Time PCR method is equal to that using the reference ISO method. This finding is demonstrated by the excellent results for the major parameters indicated in the ISO standard 16140:2003 for validation of alternative methods in food microbiology (values of 100% for relative accuracy, sensitivity and specificity, and 0% of false negative and positive ratios). As the two steps for fully validation indicated in that ISO standard (in house validation — Rodríguez-Lázaro et al., submitted for publication – and interlaboratory study – this study) has been successfully conducted, we can conclude that the Real-Time PCR method meets the requirements of a diagnostic PCR and has the potential to become a standardized method for the rapid detection of Salmonella in diagnostic laboratories. On top of that, the Real-Time PCR method is cost effective (2 € vs. 12 €) and time saving (23 h vs. N7 days for positive results) and it provided a satisfactory reproducibility when carried out by different laboratories with different Real-Time PCR platforms. However, due to the high cost of an inter-laboratory study, this trial was performed only for one food category (pork meat) among those listed in the ISO 16140. Other validation studies must be conducted for other food categories (e.g. poultry meat or molluscs) to corroborate the satisfactory results obtained in this inter-laboratory investigation using pork meat. Acknowledgments This work was supported by the EU BASELINE project. DRL and MH M.D. and N.C. acknowledge the support by the Project RTA2011-079C02-01 of the Ministry of Economy and Competitiveness, Government of Spain. References Anonymous, 2002. Microbiology of Food and Animal Feeding Stuffs. Horizontal Method for the Detection of Salmonella (ISO 6579:2002). International Organization for Standardization, Geneva, Switzerland. Anonymous, 2003. ISO 16140:2003: Microbiology of Food and Animal Feeding Stuffs — Protocol for the Validation of Alternative Methods. International Organization for Standardization, Geneva, Switzerland.
5
Anonymous, 2013. The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2011. EFSA J. 11 (4), 3129. Arguello, H., Alvarez-Ordonez, A., Carvajal, A., Rubio, P., Prieto, M., 2013. Role of slaughtering in Salmonella spreading and control in pork production. J. Food Prot. 76, 899–911. Berriman, A.D., Clancy, D., Clough, H.E., Armstrong, D., Christley, R.M., 2013. Effectiveness of simulated interventions in reducing the estimated prevalence of Salmonella in UK pig herds. PLoS One 8, e66054. COMMISSION REGULATION (EC) No 2073/2005 of 15 November 2005 on microbiological criteria for food stuffs. Davison, A.C., Hinckley, D.V., 1997. Bootstrap Methods and their Application. Cambridge University Press, Cambridge, UK. Delibato, E., Volpe, G., Romanazzo, D., De Medici, D., Toti, L., Moscone, D., Palleschi, G., 2009. Development and application of an electrochemical plate coupled with immunomagnetic beads (ELIME) array for Salmonella enterica detection in meat samples. J. Agric. Food Chem. 57, 7200–7204. Delibato, E., Fiore, A., Anniballi, F., Auricchio, B., Filetici, E., Orefice, L., Losio, M.N., De Medici, D., 2011. Comparison between two standardized cultural methods and 24 hour duplex SYBR green Real-Time PCR assay for Salmonella detection in meat samples. New Microbiol. 34, 299–306. Delibato, E., Anniballi, F., Sinibaldi Vallebona, P., Palleschi, G., Volpe, G., Losio, M.N., De Medici, D., 2013. Validation of a 1-day analytical diagnostic Real-Time PCR for the detection of Salmonella in different food meat categories. Food Anal. Methods 6, 996–1003. Hoorfar, J., Malorny, B., Abdulmawjood, A., Cook, N., Wagner, M., Fach, P., 2004. Practical considerations in design of internal amplification controls for diagnostic PCR. J. Clin. Microbiol. 42, 1863–1868. Josefsen, M.H., Krause, M., Hansen, F., Hoorfar, J., 2007. Optimization of a 12-hour TaqMan PCR-based method for detection of Salmonella bacteria in meat. Appl. Environ. Microbiol. 73, 3040–3048. Krämer, N., Löfström, C., Vigre, H., Hoorfar, J., Bunge, C., Malorny, B., 2011. A novel strategy to obtain quantitative data for modelling: combined enrichment and real-time PCR for enumeration of salmonellae from pig carcasses. Int. J. Food Microbiol. 145 (Suppl. 1), S86–S95. Langton, S.D., Chevennement, R., Nagelkerke, N., Lombard, B., 2002. Analysing collaborative trials for qualitative microbiological methods: accordance and concordance. Int. J. Food Microbiol. 79, 175–181. Löfström, C., Krause, M., Josefsen, M.H., Hansen, F., Hoorfar, J., 2009. Validation of a sameday real-time PCR method for screening of meat and carcass swabs for Salmonella. BMC Microbiol. 9, 85. Löfström, C., Hansen, F., Mansdal, S., Hoorfar, J., 2012. Detection of Salmonella in meat: comparative and interlaboratory validation of a noncomplex and cost-effective prePCR protocol. J. AOAC Int. 95, 100–104. 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 (12), 7046–7052. Malorny, B., Made, D., Teufel, P., Berghof-Jager, C., Huber, I., Anderson, A., Helmuth, R., 2007. Multicenter validation study of two blockcycler- and one capillary-based real-time PCR methods for the detection of Salmonella in milk powder. Int. J. Food Microbiol. 117, 211–218. Rodríguez-Lázaro, D., Hernandez, M., 2013. Real-time PCR in food science: introduction. Curr. Issues Mol. Biol. 15, 25–38. Rodríguez-Lázaro, Hernandez, M., Esteve, T., Hoorfar, J., Pla, M., 2003. A rapid and direct real time PCR-based method for identification of Salmonella spp. J. Microbiol. Methods 54, 381–390. Rodríguez-Lázaro, D., Lombard, B., Smith, H., Rzezutka, A., D'Agostino, M., Helmuth, R., Schroeter, A., Malorny, B., Miko, A., Guerra, B., Davison, J., Kobilinsky, A., Hernández, M., Bertheau, Y., Cook, N., 2007. Trends in analytical methodology in food safety and quality: monitoring microorganisms and genetically modified organisms. Trends Food Sci. Technol. 18, 306–319. Rodríguez-Lázaro, D., Cook, N., Hernandez, M., 2013. Real-time in food science: PCR diagnostics. Curr. Issues Mol. Biol. 15, 39–44. Scotter, S.L., Langton, S., Lombard, B., Schulten, S., Nagelkerke, N., In't Veld, P.H., Rollier, P., Lahellec, C., 2001. Validation of ISO method 11290 part 1-Detection of Listeria monocytogenes in foods. Int. J. Food Microbiol. 64, 295–306. Winter, S.E., Thiennimitr, P., Winter, M.G., Butler, B.P., Huseby, D.L., Crawford, R.W., Russell, J.M., Bevins, C.L., Adams, L.G., Tsolis, R.M., Roth, J.R., Bäumler, A.J., 2010. Gut inflammation provides a respiratory electron acceptor for Salmonella. Nature 467, 426–429.
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
Contents lists available at ScienceDirect
International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro
European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat Elisabetta Delibato a,⁎, David Rodriguez Lazaro b,c, Monica Gianfranceschi a, Alessandra De Cesare d, Damiano Comin e, Antonietta Gattuso a, Marta Hernandez b, Michele Sonnessa a, Frédérique Pasquali d, Zuzsanna Sreter-Lancz f, María-José Saiz-Abajo g, Javier Pérez-De-Juan g, Javier Butrón g, Estella Prukner-Radovcic h, Danijela Horvatek Tomic h, Gro S. Johannessen i, Džiuginta Jakočiūnė j, John E. Olsen j, Marianne Chemaly k, Francoise Le Gall k, Patricia González-García b, Antonia Anna Lettini e, Maja Lukac h, Segolénè Quesne k, Claudia Zampieron e, Paola De Santis l, Sarah Lovari l, Barbara Bertasi m, Enrico Pavoni m, Yolande T.R. Proroga n, Federico Capuano n, Gerardo Manfreda d, Dario De Medici a a
Department of Veterinary Public Health and Food Safety, Istituto Superiore di Sanità Viale Regina Elena, 299 00161 Rome, Italy Instituto Tecnológico Agrario de Castilla y León (ITACyL), Valladolid, Spain c Microbiology Section, Faculty of Sciences, University of Burgos, Burgos, Spain d Alma Mater Studiorum — Università di Bologna, Dip. Scienze e Tecnologie Agro-Alimentari, Ozzano dell'Emilia, BO, Italy e Istituto Zooprofilattico Sperimentale delle Venezie, Dept. Food Safety, Legnaro (PD), Italy f National Food Chain Safety Office, Food and Feed Safety Directorate, Food Microbiological National Reference Laboratory, Mester, Budapest, Hungary g CNTA, Centro Nacional de Tecnología y Seguridad Alimentaria, San Adrian, Navarra, Spain h University of Zagreb, Faculty of Veterinary Medicine, Dept. of Poultry Diseases with Clinic, Zagreb, Croatia i Section for Bacteriology — Food and GMO Norwegian Veterinary Institute, Oslo, Norway j Department of Veterinary Disease Biology, University of Copenhagen, Denmark k Anses, Laboratoire de Ploufragan-Plouzané, Unité Hygiène et Qualité des Produits Avicoles et Porcins, Ploufragan, France l Direzione Operativa Controllo degli Alimenti, Istituto Zooprofilattico Sperimentale del Lazio e Toscana, Rome, Italy m Istituto Zooprofilattico Sperimentale della Lombardia e dell'Emilia Romagna, Brescia, Italy n Istituto Zooprofilattico del Sperimentale Mezzogiorno Portici, NA, Italy b
a r t i c l e
i n f o
Available online xxxx Keywords: Validation Salmonella Real Time PCR ISO 6579 Pork meat Artificial contamination
a b s t r a c t The classical microbiological method for detection of Salmonella spp. requires more than five days for final confirmation, and consequently there is a need for an alternative methodology for detection of this pathogen particularly in those food categories with a short shelf-life. This study presents an international (at European level) ISO 16140-based validation study of a non-proprietary Real-Time PCR-based method that can generate final results the day following sample analysis. It is based on an ISO compatible enrichment coupled to an easy and inexpensive DNA extraction and a consolidated Real-Time PCR assay. Thirteen laboratories from seven European Countries participated to this trial, and pork meat was selected as food model. The limit of detection observed was down to 10 CFU per 25 g of sample, showing excellent concordance and accordance values between samples and laboratories (100%). In addition, excellent values were obtained for relative accuracy, specificity and sensitivity (100%) when the results obtained for the Real-Time PCR-based methods were compared to those of the ISO 6579:2002 standard method. The results of this international trial demonstrate that the evaluated Real-Time PCR-based method represents an excellent alternative to the ISO standard. In fact, it shows an equal and solid performance as well as it reduces dramatically the extent of the analytical process, and can be easily implemented routinely by the Competent Authorities and Food Industry laboratories. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Pork meat and derived products are the main sources of Salmonella infection in humans in Europe; a recent study estimated that more ⁎ Corresponding author at: Department of Food Safety and Veterinary Public Health, Istituto Superiore di Sanità, Viale Regina Elena, 299-00161 Rome, Italy. E-mail address: [email protected] (E. Delibato).
than 56% of the human salmonellosis cases could be attributable to pigs, while the contributions of total reservoirs associated with laying hens (eggs), broilers and turkeys were 17.0, 10.6 and 2.6%, respectively (EFSA, 2013). As a consequence, some EU countries have established Salmonella surveillance and control programs for pork production (Arguello et al., 2013; Berriman et al., 2013). According to the European microbiological food safety criteria Salmonella spp. must be absent in an established amount depending on the kind of the product
0168-1605/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
2
E. Delibato et al. / International Journal of Food Microbiology xxx (2014) xxx–xxx
Table 1 Interlaboratory results (mean ± relative standard error) from pork meat samples artificially contaminated with 10 CFU Salmonella spp. (low). Laboratory
1 2 3 4 5 6 7 8 9 10 11 12 13
Sample A
Sample B
Sample C
Sample D
Sample E
Sample F
Sample G
Sample H
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
+ + + + + + + + + + + + +
16.14 ± 0.13 14.25 ± 1.57 18.01 ± 0.53 21.67 ± 0.58 18.47 ± 0.14 16.13 ± 0.84 23.33 ± 0.53 15.58 ± 0.23 20.70 ± 0.20 13.16 ± 0.51 13.96 ± 0.65 19.18 ± 0.54 15.77 ± 0.65
+ + + + + + + + + + + + +
19.68 ± 0.18 15.42 ± 0.54 23.31 ± 1.64 21.62 ± 0.58 16.22 ± 0.21 17.81 ± 1.26 23.82 ± 0.66 15.52 ± 0.43 18.40 ± 0.16 12.56 ± 0.69 14.49 ± 0.50 22.58 ± 0.40 17.91 ± 0.50
+ + + + + + + + + + + + +
17.61 ± 0.99 14.21 ± 1.33 22.46 ± 1.00 26.50 ± 1.43 18.55 ± 0.10 15.57 ± 0.52 21.86 ± 0.25 15.11 ± 0.74 14.49 ± 0.23 17.97 ± 0.53 14.13 ± 0.17 17.63 ± 0.81 14.98 ± 0.86
+ + + + + + + + + + + + +
15.65 ± 0.81 13.99 ± 0.85 23.49 ± 0.61 25.26 ± 1.76 16.10 ± 0.31 16.64 ± 0.91 28.86 ± 6.38 15.79 ± 0.10 14.50 ± 4.17 16.37 ± 1.72 15.23 ± 0.06 23.03 ± 0.79 17.80 ± 0.46
+ + + + + + + + + + + + +
15.38 ± 0.21 14.88 ± 0.17 24.46 ± 2.78 31.46 ± 2.79 17.64 ± 0.28 18.31 ± 0.08 18.12 ± 0.48 15.52 ± 0.33 15.44 ± 0.48 12.41 ± 0.77 14.39 ± 0.57 18.70 ± 0.88 16.91 ± 0.05
+ + + + + + + + + + + + +
16.18 ± 0.42 14.47 ± 0.96 21.49 ± 2.02 21.22 ± 1.13 17.19 ± 0.27 17.56 ± 0.48 18.53 ± 0.18 35.02 ± 0.96 18.25 ± 2.10 13.99 ± 1.46 16.19 ± 0.18 16.19 ± 0.24 16.06 ± 0.69
+ + + + + + + + + + + + +
15.35 ± 0.14 16.72 ± 0.81 21.11 ± 0.57 18.24 ± 0.62 17.86 ± 0.53 13.93 ± 1.29 17.10 ± 0.30 16.54 ± 0.54 18.24 ± 0.17 16.79 ± 0.38 19.88 ± 0.29 19.88 ± 0.22 17.31 ± 1.03
+ + + + + + + + + + + + +
16.63 ± 0.27 15.98 ± 0.67 19.05 ± 4.04 24.02 ± 0.33 15.91 ± 0.42 17.01 ± 1.00 17.69 ± 0.53 16.29 ± 0.44 36.60 ± 0.81 13.16 ± 1.71 19.16 ± 0.39 19.16 ± 0.05 14.86 ± 0.39
(usually 25 g of sample), and a standardized ISO cultural method has to be applied to evaluate product compliance (Commission Regulation No. 2073/2005). However, the standard culture method for detecting Salmonella spp. requires up to five days to produce final results; it involves stages of pre-enrichment, selective enrichment, isolation on selective agars, and biochemical characterization of presumptive colonies, and final serological confirmation (Anonymous, 2002). As a result, this methodology is tedious and laborious, and sometimes requires more time than the established shelf-life of the product under analysis. In this context the development of rapid, cost-effective, and automated methods for the determination of Salmonella, integrated with preventive strategies such as Hazard Analysis Critical Control Points (HACCP), could significantly improve safety throughout the food chain (Delibato et al., 2009, 2011, 2013). To respond to this need, several methods, mainly involving Real-Time PCR, have been developed to detect the presence of Salmonella in foods (Malorny et al., 2003; Krämer et al., 2011; Löfström et al., 2009, 2012). The main advantages of Real-Time PCR are high sensitivity, high specificity, excellent efficiency, reduced amplicon size and no post-PCR steps that reduce risks of crosscontamination (Delibato et al., 2011; Rodríguez-Lázaro et al., 2003, 2013; Rodríguez-Lázaro and Hernandez, 2013). However, one important prerequisite for an internationally standardized and recognized methodology is that a multicenter validation is performed to show the reproducibility and robustness of the method (Josefsen et al., 2007; Malorny et al., 2007). The aim of this study was to validate an openformula, non-patented, diagnostic Real-Time PCR-based assay (Josefsen et al., 2007; Malorny et al., 2004) for detection of Salmonella spp. in artificially contaminated pork meat. The validation involved thirteen laboratories from seven European Countries and was based on a collaborative and inter-laboratory study according to the ISO 16140:2003, the International Standard for the
validation of alternative microbiological methods and the use of the ISO 6579:2002/Corr. 1:2004, as reference method. The Real-Time PCR method was validated within the FP7 European research project named BASELINE (http://www.baselineeurope.eu/). To our knowledge, this is the first study, addressing an international validation of a nonproprietary Real-Time PCR method for the detection of Salmonella, involving more than ten laboratories from different countries.
2. Materials and methods 2.1. Participating laboratories The IstitutoSuperiore di Sanità (ISS — Italy) was the organizing laboratory and led the international investigation. Thirteen laboratories from seven different European Countries participated in the trial: the University of Bologna (Italy); the National Veterinary Institute (Norway); the Centro Nacional de Tecnologia y Segurdad Alimentaria (Spain); the National Food Chain Safety Office (Hungary); the University of Zagreb (Croatia); the IstitutoTecnologicoAgrario de Castilla y León (Spain); the University of Copenhagen (Denmark); the French Agency for food, environmental and occupational health safety, Anses (France), the Istituto Zooprofilattico Sperimentale (IZS) delleVenezie (Italy); the IZS del Lazio e Toscana (Italy); the IZS della Lombardia e dell'Emilia Romagna (Italy) and the IZS del Mezzogiorno (Italy). Nine out of thirteen laboratories are accredited according to ISO 17025 (ISO standard for general requirements for the competence of testing and calibration laboratories) and two laboratories, National Food Chain Safety Office in Hungary and IZS delle Venezie in Italy, are also National Reference Laboratory for Salmonella. Each participant was provided with an operating procedure (SOP) for performance of this trial.
Table 2 Interlaboratory results (mean ± relative standard error) from pork meat samples artificially contaminated with 100 CFU Salmonella spp. (medium). Laboratory
1 2 3 4 5 6 7 8 9 10 11 12 13 a
Sample A
Sample B
Sample C
Sample D
Sample E
Sample F
Sample G
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
ISO
qPCR
Sample H ISO
qPCR
+ + + + + + + + + + + + +
15.77 ± 0.56 14.20 ± 0.66 25.67 ± 5.97 30.74 ± 5.74 17.73 ± 0.18 17.31 ± 1.30 18.58 ± 0.33 16.63 ± 0.21 18.90 ± 0.49 12.07 ± 0.87 16.49 ± 0.58 22.70 ± 1.76 17.99 ± 4.07
+ + + + + + + + + + + + +
15.76 ± 0.57 14.96 ± 1.29 28.79 ± 2.06 26.00 ± 0.84 16.80 ± 0.19 18.18 ± 0.88 20.69 ± 0.36 15.69 ± 0.35 13.53 ± 0.34 12.57 ± 2.92 15.09 ± 1.68 16.33 ± 0.15 15.37 ± 0.28
+ + + + + + + + + + + + +
15.79 ± 0.20 16.62 ± 0.40 24.98 ± 0.89 29.31 ± 0.66 19.48 ± 1.11 16.81 ± 0.43 20.10 ± 0.34 14.92 ± 2.52 20.66 ± 0.62 12.20 ± 0.62 17.10 ± 0.26 18.98 ± 0.36 19.90 ± 1.71
+ + + + + + + + + + + + +
15.96 ± 0.40 15.37 ± 1.02 28.33 ± n.a.a 27.28 ± 0.81 15.51 ± 0.16 15.49 ± 1.43 17.05 ± 0.24 15.35 ± 0.71 18.89 ± 0.38 12.05 ± 0.44 16.02 ± 0.17 20.30 ± 1.55 18.55 ± 0.63
+ + + + + + + + + + + + +
17.58 ± 0.23 14.86 ± 0.26 27.00 ± 0.99 26.07 ± 0.23 16.94 ± 0.22 17.07 ± 1.59 13.45 ± 0.64 15.30 ± 0.28 18.00 ± 0.20 12.09 ± 1.17 16.37 ± 0.47 27.74 ± 0.32 16.19 ± 0.61
+ + + + + + + + + + + + +
16.01 ± 0.32 16.61 ± 0.87 30.65 ± 1.54 27.42 ± 0.76 15.90 ± 0.47 14.81 ± 1.09 14.63 ± 0.88 14.86 ± 0.44 18.92 ± 0.30 11.87 ± 0.25 14.60 ± 0.48 21.34 ± 0.31 14.64 ± 1.35
+ + + + + + + + + + + + +
15.79 ± 0.42 14.80 ± 0.90 25.29 ± 0.30 30.16 ± 1.09 15.26 ± 0.69 18.59 ± 1.26 15.67 ± 0.72 15.82 ± 0.18 14.42 ± 0.69 11.88 ± 0.65 17.40 ± 0.11 16.28 ± 0.32 13.36 ± 0.79
+ + + + + + + + + + + + +
18.56 ± 0.03 18.10 ± 0.20 27.73 ± 1.07 28.59 ± 1.08 16.97 ± 0.41 20.54 ± 0.83 16.99 ± 0.95 16.00 ± 0.16 14.18 ± 0.16 13.88 ± 0.59 19.32 ± 1.22 15.33 ± 0.65 13.66 ± 0.32
n.a.: not applicable; only 1 out of the three PCR replicate was positive.
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
E. Delibato et al. / International Journal of Food Microbiology xxx (2014) xxx–xxx
15
3
each level of contamination (M, L, and B). One sample of pork meat not artificially contaminated and one sample of Buffered Peptone Water (BPW) without inoculated food were used as negative controls. After the experimental contamination of food samples, a ten-fold dilution of each sample in BPW was performed. Samples were homogenized for 90 s, and incubated at 37 °C ± 1 °C for 18 h ± 2 h. Samples were subsequently analyzed following the two methodological alternatives: traditional culture method (ISO6579:2002/ Corr. 1:2004) and alternative method (Real-Time PCR-based method).
10
ΔCq
5
0
5-
2.4. Real-Time PCR-based method protocol -10
-15 Fig. 1. Deviation of the data from the overall average obtained from the samples artificially inoculated with low (■) and medium (○) contamination levels of Salmonella spp.
2.2. Trial materials The organizing laboratory provided to the participating laboratories, materials (pork meat and reference material as Lenticule® discs) and all reagents to perform Real-Time PCR. Food samples were delivered by a company of northern Italy and reference materials at defined inoculum >were purchased from Public Health England (Culture Collections, Microbiology Services, Salisbury, UK). Food samples were prepared at the ISS by specific personnel not to be involved in the actual trial, and sized and weighted for preparing individual samples of 25 g. Prior to use, six samples were analyzed by culture method to check the absence of Salmonella; and all samples were found to be negative. Total aerobic counts at 30 °C, lactic acid bacteria and Enterobacteriaceae were found by standard methods to be an average of 100, 40 CFU/g and 10 CFU/g, respectively. Lenticules were prepared to include at the moment of the artificial inoculation 3 contamination levels: a medium level (M) containing around 100 CFU Salmonella enterica serovar Typhimurium (Salmonella typhimurium) NCTC 12023; a low level (L) containing around 10 CFU of S. typhimurium NCTC 12023; and a non-Salmonella contamination level (B) not containing any microorganisms. The organizing laboratory provided the foodstuffs and the lyophilized strains in ready to use, coded containers, as well as all reagents to perform the Real-Time PCR assay by a courier service. This means that the laboratories performing the analysis were blind to the actual content of Salmonella spp. in each sample. 2.3. Artificial contamination and pre-enrichment in buffered peptone water The laboratories performed the artificial contamination of 8 pork meat samples using 8 blind coded lyophilized bacterial strains for
Two mL of the pre-enrichment broth (BPW) was transferred into a clean microcentrifuge tube, and centrifuged for 10 min at 10,000 ×g at 4 °C. The supernatant was discarded carefully and the pellet was washed with 1 mL of Phosphate Buffered Solution (PBS), and centrifuged for 5 min at 10,000 ×g at 4 °C. Afterwards, the pellet was re-suspended in 200 μL of 6% Chelex 100 (Biorad, Hercules, CA, USA) by vortexing, and incubated for 20 min at 56 °C and then for 8 min at 100 °C. The suspension was immediately chilled on ice for 1 min, and centrifuged for 5 min at 10,000 ×g at 4 °C. Four μL of DNA extracted were used as template for the Salmonella-specific Real-Time PCR detection assay (Josefsen et al., 2007). The ring trial was performed using 96-well plates. PCR mix for 90 reactions was prepared using 1125 μL of Master mix (QuantiTecMultiplex PCR No Rox Master Mix — Qiagen, Hilden, Germany), 90 μL of 400 nM of each forward and reverse primers and 54 μL of 240 nM ttr5 and IAC probes (ttr5 and IAC probes were labelled with FAM and HEX, respectively). Two negative and positive PCR controls were prepared, adding 4 μL of molecular biology grade water or 4 μL of DNA standard to 21 μL of Master mix, respectively. Twenty-one μL of the internal amplification control (IAC) (about 1200 copies per reaction) were added to the remaining master mix tube. The PCR mix was mixed, centrifuged briefly and 21 μL were aliquoted in the PCR plate. For the IAC test reactions, 4 μL of water for molecular biology were added to each well. For the negative media control (NMC) 4 μL of extract of BPW in the absence of the test sample were added to each well. For the positive PCR control 4 μL of DNA standard were added to each well. PCR 96-well-plate was spinned-down, and transferred into the Real-Time PCR platform. Five different Real Time PCR platforms were used in the study: Applied BioSystems 7500 fast – Applied BioSystems, Foster City, USA – (two labs); Applied BioSystems 7900HT (one lab); Bio-Rad CFX96 – Bio-Rad – (three labs); Roche Light Cycler 96 – Roche Diagnostics, Mannheim, Germany – (one lab) and Stratagene Mx3005P – Agilent technologies, Santa Clara, USA – (six labs). The amplification was performed using an initial hot-start step at 95 °C for 15 min, followed by 40 cycles of a denaturation step at 95 °C for 30 s, an annealing step at 65 °C for 60 s and an extension step at 72 °C for 30 s. The fluorescence was recorded only at the end of annealing step. Three PCR replicates were used for each sample.
Table 3 Statistical analysis of the data obtained in the collaborative trial.
ISO
qPCR
Contamination level
Diagnostic specificity
Diagnostic sensitivity
Positive predictive value
Negative predictive value
Accordance (%)
Concordance (%)
Concordance odd ratio (COR)
Low Medium Low + medium None Low Medium Low + medium None
– – – 100.0 – – – 100.0
100.0 100.0 100.0 –
– – 100.0 – – – 100.0 –
– – – 100.00 – – – 100.00
100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
100.0 100.0 –
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
4
E. Delibato et al. / International Journal of Food Microbiology xxx (2014) xxx–xxx
Table 4 ISO 16140 evaluation parameters obtained for the Real-Time PCR-based method from the trial data. Relative accuracy (%)
Relative specificity (%)
Relative sensitivity (%)
False negative ratio (%)
False positive ratio (%)
100.0
100.0
100.0
0.0
0.0
2.5. Reporting and interpretation of data Raw data were reported by each participant to the organizing laboratory (ISS), who translated the codes and analyzed the data in collaboration with ITACyL. When an assay showed a quantification cycle (Cq) value ≤35 independently of the IAC Cq value, the result was interpreted as positive. When an assay showed a Cq value ≥ 35 with the IAC Cq value ≤ 40 was interpreted as negative. When an assay showed both the target and its IAC Cq values ≥ 37 the reaction was considered to have failed. When a participant reported that at least one of the three Salmonella PCR replicates was positive, they were considered to have identified the sample as being Salmonella contaminated. When a participant reported that the three Salmonella PCR replicates were negative, but at least one replicate IAC assay was positive, they were considered to have identified the sample as being Salmonella uncontaminated. When a participant reported that both Salmonella and IAC assays had failed, they were considered to have reported that the analysis of that sample had failed. 2.6. Statistical analysis The raw data sent by each laboratory were statistically analyzed according to the recommendations of Scotter et al. (2001) and by the methods of Langton et al. (2002). Accordance (repeatability of qualitative data), concordance (reproducibility of qualitative data) and the concordance odds ratio (COR) values were calculated. The first two values take into account different levels of replication in different laboratories by weighting results appropriately, and the latter evaluates the degree of inter-laboratory variation (Langton et al., 2002). Confidence intervals for accordance, concordance and COR were calculated by the method of Davison and Hinckley (1997); each laboratory was considered representative of all laboratories in the “population” of laboratories, not just those participating in this analysis. In addition, other important parameters included in the ISO 16140:2033 (Anonymous, 2003) were calculated: relative accuracy, diagnostic sensitivity, diagnostic specificity, positive and negative predictive values, false negative results and the relative specificity and sensitivity. 3. Results 3.1. Results of the laboratories in the collaborative trial All the laboratories did not show appreciable problems during the ring trial, and none of them was excluded. Table 1 shows the results of each participant for the analysis of pork meat samples contaminated with a low level load of Salmonella spp. (10 CFU per 25 g of sample). At this level of contamination, all the samples were considered as positive using the reference method (ISO 6579:2002/Corr. 1:2004) (Anonymous, 2002) and the RealTime PCR-based method. In the Real-Time PCR method, laboratories 4 and 7 showed higher Cq values with samples at low level of contamination than those detected in the other laboratories participating in the ring trial. Table 2 shows the results of the analysis of pork meat samples artificially contaminated with a medium level load of Salmonella spp. (100 CFU per 25 g of sample). Similarly to the results obtained for the low contamination level, all the samples
were correctly reported as contaminated when both the reference culture method and the Real-Time PCR based method were used. Finally, when the pork meat samples not contaminated with Salmonella spp., were tested, all were correctly reported as uncontaminated by both the reference and the alternative methods (the signal for IAC were in all the cases positive). As all the results obtained from artificially contaminated samples in this study showed a correct determination by both methodologies, the variation of the Cq values between samples, level of contamination and laboratories was studied. Interestingly, the average Cq values for each level of contamination were not statistically different (p N 0.5) (18.10 ± 0.42 vs 18.17 ± 0.47 for low and medium levels of contamination — mean ± relative standard error). Therefore to calculate how each single sample Cq value diverged from the mean, an overall average was calculated considering together the results of both levels (18.14 ± 0.31); then the Cq value of the sample was subtracted from that overall average. Fig. 1 shows how each single sample results varied. As expected from the average results for each level of contamination, the results point cloud did not differentiate between levels of contamination regardless to the labs involved. However, the deviation from the mean values of some values from the results observed can be only due to the differences between laboratories, regardless to the level of contamination. Interestingly, excluding the results of one laboratory, all the differences between the Cq values per each sample and the overall mean was in a range of 5 Cq units (i.e. 32-fold deviation) regardless to the laboratory and the level of contamination. This finding demonstrates that the 10-fold dilution between the two levels of contamination assayed and the different performances of the pre-enrichment and bacterial DNA extraction in the different laboratories can be corrected during the pre-enrichment step in BPW. 3.2. Statistical analysis Table 3 shows the diagnostic specificity, diagnostic sensitivity, positive and negative predictive values, accordance and concordance values and the concordance odds ratio for the collaborative trial of both analytical methods for the detection of Salmonella spp. on pork meat. Both methods showed excellent values for all these parameters. Similarly, the relevant parameters defined in ISO 16140 for validation of alternative methods in food microbiology (relative accuracy, relative specificity, relative sensitivity, and false negative and positive ratios) showed an excellent performance (Table 4). 4. Discussion Monitoring the presence of foodborne pathogens is a key issue in food safety (Rodríguez-Lázaro et al., 2007). In terms of microbiological criteria, food safety requires the reliable detection of pathogens such as Salmonella spp. along the entire food chain by appropriate analytical methods. This study presents an international effort to validate an alternative method based on an ISO-compatible pre-enrichment coupled to bacterial DNA extraction and Real-Time PCR detection. The Real-Time PCR assay used is based on the co-amplification of a specific region of the Salmonella spp. ttr gene (Malorny et al., 2004; Josefsen et al., 2007) and an internal amplification control (IAC). The choice of the ttr locus as a target specific for Salmonella spp. over other published targets might have the advantage in the identification of all Salmonella spp. strains as the ability to respire tetrathionate is significant for Salmonella survival and outgrowth in anaerobic competitive environments (Winter et al., 2010). In addition, the use of an IAC in food molecular microbiology diagnostics is becoming mandatory. An IAC indicates the presence of DNA polymerase inhibitors, errors caused by PCR components, or malfunction of the thermal cycler. The IAC used in this assay is a chimerical DNA sequence which is recognized by a specific probe. The simultaneous use in a single reaction of two differently labelled fluorescent probes makes it possible to detect the target, and if
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
E. Delibato et al. / International Journal of Food Microbiology xxx (2014) xxx–xxx
negative results are obtained, the positive IAC signal can confirm that the negative result is not due to an inhibition during the amplification (Hoorfar et al., 2004). Importantly, the Real-Time PCR-based method has previously been shown to detect Salmonella spp. robustly down to 1 CFU of Salmonella spp. in 25 g of different types of foods (pork meat, cheese and lettuce samples) and has also been successfully tested in naturally contaminated food samples (RodríguezLázaro et al., submitted for publication). The results obtained in this international trial corroborate those findings, and demonstrate that the internationally validated Real-Time PCR method represents a reliable method to verify microbiological criteria for Salmonella spp. in pork meat. The results obtained in this ring trial study shows that Real-Time PCR method not only has a robust detection limit of at least 10 CFU per 25 g of pork meat sample, but also has a superior capacity for correctly determine the nature of sample tested (values of 100% for diagnostic specificity and sensitivity and for positive and negative predictive values) regardless the analysis is performed in different food samples or laboratories (values of 100% for accordance, concordance and concordance odd ratio). On top of that, the performance of the Real-Time PCR method is equal to that using the reference ISO method. This finding is demonstrated by the excellent results for the major parameters indicated in the ISO standard 16140:2003 for validation of alternative methods in food microbiology (values of 100% for relative accuracy, sensitivity and specificity, and 0% of false negative and positive ratios). As the two steps for fully validation indicated in that ISO standard (in house validation — Rodríguez-Lázaro et al., submitted for publication – and interlaboratory study – this study) has been successfully conducted, we can conclude that the Real-Time PCR method meets the requirements of a diagnostic PCR and has the potential to become a standardized method for the rapid detection of Salmonella in diagnostic laboratories. On top of that, the Real-Time PCR method is cost effective (2 € vs. 12 €) and time saving (23 h vs. N7 days for positive results) and it provided a satisfactory reproducibility when carried out by different laboratories with different Real-Time PCR platforms. However, due to the high cost of an inter-laboratory study, this trial was performed only for one food category (pork meat) among those listed in the ISO 16140. Other validation studies must be conducted for other food categories (e.g. poultry meat or molluscs) to corroborate the satisfactory results obtained in this inter-laboratory investigation using pork meat. Acknowledgments This work was supported by the EU BASELINE project. DRL and MH M.D. and N.C. acknowledge the support by the Project RTA2011-079C02-01 of the Ministry of Economy and Competitiveness, Government of Spain. References Anonymous, 2002. Microbiology of Food and Animal Feeding Stuffs. Horizontal Method for the Detection of Salmonella (ISO 6579:2002). International Organization for Standardization, Geneva, Switzerland. Anonymous, 2003. ISO 16140:2003: Microbiology of Food and Animal Feeding Stuffs — Protocol for the Validation of Alternative Methods. International Organization for Standardization, Geneva, Switzerland.
5
Anonymous, 2013. The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2011. EFSA J. 11 (4), 3129. Arguello, H., Alvarez-Ordonez, A., Carvajal, A., Rubio, P., Prieto, M., 2013. Role of slaughtering in Salmonella spreading and control in pork production. J. Food Prot. 76, 899–911. Berriman, A.D., Clancy, D., Clough, H.E., Armstrong, D., Christley, R.M., 2013. Effectiveness of simulated interventions in reducing the estimated prevalence of Salmonella in UK pig herds. PLoS One 8, e66054. COMMISSION REGULATION (EC) No 2073/2005 of 15 November 2005 on microbiological criteria for food stuffs. Davison, A.C., Hinckley, D.V., 1997. Bootstrap Methods and their Application. Cambridge University Press, Cambridge, UK. Delibato, E., Volpe, G., Romanazzo, D., De Medici, D., Toti, L., Moscone, D., Palleschi, G., 2009. Development and application of an electrochemical plate coupled with immunomagnetic beads (ELIME) array for Salmonella enterica detection in meat samples. J. Agric. Food Chem. 57, 7200–7204. Delibato, E., Fiore, A., Anniballi, F., Auricchio, B., Filetici, E., Orefice, L., Losio, M.N., De Medici, D., 2011. Comparison between two standardized cultural methods and 24 hour duplex SYBR green Real-Time PCR assay for Salmonella detection in meat samples. New Microbiol. 34, 299–306. Delibato, E., Anniballi, F., Sinibaldi Vallebona, P., Palleschi, G., Volpe, G., Losio, M.N., De Medici, D., 2013. Validation of a 1-day analytical diagnostic Real-Time PCR for the detection of Salmonella in different food meat categories. Food Anal. Methods 6, 996–1003. Hoorfar, J., Malorny, B., Abdulmawjood, A., Cook, N., Wagner, M., Fach, P., 2004. Practical considerations in design of internal amplification controls for diagnostic PCR. J. Clin. Microbiol. 42, 1863–1868. Josefsen, M.H., Krause, M., Hansen, F., Hoorfar, J., 2007. Optimization of a 12-hour TaqMan PCR-based method for detection of Salmonella bacteria in meat. Appl. Environ. Microbiol. 73, 3040–3048. Krämer, N., Löfström, C., Vigre, H., Hoorfar, J., Bunge, C., Malorny, B., 2011. A novel strategy to obtain quantitative data for modelling: combined enrichment and real-time PCR for enumeration of salmonellae from pig carcasses. Int. J. Food Microbiol. 145 (Suppl. 1), S86–S95. Langton, S.D., Chevennement, R., Nagelkerke, N., Lombard, B., 2002. Analysing collaborative trials for qualitative microbiological methods: accordance and concordance. Int. J. Food Microbiol. 79, 175–181. Löfström, C., Krause, M., Josefsen, M.H., Hansen, F., Hoorfar, J., 2009. Validation of a sameday real-time PCR method for screening of meat and carcass swabs for Salmonella. BMC Microbiol. 9, 85. Löfström, C., Hansen, F., Mansdal, S., Hoorfar, J., 2012. Detection of Salmonella in meat: comparative and interlaboratory validation of a noncomplex and cost-effective prePCR protocol. J. AOAC Int. 95, 100–104. 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 (12), 7046–7052. Malorny, B., Made, D., Teufel, P., Berghof-Jager, C., Huber, I., Anderson, A., Helmuth, R., 2007. Multicenter validation study of two blockcycler- and one capillary-based real-time PCR methods for the detection of Salmonella in milk powder. Int. J. Food Microbiol. 117, 211–218. Rodríguez-Lázaro, D., Hernandez, M., 2013. Real-time PCR in food science: introduction. Curr. Issues Mol. Biol. 15, 25–38. Rodríguez-Lázaro, Hernandez, M., Esteve, T., Hoorfar, J., Pla, M., 2003. A rapid and direct real time PCR-based method for identification of Salmonella spp. J. Microbiol. Methods 54, 381–390. Rodríguez-Lázaro, D., Lombard, B., Smith, H., Rzezutka, A., D'Agostino, M., Helmuth, R., Schroeter, A., Malorny, B., Miko, A., Guerra, B., Davison, J., Kobilinsky, A., Hernández, M., Bertheau, Y., Cook, N., 2007. Trends in analytical methodology in food safety and quality: monitoring microorganisms and genetically modified organisms. Trends Food Sci. Technol. 18, 306–319. Rodríguez-Lázaro, D., Cook, N., Hernandez, M., 2013. Real-time in food science: PCR diagnostics. Curr. Issues Mol. Biol. 15, 39–44. Scotter, S.L., Langton, S., Lombard, B., Schulten, S., Nagelkerke, N., In't Veld, P.H., Rollier, P., Lahellec, C., 2001. Validation of ISO method 11290 part 1-Detection of Listeria monocytogenes in foods. Int. J. Food Microbiol. 64, 295–306. Winter, S.E., Thiennimitr, P., Winter, M.G., Butler, B.P., Huseby, D.L., Crawford, R.W., Russell, J.M., Bevins, C.L., Adams, L.G., Tsolis, R.M., Roth, J.R., Bäumler, A.J., 2010. Gut inflammation provides a respiratory electron acceptor for Salmonella. Nature 467, 426–429.
Please cite this article as: Delibato, E., et al., European validation of Real-Time PCR method for detection of Salmonella spp. in pork meat, Int. J. Food Microbiol. (2014), http://dx.doi.org/10.1016/j.ijfoodmicro.2014.01.005
