lable at ScienceDirect

Food Control 37 (2014) 371e379

Contents lists avai

Food Control

journal homepage: www.elsevier .com/locate/ foodcont

In-house validation of novel multiplex real-time PCR genecombination for the simultaneous detection of the main humanpathogenic vibrios (Vibrio cholerae, Vibrio parahaemolyticus, and Vibriovulnificus)

Alejandro Garrido-Maestu, María-José Chapela, Elvira Peñaranda, Juan M. Vieites,Ana G. Cabado*

Microbiology and Toxins Area, ANFACO-CECOPESCA, Campus Univ. 16, 36310 Vigo, PO, Spain

a r t i c l e i n f o

Article history:Received 25 May 2013Received in revised form5 September 2013Accepted 14 September 2013

Keywords:ompWtlhvvhAqPCRVibrio spp.Validation

* Corresponding author. Tel.: þ34 986 469 303; faxE-mail address: agcabado@anfaco.es (A.G. Cabado

0956-7135/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.foodcont.2013.09.026

a b s t r a c t

In recent years the incidence of vibriosis has greatly increased, raising the concern among consumersabout the innocuity of certain food products. Previous studies demonstrated various advantages ofmolecular methods, including qPCR, for the screening of food-borne pathogens. The new methoddeveloped in the present study allows fast and reliable detection of the main human pathogenic Vibriospecies (Vibrio cholerae, Vibrio parahaemolyticus, and Vibrio vulnificus). Specificity of the combination ofprimers and probes was successfully tested against several bacterial species and strains (44 differentstrains). Evaluation of the qPCR efficiency reported a value of 94% with the simultaneous amplification ofthe internal amplification control. The evaluation of the quality of the method was based on six pa-rameters: relative sensitivity, specificity, accuracy, positive and negative predictive values as well asKappa index of concordance. Each of the values obtained were higher than 96%. Additionally a very lowlimit of detection was determined for the developed method (less than 10 cfu/25 g). All the parameters ofthe method analyzed were obtained from the analysis of a wide variety of foodstuffs, water samples andreference material from proficiency tests, and compared against the culture reference method.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The genus Vibrio comprises more than 70 species, that areubiquitous in aquatic environments and inhabit marine animals.Thirteen different species have been associated with human path-ogenesis, however, the majority of human Vibrio infections areassociated with three species: Vibrio cholerae, Vibrio para-haemolyticus, and Vibrio vulnificus (Drake, DePaola, & Jaykus, 2007;Izumiya et al., 2011; Neogi et al., 2010; Thompson, Iida, & Swings,2004).

V. cholerae is the causative agent of Asiatic or epidemic cholera.Two serogroups (O1 andO139) are responsible for cholera epidemicsbut others (non-O1/non-O139) have been associated with sporadicgastroenteritis (Blackwell & Oliver, 2008; Jones &Oliver, 2009; Neogiet al., 2010). The first report of V. parahaemolyticus infection wasdocumented in 1950, and since then it has become the leading cause

: þ34 986 469 269.).

All rights reserved.

of seafood-derived foodpoisoningworldwide. Infection causes acutegastroenteritis, but itmayalso causewound infection and septicemia(Broberg, Calder, &Orth, 2011;Hiyoshi, Kodama, Iida, &Honda, 2010;Su & Liu, 2007). Finally, V. vulnificus has been identified as thedeadliest, in terms of case-fatality rate, food-borne pathogen in theUS, accounting for 95% of all seafood related deaths. Its fatality rateamong immunocompromised patients or individuals with underly-ing chronic disease can be as high as 50%. It may also cause gastro-enteritis, necrotizing fasciitis, and wound infections (Gulig et al.,2010; Roig, Sanjuan, Llorens, & Amaro, 2010).

European regulation does not consider analysis of pathogenicvibrios in food, not even in its latest amendments ((EC), 2005, 2007,2010). In contrast, the Canadian and American regulations specif-ically mention the main pathogenic Vibrio spp. (Canadian FoodInspection Agency; FDA, 2011).

At present several official methods exist for the screening ofpathogenic vibrios in food, like those developed by the InternalStandardization Organization (ISO) or by the Food and DrugAdministration (FDA) (Kaysner and Angelo DePaola, 2004, chap. 9;ISO 2007a, 2007b). Additionally, other newly developed methods,

Table 1Strains used for the evaluation of the specificity of primers and probes, and for the inoculation of food samples.

Bacteria Strain qPCR

ompW tlh vvhA

Vibrio parahaemolyticus CECT 511a, CECT 5271, CCUG 43362, CCUG 43363, CCUG 43364, CCUG 43365, CAIM 58 � þ �Vibrio cholerae CECT 514a (O1), CCUG 47460 (O139), 1789b (algae), 8053c (pangasius)b, 8053o (pangasius)b þ � �Vibrio vulnificus CAIM 611, CECT 529a, CECT 4869, CECT 4608 � � þVibrio alginolyticus CECT 586, CAIM 342, (unknown)b � � �Vibrio mimicus CECT 4218, BCCM/LMG 7896 � � �Aeromonas hydrophila CECT 839 � � �Pseudomonas putida CECT 324 � � �Pseudomonas aeruginosa CECT 108 � � �Pseudomonas fluorescens CECT 378 � � �Escherichia coli CECT 516, CECT 434 � � �Citrobacter freundii CECT 401 � � �Staphylococcus aureus CECT 240, CECT 435 � � �Salmonella enterica CECT 4594 � � �Salmonella spp. 311b (fishmeal), 312b (fishmeal), 313b (fishmeal), 314b (fishmeal), 315b (fishmeal) � � �Listeria monocytogenes CECT 935, 810b (mussel) � � �Listeria innocua CECT 910 � � �Listeria seeligeri CECT 917 � � �Listeria ivanovii CECT 913 � � �Shigella sonnei CECT 413 � � �Shigella flexneri CECT 4804 � � �Enterococcus faecalis CECT 481 � � �a Reference strains for the evaluation of the method.b Strain identified in our laboratory, if known, the source and/or serotype was indicated in parenthesis. CECT: Spanish Type Culture Collection, CCUG: Culture Collection

University of Göteborg, CAIM: Collection of Aquatic Important Microorganisms, BCCM/LMG: Belgian Co-Orfinated Collections Of Micro-Organsims.

A. Garrido-Maestu et al. / Food Control 37 (2014) 371e379372

like those described by Williams, Froelich, and Oliver (2013) andGriffitt and Grimes (2013) (Williams et al., 2013), have beendescribed. All these protocols are culture-based, and even thoughreliable, present several limitations like not being able to detectviable but not culturable bacteria (VBNC), are time-consuming,overgrowth of other bacteria may hide desired colonies or makethe isolation step difficult, atypical biochemical profiles may not becorrectly identified, among others (Chomvarin et al., 2007; Fedioet al., 2007; Lyon, 2001; Saravanan, Kumar, Karunasagar, &Karunasagar, 2007). All these limitations may be overcome by theapplication of molecular methods from which PCR/real-time PCR(qPCR) is the most popular (Bauer & Rorvik, 2007; Blanco-Abad,Ansede-Bermejo, Rodriguez-Castro, & Martinez-Urtaza, 2009;Malayil, Turner, Mote, Howe, & Lipp, 2011; Neogi et al., 2010;Sharma & Chaturvedi, 2006; Sheikh, Goodarzi, & Aslani, 2012).Furthermore, qPCR has been successfully used to detect andquantify food-borne pathogens, including pathogenic vibrios(Chapela et al., 2010; Gubala, 2006; Gubala & Proll, 2006; Kamio,Hara-Kudo, Miyasaka, Yahiro, & Konuma, 2008; Takahashi, Iwade,Konuma, & Hara-Kudo, 2005; Tyagi, Saravanan, Karunasagar, &Karunasagar, 2009). Even though a wide number of studies havebeen published targeting each individual bacterium, alone orcombined with several virulence factors, multiplex qPCR methodsfor simultaneous detection at the species level of V. cholerae,V. parahaemolyticus and V. vulnificus are scarce. Furthermore, thesequence of primers and probes are not always freely availablesince some are included in commercial detection kits [(Tebbs,Brzoska, Furtado, & Petrauskene), BAX� System Real-Time PCRassay for Vibrio (DuPont Qualicon, Wilmington DE)].

The aim of the present study was to develop and evaluate anopen straightforward multiplex qPCR method for simultaneousdetection of the three most important human pathogenic vibriosusing a novel combination of species-specific genes. The completemethod consisted of direct sample enrichment in Alkaline SalinePeptone Water (ASPW), and a simple DNA extraction step prior tomultiplex qPCR for the simultaneous detection of V. cholerae,V. parahaemolyticus and V. vulnificus. The method included as wellan Internal Amplification Control to evaluate possible matrix-associated inhibition of the qPCR reaction.

2. Materials and methods

2.1. Bacterial strains and culture media

V. cholerae CECT 514, V. parahemolyticus CECT 511, andV. vulnificus CECT 529 were selected as reference strains for theevaluation of the method. These microorganisms were purchasedby the Spanish Type Culture Collection (CECT). Bacteria were storedfrozen at �20 �C until use. All other organisms used in this study,and their origin are summarized in Table 1.

Fresh cultures of all strains used in the present study were ob-tained by inoculating 10 mL tubes of Tryptic Soy Broth (TSB, Bio-Mérieux S.A., France), and incubated at 37 �C overnight, except forBacillus subtilis which was incubated at 31 �C.

2.2. Culture method for detection of pathogenic Vibrio spp.

Detection of V. cholerae, V. parahemolyticus, and V. vulnificus bytraditional microbiology was carried out following the methodsdescribed in ISO/TS 21872-1: 2007 (method for detection ofV. cholerae and V. parahaemolyticus) and ISO/TS 21872-2: 2007(method for detection of potentially pathogenic vibrios otherthan V. cholerae and V. parahaemolyticus) (ISO, 2007a, 2007b).Briefly, both methods consisted of two enrichment steps inAlkaline Saline Peptone Water (ASPW, Biolife Italiana S.r.l., Italy),and plating on solid selective media. Thiosulphate Citrate Bilesalt Sucrose (TCBS, OXOID, Hampshire, England) is used in bothmethods for isolation of these bacteria. A second isolation me-dium is recommended as well, either to be chosen by the labo-ratory (ISO, 2007a) or to be selected among a list of media (ISO,2007b).

In the present study TCBS and CHROMagar� Vibrio (CHROMa-gar Microbiology, Paris, France), were used for the isolation ofV. cholerae and V. parahaemolyticus. Colistin Polymixin b-Cellobioseagar (CPC, SigmaeAldrich, St. Louis, USA) was selected as the sec-ondary isolation medium for V. vulnificus. The initial study alsoincluded HiCrome� Vibrio (HC, SigmaeAldrich, St. Louis, USA) as apossible secondary selective medium. A schematic presentation ofthe method is shown in Fig. 1.

Fig. 1. Schematic representation of ISO/TS 21872-1/2: 2007 and the newly developed qPCR method.

A. Garrido-Maestu et al. / Food Control 37 (2014) 371e379 373

2.3. DNA extraction

In order to perform the DNA extraction three different protocolswere used. A boiling procedure was applied to pure cultures, andtwo similar methods were compared for food samples (thermallysis DNA extraction was preceded by a Chelex-100 (Bio-Rad Lab-oratories, Inc., USA) purification step). DNA obtained from purecultures was quantified using a NanoDrop 1000 spectrophotometer(Thermo Fischer Scientific, Inc., USA) software ND-1000 v3.7.1. Foodsample protocols were compared according to the amount of DNAobtained (ng/mL), 260/280, and 260/230 absorbance ratios. Bothmethods were applied to 1 mL aliquots.

2.3.1. DNA extraction from pure culturesDNA extraction from pure cultures was done as previously

described by Blanco-Abad et al. (2009) (Blanco-Abad et al., 2009).The method consisted of several centrifugation and washing stepswith TriseEDTA 1X (TE 1X, SigmaeAldrich, St. Louis, USA) toeliminate food particles and inhibitory substances, and to concen-trate the bacteria. Bacteria were lysed in a boiling water bath for10 min. Finally cell debris was pelleted by centrifuging at13,000 rpm for 5 min.

2.3.2. DNA extraction from food samplesSamples were centrifuged for 2 min at 2000 rpm. The super-

natant was centrifuged again at 13,000 rpm for 5 min and bacterialpellet was resuspended in 1 mL of sterile PBS (Rossmanith,Krassnig, Wagner, & Hein, 2006), and centrifuged again under thesame conditions. Additionally, a washing step with milli-Q waterwas done for the first protocol and samples were centrifuged again(chelex 6%). Bacterial cell pellets were resuspended in 300 mL of 6%or 10% Chelex-100 and incubated at 56 �C for 15 min. After incu-bation, cells were lysed by placing the tubes in a boiling water bathfor 8 and 10min for the 6% and 10% chelex respectively, then chilledon ice for 2 min and finally centrifuged again at 13,000 rpm for5 min. Supernatant containing DNAwas stored at �20 �C until use.In order to evaluate the quality of the assays, a proficiency test wasalso analyzed.

2.4. Genes, primers and probes used for qPCR method

Genes tlh and vvhA were selected for species-specific detectionof V. parahaemolyticus, and for V. vulnificus respectively, as previ-ously described (Campbell & Wright, 2003; Nordstrom, Vickery,Blackstone, Murray, & DePaola, 2007).

A. Garrido-Maestu et al. / Food Control 37 (2014) 371e379374

Regarding V. cholerae, to our knowledge, no extensively usedqPCR primers and probes have been published for ompW, thus weredesigned in the present study. Several sequences were obtainedfrom GenBank (FJ462445, FJ462447, FJ462449, FJ462451, FJ462453,FJ462454, FJ462455, AF001009, and X51948) and aligned with CLCSequence Viewer version 6 (CLC bio) as shown in Fig. 2. Primers andprobe design was carried out with Primer3 (Untergasser et al.,2007). Specificity of primers and probe designed was verified per-forming BLAST (Basic Local Alignment Search Tool, http://blast.ncbi.nlm.nih.gov/Blast.cgi), and against the bacterial strains speci-fied in Table 1. Detailed sequences of primers and probes selectedfor the present study are summarized in Table 2.

2.5. Multiplex qPCR detection method for pathogenic Vibrio spp.

The qPCR reactionwas carried out in a final volume of 25 mL withthe following components: 12.5 mL of Brilliant III Ultra-Fast QPCRMaster Mix (Agilent Technologies, Inc., USA), 100 nM primers and100 nMprobewere used for vvhA, 75 nM primers andwere used fortlh and ompW, with 100 nM and 150 nM probe, respectively.Regarding Internal Amplification Control (IAC) 75 nM primers,45 nM probe and 8 � 102 copies of the chimerical DNA were addedper reaction. Two mL of template DNAwere added per reaction tube.

Stratagene Mx3005p thermocycler (Agilent Technologies, Inc.,USA) was used with the following thermal profile: 3 min at 95 �Cfor the activation of the polymerase (Hot Start), followed by 40cycles, each cycle consisted of a denaturation step at 95 �C for 15 s,and an annealing-extension step at 61 �C for 30 s.

2.6. Sampling and sample preparation

All food samples were received at the laboratory from externalsuppliers, either frozen or refrigerated. Samples were kept underthe same conditions until analysis. Water samples were collected insterile 500 mL plastic bottles, and stored at 4 �C prior to analysis,time no longer than 24 h.

Twenty-five grams of sample foodstuffs was weighted andmixed with 225 mL of ASPW. Samples were homogenized for 30 sin a laboratory stomacher at normal speed. Water samples wereprocessed as previously described (Garrido, Chapela, Ferreira, et al.,2012), briefly, a minimum of 500 mL, or the whole volume of thesample if less than 500 mL were available, were filtered through a0.45 mm membrane (Millipore, Ma USA), then the filter was placedin a stomacher bag with 50 mL of ASPW, and homogenized for 30 s.All samples were incubated at 35 �C for 18 � 2 h.

2.7. qPCR method evaluation

A complete evaluation of the newly developed qPCR methodwas performed taking into account three different parameters:

� Efficiency of the reaction to ensure correct amplification of thefour targets (including IAC).

Fig. 2. ompW sequence alignment. Primers and probe

� Limit of detection to evaluate the lowest concentration of eachmicroorganism that could be reliably detected.

� Method recovery from spiked samples from different originsthat was evaluated with respect to expected results.

2.7.1. qPCR efficiency determinationFor the evaluation of the qPCR efficiency, V. cholerae, V. para-

haemolyticus, and V. vulnificus reference strains were grown, DNAwas extracted as described in Materials and Methods 2.3.1respectively, and measured with the NanoDrop 1000spectrophotometer.

Initial DNA template was ten-fold serially diluted in sterile TE1X, and 2 mL of each dilution were used as template for qPCR. Ef-ficiency was determined in triplicate both, in simplex (DNA of onesingle target vibrio was used), and multiplex qPCR (template con-sisted of a mixture of all three target vibrios DNA) formats.

The Mx3005pro software automatically calculates the standardcurve for each run based on the Cycle threshold (Ct) versus amountof DNA in ng. Amplification efficiency was calculated according tothe following formula e ¼ 10�1/s�1, where “s” is the slope of thestandard curve (Blackstone et al., 2007; Nordstrom et al., 2007).

2.7.2. Evaluation of the limit of detection (LOD)The evaluation of the LOD of the method was performed by

inoculation of boiled frozen mussels or shrimps with low concen-trations of the three target bacteria. A total of 11 samples (25 geach) were analyzed, 10 spiked samples and one negative control.

Reference values of viable bacteria were obtained by growingthe microorganisms in an appropriate general broth (V. cholerae inTSB, V. parahaemolyticus, and V. vulnificus in TSB supplementedwith 2% NaCl) and incubated overnight at 37 �C. Turbidity of eachfresh culture was adjusted to 1.5e1.7 McFarland, either in TSB orTSB plus 2% NaCl. This adjustment established a theoretical startingconcentration of 5 � 108 cfu/mL. After enrichment, ten-fold serialdilutions were done in Alkaline Peptone Water (APW, Biolife Ital-iana S.r.l., Italy), and plated on Saline Nutrient Agar (SNA, Biokardiagnostics S.A., France) supplemented with 10 g/L of NaCl. Plateswere incubated overnight at 37 �C.

2.7.3. Method recoveryIn order to evaluate the quality of the method six parameters

were calculated using spiked samples. These included: relativesensitivity (SE), relative specificity (SP), relative accuracy (AC),positive and negative predictive values (PPV and NPV), and theindex kappa of concordance (k).

Each sample was classified as a Positive (PA) or Negative (NA)Accordance, whenever presenting the same result, positive ornegative, for the qPCR method and the expected result for spikedsamples. Negative Deviations (ND) include those samples expectedto be positive but with a negative result, and Positive Deviations(PD), correspond to the number of samples expected to be negativewhere a positive result was obtained.

designed for V. cholerae detection are highlighted.

Table 2Primers and probes for multiplex qPCR detection of V. cholerae, V. parahaemolyticus and V. vulnificus.

Microorganism Gene Primer/probe Sequence (50 to 30) Modifications Reference

V. cholerae ompW ompW-F TCAATGATAGCTGGTTCCTCAAC(5HEx)-(3IABkFQ) e

ompW-R CGATGATAAATACCCAAGGATTGA e This studyompW-Probe TGGTATGCCAATATTGAAACAACG 50-/5HEX//3IABkFQ/-30

V. parahaemolyticus tlh tlh-F ACTCAACACAAGAAGAGATCGACAA e

tlh-R GATGAGCGGTTGATGTCCAA e (Nordstrom et al., 2007)tlh-Probe CGCTCGCGTTCACGAAACCGT 50-/5FAM//3BHQ_1/-30

V. vulnificus vvhA vvha-F TGTTTATGGTGAGAACGGTGACA e

vvha-R TTCTTTATCTAGGCCCCAAACTTG e (Campbell & Wright, 2003)vvha-Probe CCGTTAACCGAACCACCCGCAA 50-/5Cy5//3BHQ_2/-30

IAC e IAC forward TCCAGGGCGAAAGTAAACGT e

e IAC reverse GGCGAGCCGTACGAACAC e (Calvo Martinez-Planells,Pardos-Bosch, & Garcia-Gil, 2008)

e IAC probe CCCAGTTGGCTGATCACTTTCG 50-/5TexRd-XN//3BHQ_2/-30

IABkFQ: is a trademark of IDT and stands for Iowa Black Flurophore Quencher.

A. Garrido-Maestu et al. / Food Control 37 (2014) 371e379 375

SE was defined as the percentage of positive samples giving acorrect positive signal (SE ¼ PA/(PA þ ND) � 100).

SP was defined as the percentage of negative samples giving acorrect negative signal (SP ¼ NA/(PD þ NA) � 100).

AC was defined as the degree of correspondence between theresponse obtained by the expected result and the method onidentical samples (AC¼ [(PAþNA)/N]� 100; where N¼Number ofanalyzed samples).

Positive and Negative Predictive Values (PPV and NPV) aremeasures of the performance of the method, and give the proba-bility of a sample being really positive or negative when themethod shows a positive or negative result PPV ¼ [(PA/PA) þ PD] � 100; NPV ¼ [(NA/NA þ ND) � 100].

Finally the index kappa of concordance shows the degree ofconcordance between the method and the expected resultk ¼ 2 � (PA � NA e ND � PD)/[(PA þ PD) � (PD þ NA) þ (PAþ ND)� (ND þ NA)] (Anderson et al.,2011; Tomas, Rodrigo, Hernandez, & Ferrus, 2009).

2.8. Statistical analysis

All data obtained according to the different DNA extractionprotocols were compared using ManneWhitney U test SPSS 15.0software (SPSS Inc., Chicago, IL, USA).

3. Results

3.1. Specificity of the qPCR method

All 44 strains, covering 20 different species, detailed in Table 1were correctly identified with the multiplex qPCR method devel-oped. No interference was observed among all primers and probesselected for vvhA, tlh, and ompW genes. Correct amplification of theIAC was also obtained.

3.2. DNA extraction methods

The 6% chelex method allowed to reach values of135.83 � 24.46 ng/mL and 1.81 � 0.03 for DNA concentration and

Table 3qPCR amplification evaluation in multiplex and simplex format.

Simplex

Efficiency (%) b r2

V. cholerae 101.2 � 5.9 �3.298 � 0.135 0.997 �V. parahaemolyticus 102.0 � 2.3 �3.277 � 0.054 0.999 �V. vulnificus 100.7 � 1.9 �3.306 � 0.045 1.000 �

All values are averages of three replicates � standard deviation. “b” represents the slope

260/280 ratio, meanwhile with the 10% chelex protocol valuesobtained were 187.68 � 44.84 ng/mL and 1.82 � 0.02 respectively.Even though differences were observed, specially for DNA con-centration, the statistical comparison indicated that the differenceswere not significant (p > 0.05). On the contrary, the method with6% chelex purification, showed statistically higher values for the260/230 ratio (0.88 � 0.02 with respect to 0.82 � 0.01). It wasobserved that this slight difference did not affect the qPCR perfor-mance, thus the fastest method (10% chelex) was selected forfurther analysis.

3.3. qPCR efficiency determination

Evaluation of the qPCR efficiency was performed in triplicateover a range of five orders of magnitude. It was evaluated indi-vidually for each bacterium and with a mix of DNA in equalproportions.

Individual efficiencies returned values between 100.7% and102.0% depending on the microorganism. Regarding the multiplexefficiency, values obtained were comprised between 96.7% and103.1%. Exact values are detailed in Table 3.

3.4. Evaluation of the LOD

The reference value obtained for viable bacteria returned viablecounts of 3 cfu/25 g for V. vulnificus, 7 cfu/25 g forV. parahaemolyticus, and 4 cfu/25 g for V. cholerae in the evaluationof the LOD. With these counts 100% presences in each microor-ganism were obtained.

3.5. Method recovery calculation

A total of 81 spiked samples were analyzed in order to evaluatethe quality of the method, and the following parameters wereincluded: relative sensitivity (SE), relative specificity (SP), relativeaccuracy (AC), positive and negative predictive values (PPV andNPV), and the index kappa of concordance (k). Only one discrepantresult was observed among all samples analyzed. One PD was ob-tained for V. parahaemolyticus, results are summarized in Table 5.

Multiplex

Efficiency (%) b r2

0.001 96.5 � 4.2 �3.412 � 0.108 0.992 � 0.0070.000 95.7 � 6.5 �3.437 � 0.165 0.998 � 0.0020.001 103.1 � 5.3 �3.254 � 0.118 0.900 � 0.001

and r2 the correlation coefficient.

Table 4Results of spiked samples and qPCR method evaluation.

Bivalves Crustaceans Water Fish Proficiency testsa Evaluation

PA NA PD ND PA NA PD ND PA NA PD ND PA NA PD ND PA NA PD ND SE SP AC PPV NPV k

Vc 22 7 e e 24 11 e e 3 8 e e 1 4 e e e 1 e e 100 100 100 100 100 100Vp 22 6 1 e 5 30 e e 5 6 e e e 5 e e 1 e e e 100 98 99 97 100 96Vv 22 7 e e 5 30 e e 5 6 e e e 5 e e e 1 e e 100 100 100 100 100 100

a Two different proficiency test consistent of four samples. N: number of samples analyzed. PA: Positive Agreement, NA: Negative Agreement, PD: Positive Deviation, ND:Negative Deviation. Vc: V. choleare, Vp: V. parahaemolyticus, Vv: V. vulnificus. SE: Relative Sensitivity, SP: Relative Specificity, AC: Relative Accuracy, PPV: Positive PredictiveValue, NPV: Negative Predictive Value, k: kappa index of concordance.

A. Garrido-Maestu et al. / Food Control 37 (2014) 371e379376

These data were used to calculate the SE, SP, AC, PPV, NPV, and k

index of the method. All these parameters showed values higherthan 96%, as listed in Table 4.

Additionally, 58 natural non-spiked samples from different or-igins were analyzed in parallel with bothmethods, ISO and qPCR forpathogenic Vibrio spp. V. vulnificus was not detected in any ofthe samples analyzed by using either method. RegardingV. parahaemolyticus only one positive samplewas detected with theISO method, meanwhile four were detected with the qPCR. Moresurprising results were obtained for V. cholerae finding seven pos-itive samples with the ISO procedure and 13 by using qPCR. Theseresults are summarized in Table 5.

4. Discussion

The incidence of vibriosis has increased for the last few years,consequently the interest for detecting these pathogenic bacteriahas also increased. In this context “The Cholera and Other VibrioIllness Surveillance System” (COVIS), which was initiated by CDC,FDA, and the Gulf Coast states in 1988, received data of vibriosiscases from 41 states in 2009. The number of diagnosed cases ofvibriosis in the USAduring the period comprised from 2005 to 2009showed a significant increase, from 231 up to 837 (CDC, 2012). InSeptember 2010 an expert meetingwas organized by theWHO/FAOin order to evaluate several factors related to incidence, pathoge-nicity, available methods, etc. for V. parahaemoltyticus andV. vulnificus (WHO & FAO, 2010). All these actions highlight thepotential threat posed by these bacteria to human health. Specificactions have been carried out in order to preserve food safety, suchas the inclusion of zero-tolerance regarding V. cholerae indepen-dently of its serotype. Moreover, certain recommendationsregarding the other main pathogenic Vibrio species were published,like controlling initial numbers of V. parahaemolyticus or extendingsurveillance to V. vulnificus in samples from countries at risk.

Detection of pathogenic vibrios is generally performed by offi-cial culture-dependent methods, reported previously by ISO orBAM. From an industrial point of view the disadvantages of culturemethods are critical due to the need of fast and reliable methods ascertain food products present very short shelf lives. Molecularmethods represent a possible solution that fulfills all these desired

Table 5Results comparison with ISO and qPCR methods applied to natural non-spiked samples.

Food type N ISO

Vc Vp Vv

þ � þ � þFish 32 0 32 0 32 0Bivalves 7 0 7 1 6 0Crustaceans 4 0 4 0 4 0Algae 6 4 2 0 6 0Water 9 3 6 0 9 0

N: indicates the total number of samples analyzed.þ/� sign indicates positive/negative re

requirements. Nowadays several possibilities for the detection ofpathogenic vibrios through molecular approaches have beendescribed, as NASBA, LAMP or ligation reaction, among others(Cariani et al., 2012; Drake et al., 2007; Fykse et al., 2012; Surasilpet al., 2011; Yamazaki, Seto, Taguchi, Ishibashi, & Inoue, 2008).There is no doubt that among them the most popular approach isbased on PCR or, more recently, qPCR (Blackstone et al., 2007; Fedioet al., 2007; Gubala & Proll, 2006; Izumiya et al., 2011; Takahashiet al., 2005).

In the present study three well known species-specific geneswere selected for the simultaneous detection of the main humanpathogenic vibrio species: V. cholerae (ompW) (Nandi et al., 2000),V. parahaemolyticus (tlh) (Nordstrom et al., 2007), and V. vulnificus(vvhA) (Campbell & Wright, 2003). Surprisingly, few previousstudies have approached simultaneous detection of these threebacteria at the species level (Bauer & Rorvik, 2007; Izumiya et al.,2011; Kim, Lee, Lee, & Cho, 2012; Neogi et al., 2010; Tebbs,Brzoska, Furtado, & Petrauskene, 2011), but most of them areconventional PCR-based or do not provide the sequences applied.Furthermore to our knowledge none of them have previouslyapplied the gene combination selected in the present study orperformed such an extensive in-house validation.

The newly designed primers and probe targeted an outer mem-brane protein OmpW gene (ompW) for V. cholerae detection. Highsequence homology in this genewas reported between strains (only0.2e2% variation). Furthermore, previous studies have demon-strated better detection results with ompW thanwith other targets,as toxR (Nandi et al., 2000). Additionally, this target has beenwidelyused for this purpose (Alamet al., 2006; Chatterjee et al., 2009; Goel,Ponmariappan, Kamboj, & Singh, 2007; Jain, Goel, Bhattacharya,Ghatole, & Kamboj, 2011; Phetsouvanh et al., 2008; Sharma &Chaturvedi, 2006; Sheikh et al., 2012; Shuan Ju Teh et al., 2009;Tamrakar, Jain, Goel, Kamboj, & Singh 2009). Recently, the outermembrane lipoprotein gene (lolB, previously hemH) has been pro-posed as a better target forV. cholerae species detection (Lalitha et al.2008), but it has not been extensively tested and only a few studiesreport its application (Cho et al. 2013; Chua et al., 2011; Zhang et al.,2012). Design of primers and probe for ompW detection used in thepresent study, was based on eight sequences obtained from Gen-bank. Specificitywas verified in silico through BLASTanalysis aswell

qPCR

Vc Vp Vv

� þ � þ � þ �32 5 25 1 31 0 327 0 7 3 4 0 74 0 4 0 4 0 46 5 1 0 6 0 69 3 6 0 9 0 9

sults with specifiedmethod. Vc: V. cholerae, Vp: V. parahaemolyticus, Vv: V. vulnificus.

A. Garrido-Maestu et al. / Food Control 37 (2014) 371e379 377

as against five V. cholerae strains (O1, O139 from type culture col-lections and threenatural strains isolated fromfood samples).Whencombined with primers and probes for V. parahaemolyticus,V. vulnificus and IAC, no cross-reactivity was observed with eachtarget or bacterial strain tested (16 target Vibrio spp., 5 non-tagetVibrio spp., and 23 non-vibrio strains).

For the isolation of the pathogenic vibrios of interest, severalselective agars were considered (TCBS, CHROMagar� Vibrio, CPCand HiCrome Vibrio�). HiCrome Vibrio� was discarded since itcould not differentiate between V. parahaemolyticus and Vibrioalginolyticus as both appearing bluish-green. Among the othermedia, most confirmed isolates, independently on the species,were obtained from CHROMagar� Vibrio (data not shown). Thisobservation was in good agreement with other studies (Blanco-Abad et al., 2009).

Previous reports have demonstrated that food extracts in DNAisolates may affect qPCR detection (Wright et al., 2007). This facthighlights the need of applying a DNA extraction/purification pro-cedure before the analysis. The present study proved that appli-cation of Chelex-100 at 10% was a fast (about 40 min includingbacteria concentration, purification, and lysis), simple, and eco-nomic DNA extraction protocol. Furthermore, high DNA quality andconcentration for qPCR was obtained.

Before application of the qPCR method, the evaluation of theqPCR efficiency to ensure correct amplification of all targets,including the IAC, was performed. Data obtained fitted into ex-pected values (Raymaekers, Smets, Maes, & Cartuyvels 2009),reporting an average efficiency between 95.7% and 103.1%. Novariation between amplification of one, two or the three targets,along with the IAC, was observed.

Finally, evaluation of the complete qPCRmethod was performedtaking into consideration the results obtained from spiked samples.Regarding the LOD, 100% detection was obtained by the presentnew qPCR method with a very low LOD (4 cfu/25 g for V. cholerae,7 cfu/25 g for V. parahemolyticus, and 3 cfu/25 g for V. vulnificus).These results are comparable to other methods previously devel-oped for the detection of single or multiple vibrios (Chapela et al.2010; Fedio et al. 2007; Garrido, Chapela, Ferreira, et al., 2012;Kim et al., 2012). Differences in the LOD for V. cholerae were notassociated with the type of food matrix chosen. Few deviationswere observed out of the 81 samples analyzed. These samplescovered five different food categories including proficiency tests.Only one PD in V. parahemolyticus detection was obtained withfrozen boiled mussel expected to be negative, probably due to lackof homogeneity within the batch. Even considering this slight de-viation high quality of the method was observed, as all values forSE, SP, AC, PPV, NPV returned values over 96%. With minimal dif-ferences all values obtained are comparable to previous PCR andqPCR studies (Chua et al., 2011; Garrido et al. 2013; Garrido,Chapela, Roman, et al., 2012). According to the index kappa, “verygood” concordance respect to expected results, was obtained as thevalue was comprised between 96% and 100% (Danilla et al., 2005).

Finally the newly developed method was applied to 58 naturalnon-spiked samples of five different foodstuffs categories. No dif-ferences were obtained between both methods for the detection ofV. vulnificus, since no positive samples were detected. On the con-trary, higher detection frequencies were obtained by qPCR than bytraditional culture method for V. cholerae (seven positives with ISOagainst 13 by qPCR) and V. parahaemolyticus (one positive with ISOagainst four by qPCR). These results are in agreement with theadvantages previously mentioned of molecular methods respect toculture-based ones. Previous studies had already highlightedhigher detection frequencies by PCR versus culture methods(Blanco-Abad et al., 2009; Fedio et al. 2007; Rosec, Causse, Cruz,Rauzier, & Carnat, 2012; Rosec, Simon, Causse, & Boudjemaa, 2009).

5. Conclusions

The newly developed qPCR method proved fast, economicand reliable for the simultaneous detection of the main patho-genic vibrio species (V. cholerae, V. parahaemolyticus, andV. vunificus). This facts were observed after extensive testingin vitro and against spiked and natural food and environmentalsamples.

Acknowledgments

This work is financially supported by the Secretary General forthe Sea of the Spanish Ministry of Agricultural, Land and MarineResources (MARM), by order ARM/1193/2009.

Authors also thank Victoria Docampo and Angeles Marcote fortechnical assistance.

References

Alam, M., Sadique, A., Nur, A. H., Bhuiyan, N. A., Nair, G. B., Siddique, A. K., et al.(2006). Effect of transport at ambient temperature on detection and isolation ofVibrio cholerae from environmental samples. Applied and EnvironmentalMicrobiology, 72(3), 2185e2190.

Anderson, A., Pietsch, K., Zucker, R., Mayr, A., Muller-Hohe, E., Messelhausser, U.,et al. (2011). Validation of a duplex real-time PCR for the detection of Salmonellaspp. in different food products. Food Analytical Methods, 4(3), 259e267.

Bauer, A., & Rorvik, L. M. (2007). A novel multiplex PCR for the identification ofVibrio parahaemolyticus, Vibrio cholerae and Vibrio vulnificus. Letters in AppliedMicrobiology, 45, 371e375.

Blackstone, G. M., Nordstrom, J. L., Bowen, M. D., Meyer, R. F., Imbro, P., & DePaola, A.(2007). Use of a real time PCR assay for detection of the ctxA gene of Vibriocholerae in an environmental survey of Mobile Bay. Journal of MicrobiologicalMethods, 68(2), 254e259.

Blackwell, K. D., & Oliver, J. D. (2008). The ecology of Vibrio vulnificus, Vibrio cholerae,and Vibrio parahaemolyticus in North Carolina estuaries. Journal of Microbiology,46(2), 146e153.

Blanco-Abad, V., Ansede-Bermejo, J., Rodriguez-Castro, A., & Martinez-Urtaza, J.(2009). Evaluation of different procedures for the optimized detection of Vibrioparahaemolyticus in mussels and environmental samples. International Journalof Food Microbiology, 129(3), 229e236.

Broberg, C. A., Calder, T. J., & Orth, K. (2011). Vibrio parahaemolyticus cell biology andpathogenicity determinants. Microbes and Infection, 13(12e13), 992e1001.

Calvo, L., Martinez-Planells, A., Pardos-Bosch, J., & Garcia-Gil, L. J. (2008). A newreal-time PCR assay for the specific detection of Salmonella spp. targeting thebipA gene. Food Analytical Methods, 1(4), 236e242.

Campbell, M. S., & Wright, A. C. (2003). Real-time PCR analysis of Vibrio vulnificusfrom oysters. Applied and Environmental Microbiology, 69(12), 7137e7144.

Canadian Food Inspection Agency. Fish products standards and methods manual. In.Cariani, A., Piano, A., Consolandi, C., Severgnini, M., Castiglioni, B., Caredda, G., et al.

(2012). Detection and characterization of pathogenic vibrios in shellfish by aligation detection reaction-universal array approach. International Journal ofFood Microbiology, 153(3), 474e482.

CDC. (2012). In Investigation update: Multistate outbreak of E. coli O157:H7 infectionsassociated with in-shell hazelnuts.

CLC bio, A. S. CLC sequence viewer. In (6.6.2 ed.).Chapela, M. J., Fajardo, P., Garrido, A., Cabado, A. G., Ferreira, M., Lago, J., et al. (2010).

Comparison between a TaqMan polymerase chain reaction assay and a culturemethod for ctx-positive Vibrio cholerae detection. Journal of Agricultural andFood Chemistry, 58(7), 4051e4055.

Chatterjee, S., Ghosh, K., Raychoudhuri, A., Chowdhury, G., Bhattacharya, M. K.,Mukhopadhyay, A. K., et al. (2009). Incidence, virulence factors, and clonalityamong clinical strains of non-O1, non-O139 Vibrio cholerae isolates from hos-pitalized diarrheal patients in Kolkata, India. Journal of Clinical Microbiology,47(4), 1087e1095.

Cho, M. S., Ahn, T.-Y., Joh, K., Paik, S.-Y., Kwon, O.-S., Jheong, W.-H., et al. (2013).A novel marker for the species-specific detection and quantitation of Vibriocholerae by targeting an outer membrane lipoprotein lolB gene. Journal ofMicrobiology and Biotechnology, 23(4), 555e559.

Chomvarin, C., Namwat, W., Wongwajana, S., Alam, M., Thaew-Nonngiew, K.,Sinchaturus, A., et al. (2007). Application of duplex-PCR in rapid and reliabledetection of toxigenic Vibrio cholerae in water samples in Thailand. Journal ofGeneral and Applied Microbiology, 53(4), 229e237.

Chua, A. L., Elina, H. T., Lim, B. H., Yean, C. Y., Ravichandran, M., & Lalitha, P.(2011). Development of a dry reagent-based triplex PCR for the detection oftoxigenic and non-toxigenic Vibrio cholerae. Journal of Medical Microbiology,60(4), 481e485.

Danilla, S., Andrades, P., Gomez, M. E., Chamorro, M., Leniz, P., Pineros, J. L., et al.(2005). Concordance between qualitative and quantitative cultures in burnedpatients e analysis of 2886 cultures. Burns, 31(8), 967e971.

A. Garrido-Maestu et al. / Food Control 37 (2014) 371e379378

Drake, S. L., DePaola, A., & Jaykus, L. A. (2007). An overview of Vibrio vulnificus andVibrio parahaemolyticus. Comprehensive Reviews in Food Science and Food Safety,6(4), 120e144.

(EC), C. R. (2005). Microbiological criteria for foodstuffs. Official Journal of the Eu-ropean Union, 2073.

(EC), C. R. (2007). Amending regulation (EC) No 2073/2005 on microbiologicalcriteria for foodstuffs. Official Journal of the European Union, 1441.

(EC), C. R. (2010). Amending regulation (EC) No 2073/2005 on microbiologicalcriteria for foodstuffs as regards Enterobacteriaceae in pasteurised milk andother pasteurised liquid dairy products and Listeria monocytogenes in foodgrade salt. Official Journal of the European Union, 365.

FDA.. (2011). In Fish and fishery products hazards and controls guidance (4th ed).Department of Health and Human Services, Public Health Service, Food and DrugAdministration, Center For Food Safety and Applied Nutrition & O. O. F. Safety.

Fedio, W., Blackstone, G. M., Kikuta-Oshima, L., Wendakoon, C., McGrath, T. H., &DePaola, A. (2007). Rapid detection of the Vibrio cholerae ctx gene in food en-richments using real-time polymerase chain reaction. Journal of AOAC Interna-tional, 90(5), 1278e1283.

Fykse, E. M., Nilsen, T., Nielsen, A. D., Tryland, I., Delacroix, S., & Blatny, J. M. (2012).Real-time PCR and NASBA for rapid and sensitive detection of Vibrio cholerae inballast water. Marine Pollution Bulletin, 64(2), 200e206.

Garrido, A., Chapela, M.-J., Román, B., Fajardo, P., Lago, J., Vieites, J. M., et al. (2013).A new multiplex real-time PCR developed method for Salmonella spp. andListeria monocytogenes detection in food and environmental samples. FoodControl, 30(1), 76e85.

Garrido, A., Chapela, M. J., Ferreira, M., Atanassova, M., Fajardo, P., Lago, J., et al.(2012). Development of a multiplex real-time PCR method for pathogenicVibrio parahaemolyticus detection (tdhþ and trhþ). Food Control, 24(1e2),128e135.

Garrido, A., Chapela, M. J., Roman, B., Ferreira, M., Lago, J., Vieites, J. M., et al. (2012).Development of a multiplex real-time PCR method for simultaneous detectionof Salmonella enterica, Shigella flexneri and Listeria monocytogenes in processedfood samples. European Food Research and Technology, 234(4), 571e580.

Goel, A. K., Ponmariappan, S., Kamboj, D. V., & Singh, L. (2007). Single multiplexpolymerase chain reaction for environmental surveillance of toxigenic-pathogenic O1 and non-O1 Vibrio cholerae. Folia Microbiologica, 52(1), 81e85.

Griffitt, K. J., & Grimes, D. J. (August 2013). A novel agar formulation for isolation anddirect enumeration of Vibrio vulnificus from oyster tissue. Journal of Microbio-logical Methods, 94(2), 98e102.

Gubala, A. J. (2006). Multiplex real-time PCR detection of Vibrio cholerae. Journal ofMicrobiological Methods, 65(2), 278e293.

Gubala, A. J., & Proll, D. F. (2006). Molecular-beacon multiplex real-time PCR assayfor detection of Vibrio cholerae. Applied and Environmental Microbiology, 72(9),6424e6428.

Gulig, P. A., de Crecy-Lagard, V., Wright, A. C., Walts, B., Telonis-Scott, M., &McIntyre, L. M. (2010). SOLiD sequencing of four Vibrio vulnificus genomesenables comparative genomic analysis and identification of candidate clade-specific virulence genes. BMC Genomics, 11.

Hiyoshi, H., Kodama, T., Iida, T., & Honda, T. (2010). Contribution of Vibrio para-haemolyticus virulence factors to cytotoxicity, enterotoxicity, and lethality inmice. Infection and Immunity, 78(4), 1772e1780.

ISO. (2007a). In (1st ed).Microbiology of food and animal feeding stuffs e Horizontalmethod for the detection of potentially enteropathogenic Vibrio spp.-Part 1:Detection of Vibrio parahaemolyticus and Vibrio cholerae (Vol. 21872-1; pp. 19).

ISO. (2007b). In (1 st ed).Microbiology of food and animal feeding stuffs e Horizontalmethod for the detection of potentially enteropathogenic Vibrio spp.-Part 2:Detection of species other than Vibrio parahaemolyticus and Vibrio cholerae (Vol.21872-2; pp. 25).

Izumiya, H., Matsumoto, K., Yahiro, S., Lee, J., Morita, M., Yamamoto, S., et al. (2011).Multiplex PCR assay for identification of three major pathogenic Vibrio spp.,Vibrio cholerae, Vibrio parahaemolyticus, and Vibrio vulnificus. Molecular andCellular Probes, 25(4), 174e176.

Jain, M., Goel, A. K., Bhattacharya, P., Ghatole, M., & Kamboj, D. V. (2011). Multidrugresistant Vibrio cholerae O1 El Tor carrying classical ctxB allele involved in acholera outbreak in South Western India. Acta Tropica, 117(2), 152e156.

Jones, M. K., & Oliver, J. D. (2009). Vibrio vulnificus: disease and pathogenesis.Infection and Immunity, 77(5), 1723e1733.

Kamio, A., Hara-Kudo, Y., Miyasaka, J., Yahiro, S., & Konuma, H. (2008). Efficiency ofreal-time polymerase chain reaction assay to detect Vibrio vulnificus in seawater.International Journal of Hygiene and Environmental Health, 211(5e6), 518e523.

Kaysner, C. A., & Angelo DePaola, J. (2004). Vibrio. Bacteriological analytical manual.Kim, H.-J., Lee, H.-J., Lee, K.-H., & Cho, J.-C. (2012). Simultaneous detection of

pathogenic Vibrio species using multiplex real-time PCR. Food Control, 23(2),491e498.

Lalitha, P., Suraiya, M. N. S., Lim, K. L., Lee, S. Y., Haslindawaty, A. R. N., Chan, Y. Y.,et al. (2008). Analysis of lolB gene sequence and its use in the development of aPCR assay for the detection of Vibrio cholerae. Journal of Microbiological Methods,75(1), 142e144.

Lyon, W. J. (2001). TaqMan PCR for detection of Vibrio cholerae O1, O139, non-O1,and non-O139 in pure cultures, raw oysters, and synthetic seawater. Appliedand Environmental Microbiology, 67(10), 4685e4693.

Malayil, L., Turner, J. W., Mote, B. L., Howe, K., & Lipp, E. K. (2011). Evaluation ofenrichment media for improved PCR-based detection of V. cholerae andV. vulnificus from estuarine water and plankton. Journal of Applied Microbiology,110(6), 1470e1475.

Nandi, B., Nandy, R. K., Mukhopadhyay, S., Nair, G. B., Shimada, T., & Ghose, A. C.(2000). Rapid method for species-specific identification of Vibrio cholerae usingprimers targeted to the gene of outer membrane protein OmpW. Journal ofClinical Microbiology, 38(11), 4145e4151.

Neogi, S. B., Chowdhury, N., Asakura, M., Hinenoya, A., Haldar, S., Saidi, S. M., et al.(2010). A highly sensitive and specific multiplex PCR assay for simultaneousdetection of Vibrio cholerae, Vibrio parahaemolyticus and Vibrio vulnificus. Lettersin Applied Microbiology, 51(3), 293e300.

Nordstrom, J. L., Vickery, M. C. L., Blackstone, G. M., Murray, S. L., & DePaola, A. (2007).Development of a multiplex real-time PCR assay with an internal amplificationcontrol for the detection of total and pathogenic Vibrio parahaemolyticus bacteriain oysters. Applied and Environmental Microbiology, 73, 5840e5847.

Phetsouvanh, R., Nakatsu, M., Arakawa, E., Davong, V., Vongsouvath, M., Lattana, O.,et al. (2008). Fatal bacteremia due to immotile Vibrio cholerae serogroup O21 inVientiane, Laos e a case report. Annals of Clinical Microbiology and Antimicro-bials, 7, 10.

Raymaekers, M., Smets, R., Maes, B., & Cartuyvels, R. (2009). Checklist for optimi-zation and validation of real-time PCR assays. Journal of Clinical LaboratoryAnalysis, 23(3), 145e151.

Roig, F. J., Sanjuan, E., Llorens, A., & Amaro, C. (2010). pilF Polymorphism-based PCRto distinguish Vibrio vulnificus strains potentially dangerous to public health.Applied and Environmental Microbiology, 76(5), 1328e1333.

Rosec, J.-P., Causse, V., Cruz, B., Rauzier, J., & Carnat, L. (2012). The internationalstandard ISO/TS 21872-1 to study the occurrence of total and pathogenic Vibrioparahaemolyticus and Vibrio cholerae in seafood: ITS improvement by use of achromogenic medium and PCR. International Journal of Food Microbiology,157(2), 189e194.

Rosec, J. P., Simon, M., Causse, V., & Boudjemaa, M. (2009). Detection of total andpathogenic Vibrio parahaemolyticus in shellfish: comparison of PCR protocolsusing pR72H or toxR targets with a culture method. International Journal of FoodMicrobiology, 129(2), 136e145.

Rossmanith, P., Krassnig, M., Wagner, M., & Hein, I. (2006). Detection of Listeriamonocytogenes in food using a combined enrichment/real-time PCR methodtargeting the prfA gene. Research in Microbiology, 157(8), 763e771.

Saravanan, V., Kumar, H. S., Karunasagar, I., & Karunasagar, I. (2007). Putativevirulence genes of Vibrio cholerae from seafoods and the coastal environment ofSouthwest India. International Journal of Food Microbiology, 119(3), 329e333.

Sharma, A., & Chaturvedi, A. N. (2006). Prevalence of virulence genes (ctxA, stn,OmpWand tcpA) among non-O1 Vibrio cholerae isolated from fresh water environment.International Journal of Hygiene and Environmental Health, 209(6), 521e526.

Sheikh, A. F., Goodarzi, H., & Aslani, S. (2012). Identification of Vibrio choleraepathogenicity island (ctxA, OmpW and tcpA) in non-O139 and non-O1V. cholerae strains isolated from Karun River in Ahvaz, Iran. African Journal ofMicrobiology Research, 6(6), 1185e1189.

Shuan Ju Teh, C., Lin Thong, K., Tein Ngoi, S., Ahmad, N., Balakrish Nair, G., &Ramamurthy, T. (2009). Molecular characterization of serogrouping and viru-lence genes of Malaysian Vibrio cholerae isolated from different sources. TheJournal of General and Applied Microbiology, 55(6), 419e425.

Su, Y. C., & Liu, C. C. (2007). Vibrio parahaemolyticus: a concern of seafood safety.Food Microbiology, 24(6), 549e558.

Surasilp, T., Longyant, S., Rukpratanporn, S., Sridulyakul, P., Sithigorngul, P., &Chaivisuthangkura, P. (2011). Rapid and sensitive detection of Vibrio vulnificusby loop-mediated isothermal amplification combined with lateral flow dipsticktargeted to rpoS gene. Molecular and Cellular Probes, 25(4), 158e163.

Takahashi, H., Iwade, Y., Konuma, H., & Hara-Kudo, Y. (2005). Development of aquantitative real-time PCR method for estimation of the total number of Vibrioparahaemolyticus in contaminated shellfish and seawater. Journal of Food Pro-tection, 68(5), 1083e1088.

Tamrakar, A. K., Jain, M., Goel, A. K., Kamboj, D. V., & Singh, L. (2009). Character-ization of Vibrio cholerae from deep ground water in a cholera endemic area inCentral India. Indian Journal of Microbiology, 49(3), 271e275.

Tebbs, R. S., Brzoska, P. M., Furtado, M. R., & Petrauskene, O. V. (2011). Design andvalidation of a novel multiplex real-time PCR assay for Vibrio pathogen detec-tion. Journal of Food Protection, 74(6), 939e948.

Thompson, F. L., Iida, T., & Swings, J. (2004). Biodiversity of vibrios. Microbiology andMolecular Biology Reviews, 68(3), 403e431.

Tomas, D., Rodrigo, A., Hernandez, M., & Ferrus, M. A. (2009). Validation of real-timePCR and enzyme-linked fluorescent assay-based methods for detection of Sal-monella spp. in chicken feces samples. Food Analytical Methods, 2(3), 180e189.

Tyagi, A., Saravanan, V., Karunasagar, I., & Karunasagar, I. (2009). Detection of Vibrioparahaemolyticus in tropical shellfish by SYBR green real-time PCR and evalu-ation of three enrichment media. International Journal of Food Microbiology,129(2), 124e130.

Untergasser, A., Nijveen, H., Rao, X., Bisseling, T., Geurts, R., & Leunissen, J. A. M.(2007). Primer3Plus, an enhanced web interface to Primer3. Nucleic AcidsResearch, 35(Suppl. 2), W71eW74.

WHO & FAO. (2010). In Joint FAO/WHO expert meeting on risk assessment tools forVibrio parahaemolyticus and Vibrio vulnificus associated with seafoods call fordata and experts.

Williams, T., Froelich,B., &Oliver, J.D. (June2013). Anewculture-basedmethod for theimproved identification of Vibrio vulnificus from environmental samples,reducing the need formolecular confirmation. Journal ofMicrobiological Methods,93(3), 277e283.

Wright, A. C., Garrido, V., Debuex, G., Farrell-Evans, M., Mudbidri, A. A., &Otwell, W. S. (2007). Evaluation of postharvest-processed oysters by using PCR-

A. Garrido-Maestu et al. / Food Control 37 (2014) 371e379 379

based most-probable-number enumeration of Vibrio vulnificus bacteria. Appliedand Environmental Microbiology, 73(22), 7477e7481.

Yamazaki, W., Seto, K., Taguchi, M., Ishibashi, M., & Inoue, K. (2008). Sensitive andrapid detection of cholera toxin-producing Vibrio cholerae using a loop-mediated isothermal amplification. BMC Microbiology, 8.

Zhang, X.-J., Yao, D.-R., Yan, B.-L., Bi, K.-R., Liang, L.-G., & Qin, G.-M. (2012). Identi-fication of Vibrio cholerae as a causative bacterium for an ulcer disease ofcultured loach Misgurnus anguillicaudatus in China. African Journal of Microbi-ology Research, 6(9), 2060e2070.