IdentiÞcation of Salmonella Serotypes Isolated from Cantaloupe …faz.ujed.mx/Posgrado/maos/AUTOEVALUACION/CATEGORIAS/3... · 2011-03-21 · IdentiÞcation of Salmonella Serotypes

2217

Journal of Food Protection, Vol. 71, No. 11, 2008, Pages 2217–2222

Identification of Salmonella Serotypes Isolated from Cantaloupeand Chile Pepper Production Systems in Mexico by

PCR–Restriction Fragment Length Polymorphism

MIGUEL A. GALLEGOS-ROBLES,1 ALBERTO MORALES-LOREDO,2* GENOVEVA ALVAREZ-OJEDA,3

ADRIÁN VEGA-P.,4 YAZMÍN CHEW-M.,4 SIXTO VELARDE,5 AND PINA FRATAMICO6

1Facultad de Agricultura y Zootecnia, Universidad Juarez del Estado de Durango, Venecia, Gomez Palacio, Durango, C.P. 35000, México;2INIFAP, Campo Experimental General Terán, Carretera Montemorelos-China Km. 31, General Terán, Nuevo León, C.P. 67400, México;

3INIFAP, Campo Experimental Sur de Tamaulipas, Carretera Tampico-Mante Km. 55, Estación Cuauhtémoc, México; 4INIFAP, Campo ExperimentalLa Laguna, Blvd. Jose Santos Valdez No. 1200, Col. Mariano Matamoros, C.P. 27440, México; 5INIFAP, Campo Experimental Valle de Culiacán,

Carretera Culiacán-El dorado Km. 17.5, Culiacán, Sinaloa, México; and 6U.S. Department of Agriculture, Agricultural Research Service,Microbial Food Safety Research Unit, Eastern Regional Research Center, 600 East Mermaid Lane, Wyndmoor, Pennsylvania 19038, USA

MS 08-107: Received 27 February 2008/Accepted 31 May 2008

ABSTRACT

A study was conducted in 2006 to determine the prevalence of Salmonella on three cantaloupe farms in Matamoros,Coahuila, Mexico, and on one farm that cultivates chile peppers var. Bell in Culiacán, Sinaloa, Mexico. Samples from can-taloupe farms consisted of cantaloupe rinses, irrigation water, water from furrows in the field, and workers’ hands. Samplesfrom the chile pepper farm consisted of rinses of chile peppers obtained at the field, pepper rinses obtained at the packinghouse, and irrigation water from the field. A total of 55 samples were obtained from both production systems. Twelve and 10samples from the cantaloupe and chile pepper production systems, respectively, tested positive for Salmonella according to atraditional culture method. The difference between the proportion of Salmonella-positive samples from the cantaloupe pro-duction system (12 of 28 � 0.43) and the chile pepper production system (10 of 27 � 0.37) was not statistically significant(P � 0.05). A PCR–restriction fragment length polymorphism (RFLP) method based on the fliC gene was used to determinethe serotype of the isolates. Salmonella Typhimurium was the only serotype found associated with the cantaloupe productionsystem, whereas both Salmonella Typhimurium and Enteritidis serotypes were found associated with the chile pepper produc-tion system. Results showed that 91% (20 of 22) and 9% (2 of 22) of the isolates from both agricultural systems matchedwith the Salmonella Typhimurium and Salmonella Enteritidis reference strain restriction profiles, respectively. This studydemonstrates the utility of the PCR-RFLP technique for determining the serotypes of Salmonella isolates obtained fromcantaloupe and chile pepper production systems.

The genus Salmonella comprises over 2,700 serotypesthat are found in different hosts and environments and thatcan cause human illness, including enteric fever, gastroen-teritis, and septicemia. Salmonella infection has been as-sociated with the consumption of contaminated fresh fruitsand vegetables, raw and undercooked poultry, or other redmeat and poultry products (1, 13, 24). Most studies in Mex-ico on Salmonella tend to corroborate the presence or ab-sence of Salmonella in food and water and as a cause ofoutbreaks; however, only a few studies have been per-formed investigating the prevalence of different Salmonellaserotypes in Mexico (6, 14). An important phenotypic char-acteristic of the genus is the antigenic diversity in the fla-gellar antigens, which is also observed at the genetic level(10).

The procedures used to identify Salmonella are labo-rious, time-consuming, and require a number of biochem-ical and serological tests to confirm presumptive isolates(25). On the other hand, tests using molecular tools havebeen useful in reducing the steps and the time needed for

* Author for correspondence. Tel and Fax: 52-81-83674487, Ext 132;E-mail: [email protected].

the detection, identification, and characterization of specificpathogens. The fliC gene, encoding for the flagellin protein,has been used as a target gene in assays to test the geneticdiversity in Salmonella (10). The fliC gene has a conservedterminal region and a variable central region, which deter-mines the antigenic specificity (4, 15). The goals of thisstudy were to determine the prevalence and sources of Sal-monella in cantaloupe and chile pepper production systemsin Mexico and to apply a PCR–restriction fragment lengthpolymorphism (RFLP) method based on the fliC gene toidentify the serotypes of the Salmonella isolates.

MATERIALS AND METHODS

Samples from cantaloupe and chile pepper farms. Sam-ples from cantaloupe production systems were obtained at threefarms (fields) in Matamoros, Coahuila, Mexico, during June 2006.The samples consisted of cantaloupe rinses (five composite sam-ples per farm, 125 ml per composite sample), water obtained froman irrigation channel that irrigated the three farms (four samples,25 ml per sample), water from furrows in fields (one sample perfarm, 25 ml per sample), and samples from workers’ hands (sixsamples from one farm, 1 cotton swab per worker). Samples fromthe chile pepper var. Bell production system were obtained fromonly one farm located in Culiacan, Sinaloa, Mexico, during April

J. Food Prot., Vol. 71, No. 112218 GALLEGOS-ROBLES ET AL.

TABLE 1. Salmonella reference strains used in this study andisolates from cantaloupe and chile pepper production systems

Salmonella serotype Origin/sourcea

Reference strains

1 Paratyphi A ATCC 91502 Enteritidis ATCC 130763 Typhimurium ATCC 133114 Kentucky IBT-UNAM5 Typhi IBT-UNAM6 Worthington IBT-UNAM7 Stanley IBT-UNAM8 Typhimurium IBT-UNAM

Serotype determined by PCR-RFLPCantaloupe production system

9 Typhimurium Matamoros, Coahuila/fruit rinse10 Typhimurium Matamoros, Coahuila/fruit rinse11 Typhimurium Matamoros, Coahuila/fruit rinse17 Typhimurium Matamoros, Coahuila/fruit rinse18 Typhimurium Matamoros, Coahuila/fruit rinse19 Typhimurium Matamoros, Coahuila/fruit rinse12 Typhimurium Matamoros, Coahuila/water from channel13 Typhimurium Matamoros, Coahuila/water from channel14 Typhimurium Matamoros, Coahuila/water from furrow15 Typhimurium Matamoros, Coahuila/water from furrow16 Typhimurium Matamoros, Coahuila/water from furrow20 Typhimurium Matamoros, Coahuila/workers’ hands

Chili pepper production system

21 Typhimurium Culiacán, Sinaloa/water from channel22 Typhimurium Culiacán, Sinaloa/water from channel23 Typhimurium Culiacán, Sinaloa/water from channel24 Typhimurium Culiacán, Sinaloa/water from channel25 Typhimurium Culiacán, Sinaloa/water from channel26 Enteritidis Culiacán, Sinaloa/water from channel27 Enteritidis Culiacán, Sinaloa/fruit rinse28 Typhimurium Culiacán, Sinaloa/fruit rinse29 Typhimurium Culiacán, Sinaloa/fruit rinse30 Typhimurium Culiacán, Sinaloa/fruit at packing

a ATCC, American Type Culture Collection; IBT-UNAM, Insti-tuto de Biotecnologı́a Universidad Nacional Autónoma de Mé-xico.

2006 and consisted of rinses of chile fruits obtained at the field(five composite samples, 125 ml per composite sample), fruit rins-es obtained at the packing house (eight samples, 25 ml per sam-ple), and water from the irrigation channel (14 samples, 25 ml persample). A total of 55 samples were obtained from both produc-tion systems. Cantaloupe and chile pepper rinses were obtained atboth production systems from five representative points at eachfarm, and at each point, five fruits were taken randomly within aradius of 3 m. Each fruit was washed individually by placing thefruit into a sterile Whirl-Pak bag (Nasco, Modesto, Calif.) con-taining 25 ml of 0.1% buffered peptone water (Becton Dickinson,Sparks, Md.) with a different sterile pair of gloves for each fruit.The fruit was washed by shaking and mixing the bag for at least2 min. The washings from the five cantaloupes obtained from thesame point were combined into a single glass bottle (125 ml percomposite sample). Samples from water sources were placed di-rectly into sterile glass bottles (25 ml per sample). Samples fromhands of workers in the field were obtained by rubbing the handswith a sterile cotton swab (Dequinsa, D.F. México) moistened in0.85% saline solution and then placing the cotton swab into a tubewith 5 ml of saline solution. Fruit rinses obtained at the packinghouse were obtained in a similar way as the rinses in the field,except each fruit constituted an individual sample. The sampleswere immediately transported in a cooler containing ice to thelaboratory and processed within 24 h. Each sample was mixed byshaking; 1.0 ml was removed and added to 9.0 ml of bufferedpeptone water and then incubated at 35�C for 24 h. Enrichmentand microbiological analyses were performed according to themethod described in the U.S. Department of Agriculture–FoodSafety and Inspection Service Microbiology Laboratory Guide-book (23). Briefly, 0.5 � 0.05 ml from the buffered peptone waterpreenrichment was transferred into 10 ml of tetrathionate brothand 0.1 � 0.02 ml into 10 ml of modified Rappaport-Vassiliadisbroth, which were then incubated at 42 � 0.5�C for 22 to 24 h.Next, both enrichments were streaked onto brilliant green sulfaand double-modified lysine iron agar plates and incubated at 35� 2�C for 18 to 24 h. Plates were then examined for the presenceof colonies meeting the description of suspect Salmonella colo-nies.

Salmonella strains. The control strains and isolates obtainedin this study are listed in Table 1. Salmonella enterica serotypesTyphimurium (American Type Culture Collection [ATCC] 13311,Manassas, Va.), Enteritidis (ATCC 13076), and Paratyphi A(ATCC 9150) were used as positive controls. Salmonella entericaserotypes Kentucky, Stanley, Typhi, Typhimurium, and Worthing-ton, kindly provided by Edmundo Calva (Instituto de Biotecno-logı́a, Universidad Nacional Autónoma de Mexico, Cuernavaca,Morelos, México) also were included in the study as positive con-trols. Biochemical tests and serotyping of all of the above-men-tioned strains were performed at the institution from which theywere obtained. In addition, 22 isolates obtained from the two ag-ricultural production systems that were not serotyped were in-cluded, for a total of 30 isolates subjected to PCR-RFLP analysis.

Data analyses. The test of difference between two binomialproportions (19) was performed to see whether the difference be-tween the proportions of positive samples from the agriculturalproduction systems was significant.

DNA extraction. Each isolate was streaked onto Trypticasesoy agar plates and incubated at 35�C for 24 h. Colonies fromeach agar plate were removed with a loop and then suspended in3 ml of buffered peptone water and incubated at 35�C for 24 h.The entire 3 ml of the culture was centrifuged (Sigma 1-15K,

Sigma Laborzentrifugen, Gottingen, Germany) three times, 1 mleach time, at 3000 � g for 5 min in 1.5-ml Eppendorf tubes.DNA extraction was performed on the resulting cell pellet ac-cording to the CTAB method, but omitting the use of polyvinyl-pyrrolidone and �-mercaptoethanol (5). The DNA samples werestored at �20�C until used.

PCR amplification of the fliC gene. Primers ABMS1(5-GCACAAGTCATTAATACAAACAGCC-3) and ABMS2(5-TTAACGCAGTAAAGAGAGGACG-3) described by Daugaet al. (4) were modified (by adding the bases that are underlined)on the basis of analysis with Amplify 1.2 software (University ofWisconsin, Madison) and used to amplify a fragment of approx-imately 1.5 kb from the fliC gene in each Salmonella strain. PCRamplification of the target sequence was performed in a PCR Ex-press thermal cycler (Thermo Hybaid, Ashford, Middlesex, UK).The PCR mixture contained 25 pmol of each of the primers, 250

M of a mix of deoxynucleoside triphosphates (Invitrogen, Car-slbad, Calif.), 1 mM MgCl2, 1� reaction buffer (200 mM Tris-HCl, pH 8.0, 500 mM KCl), 2.5 U Taq DNA polymerase (Bioline,

J. Food Prot., Vol. 71, No. 11 IDENTIFICATION OF SALMONELLA SEROTYPES BY PCR-RFLP 2219

TABLE 2. Samples positive for Salmonella from the cantaloupe and chile pepper production systems showing the number and type ofsamples and serotype of isolatesa

Type of sample

No. of positive samples/total analyzed(proportion)

Cantaloupe Chile pepper

Salmonella serotype found (no. of positive samples containingthe serotype/total analyzed)

Cantaloupe Chile pepper

Fruit rinses 6/15 (0.4) 3/5 (0.6) Typhimurium (6/15) Typhimurium (2/5), Enteritidis (1/5)Water from irrigation channel 2/4 (0.5) 6/14 (0.43) Typhimurium (2/4) Typhimurium (5/14), Enteritidis (1/14)Water from farm 3/3 (1.0) — Typhimurium (3/3)Workers’ hands 1/6 (0.16) — Typhimurium (1/6)Fruit rinses at packing house — 1/8 (0.12) — Typhimurium (1/8)

Total 12/28 (0.43) 10/27 (0.37) Typhimurium (12/28) Typhimurium, (8/27), Enteritidis (2/27)

a —, none of these types of samples were obtained.

Taunton, Mass.), 100 ng of DNA template, and deionized waterfor a final volume of 25 l. The reaction mixture was subjectedto PCR under the following conditions: heat denaturation at 95�Cfor 1 min and then 35 cycles of heat denaturation at 94�C for 30s, primer annealing at 60�C for 30 s, and DNA extension at 72�Cfor 30 s. After the last cycle, the PCR tubes were maintained at72�C for 10 min to complete synthesis of all strands. Five micro-liters of each PCR reaction was loaded onto a 1.5% agarose gel(Promega, Madison, Wis.), and electrophoresis was performedwith the use of 1� SB buffer (0.2 M NaOH, 0.7278 M boricacid). The gel was stained with ethidium bromide, examined witha UV transilluminator (Spectroline Transilluminator, model 7C-254R, Electronics Corporation, Westbury, N.Y.), and photo-graphed with a Polaroid camera (adapted with UV filter, filmA667). A 250-bp ladder (Invitrogen) was used as a molecularweight standard.

RFLP analysis. For the PCR-RFLP analysis, the unpurifiedamplified PCR product of the fliC gene was cleaved with Sau3AI(Promega) according to the manufacturer’s instructions. The di-gestion reaction consisted of 10 l of PCR product, 2 l of 10�reaction buffer B (60 mM Tris-HCl [pH 7.5] 500 mM NaCl, 60mM MgCl2, 10 mM dithiothreitol), 0.2 l of acetylated BSA (10

g/l), 5 U of enzyme, and deionized water for a final volumeof 20 l. The reaction was gently vortexed and then incubated at37�C for 1 h. After incubation, 6 l of sample was mixed with 3

l of loading buffer (0.25% bromophenol blue, 0.25% xylene cy-anole, 30% glycerol) and electrophoresed on a 10% acrylamidegel for 2.5 h at 100 V in 1� Tris-borate-EDTA buffer. The 100-bp Hyperladder (Bioline) was used as the molecular weight stan-dard for determining the molecular weight of the restriction frag-ments. Gels were stained and photographed as described above.The degree of variability between two strains was determined onthe basis of the Dice coefficient, and a dendrogram was made bythe unweighted pair group method with arithmetic mean (22). TheSPSS (Statistical Product for Service Solutions, v 10.0, SPSS Inc.,Chicago, Ill.) statistical package was used for the analyses.

RESULTS AND DISCUSSION

Samples positive for Salmonella. The proportions ofpositive samples, obtained by dividing the number of sam-ples positive for Salmonella by the total number of samples,were 0.43 and 0.37 from the cantaloupe and chile pepperproduction systems, respectively, and the observed differ-ence was not statistically significant (P � 0.05). Althoughthe number of samples obtained from the cantaloupe farmswas quite small, the observed differences among the three

farms were not statistically significant (P � 0.05). A totalof 22 Salmonella isolates were obtained from the two ag-ricultural production systems: 12 from the cantaloupe farmand 10 from the chile pepper farm. From the cantaloupeproduction system, at least one sample from each sourcewas positive for Salmonella, and water sources had higherproportions of positive samples (5 of 7) (P � 0.05) com-pared with the other samples (Table 2). From the chile pep-per production system, a higher number of positive samplescame from fruit rinses at the field and water from the irri-gation channel. These results demonstrate that water fromthe irrigation channel was an important source of Salmo-nella contamination. Investigations examining the environ-mental sources of Salmonella contamination indicate thatwater is an important source, particularly irrigation watercontaining manure, feces from animal wildlife, or sewageeffluent (9, 20). A likely reason why Salmonella was foundin water samples is that the water used for irrigation of thecantaloupe farms originated from the Rio Nazas and flowedthrough open channels without any protection against con-tact with animals. In addition, the cantaloupe farms wereirrigated from the same channels used to irrigate foragecrops with sewage. Castillo et al. (2) stated that Salmonellaand Escherichia coli were frequently detected in water usedfor irrigation of cantaloupe farms in South Texas becausethe water came from the Rio Grande. From the chile pepperfarm, 6 of 14 irrigation water samples were positive, and 3of 5 fruit rinse samples were positive. It is possible thatinsects or birds could also have transmitted Salmonella tothe fruits. Moore et al. (16) mentioned the possibility thatinsects belonging to the genus Chironomus were direct orindirect vectors of enteric bacteria for water and foods. Onthe other hand, Sela et al. (21) stated that direct contactwith manure-contaminated soil or dust might be a sourceof preharvest contamination, and indirect sources of con-tamination could also be due to trophic interactions betweenfruit and plant foragers like birds, mammals, and insects.The lowest numbers of positive samples were from work-ers’ hands at the cantaloupe farm and from rinses of chilepeppers obtained at the chile pepper packing house. Con-tamination in these samples could be due to worker’s poorhygiene, cross-contamination from fruit to worker, as wellas poor quality of the water used in the packing house to


FIGURE 1. PCR products obtained using ABMS1 and ABMS2primers. (A) Lane 1, Salmonella Paratyphi A (ATCC 9150); lane2, Salmonella Enteritidis (ATCC 13076); lane 3, Salmonella Ty-phimurium (ATCC 13311), lane 4, negative control. (B) Lane 1,Salmonella Kentucky; lane 2, Salmonella Typhi; lane 3, Salmo-nella Worthington; lane 4, Salmonella Stanley; lane 5, SalmonellaTyphimurium; lane 6, negative control; lanes M, 250-bp ladder(Invitrogen).

wash the fruit. The Centers for Disease Control and Pre-vention reported that in outbreaks that occurred in 2001 and2002 associated with the consumption of cantaloupes fromMexico contaminated with Salmonella Poona, possiblesources of contamination included irrigation of fields withwater contaminated with sewage, processing (cleaning andcooling) produce with Salmonella-contaminated water, poorhygienic practices of workers who harvested and processedthe cantaloupes, pests in packing facilities, and inadequatecleaning and sanitizing of equipment that came in contactwith the cantaloupes (3).

PCR amplification. The PCR product from the refer-ence strains (Fig. 1A) and from the different serotypes (Fig.1B) was of the expected size on the basis of the simulatedPCR product obtained with Amplify 1.2 software. As an-ticipated, the sizes were 1,487 bp for Salmonella ParatyphiA and Salmonella Typhimurium, 1,517 bp for SalmonellaEnteritidis (Fig. 1A), and 1,520 bp for Salmonella Typhi(Fig. 1B). Figure 1B shows the PCR products of varioussizes from the different Salmonella serotypes as evidenceof the genetic variability of the fliC gene. The PCR productsof Salmonella isolates obtained from both production sys-tems were approximately 1.5 kb in size. PCR with the prim-er set described by Kilger and Grimont (10) for amplifi-cation of the fliC gene generated products of two differentsizes: 1.24 kb for serotype Typhi and 1.5 kb for all otherserotypes. The difference in the size of the PCR productsobtained for the fliC gene can be accounted for by the var-iability of the central region of the open reading frame ofthis gene (8).

PCR-RFLP analysis. PCR-RFLP was carried out on30 strains, 8 belonging to 7 different Salmonella serotypesand 22 of unknown serotype isolated in this study (oneisolate from each positive sample) (Table 1). PCR-RFLPanalyses with Sau3AI on Salmonella strains of seven dif-ferent known serotypes yielded seven distinct restrictionprofiles for the fliC gene (Fig. 2A), demonstrating that fliCis a suitable target gene for discriminating among Salmo-nella serotypes by PCR-RFLP. Using PCR-RFLP on thegroEL gene amplicon of Salmonella digested with HaeIII,Nair et al. (18) found a low discriminatory capacity, be-cause only three different profiles were obtained, and dif-ferent serotypes of Salmonella shared the same restrictionprofile. However, other researchers have found good dis-criminatory ability with the fliC and fljB genes, particularlywith the use of a double digestion (4, 8). A DNA sequence–based approach examining the Salmonella flagellin fliCgene revealed the existence of two groups: the g-complex,which included Salmonella Enteritidis and nonmotilestrains, and the non–g-complex, which included motilestrains such as Salmonella Typhimurium (17). Results ofthis study indicated that there was a high level of sequencehomology between fliC genes of g-complex strains, and thatthe genetic basis for distinct antigens in this group of se-quences can be due to a single amino acid substitution.Also, it has been reported that differences in the fliC genesequence coding different antigenic phase 1 types could bedue to 1 to 44 nucleotide substitutions, some of which resultin changes in the amino acid sequence of the flagellin pro-tein (7).

Comparing the restriction profiles obtained with theSalmonella isolates from the cantaloupe and chile pepperproduction systems (Fig. 2B and 2C, respectively) withthose of the reference strains, all 12 isolates obtained fromthe cantaloupe production system had the Salmonella Ty-phimurium restriction profile. Six isolates were from rinsesof cantaloupe surfaces (two from each orchard), five werefrom water used for irrigation, and one was from the handsof a field worker. From the chile pepper production system,eight isolates were also identified as Salmonella Typhimu-rium: five from the water used for irrigation, two from chilepeppers, and one from chile peppers at the packing house.In addition, a second Salmonella restriction profile wasfound that matched that of Salmonella Enteritidis. It wasfound in two isolates: one from an irrigation water sampleand one from a chile pepper rinse sample.

Results from the PCR-RFLP analyses based on the fliCgene showed that 91% (20 of 22) and 9% (2 of 22) of theisolates from both agricultural systems matched with theSalmonella Typhimurium and Salmonella Enteritidis re-striction profiles, respectively, and these were the only twoserotypes found in both agricultural systems. This is inagreement with previous reports that indicate that Salmo-nella Typhimurium and Salmonella Enteritidis are the se-rotypes most commonly isolated from outbreaks of humansalmonellosis linked to the consumption of contaminatedanimal and vegetable foods (11, 12). It is important that50% of the isolates were obtained from water sources,which points out the risk associated with water used in

J. Food Prot., Vol. 71, No. 11 IDENTIFICATION OF SALMONELLA SEROTYPES BY PCR-RFLP 2221

FIGURE 2. Restriction profiles of the Salmonella fliC gene obtained with Sau3AI endonuclease digestion. (A) Lane 1, SalmonellaKentucky; lane 2, Salmonella Typhi; lane 3, Salmonella Worthington; lane 4, Salmonella Stanley; lane 5, Salmonella Paratyphi A(ATCC 9150); lane 6, Salmonella Typhimurium (ATCC 13311); lane 7, Salmonella Enteritidis (ATCC 13076). (B) Isolates obtained atthe cantaloupe production system. Lanes 1 through 7, restriction profile of Salmonella Typhimurium. (C) Isolates obtained from thechile pepper production system. Lanes 1 through 5 and 8 through 10, restriction profile of Salmonella Typhimurium; lanes 6 and 7,restriction profile of Salmonella Enteritidis; lanes M, 100-bp ladder (Invitrogen).

cantaloupe and chile pepper production systems. These re-sults partially agree with those obtained by Castillo et al.(2), who reported that most isolates they recovered werefrom water samples; however, the Salmonella serovars iso-lated from water were different from those isolated fromthe melons. It is important that Salmonella Poona, which isa serotype that often causes outbreaks of salmonellosis inthe United States associated with the consumption of mel-ons from Mexico, was not found in this study, because theexpected restriction profile for the fliC gene of this serotype(simulated with DNA Straider 1.2, GeneBank accessionAY353467 for the fliC gene of Salmonella Poona) was notobserved among the restriction profiles obtained in thisstudy. This provides some evidence that this serotype mightbe present only at melon production areas in southeasternMexico, where iguanas, which are considered a reservoirfor this serotype, are found (3). Further studies are neededto determine this possibility.

Although the number of samples analyzed in this studywas small, it was demonstrated that irrigation water couldbe an important source of contamination of produce by Sal-monella Typhimurium and Salmonella Enteritidis. Salmo-nella might have been transmitted by direct contact of thewater with the melons or by contact of the melons or thewater to other parts of the production systems, includingthe workers. It is also possible that contamination from illpersonnel during the handling of the fruit or contaminationthrough vectors, including insects and birds, could have oc-curred; however, these possibilities were not investigatedhere. The results of this investigation suggest that in pro-duction systems in which Good Agricultural Practices arenot in place, as in the farms tested in this study, contami-nation with Salmonella can occur, representing a health riskfor the farmers and the consumers of the contaminated pro-duce. Although only a small number of strains were tested,this study demonstrated the utility of the PCR-RFLP tech-nique for determining the serotypes of Salmonella isolatesobtained from cantaloupe and chile pepper production sys-tems by comparison to restriction profiles of known refer-

ence strains. The method is rapid, simple, and reproducibleand can potentially be applied for identification of isolatesobtained from other production systems. More extensivestudies need to be performed examining a larger number offarms and samples to determine the prevalence of Salmo-nella in agricultural production systems. Typing of isolatesby pulsed-field gel electrophoresis might be useful forsource tracking of Salmonella on the farms. Production sys-tems with Good Agricultural Practices and Good Manufac-turing Practices in place should be compared with those thatdo not use Good Agricultural Practices and Good Manu-facturing Practices to determine the efficacy of such prac-tices for the prevention of Salmonella contamination.

ACKNOWLEDGMENT

The Consejo Nacional de Ciencia y Tecnologı́a is acknowledged forfunding this research.

REFERENCES

1. Brenner, F. W., R. G. Villar, F. J. Angulo, R. Tauxe, and B. Swa-minathan. 2000. Salmonella nomenclature. J. Clin. Microbiol. 38:2465–2467.

2. Castillo, A., I. Mercado, L. M. Lucia, Y. Martinez-Ruiz, J. Ponce deLeon, E. A. Murano, and G. R. Acuff. 2004. Salmonella contami-nation during production of cantaloupe: a binational study. J. FoodProt. 67:713–720.

3. Centers for Disease Control and Prevention. 2002. Multistate out-breaks of Salmonella serotype Poona infections associated with eat-ing cantaloupe from Mexico—United States and Canada, 2000–2002. Morb. Mortal. Wkly. Rep. 51:1044–1047.

4. Dauga, C., A. Zabrovskaia, and P. A. D. Grimont. 1998. Restrictionfragment length polymorphism analysis of some flagellin genes ofSalmonella enterica. J. Clin. Microbiol. 36:2835–2843.

5. Doyle, J. J., and J. L. Doyle. 1987. A rapid DNA isolation procedurefor small quantities of fresh leaf tissue. Phytochem. Bull. 19:11–15.

6. Gutiérrez-Cogco, L., E. Montiel-Vázquez, P. Aguilera-Pérez, and M.C. González-Andrade. 2000. Serotipos de Salmonella identificadosen los servicios de salud de México. Salud Pública Méx. 42:490–495.

7. Herrera-Leon, S., J. R. McQuiston, M. A. Usera, P. I. Fields, J. Gar-aizar, and M. A. Echeita. 2004. Multiplex PCR for distinguishingthe most common phase 1 flagellar antigens of Salmonella spp. J.Clin. Microbiol. 42:2581–2586.

http://www.ingentaconnect.com/content/external-references?article=0095-1137()42L.2581[aid=7876999]http://www.ingentaconnect.com/content/external-references?article=0095-1137()42L.2581[aid=7876999]http://www.ingentaconnect.com/content/external-references?article=0036-3634()42L.490[aid=8517958]http://www.ingentaconnect.com/content/external-references?article=0095-1137()36L.2835[aid=2655067]http://www.ingentaconnect.com/content/external-references?article=0362-028X()67L.713[aid=6749689]http://www.ingentaconnect.com/content/external-references?article=0362-028X()67L.713[aid=6749689]http://www.ingentaconnect.com/content/external-references?article=0095-1137()38L.2465[aid=2451583]http://www.ingentaconnect.com/content/external-references?article=0095-1137()38L.2465[aid=2451583]


8. Hong, Y., T. Liu, C. Hofacre, M. Maier, D. G. White, S. Ayers, L.Wang, and J. J. Maurer. 2003. A restriction fragment length poly-morphism–based polymerase chain reaction as an alternative to se-rotyping for identifying Salmonella serotypes. Avian Dis. 47:387–395.

9. Islam, M., J. Morgan, M. P. Doyle, S. C. Phatak, P. Millner, and X.Jiang. 2004. Fate of Salmonella enterica Serovar Typhimurium oncarrots and radishes grown in fields treated with contaminated ma-nure composts or irrigation water. Appl. Environ. Microbiol. 70:2497–2502.

10. Kilger, G., and P. A. D. Grimont. 1993. Differentiation of Salmonellaphase 1 flagellar antigen types by restriction of the amplified fliCgene. J. Clin. Microbiol. 31:1108–1110.

11. Leon-Velarde, C. G., H. Y. Cai, C. Larkin, P. Bell-Rogers, R. W.Stevens, and J. A. Odumeru. 2004. Evaluation of methods for theidentification of Salmonella enterica serotype Typhimurium DT104from poultry environmental samples. J. Microbiol. Methods 58:79–86.

12. Lim, Y.-G., K. Hirose, H. Izumiya, E. Arakawa, H. Takahashi, J.Terajima, K. I. Itoh, K. Tamura, S. I. Kim, and H. Watanabe. 2003.Multiplex polymerase chain reaction assay for selective detection ofSalmonella enterica serovar Typhimurium. Jpn. J. Infect. Dis. 56:151–155.

13. Liu, G. R., A. Rahn, W.-Q. Liu, K. E. Sanderson, R. N. Johnston,and S.-L. Liu. 2002. The evolving genome of Salmonella entericaserovar Pullorum. J. Bacteriol. 184:2626–2633.

14. Mancera, M. A., N. J. Vázquez, and Z. A. Heneidi. 2004. Fagotip-ificación de aislamientos de Salmonella enteritidis obtenidos de avesde México. Téc. Pecu. Méx. 42:287–294.

15. McQuiston, J. R., R. Parrenas, M. Ortiz-Rivera, L. Gheesling, F.Brenner, and P. I. Fields. 2004. Sequencing and comparative analysisof flagellin genes fliC, fljB, and flpA from Salmonella. J. Clin. Mi-crobiol. 42: 1923–1932.

16. Moore, B. C., E. Martinez, J. M. Gay, and D. H. Rice. 2003. Survivalof Salmonella enterica in freshwater and sediments and transmissionby the aquatic midge Chironomus tentans (Chironomidae: Diptera).Appl. Environ. Microbiol. 69:4556–4560.

17. Mortimer, C. K. B., T. M. Peters, S. E. Gharbia, J. M. J. Logan, andC. Arnold. 2004. Towards the development of a DNA-sequencebased approach to serotyping of Salmonella enterica. BMC Micro-biol. 4:31.

18. Nair, S., T. K. Lin, T. Pang, and M. Altwegg. 2002. Characterizationof Salmonella serovars by PCR-single-strand conformation poly-morphism analysis. J. Clin. Microbiol. 40:2346–2351.

19. Ott, L. R. 1992. An introduction to statistical methods and data anal-ysis, 4th ed. Duxbury Press, Belmont, Calif.

20. Reilly, W. J., G. I. Forbes, G. M. Paterson, and J. C. Sharp. 1981.Human and animal salmonellosis in Scotland associated with envi-ronmental contamination, 1973–79. Vet. Rec. 108:553–555.

21. Sela, S., D. Nestel, R. Pinto, E. Nemny-Lavy, and M. Bar-Joseph.2005. Mediterranean fruit fly as a potential vector of bacterial path-ogens. Appl. Environ. Microbiol. 71:4052–4056.

22. Sneath, P. H. A., and R. R. Sokal. 1973. Numerical taxonomy. W.H. Freeman & Co., San Francisco.

23. U.S. Department of Agriculture, Food Safety and Inspection Service.2008. Isolation and identification of Salmonella from meat, poultry,and egg products. Available at: http://www.fsis.usda.gov/PDF/MLG�4�04.pdf. Accessed September 2008.

24. U.S. Food and Drug Administration. 2001. Outbreaks associatedwith fresh and fresh-cut produce. Incidence, growth, and survival ofpathogens in fresh and fresh-cut produce. Available at: http://www.cfsan.fda.gov/�comm/ift3-4a.html. Accessed 20 June 2007.

25. Wilson, R. C., P. A. Padron, and W. B. Dockstader. 1971. Salmonellascreening procedure with tests for �-galactosidase and flagellar an-tigens. Appl. Microbiol. 21:346–349.

http://www.ingentaconnect.com/content/external-references?article=0099-2240()71L.4052[aid=7993625]http://www.ingentaconnect.com/content/external-references?article=0042-4900()108L.553[aid=8517960]http://www.ingentaconnect.com/content/external-references?article=0095-1137()40L.2346[aid=8489722]http://www.ingentaconnect.com/content/external-references?article=0099-2240()69L.4556[aid=8517962]http://www.ingentaconnect.com/content/external-references?article=0095-1137()42L.1923[aid=8517963]http://www.ingentaconnect.com/content/external-references?article=0095-1137()42L.1923[aid=8517963]http://www.ingentaconnect.com/content/external-references?article=0021-9193()184L.2626[aid=8517965]http://www.ingentaconnect.com/content/external-references?article=1344-6304()56L.151[aid=7384107]http://www.ingentaconnect.com/content/external-references?article=1344-6304()56L.151[aid=7384107]http://www.ingentaconnect.com/content/external-references?article=0167-7012()58L.79[aid=8517966]http://www.ingentaconnect.com/content/external-references?article=0095-1137()31L.1108[aid=8517967]http://www.ingentaconnect.com/content/external-references?article=0099-2240()70L.2497[aid=7856316]http://www.ingentaconnect.com/content/external-references?article=0099-2240()70L.2497[aid=7856316]http://www.ingentaconnect.com/content/external-references?article=0005-2086()47L.387[aid=8517968]http://www.fsis.usda.gov/PDF/http://www.cfsan.fda.gov/comm/ift3-4a.htmlhttp://www.cfsan.fda.gov/comm/ift3-4a.htmlhttp://www.cfsan.fda.gov/comm/ift3-4a.html

________________________________________________________Revista Científica, FCV-LUZ / Vol. XIX, Nº 2, 139 - 146, 2009

CARACTERIZACIÓN DE AISLADOS DE Escherichia coli O157:H7EN CANALES DE BOVINOS Y PORCINOS MEDIANTE PCR.

Characterization of Escherichia coli O157:H7 Isolates Obtained from Bovine and PorcineCarcasses.

Miguel Gallegos1, Alberto Morales2*, Genoveva Álvarez 2, Jesús Vásquez 3, Lilia Morales4, Irma Martínez 4

y Jesús Maldonado5

1Facultad de Agricultura y Zootecnia, Universidad Juárez del Estado de Durango. 2Instituto Nacional de Investigaciones Forestales,

Agrícolas y Pecuarias (INIFAP) hasta 31 de diciembre del 2007, actualmente Consorcio Técnico del Noreste, A. C. UGRNL.

Teléfono y Fax: 01 81 83674487 Ext. 132. Km 4.5 carretera a Reynosa, Guadalupe, N. L. México.3Facultad de Ciencias Químicas, UJED. 4Facultad de Ciencias Biológicas, UANL. 5Laboratorio de Diagnóstico Molecular,

Núcleo “Héctor Ochoa Zuleta”, Universidad Centroccidental “Lisandro Alvarado”, Venezuela.

*E-mail: [email protected]

RESUMEN

Muchos de los brotes causados por Escherichia coli O157:H7se han asociado al consumo de carne bovina mal cocida, perotambién se ha reportado su presencia en la carne de otros ani-males domésticos. En México existe poca información sobre lapresencia de este patógeno en canales de res y de cerdo. Elobjetivo de este estudio fue determinar la presencia de E. coliO157:H7 en canales de res y cerdo y su caracterización me-diante PCR. De un total de 18 aislados, 12 fueron positivas porPCR para los genes rfbE y fliC que determinan el serotipoO157:H7. De estos 12, uno de canal de res y tres de canalesde cerdo fueron positivos por PCR para los genes stx1, stx2 yeaeA, por lo que fueron considerados como enterohemorrági-cos. Las diferencias encontradas en el número de canales po-sitivas para los genes caracterizados no fueron estadística-mente significativas, y los resultados señalan que E. coliO157:H7 puede ser encontrada en ambos tipos de canal, re-presentando un riesgo para la salud, por lo que se deben to-mar medidas más estrictas de higiene y manejo para evitarque canales que no cumplan con el carácter de inocuidad lle-guen a los consumidores finales.

Palabras clave: Bovinos, cerdos, E. coli O157:H7, enterohe-morrágica, caracterización molecular, PCR.

ABSTRACT

Many Escherichia coli O157:H7 outbreaks have been associ-ated to consumption of undercooked beef, but the presencehas also been reported in the meat of other domestic animals.In Mexico, little information exists on the presence of thispathogen in bovine and pork carcasses. The objective of thisstudy was to determine the presence and/or absence of E. coliserotype O157:H7 in bovine and pork carcasses and theircharacterization by means of PCR. Of 18 isolates, 12 werepositive by PCR to rfbE and fliC genes, which determineO157:H7 serotype. Of these 12, one from bovine carcass andthree of pork carcasses were positive by PCR to stx1, stx2 andeaeA genes, therefore, they were considered enterohemor-rhagic strains. The differences found in the positive carcassnumber to any of the genes were not statistically significant.The results show that E. coli O157:H7 could be found in bothcarcasses types, representing a risk for the health, so strict hy-gienic and handling measures should be taken in order toavoid that carcasses which do not fulfill the food safety aspectmight arrive to the final consumers.

Key words: Bovine and pork carcasses, E. coli O157:H7, en-terohemorrhagic, molecular characterization, PCR.

INTRODUCCIÓN

Escherichia coli O157:H7 ha emergido como un patóge-no transmitido por alimentos y es considerado de importanciaen salud pública, ya que está implicado en brotes de colitis he-

139

Recibido: 22 / 02 / 2007. Aceptado: 08 / 03 / 2008.

morrágica y posible aparición posterior del síndrome urémicohemolítico (SUH) [19]. Una característica de E. coli O157:H7es el bajo número de células requeridas para desarrollar la en-fermedad de 10 a 100 células [11] por lo que la no detecciónpor los métodos tradicionales microbiológicos no es certeza nisinónimo de seguridad del alimento. E. coli O157:H7 puedeestar presente en una gran variedad de animales silvestres ydomésticos entre los cuales se encuentran bovinos (Bos tau-rus-indicus), porcinos (Sus scrofa domestica) y ovinos (Ovisaries) [18], siendo principalmente los rumiantes y sus heces fe-cales un reservorio natural de este patógeno [17, 20]. Se hareportado que la transmisión de E. coli O157:H7 a los huma-nos en forma directa o indirecta puede ser por contaminaciónde los alimentos a partir de material fecal, agua contaminada,y contacto con personas o animales enfermos [4]. Por otrolado, microorganismos patógenos para el humano como E. coliO157:H7, que habita naturalmente en el tracto digestivo delganado bovino, puede eventualmente contaminar la canal du-rante el proceso de evisceración o en el manejo posterior de lamisma [25]. Aunque el serotipo de E. coli O157:H7 es determi-nado por los genes rfbE y fliC que codifican respectivamentepara la biosíntesis del lipopolisacárido O157 y la flagelina, loque a su vez determina respectivamente los antígenos O y H,la patogenicidad se debe a la expresión de varios genes quecodifican para factores de virulencia. Entre estos se encuen-tran los genes stx1 y stx2 que codifican para las verotoxinas 1y 2, respectivamente, los cuales se encuentran en un bacterió-fago integrado al cromosoma bacteriano y que han sido aso-ciados al desarrollo del SUH, particularmente stx2 [21]. Otrode los genes de virulencia localizado en el locus LEE (locus deefusión en enterocitos) es el gen eaeA que codifica para la inti-mina y que es necesario para el proceso de la adherencia ínti-ma a las células epiteliales del intestino humano [12, 20, 22].En México existen algunos reportes de la presencia de E. coliO157:H7 en muestras comerciales de carne de res, pero no sedefine si la carne venía contaminada desde el matadero o secontaminó en el manejo posterior [6], y siendo México un paíscon una producción promedio en los últimos cinco años de1.473.650 t de carne de bovino y de 1.050.311 t de carne decerdo [24] y un volumen de exportación combinado de ambostipos de carne de 38.677,2 t [23], es importante determinar lacalidad sanitaria de la canal al momento de salir del matadero.El objetivo del presente estudio consistió en caracterizar cepasde E. coli del serotipo O157:H7 aisladas de canales de bovi-nos y porcinos en plantas procesadoras de productos cárnicosde la Comarca Lagunera, México.

MATERIALES Y MÉTODOS

Diseño del estudio

El estudio fue de tipo longitudinal prospectivo y el perío-do comprendió de febrero a marzo del 2004. Se seleccionarondos plantas procesadoras de productos cárnicos, una de cerdo

(I) y una de bovino (II) de la Comarca Lagunera, México. Losmuestreos se realizaron en dos fechas para cada planta,20/02/2004 y 15/03/2004 para la planta I y 02/03/2004 y24/03/2004 para la planta II. A partir de estos se obtuvieron losaislados de E. coli.

Obtención de las muestras

La toma y manejo de las muestras se realizó conforme alo especificado por la Norma Oficial Mexicana NOM-109SSA1-1994 y con modificaciones a lo propuesto por Gill y Jo-nes [15]. Las canales se muestrearon inmediatamente a suarribo al área de inspección y sellado (15 y 10 minutos en pro-medio después del inicio del sacrificio del animal hasta la tomade las muestras, respectivamente para canal de bovino y cer-do). En cada planta procesadora se procedió a la selección de15 canales al azar, para el análisis de superficies vivas, de lascuales se recolectaron las muestras con esponja estéril NAS-CO®, Whirl-Pak, Speci-Sponge, B01324WA (5x10x1 cm) (Nas-co EUA) en tres sitios diferentes: falda (sitio A), costado (sitioB) y cuello (sitio C), para un total de 45 muestras por planta.De las 15 muestras de cada sitio de la canal (A, B, C) se for-maron tres grupos de cinco muestras cada uno, para tenernueve muestras compuestas por planta, lo que dio un tamañofinal de 18 muestras compuestas.

Aislamiento y confirmación de E. coli O157:H7en muestras compuestas

Cada muestra compuesta se pasó a enriquecimiento encaldo Reveal (NEOGEN®, EUA) y se incubó durante 8 h a42°C, luego de este tiempo se tomaron 120 µL y se deposita-ron sobre la placa de lectura del kit. Después de 15 minutos serealizaron las lecturas y de las muestras presuntivas se tomóuna asada y se estrío en medio cromogénico CHROmagar®

O157 (CHROmagar, Francia). Las colonias características seconfirmaron con pruebas bioquímicas utilizando el API® 20 E(Biomereux, Francia) y el serotipo utilizando un kit de serologíaDifco™ E. coli Antiserum (DIFCO, EUA). En las muestras quese confirmó la presencia de E. coli O157:H7 se procedió a rea-lizar la obtención de ADN para la caracterización molecularmediante la Reacción en Cadena de la Polimerasa (PCR).

Cepa de referencia, condiciones y medio de cultivo

La cepa de referencia de E. coli O157:H7 que se usócomo control positivo en los ensayos de PCR fue proporciona-da por el Centro para el Control y la Prevención de Enferme-dades (CDC) a través del Centro Nacional de Servicios deConstatación en Salud Animal (CENAPA), México. La cepa fueactivada en caldo infusión cerebro corazón a 35°C por 24 h.

Extracción del ADN

Los aislados de E. coli O157:H7 se incubaron en caldoinfusión cerebro corazón por 24 h a 37°C. Después de la incu-bación y a partir de este medio se tomó tres veces 1 mL y se

140

Caracterización de aislados de E. coli O157:H7 en canales de bovinos y porcinos mediante PCR / Gallegos, M. y col.__________________

centrifugó a 3000 rpm durante 1 min (Centrifuga Sigma 1-15K,Alemania) para formar una pastilla en tubos Eppendorf de 1,5mL y hacer la extracción de ADN con el método CTAB [7] peroomitiendo el uso de polivinilpirrolidona y 2�-mercaptoetanol.

Reacciones de PCR

Las reacciones de PCR se realizaron utilizando los oligo-nucleótidos mencionados en la TABLA I. Primero se realizóuna reacción para amplificar simultáneamente dos fragmentoscorrespondientes a los genes rfbE y fliC [5, 13]. En los aisla-dos que resultaron positivos para los genes anteriores, se rea-lizaron reacciones por separado para amplificar un fragmentoen cada uno de los siguientes genes: stx1, stx2 y eaeA [1, 14].Las reacciones de PCR se realizaron en volúmenes de 25 µL,utilizando 1 µL (25 pmoles) de cada uno de los oligonucleóti-dos, 2,5 µL de los dNTP´s (200 µM) (GIBCO-BRL), 0,5 µL deMgCl2 (1,5 mM), 2,5 µL de buffer para PCR (10X) (200 mMTris-HCl pH 8,0, 500 mM KCl), 2,5 unidades de la enzimaTaq-DNA polimerasa (Bioline) y 1 µL de ADN (100 ng). El con-trol negativo de la PCR tenía los mismos ingredientes, exceptoel ADN, sustituyéndose este volumen con agua miliQ estéril.La amplificación se llevó a cabo en un termociclador PCR Ex-press (ThermoHybaid, Reino Unido) y las condiciones de corri-da para los genes rfbE y fliC fueron de un ciclo de 1 min a95°C, seguido de 35 ciclos de tres pasos: desnaturalización a95°C por 30 s, alineamiento a 66°C por 30 s y una extensión a72°C por 75 s, con una extensión final de 10 min a 72°C. Paralos genes stx1 y stx2 se modificó el paso de alineamiento a58°C durante 30 s. Para el gen eaeA se modificó el paso dealineamiento a 64°C por 30 s. Las demás temperaturas y tiem-pos fueron iguales. Los productos de PCR fueron visualizadosen geles de agarosa al 1,5% teñidos con bromuro de etidio(0,5 µg mL–1), y visualizados en un transiluminador de luz UV(Spectroline, EUA). Posteriormente fueron fotografiados conuna cámara polaroid (película A667, EUA) adaptada con filtropara luz ultravioleta.

Prueba estadística

Se usó la prueba de Ji cuadrado [3] para probar si las di-ferencias observadas en el número de canales positivas al pa-tógeno de cerdo y de bovino eran estadísticamente significati-vas.

RESULTADOS Y DISCUSIÓN

De un total de 18 muestras (nueve de cerdo y nueve deres), seis de cerdo (66,66%) y seis de bovino (66,66%) fueronpositivas por PCR para la presencia de E. coli serotipoO157:H7, demostrándose así que este patógeno puede estarpresente en ambos tipos de canal (TABLA II y FIG. 1). E. coliO157:H7 se ha encontrado en una gran variedad de alimentos,tanto de origen vegetal como animal, pero se ha asociado par-ticularmente a la carne del ganado bovino, ya que esta espe-cie se ha considerado como un reservorio natural de este pa-tógeno. Sin embargo existen reportes que mencionan su pre-sencia en otras especies animales como cerdos, ovejas, caba-llos (Equs caballus), venados, perros (Canis familiaris) y aves[2, 8, 16]. En relación con la planta I, en su primera fecha demuestreo (20/02/2004) se encontró que, si bien dos aisladosresultaron positivos para los genes que codifican para los antí-genos O157 y H7, éstos no eran portadores de ninguno de losdos genes que codifican para las verotoxinas; sin embargo, enla segunda fecha de muestreo (15/03/2004), de cuatro aisla-dos positivos para los genes rfbE y fliC, tres fueron positivospara los genes de las verotoxinas 1 y 2 y el de la intimina. Esimportante mencionar que cada uno de estos tres aislados fue-ron encontrados en cada uno de los tres diferentes sitios demuestreo en la canal, es decir, uno en la cadera, otro en elcostado y otro en el cuello, lo que plantea la posibilidad que setratara de un mismo animal, o bien de animales diferentespero procedentes de una misma granja o lugar y en la cualexiste la presencia de E. coli O157:H7 enterohemorrágica.

141

_______________________________________________________________Revista Científica, FCV-LUZ / Vol. XIX, Nº 2, 139 - 146, 2009

TABLA I

INICIADORES USADOS EN LAS REACCIONES DE PCR PARA LA IDENTIFICACIÓN Y CARACTERIZACIÓN MOLECULARDE E. coli O157:H7 / PRIMERS USED IN THE PCR REACTIONS FOR THE IDENTIFICATION AND MOLECULAR CHARACTERIZATION

OF E. coli O157:H7

Gen Secuencia5´ � 3´

Tamaño (pb)

rfbE AAGATTGCGCTGAAGCCTTTGCATTGGCATCGTGTGGACAG

497

fliC GCGCTGTCGAGTTCTATCGAGCCAACGGTGACTTTATCGCCATTCC

625

stx1 CGCTGAATGTCATTCGCTCTGCCGTGGTATAGCTACTGTCACC

302

stx2 CTTCGGTATCCTATTCCC a

CTGCTGTGACAGTGACAAAACG b518

eaeA CAGGTCGTCGTGTCTGCTAAATCAGCGTGGTTGGATCAACCT

1087

a Se eliminaron dos guaninas en el extremo 3´. bSe eliminó una citosina en el extremo 3´.

Respecto a los seis aislados que fueron encontrados enla planta II y que amplificaron por PCR para los genes que co-difican para los antígenos O157 y H7, cuatro fueron portadoresdel gen de la verotoxina 1, y de estos sólo uno fue portador delgen de verotoxina 2 y el de intimina, y fue obtenido del cuellodel animal.

Si bien lo anterior señala el riesgo de encontrar a estepatógeno en ambos tipos de canal y representa un riesgo parala salud, no necesariamente implica que el animal sea el porta-dor, sino que pone de manifiesto la ineficiencia de las buenasprácticas higiénicas en el manejo de la canal, puesto que unanimal sano puede portar el patógeno en su pelo, piel y tracto

142


TABLA II

RESUMEN DE LA CARACTERIZACIÓN DE AISLADOS DE E. coli / SUMMARY OF THE CHARACTERIZATION OF E. coli ISOLATES

Aislado Especie Planta Sitio de muestreoen la canal

Fecha demuestreo

Genotipo

rfbE fliC stx1 stx2 eaeA

1.1 Cerdo I B 20/02/2004 – – – – –

2.1 Cerdo I B 20/02/2004 – – – – –

3.1 Cerdo I C 20/02/2004 + + – – –

4.1 Cerdo I C 20/02/2004 + + – – –

5.1 Bovino II B 2/03/2004 + + + – –

6.1 Bovino II C 2/03/2004 + + + + +

7.1 Bovino II A 2/03/2004 + + + – –

8.1 Bovino II B 2/03/2004 + + + – –

10.1 Cerdo I A 15/03/2004 + + + + +

11.1 Cerdo I B 15/03/2004 + + + + +

12.1 Cerdo I C 15/03/2004 + + – – –

13.1 Cerdo I C 15/03/2004 + + + + +

14.1 Bovino II B 24/03/2004 + + – – –

15.1 Bovino II C 24/03/2004 + + – – –

A = falda. B = costado. C = cuello. + y – = respectivamente portadores y no portadores del gen en cuestión.

497625

1 2 3 4 5 6 7 8 9 10 11 12 13 M

FIGURA 1. PRODUCTOS DE PCR A PARTIR DE LOS GENES fliC (625 pb) Y rfbE (497 pb) EN AISLADOS DE E. coliOBTENIDOS DE CANALES DE CERDO Y BOVINO. CARRILES 1-3 y 8-11 = AISLADOS A PARTIR DE CANALES DE CERDO.CARRILES 4-7 = AISLADOS A PARTIR DE CANALES DE BOVINO. CARRIL 12 = E. coli O157:H7. CARRIL 13 = CONTROL

NEGATIVO. CARRIL M = MARCADOR DE PESO MOLECULAR HYPERLADDER 100 PB (BIOLINE)/ PCR PRODUCTS OF fliC (625pb) AND rfbE (497 pb) GENES OF E. coli ISOLATES OBTAINED OF PORCINE AND BOVINE CARCASSES. LINES 1-3 AND 8-11 = ISOLATES

OBTAINED OF PORCINE CARCASSES. LINES 4-7 = ISOLATES OBTAINED OF BOVINE CARCASSES. LINES 12 = E. coli O157:H7. LINE 13 =

NEGATIVE CONTROL. LINE M = HYPERLADDER MARKER 100 PB (BIOLINE).

intestinal [2]; o bien la canal contaminada en el proceso de sa-crificio contamine a otras canales debido a un deficiente proce-so de evisceración. A pesar que el proceso de sacrificio delganado bovino y porcino es diferente, las diferencias observa-das entre el número de canales positivas de cerdo y bovinopara cualquiera de los genes no fueron significativas estadísti-camente (TABLA III), lo que sugiere la posibilidad de encontrara E. coli O157:H7 enterohemorrágica indistintamente en am-bos tipos de canal, sin embargo, estos datos deben conside-rarse con reserva, dado el tamaño de muestras analizadas apesar que se tiene concordancia con otros autores [4] quienesmencionan la posibilidad que los cerdos sean hospederos bio-lógicamente competentes para E. coli O157:H7 y otras cepasde E. coli verotoxigénicas.

En algunos países como Estados Unidos de Norteaméri-ca, la presencia del serotipo O157:H7 en los alimentos, no im-porta si es o no verotoxigénica, es motivo de preocupación porel riesgo que ello implica en la salud de los consumidores. Porlo tanto, además de determinar la presencia de E. coliO157:H7 en los alimentos es muy importante caracterizarlosen relación con otros factores de patogenicidad. La importan-cia clínica que tiene E. coli O157:H7 como patógeno radica enel hecho que puede ser portadora de uno o ambos genes quecodifican para las verotoxinas, así como el gen de la intimina

lo que determina que la cepa sea considerada enterohemorrá-gica [19]. La FIG. 2 muestra los productos de PCR a partir delgen stx1 de aislados que fueron positivos para los genes rfbEy fliC. Se observa que no todos los aislados presentaron estegen, sólo siete de ellos, correspondiendo cuatro a aislados ob-tenidos a partir de canales de bovino y tres a partir de canalesde cerdo. Puesto que los genes stx1 y stx2 se encuentrancada uno en un bacteriófago temperado lisogénico, los cualesintegran su ADN al cromosoma de E. coli O157:H7 [26], sepuede explicar porque algunas de las cepas de E. coliO157:H7 expresaron sólo una o ambas verotoxinas, lo cual yaha sido documentado [22].

En otro estudio [10] sobre la presencia de E. coliO157:H7 en cerdos, se encontró además del genotipoO157:H7 portador de los genes stx1, stx2 y eaeA, la presenciadel genotipo portador de stx1 y los genes de virulencia eaeA yhly, o bien, genotipos portadores de eaeA, stx1 y stx2, pero nocon los cuatro genes stx1, stx2, eaeA y hly. Así mismo otrosautores reportaron en canales de bovino, un aislado de E. coliO157:H7 que fue negativa por PCR para los genes stx peropositiva para los genes de virulencia ehx y eaeA [9]. De losaislados que fueron caracterizados como portadores del genstx1, cuatro fueron portadores del gen stx2, siendo tres de losaislados obtenidos de canales de cerdo y uno de canal de bo-

143


TABLA III

FRECUENCIA DE AISLADOS DE E. coli O157:H7 PORTADORES DE CADA UNO DE LOS GENES rfbE, fliC, stx1, stx2, eaeA,Y DE LOS CINCO GENES POR TIPO DE CANAL / FREQUENCY OF E. coli O157:H7 ISOLATES HARBORING EACH OF THE GENES rfbE,

fliC, stx1, stx2, eaeA, AND THE FIVE GENES BY KIND OF CARCASS

Tipo de canal rfbE fliC stx1 stx2 eaeA Cinco genes

Bovino 6/9 = 0,666 6/9 = 0,666 4/9 = 0,444 1/9 = 0,111 1/9 = 0,111 1/9 = 0,1111

Cerdo 6/9 = 0,666 6/9 = 0,666 3/9 = 0,333 3/9 = 0,333 3/9 = 0,333 3/9 = 0,3333

Significancia NS NS NS NS NS NS

NS = diferencia no significativa (P � 0,05).

1 2 3 4 5 6 M 7 8 9 10 11 12

302 pb

FIG. 2. PRODUCTOS DE PCR DEL sxt1 (302 PB) DE SUPUESTOS AISLADOS DE E. coli O157:H7.CARRILES 1-2 Y 7-10 = AISLADOS A PARTIR DE CANALES DE CERDOS. CARRILES 3-6 = AISLADOS A PARTIR

DE CANALES DE BOVINO. CARRIL 11 = E. coli O157:H7. CARRIL 12 = CONTROL NEGATIVO. CARRIL M = MARCADORDE PESO MOLECULAR HYPERLADDER 100 PB (BIOLINE) / PCR PRODUCTS OF stx1 GENE (302 pb) OF PUTATIVE E. coli O157:H7

ISOLATES. LINES 1-2 AND 8-11 = ISOLATES FROM PORCINE CARCASSES. LINES 3-6 = ISOLATES FROM BOVINE CARCASSES. LINE 12 = E. coli

O157:H7. LINE 13 = NEGATIVE CONTROL. LINE M = HYPERLADDER MARKER 100 PB (BIOLINE).

vino (FIG. 3). No obstante se ha reportado una mayor propor-ción de cepas portadoras del gen stx2 que del gen stx1 [9], enel presente trabajo no se encontró la misma proporción, perosí hubo coincidencia con otros reportes [16] donde mencionanque la mayoría de las cepas portadoras del gen stx2, tambiénlo son para el gen eaeA, ya que los cuatro aislados portadoresdel gen stx2 también fueron portadores del gen eaeA (FIG. 4).

CONCLUSIONES

Las canales de bovino y las de cerdo pueden ser porta-doras de E. coli O157:H7, lo que refleja la habilidad de estepatógeno para colonizar también a la especie porcina. Los re-sultados obtenidos en este trabajo a través de la PCR constitu-yen un aporte valioso en materia de inocuidad alimentaria ysalud pública al demostrar la presencia de E. coli O157:H7 en-terohemorrágica en estos tipos de alimentos. De gran utilidadresultó la PCR para determinar el grado de virulencia de losaislados ya que permitió identificar los que fueron portadoresde los genes de virulencia verotoxinas 1 y 2 y del gen de la in-timina, siendo además muy importante desde el punto de vistaepidemiológico el hecho de haber encontrado cepas portado-ras de los factores de virulencia por ser consideradas patóge-nas al humano. Debido a que no se obtuvieron muestras am-bientales, ni de los trabajadores dentro de los mataderos, asícomo de las herramientas de corte que ellos utilizan, serán ne-cesarios estudios posteriores para descartar la existencia defuentes externas de contaminación, así como la posibilidad decontaminación cruzada al momento de la evisceración. Deacuerdo con los resultados encontrados, se abre la posibilidadpara realizar estudios más a fondo del destino que tendrían eltipo de cepas de E. coli O157:H7 sobre la incidencia de enfer-medades transmitidas por los alimentos, o si bien, el procesode cocción es suficiente para eliminar el riesgo de contamina-ción al momento del consumo de la carne contaminada con di-cha bacteria.

144


1 2 3 4 5 M 6 7 8 9

518 pb

FIG. 3. PRODUCTOS DE PCR DEL GEN sxt2 (518 PB) DE SUPUESTO AISLADOS DE E. coli O157:H7.CARRILES 1-4 = AISLADOS A PARTIR DE CANALES DE BOVINO. CARRILES 5 Y 6-7 = AISLADOS A PARTIR DE CANALES

DE CERDO. CARRIL 8 = E. coli O157:H7. CARRIL 9 = CONTROL NEGATIVO. CARRIL M = MARCADOR DE PESOMOLECULAR HYPERLADDER 100 PB (BIOLINE) / PCR PRODUCTS OF stx2 GENE (518 PB) OF PUTATIVE E. coli O157:H7 ISOLATES.

LINES 1-4 = ISOLATES FROM BOVINE CARCASSES. LINES 5 AND 7-8 = ISOLATES FROM PORK CARCASSES. LINE 9 = E. coli O157:H7. LINE 10 =

NEGATIVE CONTROL. LINE M = HYPERLADDER MOLECULAR MARKER 100 PB (BIOLINE).

M 1 2 3 4 5 6

1087 pb

FIG. 4. PRODUCTOS DE PCR DEL GEN eaeA (1087 PB) DESUPUESTOS AISLADOS DE E. coli O157:H7. CARRIL M =MARCADOR DE PESO MOLECULAR HYPERLADDER 100PB (BIOLINE). CARRIL 1 = AISLADO A PARTIR DE CANALDE BOVINO. CARRILES 2-4 = AISLADOS A PARTIR DECANALES DE CERDO. CARRIL 5 = E. coli O157:H7. CA-RRIL 6 = CONTROL NEGATIVO / PCR PRODUCTS OF eaeAGENE (1087 PB) OF PUTATIVE E. coli O157:H7 ISOLATES. LINE M =

HYPERLADDER MOLECULAR MARKER 100 PB (BIOLINE). LINE 2 =

ISOLATE FORM BOVINE CARCASS. LINES 3-5 = ISOLATES FROM

PORK CARCASSES. LINE 6 = E. coli O157:H7. LINE 7 = NEGATIVE

CONTROL.

REFERENCIAS BIBLIOGRÁFICAS

[1] BLANCO, M.; BLANCO, J.E.; MORA, A.; REY, J.;ALONSO, M.; HERMOSO, J.; ALONSO, M.P.; DHABI,G.; GONZALEZ, E.A.; BERNARDEZ, M.I.; BLANCO, J.Serotypes, virulence genes and intimin types of shigatoxins (verotoxin)-producing Escherichia coli isolatesfrom healthy sheep in Spain. J. Clin. Microbiol. 41:1351-1356. 2003.

[2] BOUVET, J.; BAVAI, C.; ROSSEL, R.; LE ROUX, A.;MONTET, M.P.; RAY-GUENIOT, S.; MAZUY, C.; AR-QUILLIE´RE, C.; VERNOZY-ROZAND, C. Prevalence ofverotoxin-producing Escherichia coli and E. coli O157:H7in pig carcasses from three French slaughterhouses. Int.J. Food Microbiol. 71: 249-255. 2001.

[3] CHRISTENSEN, H. B. Datos categóricos y sus pruebas.Novena unidad. En: Houghton Mifflin Co. (Ed) Estadísti-ca paso a paso. 2a. Ed. Trillas, México. Pp 459-477.1989.

[4] CORNICK, N.A.; HELGERSON, A.F. Transmission andInfectious Dose of Escherichia coli O157:H7 in Swine.Appl Environ Microbiol. 70: 5331-5335. 2004.

[5] DESMARCHELIER, P.M.; BILGE, S.S.; FEGAN, N.;MILLS, L.; VARY JR, J.C.; TARR, P.I. A PCR specific forEscherichia coli O157:H7 based on the rfb locus encod-ing O157 lipopolisaccharide. J Clin Microbiol. 36:1801-1804. 1998.

[6] DIAZ-CINCO, M.E.; GARCIA, A.; ACEDO, E.; GASTE-LUM, A. Recuperación e Incidencia de Escherichia coliO157:H7 en carne molida expendida en el mercado localde la ciudad de Hermosillo, Sonora. In: Resúmenes detrabajos libres: microbiología y toxicología. V Congresodel Noroeste, I Nacional, en Ciencias Alimentarias y Bio-tecnología. Centro de las Artes de la Universidad de So-nora Hermosillo, Sonora. México. 7-12 de noviembre del2005. En línea: http://www.dipa.uson.mx/wb2/dipa/dipa_congresodelnoroeste. 10/02/2006.

[7] DOYLE, J.J.; DOYLE, J.L. A rapid DNA isolation proce-dure for small quantities of fresh leaf tissue. Phytochem.Bull. 19:11-15. 1987.

[8] DOYLE, M.P. Escherichia coli O157:H7 and its signifi-cance in foods. International J. Food Microbiol. 12:289-301. 1991.

[9] ELDER, R.O.; KEEN, J.E.; SIRAGEUA, G.R.;BARKOCY-GALLAGHER, G.A.; KOOHMARAIE, M.;LAEGREID, W.W. Correlation of enterohemorrhagicEscherichia coli O157 prevalence in feces, hides, andcarcasses of beef cattle during processing. Proceedingsof the National Academic of Sciences. 97: 2999-3003.2000.

[10] FEDER, I.; MORGAN, F.; GRAY, J.T.; FRATAMICO, P.;FEDORKA-CRAY, P.J.; PEARCE, R.A.; CALL, J.E.;PERRINE, R.; LUCHANSKY, J.B. Isolation of Escheri-chia coli O157:H7 from Intact Colon Fecal Samples ofSwine. Emerg. Infect. Dis. 9: 380-383. 2003.

[11] FENG, P.; WEAGANT, S.D. Diarrheagenic Escherichiacoli. 2002. Bacteriological Analytical Manual. Chapter 4a.U.S. Food and Drug Administration. Center for FoodSafety and Applied Nutrition. U.S.A. Online: http://www.cfsan.fda.gov/~ebam/bam-4a.html#fn1. 05/01/2006.

[12] FRATAMICO, P.; SACKITEY, S.K.; WIEDMANN, M.;DENG, M.Y. Detection of Escherichia coli O157:H7 bymultiplex PCR. J. Clin. Microbiol. 33: 2188-2191. 1995.

[13] GANNON, V.P.J.; D´SOUZA,S.; GRAHAM, T.; KING,R.K.; RAHN, K.; READ, S. Use of the Flagellar H7 Geneas a Target in Multiplex PCR Assays and ImprovedSpecificity in Identification of Enterohemorrhagic Escheri-chia coli Strains. J. Clin. Microbiol. 35: 656-662. 1997.

[14] GANNON, V.P.J.; RASHED, M.; KING, R.K.; GOSTEYN,E.J. Detection and Characterization of the eae Gene ofShiga-Like Toxin-Producing Escherichia coli Using Po-lymerase Chain Reaction. J. Clin. Microbiol. 31: 1268-1274. 1993.

[15] GILL, C.O.; JONES T. Microbiological sampling of car-casses by excision or swabbing. J. Food Prot. 63: 167-173. 2000.

[16] HEUVELINK, A.E.; SWARTKRUIS-NAHUIS, J.T.; BEU-MER, R.R. DE BOER, E. Occurrence and survival ofverocytotoxin-producing Escherichia coli O157 in meatsobtained from retail outlets in The Netherlands. J. FoodProtec. 62: 1115-1122. 1999.

[17] HEUVELINK, A.E.; VAN DEN BIGGELAAR, F.L.A.M.;ZWARTKRUIS.NAHUIS, J.T.M.; HERBES, R.G.; HUY-BEN, R.; NAGELKERKE, N.; MELCHERS, W.J.G.;MONNENS, L.A.H.; DE BOER, E. Occurrence ofverocytotoxin-producing Escherichia coli O157 on Dutchdairy farms. J. Clin. Microbiol. 36: 3480-3487. 1998.

[18] HUTCHISON, M.L.; NICHOLSON, F.A.; SMITH, K.A.;KEEVIL, C.W.; MOORE, A. A study on farm manureapplications to agricultural land and an assessmentof the risks of pathogen transfer into the food chain.HMSO: MAFF Publications. U.K. Pp. 1-192. 2000.

[19] MEAD, P.S.; GRIFFIN, P.M. Escherichia coli O157:H7.Lancet 352: 1207-1212. 1998.

[20] NATARO, J P.; KAPER, J. B. Diarrheagenic Escherichiacoli. Clin. Microbiol. Rev. 11: 142-201. 1998.

[21] PATON, A.W.; PATON, J.C. Detection and Characteriza-tion of Shiga Toxigenic Escherichia coli by Using Multi-plex PCR Assays for stx1, stx2, eaeA, Enterohemor-

145


rhagic E. coli hlyA, rfb O111, and rfb O157. J. Clin. Micro-biol. 36: 598-602. 1998.

[22] ROGERIE, F.A.; MARECAT, A.; GAMBADE, S.; DU-POND, F.; BEAUBOIS, P.; LANGE, M. Characterizationof Shiga toxin producing E. coli and O157 serotype E.coli isolated in France from healthy domestic cattle. In-ternat. J. Food Microbiol. 63: 217-223. 2001.

[23] SAGARPA (SECRETARÍA DE AGRICULTURA, GANA-DERÍA, DESARROLLO RURAL, PESCA Y ALIMENTA-CIÓN). Exportaciones mexicanas de carnes frescas,refrigeradas o congeladas (toneladas). 2002. México. Enlínea: http://www.sagarpa.gob.mx/Dgg/expocar.htm. 21/02/2006.

[24] SAGARPA (SECRETARÍA DE AGRICULTURA, GANA-DERÍA, DESARROLLO RURAL, PESCA Y ALIMENTA-CIÓN). Estimación del Consumo Nacional Aparente1990-2004. México. 2005. En línea: http://www.sagar-pa.gob. mx/Dgg/ganind3.htm. 21/02/2006.

[25] SCHROEDER, C.M.; WHITE, D.G.; GE, B.; ZHANG, Y.;MCDERMOTT, P.F.; AYERS, S.; ZHAO, S.; MENG, J.Isolation of antimicrobial-resistant Escherichia coli fromretail meats purchased in Greater Washington, DC, EUA.Internat. J. Food Microbiol. 85: 197-202. 2003.

[26] SHAIKH, N.; TARR, P.I. Escherichia coli O157:H7 ShigaToxin-Encoding Bacteriophages: Integrations, Excisions,Truncations, and Evolutionary Implications. J. Bacteriol.185: 3596-3605. 2003.

146


M:Fo

odMi

crob

iolog

y&

Safet

y

JFS M: Food Microbiology and Safety

PCR Detection and Microbiological Isolation ofSalmonella spp. from Fresh Beef and CantaloupesM.A. GALLEGOS-ROBLES, A. MORALES-LOREDO, G. ÁLVAREZ-OJEDA, J.A. OSUNA-GARCÍA, I.O. MART́INEZ,L.H. MORALES-RAMOS, AND P. FRATAMICO

ABSTRACT: Species belonging to the genus Salmonella are an important cause of enteric fevers, gastroenteritis,and septicemia, and the pathogens are commonly transmitted through contaminated food. In this study, poly-merase chain reaction (PCR) amplification of a 287-bp region of the invA gene was compared to a microbiologicaltechnique to determine the presence of Salmonella in retail beef and in cantaloupe rinse samples. Both methodsshowed the same level of sensitivity, detecting 1 CFU/25 g of meat after enrichment for 24 h at 42 ◦C. The presenceof Salmonella was determined in 50 commercial top sirloin beef samples that were not artificially inoculated. Threesamples were positive by the microbiological method, and these samples and an additional sample were positive bythe PCR. Both methods were also used to test surface rinses of cantaloupes collected from 4 farms in Nayarit, Mex-ico. Salmonella was detected by the microbiological method in 9 of 20 samples (45%), whereas the pathogen wasdetected by the PCR in 11 samples (55%). This study demonstrates the utility of the PCR targeting the invA gene todetermine the presence of Salmonella spp. in beef and cantaloupe samples.

Keywords: cantaloupe, detection, meat, PCR, Salmonella

Introduction

Salmonellosis caused by species in the genus Salmonella wasdescribed in 1984 as a “new and significant threat to the pub-lic health” by the World Health Organization (FAO 1984), andSalmonella has remained a major foodborne pathogen associatedwith different types of food. Gutiérrez and others (2000) reportedthe isolation of Salmonella spp. in Mexico in 51% of fast foodsamples, 23% of processed meat products (ham, chorizo, and ba-con), 22% of ground food samples (beef, chicken, fish), 3% of milkproducts, and in 1% of both fresh and powdered eggs. Salmonel-losis outbreaks in the United States linked to the consumption ofcantaloupes implicated Salmonella serotypes Saphra and Poonaas the causative agents from cantaloupes that originated fromMexico (Mohle and others 1999; CDC 2002). Analyses of fruits andvegetables imported to the United States in 1999 showed that of1003 analyzed samples, 35 (3.5%) tested positive for Salmonellaspp., and of these, 8 were cantaloupe samples (22.9%), indicatingthat cantaloupe was the 2nd most contaminated type of productafter cilantro (FDA 2001). Currently in Mexico, the official proce-dure for detection of Salmonella spp. is a cultural method, and thisprocedure could take from 3 to 5 d for confirmation, which is adisadvantage when the results are needed promptly (SSA 1994; Pe-plow and others 1999). Molecular methods, such as the polymerase

MS 20080484 Submitted 6/29/2008, Accepted 10/1/2008. Author Gallegos-Robles is with Facultad de Agricultura y Zootecnia, Univ. Juárez del Es-tado de Durango, Venecia, Gomez Palacio, Durango, México, C.P. 35000.Author Morales-Loredo is with Consorcio Técnico del Noreste de México,A.C. UGRNL. Km. 4.5, Carretera a Reynosa. Guadalupe, N.L. México., C.P.67100. Authors Álvarez-Ojeda and Osuna-Garcı́a are with Inst. Nacionalde Investigaciones, Forestales, Agrı́colas y Pecuarias (INIFAP), Guadalupe,Nuevo León, México, C.P. 67100. Authors Martı́nez and Morales-Ramosare with Facultad de Ciencias Biológicas, Univ. Autónoma de NuevoLeón, San Nicolás de los Garza, Nuevo León, México. Author Fratam-ico is with USDA, Agricultural Research Service, Eastern Regional Re-search Center, Microbial Food Safety Research Unit, 600 E. Mermaid Lane,Wyndmoor, PA 19038 U.S.A. Direct inquiries to author Morales-Loredo(E-mail: [email protected]).

chain reaction (PCR), have shown high sensitivity and specificity fordetecting target pathogens, including Salmonella, in different typesof foods, and the time required to obtain results can be as short as12 h (Ferretti and others 2001; Croci and others 2004). However, mi-crobiological techniques are used as reference methods to demon-strate the efficacy and validity of new techniques (Fernandez 2000).The objective of this study was to compare the sensitivity of a PCRassay to a microbiological method and to evaluate the 2 methodsfor the detection of Salmonella spp. in naturally contaminated beefand cantaloupe rinse samples.

Materials and Methods

Bacterial strain, growth conditions, and preparationof inoculum

S. Typhimurium ATCC 13311 was grown on trypticase soy agar(TSA) (Becton Dickinson Co., Sparks, Md., U.S.A.) at 37 ◦C for 24 h.Afterward, 10 mL of physiological saline solution (0.85%) wereadded to the plate to obtain a homogeneous suspension of bacte-ria. An aliquot of the suspension was diluted to a concentration ofthe nr 5 tube of the MacFarland scale (1.5 × 108 CFU/mL). Ten-foldserial dilutions were prepared in saline solution to give suspensionscontaining 100 to 104 CFU/mL.

Artificial inoculation of top sirloin meat samplesThe top sirloin meat samples were obtained on the same date

from supermarkets in Monterrey, Nuevo León, Mexico. Meat sam-ples were placed into a cooler (at 4 ◦C) and transported to thelaboratory. The samples were immediately separately inoculatedwith the previously mentioned dilutions as described subsequently.Three sterile plastic bags (Whirl-Pak

R©, Nasco, Modesto, Calif.,

U.S.A.) with the capacity of 500 mL were labeled for each dilu-tion, and 25 g of top sirloin meat were weighed and placed ineach plastic bag with 225 mL sterile buffered peptone water (BPW)(Becton Dickinson and Co.). An uninoculated control sample was

C© 2008 Institute of Food Technologists R© Vol. 74, Nr. 1, 2009—JOURNAL OF FOOD SCIENCE M37doi: 10.1111/j.1750-3841.2008.01006.xFurther reproduction without permission is prohibited

M:FoodMicrobiology&

Safety

Detection of Salmonella spp. in food . . .

included to ensure that the meat was not naturally contaminatedwith Salmonella. One milliliter of each bacterial dilution (100 to 104

CFU/mL) was added to the corresponding plastic bag, which wasthen mixed by hand for 2 min followed by incubation at 35 ◦C for24 h. The enrichment and microbiological analyses were performedusing tetrathionate (TT) broth (Becton Dickinson Co.) according tothe method described in the Microbiology Laboratory Guidebook(FSIS-USDA 2004). The experiments were performed in triplicate.

DNA extraction from inoculated top sirloin meatsamples

Three milliliters from the Tetrathionate broth enrichments wereused to form cell pellets by centrifuging at 3000 rpm for 5 min.DNA extraction from the cell pellets was performed using the CTAB(cetyl trimethyl ammonium bromide) method; however, the use ofpolyvinylpyrrolidone and ß-mercaptoethanol was omitted (Doyleand Doyle 1987). The extracted DNA was stored at −20 ◦C.

PCR amplification conditionsAmplification of the target sequence was performed using a PCR

Express thermal cycler (PCR Express; Thermo Hybaid, Middlesex,U.K.). The PCR mixture contained 25 pmoles of each of the primerstargeting the invA gene (Rahn and others 1992), 200 μM of eachof the 4 deoxynucleoside triphosphates (Bioline Inc., Randolph,Mass., U.S.A.), 1 mM MgCl2, 1× Reaction Buffer (200 mM Tris-HClpH 8, 500 mM KCl), 2.5 U of Taq DNA polymerase (Promega, Madi-son, Wis., U.S.A.), 100 ng of DNA template, and deionized water fora final volume of 25 μL. The reaction mixture was subjected to thefollowing thermal cycling conditions: heat denaturation at 95 ◦Cfor 1 min, and then 35 cycles with heat denaturation at 95 ◦C for30 s, primer annealing at 58 ◦C for 30 s, and DNA extension at72 ◦C for 30 s. After the last cycle, samples were maintained at 72 ◦Cfor 10 min to complete synthesis of all strands. The PCR productswere subjected to gel electrophoresis (1.5% agarose; Promega), andthen stained with ethidium bromide (0.5 μg/mL), visualized with aUV transilluminator (Spectroline Transilluminator, Model 7C-254R.Electronics Corp., Westbury, New York, U.S.A.), and photographed.

Detection of Salmonella spp. in samples ofcommercial bovine meat

Fifty different samples of bovine meat (New York Strip andRib Eye) obtained from different supermarkets located in themetropolitan area of Monterrey, N. L. Mexico, were collected dur-ing the month of May in 2005 to determine if the meat was natu-rally contaminated with Salmonella spp. Handling of the samplesand the microbiological and PCR methods were performed as de-scribed previously.

Detection of Salmonella spp. from rinses ofcantaloupe surfaces

Cantaloupe melons were obtained at 4 fields in Nayarit, Mexico,on 4 different dates from February 25 to April 11 of 2005 (1 field perdate). Each field was divided into 5 representative quadrants (I toV). Twenty-five melons were collected from each field, 5 melons perquadrant. The melons were randomly collected, and adhered par-ticles of soil were not removed. Melon surface washings were doneat the field. Each melon was placed in a sterile Whirl-Pak bag usinggloves that were changed with each melon. Twenty-five milliliters of0.1% sterile buffered peptone water were added to each plastic bag,and the fruit was washed thoroughly by shaking and mixing the bagby hand for at least 2 min. The washings from the 5 melons ob-tained from the same quadrant were combined into a single sterileglass bottle, and then placed into a cooler (at 4 ◦C) and transported

to the laboratory. Processing of the samples was performed within24 h of obtaining the rinses. At the laboratory, each sample wasmixed by shaking, and 25 mL were removed and added to 225 mLof BPW to continue with the microbiological method and the PCRassay as described previously.

Results and Discussion

Isolation of Salmonella spp. from artificiallyinoculated meat samples and detection by the PCR

The PCR assay and the microbiological method showed an equallevel of sensitivity with a limit of detection for S. Typhimuriumby both methods of 1 CFU/25 g of meat (initial inoculum level)(data not shown). The primer pair targeting the invA gene, de-signed by Rahn and others (1992), was tested using a collection of630 Salmonella strains and showed 99.4% specificity and no am-plification of DNA from non-Salmonella strains. Moreover, the ad-equacy of invA as a target gene in PCR assays was tested by Daumand others (2002) in a fluorogenic TaqMan PCR assay to confirmthe presence of Salmonella directly from chicken in less than 3 h.Although in the current study a single serovar was tested using thisassay, we have also amplified the PCR product of the invA gene inserovar S. Enteritidis using the same primer pair in assays for thedetection of Salmonella in other types of food samples (unpub-lished data). Thus, the PCR assay targeting the invA gene can po-tentially be used to detect Salmonella in raw beef samples as analternative to the cultural method. However, enrichment is a nec-essary step if fresh products are tested by either the microbiolog-ical method or the PCR, since it is possible that a low level of thepathogen of interest may be present. Growth of the target organ-ism, however, could be inhibited somewhat by the accompanyingflora, which might give a negative result by the cultural method, orthe sensitivity of the PCR assay could be decreased. As such, it is im-portant to select a suitable enrichment medium to inhibit the back-ground flora and the appropriate enrichment time because PCRsensitivity tends to increase with increases in enrichment times(Guo and others 2000).

Microbiological assay and PCR of commercial meatsamples

The data in Table 1 show results of 4 samples, which were pos-itive by the PCR assay and the results of the same samples testedusing the cultural method for the detection of Salmonella spp. Theother 46 meat samples analyzed were negative using both meth-ods. Three samples out of 50 (6%) were detected as positives bythe microbiological method, while the PCR assay detected an ad-ditional sample (8%). Several reasons could explain why a samplewas detected as positive by the PCR (sample nr 12) and not de-tected as positive by the microbiological method. It is possible thatthe cells in the meat sample were injured due to the use of sani-tation products or due to stressful storage conditions, which candamage or even kill the Salmonella. This in turn would affect theability to detect the pathogen using the cultural method, since thismethod is dependent on growth of the cells. Figure 1 shows the

Table 1 --- Meat samples positive for Salmonella spp., byboth the cultural and PCR methods.

Sample nr Cultural method PCR assay

3 + +12 − +33 + +41 + +

M38 JOURNAL OF FOOD SCIENCE—Vol. 74, Nr. 1, 2009

M:Fo

odMi

crob

iolog

y&

Safet

y


PCR products obtained following amplification of the positive meatsamples. Sample nr 12 (lane 3) shows a weak band. Thus, it is pos-sible that there was less PCR product from this sample becausethere were a lower number of cells after enrichment compared tothe other samples. This was likely the reason why the pathogen wasnot detected by plating. Although, the sensitivity of both the mi-crobiological and PCR methods was the same (1 CFU/25 g) whenmeat samples were inoculated with nonstressed cells, in naturallycontaminated samples, the cells may be stressed affecting theirability to grow as rapidly as nonstressed cells in the enrichmentmedium.

Cultural and PCR assays using cantaloupe rinsesamples

The data in Table 2 show that in the 1st field, 4 positive sam-ples were detected (quadrants I to III and V) by the microbiologicalmethod, and there were 4 positives from the same 4 quadrants bythe PCR. However, for quadrants III and IV, the results of both meth-ods differed. In the 2nd field, the results of both methods were thesame, with positive results from samples in all 5 quadrants. In the3rd and 4th fields, all results using the microbiological assay werenegative, while using the PCR, results were positive from samplesfrom the 3rd field from quadrants II and III and were negative fromthe 4th field from all 5 quadrants. With the exception of samplesfrom quadrant III and IV from the 1st field where results differed,results of the PCR assay agreed with those of the cultural method,and in some cases the PCR assay was more sensitive for detection ofSalmonella (quadrants II and III of the 3rd sampling). In summary,Salmonella spp. were detected by the microbiological method in 9of 20 samples (45%), whereas the pathogen was detected by the PCRin 11 samples (55%). These results are partially in agreement withthose of Espinoza-Medina and others (2006) who found that bythe PCR method, 25.7% of samples from in-field cantaloupes were

Figure 1 --- PCR products from the Salmonella invA gene inbeef samples positive for Salmonella spp. Lane 1, molecu-lar weight markers, 100-bp ladder (BIOLINE); lanes 2 to 5,samples 3, 12, 33, and 41, respectively; lane 6, negativecontrol; and lane 7, positive control --- S. Typhimurium.

Table 2 --- Positive results for Salmonella spp. from cantaloupe surface washings using the cultural method and thePCR assay.

Assay results

First field Second field Third field Fourth field

Quadrant Microbiological PCR Microbio logical PCR Microbiological PCR Microbiological PCR

I + + + + − − − −II + + + + − + − −III + − + + − + − −IV − + + + − − − −V + + + + − − − −

positive for Salmonella, whereas no positive samples were detectedby the standard method.

The detection of this pathogen by the PCR was done from an en-richment culture; therefore, in addition to growth of Salmonella,the microflora from the melon samples also grew. Thus, the PCRwas sensitive and specific, since Salmonella was detected in thepresence of other microorganisms found in the melon produc-tion environment. Previous microbiological studies conducted inthe Lagunera Region of Mexico (Froto and others 2004) found thepresence of bacteria that belonged to the Enterobacteriaceae fam-ily and other microorganisms in the cuticle of melons, as well asplant pathogens such as Fusarium spp., Verticillium spp., and Rhi-zoctonia solani, and saprophytes such as Aspergillus spp., Rhizo-pus spp., and Penicillium spp. Also, human pathogens, includingClostridium botulinum, Listeria monocytogenes, Vibrio cholerae,Brucella melitensis, Salmonella Typhi, Salmonella Paratyphi, hep-atitis A virus, Escherichia coli, and Shigella dysenteriae, were found.Most of these pathogens could be found in the soil coming frombovine and avian manure and from human feces (Froto andothers 2004). The absence of the Salmonella invA sequence in otherinvasive bacteria such as Yersinia spp., Shigella spp., and enteroin-vasive E. coli, which also have the capacity to invade epithelial cells,demonstrates the particular specificity and utility of this primerpair for detection of Salmonella spp. (Galán and Curtiss 1991). TheinvA gene has been used as the target in PCR assays mainly for de-tecting Salmonella in poultry, meats, and dairy products, and invegetables and fruits (Guo and others 2000). The PCR results of thecurrent study shown in Figure 2 indicate that the primers could beused for detection of Salmonella from cantaloupe surface washingsand potentially in other types of fruits and vegetable samples, aswell.

The differences in results from field to field may have been dueto differences in levels of Salmonella contamination. This may havebeen influenced in part by changes in the environment, includingloss of specific nutrients and fluctuations in humidity, temperature,and ultraviolet light, all of which could damage bacterial cells and

Figure 2 --- Detection of Salmonella spp., by the PCR fromcantaloupe surface washings. Lanes 1 to 4, quadrants Ito IV (3rd field); lanes 5 to 6 and 8 to 10, quadrants I to V(4th field); lane 7, molecular weight marker, 100-bp lad-der (BIOLINE); lane 11, positive control --- S. Typhimurium;lane 12, negative control.

Vol. 74, Nr. 1, 2009—JOURNAL OF FOOD SCIENCE M39

M:FoodMicrobiology&

Safety


combined with the low probability that human pathogens developstress resistance (Dickinson 1986; O’Brien and Lindow 1988). Fur-thermore, each field was sampled only once during the 1.5 moperiod, and Good Agricultural Practices (GAP) were not appliedin the first 3 fields. For the 4th field, in which no positive resultswere obtained, the cantaloupes were obtained from a field in whichGAP were applied. These included the use of plastic mulch, fertir-rigation, water without evidence of microbiological contamination,portable restrooms in the fields, training of field workers in GAP, useof authorized pesticides, designation of areas where workers mayeat and take breaks, and availability of potable water for workers,which according to HACCP (Hazard Analysis and Critical ControlPoint) plans are basic tools for reducing physical, chemical, and mi-crobiological hazards in agricultural production.

Conclusions

Although meat is usually not consumed raw, there is risk ofSalmonella infection if the meat is improperly cooked, andthere is also the possibility of cross contamination of Salmonellawith foods that are consumed raw. Produce may become con-taminated in the field through the use of contaminated irrigationwater or manure or also from animals or inadequate worker hy-giene. Therefore, the ability to rapidly detect Salmonella in meat,fruit, and other foods could lower the risk of contaminated foodreaching the consumer. Use of a sensitive assay for detection ofSalmonella in melons is also very important, since this food is eatenraw. The PCR assay evaluated in the current study could be used asa screening test, since results would be available in less time thanwith the cultural method. PCR-positive results could then be con-firmed by the cultural method. Because the invA gene is presentin pathogenic Salmonella serotypes, the PCR assay based on theprimer pair targeting this gene could be applied for detection ofSalmonella spp. that may be associated with particular food prod-ucts, including poultry and food products that are consumed rawsuch as fruits and vegetables and/or ready-to-eat food. Further re-search will focus on validating the robustness of the SalmonellaPCR assay in approved laboratories in Mexico for its use as a screen-ing test using different types of samples, along with confirmationof the pathogen by the microbiological method. Additional stud-ies to determine the most prevalent Salmonella serotypes found inbeef and in cantaloupes and to determine the effect of environmen-tal changes and the use of GAP on the prevalence of Salmonella inthese foods are warranted.

AcknowledgmentsThe Consejo Nacional de Ciencia y Tecnologı́a is acknowledged forfunding this research.

References[CDC] Centers for Disease Control and Prevention. 2002. Multistate outbreaks of

Salmonella serotype Poona infections associated with eating cantaloupe fromMexico—United States and Canada, 2000–2002. Morb Mortal Wkly Rep 51:1044–7.

Croci L, Delibato E, Volpe G, Medici DD, Palleschi G. 2004. Comparison of PCR,electrochemical enzyme-linked immunosorbent assays, and the standard cul-ture method for detecting Salmonella in meat products. Appl Environ Microbiol70:1393–6.

Daum LT, Barnes WJ, McAvin JC, Neidert MS, Cooper LA, Huff WB, Gaul L, Riggins WS,Morris S, Salmen A, Lohman KL. 2002. Real-time PCR detection of Salmonella insuspect foods from a gastroenteritis outbreak in Kerr County, Texas. J Clin Microbial40:3050–2.

Dickinson C. 1986. Adaptations of micro-organisms to climatic conditions affectingaerial plant surfaces. In: Fokkema NJ, van den Heu

Documents

IdentiÞcation of Salmonella Serotypes Isolated from Cantaloupe …faz.ujed.mx/Posgrado/maos/AUTOEVALUACION/CATEGORIAS/3... · 2011-03-21 · IdentiÞcation of Salmonella Serotypes