880 Journal o f Food Protection, Vol. 77, No. 6, 2014, Pages 880-884 doi: 10.4315/0362-028X.JFP-13-368 Copyright © , International Association for Food Protection
A rc o b a c te r: Comparison of Isolation Methods, Diversity, and
Potential Pathogenic Factors in Commercially Retailed Chicken Breast Meat from Costa Rica KAROLINA L. FALLAS-PADILLA, 1 CARLOS E. RODRIGUEZ-RODRIGUEZ, HERIBERTO FERNANDEZ JARAMILLO,3 a n d MARIA LAURA ARIAS ECHANDI1* 1Tropical Disease Research Center and Microbiology Faculty and 2Research Center o f Environmental Pollution (C1CA), Universidad de Costa Rica, 2060 San Jost, Costa Rica; and 3Institute o f Clinical Microbiology, Universidad Austral de Chile, 5090000 Valdivia, Chile MS 13-368: Received 6 September 2013/Accepted 20 January 2014
ABSTRACT Arcobacter species have been recognized as potential food- and waterborne pathogens. The lack of standardized isolation methods and the relatively scarce knowledge about their prevalence and distribution as emerging pathogens are due to the limitations in their detection and identification. This study aimed to determine the presence and the identification of Arcobacter in chicken breast samples commercially retailed in San Jose, Costa Rica, as well as to describe the adherence and invasive potential of the strains to human cells (HEp-2). Fifty chicken breast samples were collected from retail markets in the metropolitan area of the country. Six different isolation methodologies were applied for the isolation of Arcobacter. Isolation strategies consisted of combinations of enrichments in de Boer or Houf selective broths and subsequent isolation in blood agar (directly or with a previous passive membrane filtration step) or Arcobacter selective agar. Suspicious colonies were identified with a genus-specific PCR, whereas species-level identification was achieved with a multiplex PCR. The overall isolation frequency of Arcobacter was 56%. From the isolation strategies, the combination of enrichment in Houf selective broth followed by filtration on blood agar showed the best performance, with a sensitivity of 89% and a specificity of 84%. A total of 46 isolates were confirmed as Arcobacter with the genus-specific PCR, from which 27 (59%) corresponded to Arcobacter butzleri, 9 (19%) to Arcobacter cryaerophilus, and 10 (22%) were not identified with this multiplex PCR. Regarding the potential pathogenicity, 75% of the isolates presented adherence to HEp-2 cells, while only 22% were invasive to that cell line. All invasive strains were A. butzleri or nonidentified strains. The results show the presence of potentially pathogenic Arcobacter in poultry and recognize the importance it should receive as a potential foodbome pathogen from public health authorities.
The genus Arcobacter comprises, at least, 17 species of gram-negative, non-spore-forming, curved rods (25) be longing to the family Campylobacteriaceae, first isolated in the late 1970s from aborted bovine fetuses (13) and formerly known as aerotolerant Campylobacter-Mke bacteria (28, 29). Habitats occupied by Arcobacter are diverse, with some species living in association with animals, while others are found in seawater or sludge (17). Significance of Arcobacter in veterinary microbiology is related to its association with mastitis and abortions in animals (19). On the other hand, three species, Arcobacter butzleri, Arco bacter cryaerophilus, and Arcbacter skirrowii are reported as human pathogens (34), with A. butzleri and A. cryaerophilus associated with diarrhea and bacteremia (19, 30) and A. skirrowii with chronic diarrhea in elderly patients (39). Furthermore, A. butzleri was recently recognized as an etiological agent of traveler’s diarrhea (14). Despite its relation to disease, few studies on pathogenicity and virulence factors have been reported (11, 20, 26). * Author for correspondence. Tel: 506-2511 8643; Fax: 506-2225 2374; E-mail: [email protected].
Arcobacter species have been recognized as potential food- and waterborne pathogens (35). Reports of their presence in diverse products of animal origin include poultry (3), beef (32), pork (37), shellfish (7), and milk (33). Similarly, isolation from drinking and fecally polluted water has been reported (8, 24). Diverse culture-based methods have been used in the isolation of Arcobacter from fecal and food samples, ranging from Leptospira and Campylobacter techniques to those that use Arcobacterspecific selective media (28). Nonetheless, standard meth ods for the detection of Arcobacter have not been established in such matrices (34), making the comparison of different studies a difficult task. Moreover, the lack of an appropriate standard detection method may lead to the underestimation of the pathogen (34). Although little information is available in Latin America, isolation of Arcobacter has been reported in several countries in Central and South America (10,15,16). This study aimed to determine the frequency and diversity of Arcobacter species in fresh chicken breast samples obtained from retailed markets in San Jose, Costa Rica.
J. Food Prot., Vol. 77, No. 6
ARCOBACTER ISOLATION FROM CHICKEN BREAST SAMPLES
Several isolation methods were used to avoid the underes timation of the real frequency of Arcobacter. Adherence and invasion assays were done to determine the potential pathogenicity of the isolates. M A TER IA LS A N D M E TH O D S Sampling and isolation. Fifty samples of raw chicken breast were randomly collected from retail markets in the metropolitan area of San Jose, Costa Rica, during January to August 2012. Samples were transported at 4 to -6 °C to the Food Microbiology Laboratory, University of Costa Rica, and processed within 24 h. Six different isolation strategies consisting of modifications from the original methodology by de Boer et al. (9) were used. Each sample was rinsed with 90 ml of sterile peptone water 0.1%, from which 1 ml was transferred to tubes containing either 10 ml of de Boer selective broth (9) or 10 ml of Houf selective broth (21) and incubated under aerobic conditions for 48 h at 25°C. Then, one loopful of incubated mixture from each tube was used to inoculate plates of blood agar (BA) and Arcobacter selective agar (AA). Additionally, 100 pi from the broths were placed on 0.45-pm-poresize cellulose acetate sterile filters placed on BA plates. The filters rested on the plates for 3 h at 25°C before being removed. All the plates were incubated at 25°C for 5 to 7 days under aerobic conditions. Presumptive colonies (pointed, waterdrop-like, defined edge) were tested by Gram stain, and those that showed gram negative, curved, or spiral rods were streaked onto fresh BA plates for purification. A dense inoculum (two to three colonies) was placed in Eppendorf tubes containing 1 ml of Brucella broth (Oxoid, Neapan, Ontario, Canada) supplemented with fetal bovine serum (10%) and glycerol (15%) for storage at -20°C . Codification of the isolation strategies was as follows: B or H (depending on the enrichment medium, de Boer or Houf selective broth, respectively) and BA, AA, or F (depending on the isolation medium, BA, AA, or filtration on BA, respectively). For example, B-F indicates enrichment in de Boer selective broth and filtration on BA. The sensitivity of each method was calculated as the ability of the method to detect Arcobacter-positive samples. The number of Arcobacter-positive samples was taken as the total number of samples testing positive for the presence of Arcobacter isolates (confirmed by genus-specific PCR, as described below), using all the isolation methods. Sensitivity was determined as the percentage of Arcobacfer-positive samples obtained with a particular method, with respect to the total number of positive samples. Specificity was calculated as the percentage of Arcobacter r isolates obtained using each method, respective of the total isolates retrieved overall (28). Molecular identification of Arcobacter. The Arcobacter genus-specific 16S rRNA fragment was amplified from total genomic DNA using a Taq polymerase kit (Oxoid). For DNA extraction, the target colonies were suspended in Eppendorf tubes containing 1 ml of sterile distilled water, boiled for 20 min, and centrifuged at 4,040 x g for 10 min to obtain the genomic DNA. The supernatant was used as the template in the genus-specific PCR. PCR amplification was performed in a reaction mixture (50 pi) containing Tris HC1 buffer (pH 7.4), 0.2 mM deoxynucleoside triphosphate, 1.5 mM MgCl2, 1 pM each primer, 1.5 U/pl Taq polymerase (Oxoid), and 5 pi of template DNA. The Arcobacter genus-specific 16S rRNA fragment was amplified using the forward primer Arco I (5'-AGA GAT TAG CCT GTA TTG TAT C-3') and the reverse primer Arco n (5'-TAG CAT CCC CGC TTC GAA TGA-3') (19). The thermocycling program was 94°C for 4 min, 94°C for 1 min, 56°C for 1 min and 72°C for 1 min (25 cycles), and 72°C for 7 min. Analysis of PCR products was
881
done by electrophoresis (60 V, 1.5 h) in 1.5% agarose gels (wt/vol) with a mass ruler (100 to 1,000 bp) and fast red staining (Thermo Fisher Scientific, Waltham, MA). Positive Arcobacter identifica tion was reported when a product of 1,223 bp was obtained. Positive isolates were then analyzed to species level with a multiplex PCR. Amplification of the specific fragments was done with the forward primer Arco F (5'-GCT AGA GGA AGA GAA ATC AA-3'), the reverse primers ButR (5'-TCC TGA TAC AAG ATA ATT GTA CG-3'), TherR (5'-GCA ACC TCT TTG GCT TAC GAA-3'), CibR (5'-CGA ACA GGA TTC TCA CCT GT3'), and SkiR (5'-TCA GGA TAC CAT TAA AGT TAT TGA TG-3') for A. butzleri (2,061 bp), A. thereius (1,590 bp), A. cibarius (1,125 bp), and A. skirrowii (198 bp), respectively (12). Amplification of the 395-bp fragment from A. cryaerophilus was done with the primers CriF (5'-CAG AGG AAG AGA AAT CAA AT-3') and CriR (5'-CCC ACT ATT CCA TCA GTG AG-3') (12). Template preparation and reaction mixture were done as described above. The thermocycling program was 94°C for 4 min; 94°C for 45 s, 58°C for 45 s, and 72°C for 2 min (30 cycles); and 72°C for 7 min. The size of the products was estimated by electrophoresis (120 V, 2 h) in 2% agarose gels (wt/vol) using the mass ruler (100 to 1,000 bp) and fast red staining. In vitro adherence and invasion assays. The assays of cell adherence and invasion were performed according to Fernandez et al. (14). Briefly, confluent HEp-2 cell monolayers on slide covers placed in 12-well plates and maintained in RPMI medium (supplemented with 10% fetal bovine serum, without antibiotics) were exposed to a standardized bacterial inoculum (0.5 ml, McFar land 2; BBL, BD, Franklin Lakes, NJ) for 2 h at 37°C and 5% C 0 2. After three washes with sterile phosphate-buffered saline (PBS), the slides were fixed with pure methanol for 15 min and stained with Giemsa. Results were expressed as the percentage of HEp-2 cells with bacterial cell adhered, examining 10 to 15 microscopic fields, as follows: 0 to 25%, negative; 26 to 50%, positive (+); 51 to 76%, positive (++); and 76 to 100%, positive (+++). In the invasion assay, HEp-2 cells were exposed to bacterial inocula, as described above. Then, monolayers were washed three times with PBS and incubated (same conditions) after adding 1 ml of RPMI medium to each well, following fixation with pure methanol in the same conditions described above. Finally, monolayers were stained with 0.01% acridine orange and 0.05% crystal violet. A positive result was defined as invasion of one or more bacteria into at least 25% HEp-2 cells. Assays were done in duplicates. One uninoculated well and one inoculated with Escherichia coli 0157:H7 ATCC B9831 of known adherence were used as negative and positive controls, respectively (14). RESULTS
From the 50 chicken breasts samples analyzed, a total of 28 samples were positive for Arcobacter as determined by the genus-specific PCR, representing an isolation frequency of 56%. From the 28 positive samples, 25 (89%) were obtained by the combination of an enrichment in Houf selective broth and filtration on BA, as the best strategy/method to use (Table 1). Although isolation strategies using AA also showed high specificity (>80%), their sensitivity was very low (
A rc o b a c te r: Comparison of Isolation Methods, Diversity, and
Potential Pathogenic Factors in Commercially Retailed Chicken Breast Meat from Costa Rica KAROLINA L. FALLAS-PADILLA, 1 CARLOS E. RODRIGUEZ-RODRIGUEZ, HERIBERTO FERNANDEZ JARAMILLO,3 a n d MARIA LAURA ARIAS ECHANDI1* 1Tropical Disease Research Center and Microbiology Faculty and 2Research Center o f Environmental Pollution (C1CA), Universidad de Costa Rica, 2060 San Jost, Costa Rica; and 3Institute o f Clinical Microbiology, Universidad Austral de Chile, 5090000 Valdivia, Chile MS 13-368: Received 6 September 2013/Accepted 20 January 2014
ABSTRACT Arcobacter species have been recognized as potential food- and waterborne pathogens. The lack of standardized isolation methods and the relatively scarce knowledge about their prevalence and distribution as emerging pathogens are due to the limitations in their detection and identification. This study aimed to determine the presence and the identification of Arcobacter in chicken breast samples commercially retailed in San Jose, Costa Rica, as well as to describe the adherence and invasive potential of the strains to human cells (HEp-2). Fifty chicken breast samples were collected from retail markets in the metropolitan area of the country. Six different isolation methodologies were applied for the isolation of Arcobacter. Isolation strategies consisted of combinations of enrichments in de Boer or Houf selective broths and subsequent isolation in blood agar (directly or with a previous passive membrane filtration step) or Arcobacter selective agar. Suspicious colonies were identified with a genus-specific PCR, whereas species-level identification was achieved with a multiplex PCR. The overall isolation frequency of Arcobacter was 56%. From the isolation strategies, the combination of enrichment in Houf selective broth followed by filtration on blood agar showed the best performance, with a sensitivity of 89% and a specificity of 84%. A total of 46 isolates were confirmed as Arcobacter with the genus-specific PCR, from which 27 (59%) corresponded to Arcobacter butzleri, 9 (19%) to Arcobacter cryaerophilus, and 10 (22%) were not identified with this multiplex PCR. Regarding the potential pathogenicity, 75% of the isolates presented adherence to HEp-2 cells, while only 22% were invasive to that cell line. All invasive strains were A. butzleri or nonidentified strains. The results show the presence of potentially pathogenic Arcobacter in poultry and recognize the importance it should receive as a potential foodbome pathogen from public health authorities.
The genus Arcobacter comprises, at least, 17 species of gram-negative, non-spore-forming, curved rods (25) be longing to the family Campylobacteriaceae, first isolated in the late 1970s from aborted bovine fetuses (13) and formerly known as aerotolerant Campylobacter-Mke bacteria (28, 29). Habitats occupied by Arcobacter are diverse, with some species living in association with animals, while others are found in seawater or sludge (17). Significance of Arcobacter in veterinary microbiology is related to its association with mastitis and abortions in animals (19). On the other hand, three species, Arcobacter butzleri, Arco bacter cryaerophilus, and Arcbacter skirrowii are reported as human pathogens (34), with A. butzleri and A. cryaerophilus associated with diarrhea and bacteremia (19, 30) and A. skirrowii with chronic diarrhea in elderly patients (39). Furthermore, A. butzleri was recently recognized as an etiological agent of traveler’s diarrhea (14). Despite its relation to disease, few studies on pathogenicity and virulence factors have been reported (11, 20, 26). * Author for correspondence. Tel: 506-2511 8643; Fax: 506-2225 2374; E-mail: [email protected].
Arcobacter species have been recognized as potential food- and waterborne pathogens (35). Reports of their presence in diverse products of animal origin include poultry (3), beef (32), pork (37), shellfish (7), and milk (33). Similarly, isolation from drinking and fecally polluted water has been reported (8, 24). Diverse culture-based methods have been used in the isolation of Arcobacter from fecal and food samples, ranging from Leptospira and Campylobacter techniques to those that use Arcobacterspecific selective media (28). Nonetheless, standard meth ods for the detection of Arcobacter have not been established in such matrices (34), making the comparison of different studies a difficult task. Moreover, the lack of an appropriate standard detection method may lead to the underestimation of the pathogen (34). Although little information is available in Latin America, isolation of Arcobacter has been reported in several countries in Central and South America (10,15,16). This study aimed to determine the frequency and diversity of Arcobacter species in fresh chicken breast samples obtained from retailed markets in San Jose, Costa Rica.
J. Food Prot., Vol. 77, No. 6
ARCOBACTER ISOLATION FROM CHICKEN BREAST SAMPLES
Several isolation methods were used to avoid the underes timation of the real frequency of Arcobacter. Adherence and invasion assays were done to determine the potential pathogenicity of the isolates. M A TER IA LS A N D M E TH O D S Sampling and isolation. Fifty samples of raw chicken breast were randomly collected from retail markets in the metropolitan area of San Jose, Costa Rica, during January to August 2012. Samples were transported at 4 to -6 °C to the Food Microbiology Laboratory, University of Costa Rica, and processed within 24 h. Six different isolation strategies consisting of modifications from the original methodology by de Boer et al. (9) were used. Each sample was rinsed with 90 ml of sterile peptone water 0.1%, from which 1 ml was transferred to tubes containing either 10 ml of de Boer selective broth (9) or 10 ml of Houf selective broth (21) and incubated under aerobic conditions for 48 h at 25°C. Then, one loopful of incubated mixture from each tube was used to inoculate plates of blood agar (BA) and Arcobacter selective agar (AA). Additionally, 100 pi from the broths were placed on 0.45-pm-poresize cellulose acetate sterile filters placed on BA plates. The filters rested on the plates for 3 h at 25°C before being removed. All the plates were incubated at 25°C for 5 to 7 days under aerobic conditions. Presumptive colonies (pointed, waterdrop-like, defined edge) were tested by Gram stain, and those that showed gram negative, curved, or spiral rods were streaked onto fresh BA plates for purification. A dense inoculum (two to three colonies) was placed in Eppendorf tubes containing 1 ml of Brucella broth (Oxoid, Neapan, Ontario, Canada) supplemented with fetal bovine serum (10%) and glycerol (15%) for storage at -20°C . Codification of the isolation strategies was as follows: B or H (depending on the enrichment medium, de Boer or Houf selective broth, respectively) and BA, AA, or F (depending on the isolation medium, BA, AA, or filtration on BA, respectively). For example, B-F indicates enrichment in de Boer selective broth and filtration on BA. The sensitivity of each method was calculated as the ability of the method to detect Arcobacter-positive samples. The number of Arcobacter-positive samples was taken as the total number of samples testing positive for the presence of Arcobacter isolates (confirmed by genus-specific PCR, as described below), using all the isolation methods. Sensitivity was determined as the percentage of Arcobacfer-positive samples obtained with a particular method, with respect to the total number of positive samples. Specificity was calculated as the percentage of Arcobacter r isolates obtained using each method, respective of the total isolates retrieved overall (28). Molecular identification of Arcobacter. The Arcobacter genus-specific 16S rRNA fragment was amplified from total genomic DNA using a Taq polymerase kit (Oxoid). For DNA extraction, the target colonies were suspended in Eppendorf tubes containing 1 ml of sterile distilled water, boiled for 20 min, and centrifuged at 4,040 x g for 10 min to obtain the genomic DNA. The supernatant was used as the template in the genus-specific PCR. PCR amplification was performed in a reaction mixture (50 pi) containing Tris HC1 buffer (pH 7.4), 0.2 mM deoxynucleoside triphosphate, 1.5 mM MgCl2, 1 pM each primer, 1.5 U/pl Taq polymerase (Oxoid), and 5 pi of template DNA. The Arcobacter genus-specific 16S rRNA fragment was amplified using the forward primer Arco I (5'-AGA GAT TAG CCT GTA TTG TAT C-3') and the reverse primer Arco n (5'-TAG CAT CCC CGC TTC GAA TGA-3') (19). The thermocycling program was 94°C for 4 min, 94°C for 1 min, 56°C for 1 min and 72°C for 1 min (25 cycles), and 72°C for 7 min. Analysis of PCR products was
881
done by electrophoresis (60 V, 1.5 h) in 1.5% agarose gels (wt/vol) with a mass ruler (100 to 1,000 bp) and fast red staining (Thermo Fisher Scientific, Waltham, MA). Positive Arcobacter identifica tion was reported when a product of 1,223 bp was obtained. Positive isolates were then analyzed to species level with a multiplex PCR. Amplification of the specific fragments was done with the forward primer Arco F (5'-GCT AGA GGA AGA GAA ATC AA-3'), the reverse primers ButR (5'-TCC TGA TAC AAG ATA ATT GTA CG-3'), TherR (5'-GCA ACC TCT TTG GCT TAC GAA-3'), CibR (5'-CGA ACA GGA TTC TCA CCT GT3'), and SkiR (5'-TCA GGA TAC CAT TAA AGT TAT TGA TG-3') for A. butzleri (2,061 bp), A. thereius (1,590 bp), A. cibarius (1,125 bp), and A. skirrowii (198 bp), respectively (12). Amplification of the 395-bp fragment from A. cryaerophilus was done with the primers CriF (5'-CAG AGG AAG AGA AAT CAA AT-3') and CriR (5'-CCC ACT ATT CCA TCA GTG AG-3') (12). Template preparation and reaction mixture were done as described above. The thermocycling program was 94°C for 4 min; 94°C for 45 s, 58°C for 45 s, and 72°C for 2 min (30 cycles); and 72°C for 7 min. The size of the products was estimated by electrophoresis (120 V, 2 h) in 2% agarose gels (wt/vol) using the mass ruler (100 to 1,000 bp) and fast red staining. In vitro adherence and invasion assays. The assays of cell adherence and invasion were performed according to Fernandez et al. (14). Briefly, confluent HEp-2 cell monolayers on slide covers placed in 12-well plates and maintained in RPMI medium (supplemented with 10% fetal bovine serum, without antibiotics) were exposed to a standardized bacterial inoculum (0.5 ml, McFar land 2; BBL, BD, Franklin Lakes, NJ) for 2 h at 37°C and 5% C 0 2. After three washes with sterile phosphate-buffered saline (PBS), the slides were fixed with pure methanol for 15 min and stained with Giemsa. Results were expressed as the percentage of HEp-2 cells with bacterial cell adhered, examining 10 to 15 microscopic fields, as follows: 0 to 25%, negative; 26 to 50%, positive (+); 51 to 76%, positive (++); and 76 to 100%, positive (+++). In the invasion assay, HEp-2 cells were exposed to bacterial inocula, as described above. Then, monolayers were washed three times with PBS and incubated (same conditions) after adding 1 ml of RPMI medium to each well, following fixation with pure methanol in the same conditions described above. Finally, monolayers were stained with 0.01% acridine orange and 0.05% crystal violet. A positive result was defined as invasion of one or more bacteria into at least 25% HEp-2 cells. Assays were done in duplicates. One uninoculated well and one inoculated with Escherichia coli 0157:H7 ATCC B9831 of known adherence were used as negative and positive controls, respectively (14). RESULTS
From the 50 chicken breasts samples analyzed, a total of 28 samples were positive for Arcobacter as determined by the genus-specific PCR, representing an isolation frequency of 56%. From the 28 positive samples, 25 (89%) were obtained by the combination of an enrichment in Houf selective broth and filtration on BA, as the best strategy/method to use (Table 1). Although isolation strategies using AA also showed high specificity (>80%), their sensitivity was very low (
