© 2015 APMIS. Published by John Wiley & Sons Ltd. DOI 10.1111/apm.12367
APMIS
Comparison of four molecular methods to type Salmonella Enteritidis strains PITONDO-SILVA,1 ALZIRA M.M. BERGAMINI2 and FABIO CAMPIONI,1 ANDRE ~ 1 JULIANA P. FALCAO 1
Faculdade de Ci^encias Farmac^euticas de Ribeir~ ao Preto, Departamento de An alises Clınicas, Toxicol ogicas e ao Bromatol ogicas, Universidade de S~ ao Paulo (USP), Ribeir~ ao Preto, SP; 2Instituto Adolfo Lutz de Ribeir~ Preto, Ribeir~ ao Preto, SP, Brazil
Campioni F, Pitondo-Silva A, Bergamini AMM, Falc~ ao JP. Comparison of four molecular methods to type Salmonella Enteritidis strains. APMIS 2015. This study compared the pulsed-field gel electrophoresis (PFGE), enterobacterial repetitive intergenic consensus-PCR (ERIC-PCR), multilocus variable-number of tanden-repeat analysis (MLVA), and multilocus sequence typing (MLST) methods for typing 188 Salmonella Enteritidis strains from different sources isolated over a 24-year period in Brazil. PFGE and ERIC-PCR were more efficient than MLVA for subtyping the strains. However, MLVA provided additional epidemiological information for those strains. In addition, MLST showed the Brazilian strains as belonging to the main clonal complex of S. Enteritidis, CC11, and provided the first report of two new STs in the S. enterica database but could not properly subtype the strains. Our results showed that the use of PFGE or ERIC-PCR together with MLVA is suitable to efficiently subtype S. Enteritidis strains and provide important epidemiological information. Key words: Salmonella Enteritidis; PFGE; ERIC-PCR; MLVA; MLST; molecular epidemiology. Juliana P. Falc~ ao, Departamento de An alises Clınicas, Toxicol ogicas e Bromatol ogicas, Faculdade de Ci^encias Farmac^euticas de Ribeir~ ao Preto–USP, Av. do Cafe, s/no Universidade de S~ ao Paulo-USP, Bloco S, Sala 41, Ribeir~ao Preto, SP 14040-903, Brazil. E-mail: [email protected]
Salmonella Enteritidis (S. Enteritidis) is one of the most frequent isolated serovar of Salmonella enterica from foodborne outbreaks in several countries, resulting in high rates of morbidity and mortality and generating public health spending (1–3). Epidemiological studies have been increasingly used to elucidate contamination routes, to differentiate strains involved in the same outbreak from strains isolated in sporadic cases, and to provide data about the global dissemination of a specific pathogen and in population genetic studies, among others (4). Specifically, these studies of S. Enteritidis provided important information about the upsurge of this pathogen worldwide in the mid1980s related to the consumption of raw or undercooked eggs and poultry meat (5–8). The classical epidemiological methods used to type S. Enteritidis include phenotypic methodologies, such as serotyping, phage typing, and antimicrobial
Received 4 November 2014. Accepted 23 December 2014
susceptibility profiling. However, these methodologies have shown limited capacity to discriminate related strains from unrelated strains and to differentiate subtypes among strains from the same species, and have also presented drawbacks related to reproducibility (9–12). Molecular typing methodologies, such as the ‘gold standard’ pulsed-field gel electrophoresis (PFGE) plus enterobacterial repetitive intergenic consensus-PCR (ERIC-PCR), multilocus variablenumber of tanden-repeat analysis (MLVA), and multilocus sequence typing (MLST), have been successfully used to type S. Enteritidis strains, complementing the phenotypic assays. However, some studies have presented variable results related to the ability of specific techniques to differentiate the relatively genetically homogeneous serovar Enteritidis (10, 12–28). In light of the several molecular typing methodologies presented in the literature used to type S. Enteritidis, we decided to compare our findings using some of the most employed and modern 1
CAMPIONI et al.
Fig 1. Diagram of genetic similarity generated by the software eBURSTv3 with the 48 S. Enteritidis strains studied and the other strains of this serovar in the database. Each ST is represented by a dot. In pink, the ST11 found for 44 of the 46 strains of S. Enteritidis studied; green two new STs found (ST1632 and 1633). The dot representing the predicted primary founder of the clonal complex 11 (CC11) is blue colored. The areas of each of the dots indicate the prevalence of the ST in the input data. Black, other STs of S. Enteritidis.
methodologies to type strains of this serotype isolated in Brazil. Thus, we hope to provide a description of our results focused on the methodologies to help other researchers choose the method that better fits the purpose of their study. MATERIAL AND METHODS Bacterial strains A total of 188 strains of S. Enteritidis isolated in Brazil from humans (67), food (61), chickens (44), and farm environments (16), from outbreaks and sporadic cases, between 1986 and 2010 were typed. Detailed description of the strains can be found in the works of Campioni and colleagues (16–18).
Molecular typing methods The ability of PFGE, ERIC-PCR, and MLVA to type S. Enteritidis strains isolated in Brazil was compared based on the results reported in previous studies from our research group (16–18). Briefly, for PFGE, agarose blocks were prepared with each strain studied followed by digestion with 30U of XbaI (Invitrogen - Life Technologies, Carlsbad, CA, USA) overnight at 37 °C. Macrorestriction fragments were resolved by counter-clamped homogeneous electric field electrophoresis in a CHEF-DR III apparatus (Bio-Rad, Hercules, CA, USA) with an electric field of 6 V cm 1 and an angle of 120°. The migration of fragments was performed at 14 °C in 0.5 9 TBE buffer (4.5 mM Tris; 45 mM boric acid; 1 mM EDTA, pH 8.0) and 1.0% ultra-pure pulsed-field agarose (Bio-Rad, Hercules, CA, USA). The pulse times were ramped from 2.2 to 63.8 s over 19 h (16, 18). The relatedness among the PFGE profiles was analyzed with the software package BioNumerics 5.1 (Applied Maths, Austin, TX, USA).
2
Only bands representing fragments between 48.5 and 630.0 kb in size were included in the analysis. A similarity dendrogram was constructed by the UPGMA method, using the DICE similarity coefficient and a position tolerance of 1.2 (16, 18). For ERIC-PCR, reactions were prepared with specific primers to the ERIC sequences, followed by PCR cycling: initial incubation at 94 °C for 7 min; 30 cycles of 30 s at 94 °C, 1 min at 52 °C, and 8 min at 65 °C; and a final incubation at 65 °C for 10 min. The ERICPCR amplicons were resolved in 1.5% agarose gel electrophoresis into bands, which were stained with ethidium bromide (0.5 lg mL 1) and visualized under UV light. The data were analyzed with the software package BioNumerics 5.1 (Applied Maths, Austin). A similarity dendrogram was constructed by the UPGMA method, using the DICE similarity coefficient and a position tolerance of 1.2 (16, 18). For MLVA, multiplex PCRs were prepared using specific primers to 9 VNTR loci followed by capillary electrophoresis on an ABI-3730 DNA analyzer using POP-7 polymer. Each peak was identified according to color and size using GeneMarker software (Softgenetics LLC). The fragment sizes for each VNTR locus were converted into numbers of predicted repeats and analyzed with the software BioNumerics (Applied Maths, Keistraat, Belgium). A similarity dendrogram was constructed by the UPGMA method, using the categorical coefficient (17). The use of MLST to type the strains mentioned above is described for the first time in this study. For this, we chose 46 S. Enteritidis strains among the 188 studied, covering strains of humans, food, chickens, and farm environments of each year of the period studied that was from 1986 to 2010. These strains were typed using the standard protocol available at the MLST databases from the University of Warwick (http://mlst.warwick.ac.uk/mlst/dbs/Senterica/ documents/primersEnterica_html) using sequences from the following housekeeping genes: thrA, purE, sucA, hisD, aroC, hemD, and dnaN. The PCRs for gene amplification and © 2015 APMIS. Published by John Wiley & Sons Ltd
MOLECULAR METHODS TO TYPE S. ENTERITIDIS STRAINS
sequencing were carried out according to the description of Souza and collaborators (29) using the annealing temperatures recommended in the MLST database cited above. All the studied strains were registered, and their data are available in the database.
RESULTS Molecular typing methods
Multilocus sequence typing (MLST) showed 44 among the 46 strains studied as belonging to the same sequence typing (ST), the ST 11. The strain SE 231 belonged to the new ST 1632 and the strain SE 11 to the new ST 1633 (Fig. 1). The methodologies PFGE and ERIC-PCR discriminated the 128 strains of S. Enteritidis isolated from humans (67) and food (61) in 68 and 55 subtypes, respectively. The 60 strains isolated from chickens (44) and farm environments (16) were discriminated in 17 subtypes by PFGE and 24 by ERIC-PCR. Multilocus variable-number of tandenrepeat analysis (MLVA) discriminated all the strains mentioned above in 14 subtypes (16–18). DISCUSSION Salmonella Enteritidis remains a major pathogen isolated from foodborne outbreaks (2). Several molecular typing methods have been described to provide fast and discriminatory results that permit the differentiation of strains isolated from outbreaks and sporadic cases, as well as the elucidation of contamination routes. The gold standard methodology remains PFGE; however, other methodologies have been used successfully to type the genetically homogeneous serotype Enteritidis (12, 30). On typing of S. Enteritidis strains isolated in Brazil from outbreaks and sporadic cases over a 24-year period, the PFGE and ERIC-PCR methodologies were the most discriminatory and showed similar discrimination power (16). Despite the efficiency of these methodologies to differentiate the strains studied into subtypes, the results suggested that a prevalent subtype of S. Enteritidis may have been contaminating food, humans, and animals in Brazil, because the S. Enteritidis strains studied, which were isolated over more than two decades, presented genetic similarity greater than 79.7% by ERIC-PCR and greater than 73.1% by PFGE. When MLVA was used for typing, the 188 S. Enteritidis strains isolated in Brazil were compared to 100 strains isolated in North America from humans (52) and poultry (48). The Brazilian strains were discriminated into 14 subtypes, less than by
© 2015 APMIS. Published by John Wiley & Sons Ltd
PFGE or ERIC-PCR. However, MLVA clearly showed the genetic diversity among strains isolated before and after the beginning of the pandemic in Brazil, which occurred in the mid-1990s, and suggested the possibility that pre-pandemic strains were more likely to be genetically diverse and also that after 1993, a new and prevalent subtype of S. Enteritidis was introduced and still circulates in this country. Moreover, North American strains have been shown to be more genetically diverse (33 subtypes) in comparison with Brazilian strains (14 subtypes) that were more genetically clonal (17). The typing by MLST of 46 representative isolates from the entire 24-year period studied strains of each year, showed 44 strains as belonging to ST 11, which is the most representative ST for S. Enteritidis encompassing the majority of strains of this serovar deposited in the database and isolated in several countries around the world. The remaining two strains were the first isolates of two new STs in the database. Specifically, strain SE 231 presented a new allele of the hemD gene (307), which generated the new ST 1632, and strain SE 11 presented a new allele of the sucA gene (390), which generated the new ST 1633. Thus, our study inserted two new STs into the S. enterica database, which were single-locus variants (SLV) of ST 11 (Fig. 1). Although these two strains represent new STs, they belong to clonal complex 11 (CC11), which is predominantly composed of Salmonella Enteritidis. Moreover, an analysis of the S. enterica MLST database showed that strains of the different serovars are highly clonal because in general, each clonal complex is represented by a predominant serovar, which was observed by Achtman and colleagues (13). In conclusion, PFGE and ERIC-PCR were more efficient for subtyping the strains studied than MLVA. However, MLVA provided additional epidemiological information compared to PFGE and ERIC-PCR. The MLST results typed the strains studied as belonging to clonal complex 11, which is the predominant type found for the majority of the strains from this serovar deposited in the database, which were isolated from different sources and countries. Moreover, it provided the first report of two new STs in the S. enterica database but could not properly subtype the strains. Our results showed that the use of PFGE or ERIC-PCR together with MLVA is suitable for efficiently subtyping S. Enteritidis strains and provide important epidemiological information.
We thank S~ ao Paulo Research Foundation (FAPESP) (Processo 2008/57478-1) for financial support. During the
3
MOLECULAR METHODS TO TYPE S. ENTERITIDIS STRAINS
sequencing were carried out according to the description of Souza and collaborators (29) using the annealing temperatures recommended in the MLST database cited above. All the studied strains were registered, and their data are available in the database.
RESULTS Molecular typing methods
Multilocus sequence typing (MLST) showed 44 among the 46 strains studied as belonging to the same sequence typing (ST), the ST 11. The strain SE 231 belonged to the new ST 1632 and the strain SE 11 to the new ST 1633 (Fig. 1). The methodologies PFGE and ERIC-PCR discriminated the 128 strains of S. Enteritidis isolated from humans (67) and food (61) in 68 and 55 subtypes, respectively. The 60 strains isolated from chickens (44) and farm environments (16) were discriminated in 17 subtypes by PFGE and 24 by ERIC-PCR. Multilocus variable-number of tandenrepeat analysis (MLVA) discriminated all the strains mentioned above in 14 subtypes (16–18). DISCUSSION Salmonella Enteritidis remains a major pathogen isolated from foodborne outbreaks (2). Several molecular typing methods have been described to provide fast and discriminatory results that permit the differentiation of strains isolated from outbreaks and sporadic cases, as well as the elucidation of contamination routes. The gold standard methodology remains PFGE; however, other methodologies have been used successfully to type the genetically homogeneous serotype Enteritidis (12, 30). On typing of S. Enteritidis strains isolated in Brazil from outbreaks and sporadic cases over a 24-year period, the PFGE and ERIC-PCR methodologies were the most discriminatory and showed similar discrimination power (16). Despite the efficiency of these methodologies to differentiate the strains studied into subtypes, the results suggested that a prevalent subtype of S. Enteritidis may have been contaminating food, humans, and animals in Brazil, because the S. Enteritidis strains studied, which were isolated over more than two decades, presented genetic similarity greater than 79.7% by ERIC-PCR and greater than 73.1% by PFGE. When MLVA was used for typing, the 188 S. Enteritidis strains isolated in Brazil were compared to 100 strains isolated in North America from humans (52) and poultry (48). The Brazilian strains were discriminated into 14 subtypes, less than by
© 2015 APMIS. Published by John Wiley & Sons Ltd
PFGE or ERIC-PCR. However, MLVA clearly showed the genetic diversity among strains isolated before and after the beginning of the pandemic in Brazil, which occurred in the mid-1990s, and suggested the possibility that pre-pandemic strains were more likely to be genetically diverse and also that after 1993, a new and prevalent subtype of S. Enteritidis was introduced and still circulates in this country. Moreover, North American strains have been shown to be more genetically diverse (33 subtypes) in comparison with Brazilian strains (14 subtypes) that were more genetically clonal (17). The typing by MLST of 46 representative isolates from the entire 24-year period studied strains of each year, showed 44 strains as belonging to ST 11, which is the most representative ST for S. Enteritidis encompassing the majority of strains of this serovar deposited in the database and isolated in several countries around the world. The remaining two strains were the first isolates of two new STs in the database. Specifically, strain SE 231 presented a new allele of the hemD gene (307), which generated the new ST 1632, and strain SE 11 presented a new allele of the sucA gene (390), which generated the new ST 1633. Thus, our study inserted two new STs into the S. enterica database, which were single-locus variants (SLV) of ST 11 (Fig. 1). Although these two strains represent new STs, they belong to clonal complex 11 (CC11), which is predominantly composed of Salmonella Enteritidis. Moreover, an analysis of the S. enterica MLST database showed that strains of the different serovars are highly clonal because in general, each clonal complex is represented by a predominant serovar, which was observed by Achtman and colleagues (13). In conclusion, PFGE and ERIC-PCR were more efficient for subtyping the strains studied than MLVA. However, MLVA provided additional epidemiological information compared to PFGE and ERIC-PCR. The MLST results typed the strains studied as belonging to clonal complex 11, which is the predominant type found for the majority of the strains from this serovar deposited in the database, which were isolated from different sources and countries. Moreover, it provided the first report of two new STs in the S. enterica database but could not properly subtype the strains. Our results showed that the use of PFGE or ERIC-PCR together with MLVA is suitable for efficiently subtyping S. Enteritidis strains and provide important epidemiological information.
We thank S~ ao Paulo Research Foundation (FAPESP) (Processo 2008/57478-1) for financial support. During the
3
MOLECULAR METHODS TO TYPE S. ENTERITIDIS STRAINS
28. Suh DK, Song JC. Analysis of Salmonella enterica serotype Enteritidis isolated from human and chickens by repetitive sequence-PCR fingerprinting, antibiotic resistance and plasmid profiles. J Vet Sci 2006;7:37– 41. 29. Souza RA, Pitondo-Silva A, Falc~ ao DP, Falc~ ao JP. Evaluation of four molecular typing methodologies as tools for determining taxonomy relations and for
© 2015 APMIS. Published by John Wiley & Sons Ltd
identifying species among Yersinia isolates. J Microbiol Methods 2010;82:141–50. 30. Hyeon JY, Chon JW, Park JH, Kim MS, Oh YH, Choi IS, et al. A Comparison of subtyping methods for differentiating Salmonella enterica serovar Enteritidis isolates obtained from food and human sources. Osong Public Health Res Perspect 2013;4:27–33.
5
APMIS
Comparison of four molecular methods to type Salmonella Enteritidis strains PITONDO-SILVA,1 ALZIRA M.M. BERGAMINI2 and FABIO CAMPIONI,1 ANDRE ~ 1 JULIANA P. FALCAO 1
Faculdade de Ci^encias Farmac^euticas de Ribeir~ ao Preto, Departamento de An alises Clınicas, Toxicol ogicas e ao Bromatol ogicas, Universidade de S~ ao Paulo (USP), Ribeir~ ao Preto, SP; 2Instituto Adolfo Lutz de Ribeir~ Preto, Ribeir~ ao Preto, SP, Brazil
Campioni F, Pitondo-Silva A, Bergamini AMM, Falc~ ao JP. Comparison of four molecular methods to type Salmonella Enteritidis strains. APMIS 2015. This study compared the pulsed-field gel electrophoresis (PFGE), enterobacterial repetitive intergenic consensus-PCR (ERIC-PCR), multilocus variable-number of tanden-repeat analysis (MLVA), and multilocus sequence typing (MLST) methods for typing 188 Salmonella Enteritidis strains from different sources isolated over a 24-year period in Brazil. PFGE and ERIC-PCR were more efficient than MLVA for subtyping the strains. However, MLVA provided additional epidemiological information for those strains. In addition, MLST showed the Brazilian strains as belonging to the main clonal complex of S. Enteritidis, CC11, and provided the first report of two new STs in the S. enterica database but could not properly subtype the strains. Our results showed that the use of PFGE or ERIC-PCR together with MLVA is suitable to efficiently subtype S. Enteritidis strains and provide important epidemiological information. Key words: Salmonella Enteritidis; PFGE; ERIC-PCR; MLVA; MLST; molecular epidemiology. Juliana P. Falc~ ao, Departamento de An alises Clınicas, Toxicol ogicas e Bromatol ogicas, Faculdade de Ci^encias Farmac^euticas de Ribeir~ ao Preto–USP, Av. do Cafe, s/no Universidade de S~ ao Paulo-USP, Bloco S, Sala 41, Ribeir~ao Preto, SP 14040-903, Brazil. E-mail: [email protected]
Salmonella Enteritidis (S. Enteritidis) is one of the most frequent isolated serovar of Salmonella enterica from foodborne outbreaks in several countries, resulting in high rates of morbidity and mortality and generating public health spending (1–3). Epidemiological studies have been increasingly used to elucidate contamination routes, to differentiate strains involved in the same outbreak from strains isolated in sporadic cases, and to provide data about the global dissemination of a specific pathogen and in population genetic studies, among others (4). Specifically, these studies of S. Enteritidis provided important information about the upsurge of this pathogen worldwide in the mid1980s related to the consumption of raw or undercooked eggs and poultry meat (5–8). The classical epidemiological methods used to type S. Enteritidis include phenotypic methodologies, such as serotyping, phage typing, and antimicrobial
Received 4 November 2014. Accepted 23 December 2014
susceptibility profiling. However, these methodologies have shown limited capacity to discriminate related strains from unrelated strains and to differentiate subtypes among strains from the same species, and have also presented drawbacks related to reproducibility (9–12). Molecular typing methodologies, such as the ‘gold standard’ pulsed-field gel electrophoresis (PFGE) plus enterobacterial repetitive intergenic consensus-PCR (ERIC-PCR), multilocus variablenumber of tanden-repeat analysis (MLVA), and multilocus sequence typing (MLST), have been successfully used to type S. Enteritidis strains, complementing the phenotypic assays. However, some studies have presented variable results related to the ability of specific techniques to differentiate the relatively genetically homogeneous serovar Enteritidis (10, 12–28). In light of the several molecular typing methodologies presented in the literature used to type S. Enteritidis, we decided to compare our findings using some of the most employed and modern 1
CAMPIONI et al.
Fig 1. Diagram of genetic similarity generated by the software eBURSTv3 with the 48 S. Enteritidis strains studied and the other strains of this serovar in the database. Each ST is represented by a dot. In pink, the ST11 found for 44 of the 46 strains of S. Enteritidis studied; green two new STs found (ST1632 and 1633). The dot representing the predicted primary founder of the clonal complex 11 (CC11) is blue colored. The areas of each of the dots indicate the prevalence of the ST in the input data. Black, other STs of S. Enteritidis.
methodologies to type strains of this serotype isolated in Brazil. Thus, we hope to provide a description of our results focused on the methodologies to help other researchers choose the method that better fits the purpose of their study. MATERIAL AND METHODS Bacterial strains A total of 188 strains of S. Enteritidis isolated in Brazil from humans (67), food (61), chickens (44), and farm environments (16), from outbreaks and sporadic cases, between 1986 and 2010 were typed. Detailed description of the strains can be found in the works of Campioni and colleagues (16–18).
Molecular typing methods The ability of PFGE, ERIC-PCR, and MLVA to type S. Enteritidis strains isolated in Brazil was compared based on the results reported in previous studies from our research group (16–18). Briefly, for PFGE, agarose blocks were prepared with each strain studied followed by digestion with 30U of XbaI (Invitrogen - Life Technologies, Carlsbad, CA, USA) overnight at 37 °C. Macrorestriction fragments were resolved by counter-clamped homogeneous electric field electrophoresis in a CHEF-DR III apparatus (Bio-Rad, Hercules, CA, USA) with an electric field of 6 V cm 1 and an angle of 120°. The migration of fragments was performed at 14 °C in 0.5 9 TBE buffer (4.5 mM Tris; 45 mM boric acid; 1 mM EDTA, pH 8.0) and 1.0% ultra-pure pulsed-field agarose (Bio-Rad, Hercules, CA, USA). The pulse times were ramped from 2.2 to 63.8 s over 19 h (16, 18). The relatedness among the PFGE profiles was analyzed with the software package BioNumerics 5.1 (Applied Maths, Austin, TX, USA).
2
Only bands representing fragments between 48.5 and 630.0 kb in size were included in the analysis. A similarity dendrogram was constructed by the UPGMA method, using the DICE similarity coefficient and a position tolerance of 1.2 (16, 18). For ERIC-PCR, reactions were prepared with specific primers to the ERIC sequences, followed by PCR cycling: initial incubation at 94 °C for 7 min; 30 cycles of 30 s at 94 °C, 1 min at 52 °C, and 8 min at 65 °C; and a final incubation at 65 °C for 10 min. The ERICPCR amplicons were resolved in 1.5% agarose gel electrophoresis into bands, which were stained with ethidium bromide (0.5 lg mL 1) and visualized under UV light. The data were analyzed with the software package BioNumerics 5.1 (Applied Maths, Austin). A similarity dendrogram was constructed by the UPGMA method, using the DICE similarity coefficient and a position tolerance of 1.2 (16, 18). For MLVA, multiplex PCRs were prepared using specific primers to 9 VNTR loci followed by capillary electrophoresis on an ABI-3730 DNA analyzer using POP-7 polymer. Each peak was identified according to color and size using GeneMarker software (Softgenetics LLC). The fragment sizes for each VNTR locus were converted into numbers of predicted repeats and analyzed with the software BioNumerics (Applied Maths, Keistraat, Belgium). A similarity dendrogram was constructed by the UPGMA method, using the categorical coefficient (17). The use of MLST to type the strains mentioned above is described for the first time in this study. For this, we chose 46 S. Enteritidis strains among the 188 studied, covering strains of humans, food, chickens, and farm environments of each year of the period studied that was from 1986 to 2010. These strains were typed using the standard protocol available at the MLST databases from the University of Warwick (http://mlst.warwick.ac.uk/mlst/dbs/Senterica/ documents/primersEnterica_html) using sequences from the following housekeeping genes: thrA, purE, sucA, hisD, aroC, hemD, and dnaN. The PCRs for gene amplification and © 2015 APMIS. Published by John Wiley & Sons Ltd
MOLECULAR METHODS TO TYPE S. ENTERITIDIS STRAINS
sequencing were carried out according to the description of Souza and collaborators (29) using the annealing temperatures recommended in the MLST database cited above. All the studied strains were registered, and their data are available in the database.
RESULTS Molecular typing methods
Multilocus sequence typing (MLST) showed 44 among the 46 strains studied as belonging to the same sequence typing (ST), the ST 11. The strain SE 231 belonged to the new ST 1632 and the strain SE 11 to the new ST 1633 (Fig. 1). The methodologies PFGE and ERIC-PCR discriminated the 128 strains of S. Enteritidis isolated from humans (67) and food (61) in 68 and 55 subtypes, respectively. The 60 strains isolated from chickens (44) and farm environments (16) were discriminated in 17 subtypes by PFGE and 24 by ERIC-PCR. Multilocus variable-number of tandenrepeat analysis (MLVA) discriminated all the strains mentioned above in 14 subtypes (16–18). DISCUSSION Salmonella Enteritidis remains a major pathogen isolated from foodborne outbreaks (2). Several molecular typing methods have been described to provide fast and discriminatory results that permit the differentiation of strains isolated from outbreaks and sporadic cases, as well as the elucidation of contamination routes. The gold standard methodology remains PFGE; however, other methodologies have been used successfully to type the genetically homogeneous serotype Enteritidis (12, 30). On typing of S. Enteritidis strains isolated in Brazil from outbreaks and sporadic cases over a 24-year period, the PFGE and ERIC-PCR methodologies were the most discriminatory and showed similar discrimination power (16). Despite the efficiency of these methodologies to differentiate the strains studied into subtypes, the results suggested that a prevalent subtype of S. Enteritidis may have been contaminating food, humans, and animals in Brazil, because the S. Enteritidis strains studied, which were isolated over more than two decades, presented genetic similarity greater than 79.7% by ERIC-PCR and greater than 73.1% by PFGE. When MLVA was used for typing, the 188 S. Enteritidis strains isolated in Brazil were compared to 100 strains isolated in North America from humans (52) and poultry (48). The Brazilian strains were discriminated into 14 subtypes, less than by
© 2015 APMIS. Published by John Wiley & Sons Ltd
PFGE or ERIC-PCR. However, MLVA clearly showed the genetic diversity among strains isolated before and after the beginning of the pandemic in Brazil, which occurred in the mid-1990s, and suggested the possibility that pre-pandemic strains were more likely to be genetically diverse and also that after 1993, a new and prevalent subtype of S. Enteritidis was introduced and still circulates in this country. Moreover, North American strains have been shown to be more genetically diverse (33 subtypes) in comparison with Brazilian strains (14 subtypes) that were more genetically clonal (17). The typing by MLST of 46 representative isolates from the entire 24-year period studied strains of each year, showed 44 strains as belonging to ST 11, which is the most representative ST for S. Enteritidis encompassing the majority of strains of this serovar deposited in the database and isolated in several countries around the world. The remaining two strains were the first isolates of two new STs in the database. Specifically, strain SE 231 presented a new allele of the hemD gene (307), which generated the new ST 1632, and strain SE 11 presented a new allele of the sucA gene (390), which generated the new ST 1633. Thus, our study inserted two new STs into the S. enterica database, which were single-locus variants (SLV) of ST 11 (Fig. 1). Although these two strains represent new STs, they belong to clonal complex 11 (CC11), which is predominantly composed of Salmonella Enteritidis. Moreover, an analysis of the S. enterica MLST database showed that strains of the different serovars are highly clonal because in general, each clonal complex is represented by a predominant serovar, which was observed by Achtman and colleagues (13). In conclusion, PFGE and ERIC-PCR were more efficient for subtyping the strains studied than MLVA. However, MLVA provided additional epidemiological information compared to PFGE and ERIC-PCR. The MLST results typed the strains studied as belonging to clonal complex 11, which is the predominant type found for the majority of the strains from this serovar deposited in the database, which were isolated from different sources and countries. Moreover, it provided the first report of two new STs in the S. enterica database but could not properly subtype the strains. Our results showed that the use of PFGE or ERIC-PCR together with MLVA is suitable for efficiently subtyping S. Enteritidis strains and provide important epidemiological information.
We thank S~ ao Paulo Research Foundation (FAPESP) (Processo 2008/57478-1) for financial support. During the
3
MOLECULAR METHODS TO TYPE S. ENTERITIDIS STRAINS
sequencing were carried out according to the description of Souza and collaborators (29) using the annealing temperatures recommended in the MLST database cited above. All the studied strains were registered, and their data are available in the database.
RESULTS Molecular typing methods
Multilocus sequence typing (MLST) showed 44 among the 46 strains studied as belonging to the same sequence typing (ST), the ST 11. The strain SE 231 belonged to the new ST 1632 and the strain SE 11 to the new ST 1633 (Fig. 1). The methodologies PFGE and ERIC-PCR discriminated the 128 strains of S. Enteritidis isolated from humans (67) and food (61) in 68 and 55 subtypes, respectively. The 60 strains isolated from chickens (44) and farm environments (16) were discriminated in 17 subtypes by PFGE and 24 by ERIC-PCR. Multilocus variable-number of tandenrepeat analysis (MLVA) discriminated all the strains mentioned above in 14 subtypes (16–18). DISCUSSION Salmonella Enteritidis remains a major pathogen isolated from foodborne outbreaks (2). Several molecular typing methods have been described to provide fast and discriminatory results that permit the differentiation of strains isolated from outbreaks and sporadic cases, as well as the elucidation of contamination routes. The gold standard methodology remains PFGE; however, other methodologies have been used successfully to type the genetically homogeneous serotype Enteritidis (12, 30). On typing of S. Enteritidis strains isolated in Brazil from outbreaks and sporadic cases over a 24-year period, the PFGE and ERIC-PCR methodologies were the most discriminatory and showed similar discrimination power (16). Despite the efficiency of these methodologies to differentiate the strains studied into subtypes, the results suggested that a prevalent subtype of S. Enteritidis may have been contaminating food, humans, and animals in Brazil, because the S. Enteritidis strains studied, which were isolated over more than two decades, presented genetic similarity greater than 79.7% by ERIC-PCR and greater than 73.1% by PFGE. When MLVA was used for typing, the 188 S. Enteritidis strains isolated in Brazil were compared to 100 strains isolated in North America from humans (52) and poultry (48). The Brazilian strains were discriminated into 14 subtypes, less than by
© 2015 APMIS. Published by John Wiley & Sons Ltd
PFGE or ERIC-PCR. However, MLVA clearly showed the genetic diversity among strains isolated before and after the beginning of the pandemic in Brazil, which occurred in the mid-1990s, and suggested the possibility that pre-pandemic strains were more likely to be genetically diverse and also that after 1993, a new and prevalent subtype of S. Enteritidis was introduced and still circulates in this country. Moreover, North American strains have been shown to be more genetically diverse (33 subtypes) in comparison with Brazilian strains (14 subtypes) that were more genetically clonal (17). The typing by MLST of 46 representative isolates from the entire 24-year period studied strains of each year, showed 44 strains as belonging to ST 11, which is the most representative ST for S. Enteritidis encompassing the majority of strains of this serovar deposited in the database and isolated in several countries around the world. The remaining two strains were the first isolates of two new STs in the database. Specifically, strain SE 231 presented a new allele of the hemD gene (307), which generated the new ST 1632, and strain SE 11 presented a new allele of the sucA gene (390), which generated the new ST 1633. Thus, our study inserted two new STs into the S. enterica database, which were single-locus variants (SLV) of ST 11 (Fig. 1). Although these two strains represent new STs, they belong to clonal complex 11 (CC11), which is predominantly composed of Salmonella Enteritidis. Moreover, an analysis of the S. enterica MLST database showed that strains of the different serovars are highly clonal because in general, each clonal complex is represented by a predominant serovar, which was observed by Achtman and colleagues (13). In conclusion, PFGE and ERIC-PCR were more efficient for subtyping the strains studied than MLVA. However, MLVA provided additional epidemiological information compared to PFGE and ERIC-PCR. The MLST results typed the strains studied as belonging to clonal complex 11, which is the predominant type found for the majority of the strains from this serovar deposited in the database, which were isolated from different sources and countries. Moreover, it provided the first report of two new STs in the S. enterica database but could not properly subtype the strains. Our results showed that the use of PFGE or ERIC-PCR together with MLVA is suitable for efficiently subtyping S. Enteritidis strains and provide important epidemiological information.
We thank S~ ao Paulo Research Foundation (FAPESP) (Processo 2008/57478-1) for financial support. During the
3
MOLECULAR METHODS TO TYPE S. ENTERITIDIS STRAINS
28. Suh DK, Song JC. Analysis of Salmonella enterica serotype Enteritidis isolated from human and chickens by repetitive sequence-PCR fingerprinting, antibiotic resistance and plasmid profiles. J Vet Sci 2006;7:37– 41. 29. Souza RA, Pitondo-Silva A, Falc~ ao DP, Falc~ ao JP. Evaluation of four molecular typing methodologies as tools for determining taxonomy relations and for
© 2015 APMIS. Published by John Wiley & Sons Ltd
identifying species among Yersinia isolates. J Microbiol Methods 2010;82:141–50. 30. Hyeon JY, Chon JW, Park JH, Kim MS, Oh YH, Choi IS, et al. A Comparison of subtyping methods for differentiating Salmonella enterica serovar Enteritidis isolates obtained from food and human sources. Osong Public Health Res Perspect 2013;4:27–33.
5
