- Email: [email protected]
H- strains and hierarchy of their phylogenetic relationships
Infection, Genetics and Evolution 12 (2012) 1724–1728
Contents lists available at SciVerse ScienceDirect
Infection, Genetics and Evolution journal homepage: www.elsevier.com/locate/meegid
Clade analysis of enterohemorrhagic Escherichia coli serotype O157:H7/H- strains and hierarchy of their phylogenetic relationships Eiji Yokoyama ⇑, Shinichiro Hirai, Ruiko Hashimoto, Masako Uchimura Division of Bacteriology, Chiba Prefectural Institute of Public Health, Chiba, Japan
a r t i c l e
i n f o
Article history: Received 27 January 2012 Received in revised form 12 June 2012 Accepted 7 July 2012 Available online 27 July 2012 Keywords: Enterohemorrhagic Escherichia coli serotype O157:H7/HClade Lineage-specific polymorphism assay 6 (LSAP6) Cluster Subgroup Evolutionary model
a b s t r a c t Enterohemorrhagic Escherichia coli serotype O157:H7/H-(O157) strains isolated in Chiba prefecture, Japan, during 2002–2009 were studied by lineage, subgroup, cluster, and clade analysis. Lineage analysis of 470 O157 strains with no known epidemiological relationships using lineage specific polymorphism assay-6 showed that there were 242 lineage I strains, 160 lineage I/II strains, 67 lineage II strains, and 1 atypical strain. Clade analysis of these strains by single nucleotide polymorphism in eight loci showed that lineage I contained all the clade 1, clade 2, and clade 3 strains, and some of the clade 4/5 strains. In contrast, clade 7, clade 8, and the remaining clade 4/5 strains were divided between lineage I/II and II, and clade 6 was in lineage I/II, suggesting paraphyletic evolution of these lineages. Cluster and subgroup analysis of the stx phage insertion site showed that all lineage I strains were cluster 3 and all lineage I/II and II strains, with the exception of clade 9, were in cluster 1. Clade analysis also indicated that there were three phylogenetic groups of clade 4/5 strains: ancestral groups containing lineage I/IIand II strains and a descendant group containing lineages I. Analysis of stx2c gene distribution showed that stx2c was in ancestral clade 4/5 strains but not in descendant 4/5 strains, suggesting that the ancestral group may be clade 4 as reported by Manning et al. The results with the markers used in this study suggested that the hierarchy of O157 phylogenetic relationships was lineage as the upper level, followed by subgroup and then cluster, and clade as the lowest level. The need for refinement of clade definition and modification of the model of the O157 evolution have been discussed. Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction Aspects of the evolution of enterohemorrhagic Escherichia coli serotype O157:H7/H- (O157) have been elucidated during the past two decades. E. coli serotype O55:H7 (O55) was thought to be the ancestor of O157, and O157 were divided into three groups, (i.e., subgroup A, B, and C) based on multilocus enzyme electrophoresis analysis (Whittam and Wilson, 1988). Subgroup A contained O55 strains, subgroup B contained sorbitol-fermenting O157 strains, and subgroup C contained sorbitol-nonfermenting O157 strains. The O157 strains in subgroup C were divided into three groups (designated as cluster 1, 2, and 3) by analysis of stx phage insertion site (Shaikh and Tarr, 2003). These three clusters were thought to have evolved sequentially. This hypothesis was subsequently confirmed by single nucleotide polymorphism (SNP) analysis of whole genomes of O157 strains (Shaikh and Tarr, 2003; Leopold et al., 2009). O157 Lineages have been proposed based on a specific inserted sequence in the folD–sfmA region (Kim et al., 2001). Yang et al. (2004) then proposed that lineage specific polymorphism assay-6 ⇑ Corresponding author. Address: Division of Bacteriology, Chiba Prefectural Institute of Public Health, 666-2 Nitona, Chuo, Chiba 260-8715, Japan. Tel.: +81 43 266 6723; fax: +81 43 265 5544. E-mail address: [email protected] (E. Yokoyama). 1567-1348/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.meegid.2012.07.003
(LSPA-6) could determine lineages based on insertions/deletions in six loci. Lineages defined by LSAP-6 were used to demonstrate differences in pathogenicity among different lineage strains (Dowd and Williams, 2008). However, there are some discrepancies in the interpretation of LSPA-6 codes in terms of lineages in previous descriptions (Franz et al., 2012; Lee et al., 2012; Sharma et al., 2009; Whitworth et al., 2010; Yokoyama et al., 2011b). Manning et al. (2008) proposed the clade phylogenetic groups as defined by SNP analysis. In that report, clade analysis demonstrated strong pathogenicity for clade 8 strains, which were a greater fraction of O157 isolates from patients with hemolytic– uremic syndrome (HUS) than strains of other clades, suggesting that clade analysis might be useful in indicating emergence of specific O157 clones. However, clade analysis was partially discordant with the monophyletic evolutionary scenario for O157 reported by Shaikh and Tarr (2003) and Leopold et al. (2009). The phylogenetic network model described by Manning et al. (2008) showed networks between several clades, indicating that O157 did not evolve by clonal evolution. SNP groups 21, 22, and 27 were subsequently deleted from phylogenetic analyses because strains in these groups were found to be mixed cultures, and clade 5 was consequently also deleted from these analyses (Whittam et al., 2010). However, these modifications have not yet been sufficiently evaluated.
E. Yokoyama et al. / Infection, Genetics and Evolution 12 (2012) 1724–1728
The hierarchy of the O157 phylogenetic levels (e.g., clusters and lineages) of the strains in the clades described by Manning et al. (2008) has not been adequately elucidated. Eppinger et al. (2011) suggested that clade 7 strains diverged after lineage II conversion, indicating that all clade 7 strains may be lineage II. In contrast, Liu et al. (2009) reported that clade 7 strains belonged to lineage I/II. A common limitation of these contradictory reports is the small number of O157 strains tested. In this study, we investigated the annual clade distribution of clinical O157 strains in isolates from Japan from 2002 to 2009. Moreover, we investigated the relationship of phylogenetic groups such as clade, lineage, and cluster. Based on these data, we have discussed modification of the model of O157 evolution. 2. Materials and methods
1725
(Yokoyama et al., 2011b). The LSPA-6 codes of the O157 lineages were previously described (Yokoyama et al., 2011b). Briefly, strains with LSPA-6 code 1N111N (‘‘N’’ denotes any allele) were lineage I, those with LSPA-6 code 2N111N were lineage I/II, and strains with other LSPA-6 codes were lineage II. LSPA-6 codes were also investigated for two O55 strains by in silico analysis. O55 CB9165 and RIM12579 were analyzed by a blastn search with primers for LSPA-6 to determine the lineages of these two strains. Analysis of stx phage insertion sites was done by the method described by Shaikh and Tarr (2003) as previously reported (Yokoyama et al., 2011a). From these results, the O157 strains were divided into subgroups and clusters as described by Leopold et al. (2009). Variation of the stx phage gene in the O157 strains in this study was investigated as described by Manning et al. (2008).
2.1. Bacterial strains During 2002–2009, 985 O157 strains were isolated from patients and asymptomatic carriers in Chiba prefecture, Japan. Conventional epidemiological analyses were carried out by the local health centre of the Chiba prefectural government to investigate the epidemiological relationships among the patients and carriers. Of the 985 O157 isolates, 470 epidemiologically unlinked strains were used in this study. Production of verocytotoxin (VT) by these 470 strains was investigated using a Verotox-F RPLA kit (Denka Seiken, Niigata, Japan): 13 strains were found to produce only VT1, 183 strains were found to produce only VT2, and 274 strains were found to produce both VT1 and VT2. DNA was extracted from the O157 strains using InstaGene Matrix (Bio-Rad, Hercules, CA) and adjusted to 2 ng/ll. 2.2. Clade analysis SNPs in the eight loci described by Manning et al. (2008), Riordan et al. (2008), and Whittam et al. (2010) were selected for clade analysis and are shown in Supplementary Table S1. A phylogenetic network model was reconstructed by neighbor net analysis using Splits Tree 4 Ver. 4.13 software (Hunson and Bryant, 2006) to investigate the agreement of the phylogeny determined by this SNP analysis with that previously reported (Manning et al., 2008; Whittam et al., 2010). Analysis of the SNPs in each locus was carried out by amplification refractory mutation system PCR (ARMS–PCR) (Newton et al., 1989). The primers used for ARMS–PCR are listed in Supplementary Table 2. Amplification was done by touchdown PCR using FastStart Taq DNA Polymerase (Roche, Basel, Switzerland). Briefly, the first cycle was at 95 °C for 5 min; followed by 5 cycles of 95 °C for 30 s, annealing at Tm+5 °C for 30 s (with the annealing temperature decreased 1 °C for each cycle), and 72 °C for 30 s; 35 cycles of 95 °C for 30 s, annealing at Tm for 30 s, and 72 °C for 30 s; and 1 cycle at 72 °C for 7 min. After amplification, the presence of amplified DNA was confirmed using Sybre Green I (TaKaRa, Kyoto, Japan) as described by Hirotsu et al. (2010). Analysis by agarose gel electrophoresis was carried out if it was difficult to confirm PCR amplification with Sybre Green I. If clade 9 strains were identified by clade analysis, b-glucuronidase activity was investigated on CLIG medium (Kyokuto, Tokyo, Japan). 2.3. Analysis of LSPA-6 lineages, stx phage insertion sites, and stx gene variation The O157 strains were divided into lineages by the LSPA-6 method described by Yang et al. (2004) as previously reported
3. Results 3.1. Clade analysis Neighbor net analysis using the SNP data in this study showed monophyletic evolution of the O157 strains from clade 9 to clade 1 (Fig. 1). The branching order of the O157 clades in this phylogenetic network model was consistent with the previously reported O157 minimum evolutionary tree (Manning et al., 2008; Whittam et al., 2010). The clade analysis results are shown in Table 1. The two clade 9 strains had positive b-glucuronidase reactions. Thirty percent of the strains were in clade 7, 26% were in clade 3, and 22% were in clade 2, indicating that the clades 7, 3, 2 were dominant, comprising 78% of the O157 strains isolated in Chiba prefecture.
3.2. Hierarchy of O157 phylogenetic levels The 470 O157 strains were divided into three lineages by LSPA6 analysis: 242 were lineage I, 160 were lineage I/II, 67 were lineage II, and one had an atypical LSPA-6 code. The strain with an atypical code produced an Z5935 amplicon with a size intermediate between allele 1 and 2, and no amplification in yhcG. For comparison, when in silico LSPA-6 analyses were investigated on O55 strains that were thought to be ancestral strains of O157, an O55 strain (CB9165) with a 272111 LSPA-6 code was lineage II, and another O55 strain (RIM12579) with a 252111 LSPA-6 code was also lineage II. Strains in subgroup C, which contains clade 1–8, can be divided by stx phage insertion site analysis, but clade 9 strains were excluded from this analysis because clade 9 is not in subgroup C (Shaikh and Tarr, 2003; Leopold et al., 2009). The stx analysis showed that all lineage I strains had an ‘‘Occupied’’ yehV and ‘‘Occupied‘‘ wrbA (data not shown), indicating that all lineage I strains were in cluster 3 of subgroup C as described by Leopold et al. (2009). All lineage I/II and II strains had an ‘‘Occupied’’ yehV and ‘‘Intact’’ wrbA, indicating that all lineage I/II and II strains were in cluster 1 of subgroup C. Analysis of stx phage gene variation showed that the stx2c gene was not present in lineage I, cluster 3 strains. Of the 67 lineage II strains, 61 had the stx2c gene (Table 1). All clade 1, clade 2, and clade 3 strains and 8 of 11 clade 4/5 strains were lineage I, cluster 3 strains. All clade 6 strains, 80of 139 clade 7 strains, and 59 of 66 clade 8, and 2 of 11 clade 4/5 strains were lineage I/II, cluster 1 strains. The remaining clade 7, 8, and 4/5 strains were lineage II, cluster 1 strains. One of the two clade 9 strains was in lineage I/II, and the other clade 9 strain
1726
E. Yokoyama et al. / Infection, Genetics and Evolution 12 (2012) 1724–1728 0.1
Clade 2 Clade 1
Clade 3
Clade 4/5
Clade 7 Clade 9
Clade 6
Clade 8
Fig. 1. Phylogenetic network model of enterohemorrhagic Escherichia coli serotype O157:H7/H-strains in this study reconstructed from SNP data. SNP group 19 in clade 4 and SNP group 13 in clade 3 were not included in this analysis to more clearly show the relationships between the clades.
Table 1 Distribution of O157 strains and stx variants in each clade and lineage. Clade
1 2 3 4/5
6 7 8 9
Lineage
I I I I I/II II I/II I/II II I/II II I/II Atypical
Total
Number of strains
Number of strains with stx2 and stx2c variants stx2
stx2c
stx2 and stx2c
UTa
NDb
7 103 124 8 2 1 18 80 59 59 7 1 1
7 103 121 8 0 1 2 39 0 33 1 0 0
0 0 0 0 1 0 1 29 56 13 0 0 0
0 0 0 0 1 0 14 10 3 13 2 0 0
0 0 0 0 0 0 1 1 0 0 1 1 1
0 0 3 0 0 0 0 1 0 0 3 0 0
470
315
100
43
5
7
a
Unable to type. These strains did not produce amplicons by PCR for an stx2 variant, but did showed positive reactions for VT2 by RPLA. b Not detected. These strains, which had negative reactions for VT2 by RPLA, did not produce amplicons by PCR for an stx2 variant.
had an atypical LSPA-6 code and could not be assigned to a lineage (Table 1). 4. Discussion The hierarchy of O157 phylogenetic relationships in this study suggested a modification of model of O157 evolution as shown in Fig. 2. Our study demonstrated that lineage I and I/II could be included in the O157 evolutionary model reported previously (Leopold et al., 2009). The cluster 3 strains in this study were concordant with lineage I strains, and the cluster 1 strains were concordant with lineage I/II and II strains. A similar result was reported by Whitworth et al. (2010), although lineage II strains were not all in any one cluster. In this study, inclusion of lineage II in the model of O157 evolution was resolved by LSPA-6 analyses, including in silico analysis. When the lineages of the two O55 strains were analyzed by in silico LSPA-6 analysis, both strains were lineage II and had allele 2 in the yhcG locus. We previously showed that changes in the LSPA-6 code in 5 out of 6 loci were due to changes in the number of tandem repeat, changes in yhcG were due to deletion of a 79 bp sequence, not to a change in the number of tandem repeats
(Yokoyama et al., 2011b). The O55 strains in this study had the same 79 bp sequence as that in previously investigated lineage II O157 strains (data not shown). These results indicated that lineage I/II diverged from lineage II with deletion of the 79 bp sequence. This is discordant with the recent report by Eppinger et al. (2011). This discordance may be due to the difference in number of strains tested in these two studies. In the stepwise model of O157 evolution, lineage I/II would have diverged from lineage II before or within A5 group. Two strains were sorbitol-nonfermenting and positive for b-glucuronidase reactions in this study: these were clade 9 strains, one in lineage I/II and the other with an atypical LSPA-6 code. In addition, Liu et al. (2009) reported a clade 9 strain that belonged to lineage II. Since a recent report demonstrated that the revised model of O157 evolution that were sorbitol-nonfermenting and positive for b-glucuronidase were in A5 (Kyle et al., 2012), clade 9 strains of this study were classified in A5. Although this is discordant with the report by Kim et al. (2001), this discordance may be explained by the recent modification of the model of O157 evolution discussed above. Since A4 and A5 groups diverged from A3 (Kyle et al., 2012), O157 strains in A3 should be studied to indicate when lineage I/II diverged from lineage II. The organization of clades in the model of O157 evolution is complicated. Inclusion of clades in each lineage indicated that clade 4/5 included strains in several different phylogenetic groups, a lineage I/II or II ancestral group and a lineage I descendant group. The SNP loci used in this study could not differentiate clade 4 and 5 strains, so they were designated clade 4/5 strains in this study. Since clade 5 had to be excluded from SNP analysis due to mixed cultures (Whittam et al., 2010), clade 4/5 strains in this study must include clade 4 strains only. Clade 4 strains have the stx2c gene (Manning et al., 2008), so the ancestral clade 4/5 strains in this study, but not the descendant clade 4/5 strains, may be clade 4 strains as described by Manning et al. (2008). Whittam et al. (2010) suggested that one of the SNP types included in clade 4 should be separated from other clade 4 strains. We have not confirmed whether ancestral clade 4/5 was the same as that SNP type because, at present, no information has been published for differentiating that SNP type from other clades. Clade analysis in this study suggested paraphyletic evolution of lineage II and lineage I/II. Several clades (i.e., ancestral 4/5 and clade 7 and 8) were found in both lineage II and I/II in cluster 1. Previous descriptions showed that Lineage II is significantly different from Lineage I/II, and lineage I/II is more similar to lineage I than lineage II (Bono et al., 2012; Eppinger et al., 2011; Zhang et al., 2006), which may support the paraphyletic evolution
1727
E. Yokoyama et al. / Infection, Genetics and Evolution 12 (2012) 1724–1728 A4 Subgroup B
Clade 8
O55:H7
A3
Subgroup C Cluster 3
Subgroup C Cluster 1
A5 Clade 9
Subgroup C Cluster 2
Clade 6
Clade 7
Descendant clade 4/5
Ancestral clade 4/5
Clade 3
Clade 8 Lineage I
Ancestral clade 4/5
Clade 2
Lineage I/II Clade 1
Lineage II
Clade 7
Lineage has not been determined
Fig. 2. Model of O157 evolution described by Leopold et al. (2009) and Kyle et al. (2012) as modified in this study. The dotted lines denote the evolutionary route to be clarified.
hypothesis. A comparative study of whole genomes using more strains in both lineages is needed to elucidate paraphyletic evolution of lineages. In this study, the number of lineage II strains was smaller than that of lineage I strains. This may have been influenced by the source of O157 isolates, because more lineage II strains have been isolated from bovine samples than from human samples (Yang et al., 2004). In conclusion, the hierarchy of O157 phylogenetic levels, based on experiments with several phylogenetic markers, were (in descending order) lineage, subgroup, cluster and clade. Lineage I/II may have diverged from lineage II before or within the A5 group and those was paraphyletic. Clade 4/5 strains were found in several phylogenetic groups, indicating that the clade definition needs to be clarified.
Acknowledgement Part of this study was supported by a Grant from the Daido Life Welfare Foundation for Regional Health and Welfare Research in 2011.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.meegid.2012. 07.003. References Bono, J.L., Smith, P.L., Keen, J.E., Harhay, G.P., McDaneld, T.G., Mandrell, R.E., Jung, W.K., Besser, T.E., Gerner-Smidt, P., Nielaszewska, M., Karch, H., Clawson, M.L., 2012. Phylogeny of Shiga toxin-producing Escherichia coli O157 isolated from cattle and clinically ill humans. Mol. Biol. Evol 29, 2047–2062. Dowd, S.E., Williams, J.B., 2008. Comparison of Shiga-like toxin II expression between two genetically diverse lineages of Escherichia coli O157:H7. J. Food Prot. 71, 1673–1678. Eppinger, M., Mammel, M.K., Leclerc, J.E., Ravel, J., Cebula, T.A., 2011. Genomic anatomy of Escherichia coli O157:H7 outbreaks. Proc. Natl. Acad. Sci. USA 108, 20142–20147. Franz, E., van Hoek, A.H.A.M., van der Wal, F.J., de Boer, A., Zwartkruis-Nahuis, A., van der Zwaluw, K., Aarts, H.J.M., Heuvelink, A.E., 2012. Genetic features differentiating bovine, food, and human isolates of Shiga toxin-producing Escherichia coli O157 in the Netherlands. J. Clin. Microbiol. 50, 772–780.
Hirotsu, N., Murakami, N., Kashiwagi, T., Ujiie, K., Ishimaru, K., 2010. Protocol: a simple gel-free method for SNP genotyping using allele-specific primers in rice and other plant species. Plant Methods 6, 12. Hunson, D.H., Bryant, D., 2006. Application of phylogenetic networks in evolutionary studies. Mol. Biol. Evol. 23, 254–267. Kim, J., Nietfeldt, J., Ju, J., Wise, J., Fegan, N., Desmarchelier, P., Benson, A.K., 2001. Ancestral divergece, genome diversification, and phylogeographic variation in subpopulation of sorbitole-negative, b-glucuronidase-negative enterohemorrhagic Escherichia coli O157. J. Bacteriol. 183, 6885–6897. Kyle, J.L., Cummings, C.A., Parker, C.T., Quiñones, B., Vatta, P., Newton, Elizabeth, Huynh, S., Swimley, M., Degoricija, L., Barker, M., Fontanoz, S., Nguyen, K., Patel, R., Fang, R., Tebbs, R., Petrauskene, O., Furtado, M., Mandrell, R.E., 2012. Escherichia coli serotype O55:H7 diversity supports pararell acquisition of bacteriophage at Shiga toxin phage insertion sites during evolution of O157:H7 lineage. J. Bacteriol., 1885–1896. Lee, K.-I., French, L.N., Jones, G., Hara-Kudo, Y., Iyoda, S., Koyabashi, H., SugitaKonishi, Y., Tsubone, H., Kumagai, S., 2012. Variation in stress resistance patterns among stx genotypes and genetic lineages of Shiga toxin-producing Escherichia coli O157. Appl. Environ. Microbiol. 78, 3361–3368. Leopold, S.R., Magrini, V., Holt, N.J., Shaikh, N., Mardis, E.R., Cagno, J., Ogura, Y., Iguchi, A., Hayashi, T., Mellmann, A., Karch, H., Besser, T.E., Sawyer, S.A., Whittam, T.S., Tarr, P.I., 2009. A precise reconstruction of the emergence and constrained radiations of Escherichia coli O157 portrayed by backbone concatenomic analysis. Proc. Natl. Acad. Sci. USA 106, 8713–8718. Liu, K., Knabel, S.J., Dudley, E.G., 2009. Rhs genes are potential markers for multilocus sequence typing of Escherichia coli O157:H7 strains. Appl. Environ. Microbiol. 75, 5853–5862. Manning, S.D., Motiwala, A.S., Springman, A.C., Qi, W., Lacher, D.W., Ouellette, L.M., Mladonicky, J.M., Somsel, P., Rudrik, J.T., Dietrich, S.E., Zhang, W., Swaminathan, B., Alland, D., Whittam, T.S., 2008. Variation in virulence among clades of Escherichia coli O157:H7 associated with disease outbreaks. Proc. Nat. Acad. Sci. USA 105, 4868–4873. Newton, C.R., Graham, A., Heptinstall, L.E., Powell, S.J., Summers, C., Kalsheker, N., Smith, J.C., Markham, A.F., 1989. Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS). Nucleic Acids Res. 17, 2503– 2516. Riordan, J.T., Viswanath, S.B., Manning, S.D., Whittam, T.S., 2008. Genetic differentiation of Escherichia coli O157:H7 clades associated with human disease by real-time PCR. J. Clin. Microbiol. 46, 2070–2073. Sharma, R., Stanford, K., Louie, M., Munns, K., John, S.J., Zhang, Y., Gannon, V., Chui, L., Read, R., Topp, E., McAllister, T., 2009. Escherichia coli O157:H7 lineages in healthy beef and dairy cattle and clinical human cases in Alberta. Canada J. Food Prot. 72, 601–607. Shaikh, N., Tarr, P.I., 2003. Escherichia coli O157:H7 Shiga toxin-producing bacteriophages: integrations, excisions, truncations, and evolutionary implications. J. Bacteriol. 185, 3596–3605. Whittam, T., Manning, S., Alland, D., Motiwala, A. 2010. Methods of detecting and genotyping Escherichia coli O157:H7. United States Patent Application Publication. US 2010/0279294 A1. Whittam, T.S., Wilson, R.A., 1988. Genetic relationships among pathogenic Escherichia coli of serogroup O157. Infect. Immun. 56, 2467–2473. Whitworth, J., Zhang, Y., Bono, J., Pleydell, E., French, N., Besser, T., 2010. Diverse genetic markers concoedantly identified bovine origin Escherichia coli O157 genotypes underrepresented in human disease. Appl. Environ. Microbiol. 76, 361–365.
1728
E. Yokoyama et al. / Infection, Genetics and Evolution 12 (2012) 1724–1728
Yang, Z., Kovar, J., Kim, J., Nietfeldt, J., Smith, D.R., Moxley, R.A., Olson, M.E., Fey, P.D., Benson, A.K., 2004. Identification of common subpopulations of non-sorbitolfermenting, beta-glucuronidase-negative Escherichia coli O157:H7 from bovine production environments and human clinical samples. Appl. Environ. Microbiol. 70, 6846–6854. Yokoyama, E., Etoh, Y., Ichihara, S., Horikawa, K., Konishi, N., Kai, A., Matsumoto, Y., Kurosaki, M., Kasahara, H., Kurazono, T., Yoda, K., 2011a. Emergence of enterohemorrhagic Escherichia coli serovar O157 strains in
clade 8 with highly similar pulsed-field gel electrophoresis patterns. J. Food Prot. 74, 1324–1327. Yokoyama, E., Hashimoto, R., Etoh, Y., Ichihara, S., Horikawa, K., Uchimura, M., 2011b. Biased distribution of IS629 among strains in different lineages of enterohemorrhagic Escherichia coli serovar O157. Infect. Genet. Evol. 11, 78–82. Zhang, Y., Laing, C., Steele, M., Ziebell, K., Johnson, R., Benson, A.K., Taboada, E., Gannon, V.P.J., 2006. Genome evolution in major Escherichia coli O157:H7 lineages. BMC Genomics 8, 121.
Contents lists available at SciVerse ScienceDirect
Infection, Genetics and Evolution journal homepage: www.elsevier.com/locate/meegid
Clade analysis of enterohemorrhagic Escherichia coli serotype O157:H7/H- strains and hierarchy of their phylogenetic relationships Eiji Yokoyama ⇑, Shinichiro Hirai, Ruiko Hashimoto, Masako Uchimura Division of Bacteriology, Chiba Prefectural Institute of Public Health, Chiba, Japan
a r t i c l e
i n f o
Article history: Received 27 January 2012 Received in revised form 12 June 2012 Accepted 7 July 2012 Available online 27 July 2012 Keywords: Enterohemorrhagic Escherichia coli serotype O157:H7/HClade Lineage-specific polymorphism assay 6 (LSAP6) Cluster Subgroup Evolutionary model
a b s t r a c t Enterohemorrhagic Escherichia coli serotype O157:H7/H-(O157) strains isolated in Chiba prefecture, Japan, during 2002–2009 were studied by lineage, subgroup, cluster, and clade analysis. Lineage analysis of 470 O157 strains with no known epidemiological relationships using lineage specific polymorphism assay-6 showed that there were 242 lineage I strains, 160 lineage I/II strains, 67 lineage II strains, and 1 atypical strain. Clade analysis of these strains by single nucleotide polymorphism in eight loci showed that lineage I contained all the clade 1, clade 2, and clade 3 strains, and some of the clade 4/5 strains. In contrast, clade 7, clade 8, and the remaining clade 4/5 strains were divided between lineage I/II and II, and clade 6 was in lineage I/II, suggesting paraphyletic evolution of these lineages. Cluster and subgroup analysis of the stx phage insertion site showed that all lineage I strains were cluster 3 and all lineage I/II and II strains, with the exception of clade 9, were in cluster 1. Clade analysis also indicated that there were three phylogenetic groups of clade 4/5 strains: ancestral groups containing lineage I/IIand II strains and a descendant group containing lineages I. Analysis of stx2c gene distribution showed that stx2c was in ancestral clade 4/5 strains but not in descendant 4/5 strains, suggesting that the ancestral group may be clade 4 as reported by Manning et al. The results with the markers used in this study suggested that the hierarchy of O157 phylogenetic relationships was lineage as the upper level, followed by subgroup and then cluster, and clade as the lowest level. The need for refinement of clade definition and modification of the model of the O157 evolution have been discussed. Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction Aspects of the evolution of enterohemorrhagic Escherichia coli serotype O157:H7/H- (O157) have been elucidated during the past two decades. E. coli serotype O55:H7 (O55) was thought to be the ancestor of O157, and O157 were divided into three groups, (i.e., subgroup A, B, and C) based on multilocus enzyme electrophoresis analysis (Whittam and Wilson, 1988). Subgroup A contained O55 strains, subgroup B contained sorbitol-fermenting O157 strains, and subgroup C contained sorbitol-nonfermenting O157 strains. The O157 strains in subgroup C were divided into three groups (designated as cluster 1, 2, and 3) by analysis of stx phage insertion site (Shaikh and Tarr, 2003). These three clusters were thought to have evolved sequentially. This hypothesis was subsequently confirmed by single nucleotide polymorphism (SNP) analysis of whole genomes of O157 strains (Shaikh and Tarr, 2003; Leopold et al., 2009). O157 Lineages have been proposed based on a specific inserted sequence in the folD–sfmA region (Kim et al., 2001). Yang et al. (2004) then proposed that lineage specific polymorphism assay-6 ⇑ Corresponding author. Address: Division of Bacteriology, Chiba Prefectural Institute of Public Health, 666-2 Nitona, Chuo, Chiba 260-8715, Japan. Tel.: +81 43 266 6723; fax: +81 43 265 5544. E-mail address: [email protected] (E. Yokoyama). 1567-1348/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.meegid.2012.07.003
(LSPA-6) could determine lineages based on insertions/deletions in six loci. Lineages defined by LSAP-6 were used to demonstrate differences in pathogenicity among different lineage strains (Dowd and Williams, 2008). However, there are some discrepancies in the interpretation of LSPA-6 codes in terms of lineages in previous descriptions (Franz et al., 2012; Lee et al., 2012; Sharma et al., 2009; Whitworth et al., 2010; Yokoyama et al., 2011b). Manning et al. (2008) proposed the clade phylogenetic groups as defined by SNP analysis. In that report, clade analysis demonstrated strong pathogenicity for clade 8 strains, which were a greater fraction of O157 isolates from patients with hemolytic– uremic syndrome (HUS) than strains of other clades, suggesting that clade analysis might be useful in indicating emergence of specific O157 clones. However, clade analysis was partially discordant with the monophyletic evolutionary scenario for O157 reported by Shaikh and Tarr (2003) and Leopold et al. (2009). The phylogenetic network model described by Manning et al. (2008) showed networks between several clades, indicating that O157 did not evolve by clonal evolution. SNP groups 21, 22, and 27 were subsequently deleted from phylogenetic analyses because strains in these groups were found to be mixed cultures, and clade 5 was consequently also deleted from these analyses (Whittam et al., 2010). However, these modifications have not yet been sufficiently evaluated.
E. Yokoyama et al. / Infection, Genetics and Evolution 12 (2012) 1724–1728
The hierarchy of the O157 phylogenetic levels (e.g., clusters and lineages) of the strains in the clades described by Manning et al. (2008) has not been adequately elucidated. Eppinger et al. (2011) suggested that clade 7 strains diverged after lineage II conversion, indicating that all clade 7 strains may be lineage II. In contrast, Liu et al. (2009) reported that clade 7 strains belonged to lineage I/II. A common limitation of these contradictory reports is the small number of O157 strains tested. In this study, we investigated the annual clade distribution of clinical O157 strains in isolates from Japan from 2002 to 2009. Moreover, we investigated the relationship of phylogenetic groups such as clade, lineage, and cluster. Based on these data, we have discussed modification of the model of O157 evolution. 2. Materials and methods
1725
(Yokoyama et al., 2011b). The LSPA-6 codes of the O157 lineages were previously described (Yokoyama et al., 2011b). Briefly, strains with LSPA-6 code 1N111N (‘‘N’’ denotes any allele) were lineage I, those with LSPA-6 code 2N111N were lineage I/II, and strains with other LSPA-6 codes were lineage II. LSPA-6 codes were also investigated for two O55 strains by in silico analysis. O55 CB9165 and RIM12579 were analyzed by a blastn search with primers for LSPA-6 to determine the lineages of these two strains. Analysis of stx phage insertion sites was done by the method described by Shaikh and Tarr (2003) as previously reported (Yokoyama et al., 2011a). From these results, the O157 strains were divided into subgroups and clusters as described by Leopold et al. (2009). Variation of the stx phage gene in the O157 strains in this study was investigated as described by Manning et al. (2008).
2.1. Bacterial strains During 2002–2009, 985 O157 strains were isolated from patients and asymptomatic carriers in Chiba prefecture, Japan. Conventional epidemiological analyses were carried out by the local health centre of the Chiba prefectural government to investigate the epidemiological relationships among the patients and carriers. Of the 985 O157 isolates, 470 epidemiologically unlinked strains were used in this study. Production of verocytotoxin (VT) by these 470 strains was investigated using a Verotox-F RPLA kit (Denka Seiken, Niigata, Japan): 13 strains were found to produce only VT1, 183 strains were found to produce only VT2, and 274 strains were found to produce both VT1 and VT2. DNA was extracted from the O157 strains using InstaGene Matrix (Bio-Rad, Hercules, CA) and adjusted to 2 ng/ll. 2.2. Clade analysis SNPs in the eight loci described by Manning et al. (2008), Riordan et al. (2008), and Whittam et al. (2010) were selected for clade analysis and are shown in Supplementary Table S1. A phylogenetic network model was reconstructed by neighbor net analysis using Splits Tree 4 Ver. 4.13 software (Hunson and Bryant, 2006) to investigate the agreement of the phylogeny determined by this SNP analysis with that previously reported (Manning et al., 2008; Whittam et al., 2010). Analysis of the SNPs in each locus was carried out by amplification refractory mutation system PCR (ARMS–PCR) (Newton et al., 1989). The primers used for ARMS–PCR are listed in Supplementary Table 2. Amplification was done by touchdown PCR using FastStart Taq DNA Polymerase (Roche, Basel, Switzerland). Briefly, the first cycle was at 95 °C for 5 min; followed by 5 cycles of 95 °C for 30 s, annealing at Tm+5 °C for 30 s (with the annealing temperature decreased 1 °C for each cycle), and 72 °C for 30 s; 35 cycles of 95 °C for 30 s, annealing at Tm for 30 s, and 72 °C for 30 s; and 1 cycle at 72 °C for 7 min. After amplification, the presence of amplified DNA was confirmed using Sybre Green I (TaKaRa, Kyoto, Japan) as described by Hirotsu et al. (2010). Analysis by agarose gel electrophoresis was carried out if it was difficult to confirm PCR amplification with Sybre Green I. If clade 9 strains were identified by clade analysis, b-glucuronidase activity was investigated on CLIG medium (Kyokuto, Tokyo, Japan). 2.3. Analysis of LSPA-6 lineages, stx phage insertion sites, and stx gene variation The O157 strains were divided into lineages by the LSPA-6 method described by Yang et al. (2004) as previously reported
3. Results 3.1. Clade analysis Neighbor net analysis using the SNP data in this study showed monophyletic evolution of the O157 strains from clade 9 to clade 1 (Fig. 1). The branching order of the O157 clades in this phylogenetic network model was consistent with the previously reported O157 minimum evolutionary tree (Manning et al., 2008; Whittam et al., 2010). The clade analysis results are shown in Table 1. The two clade 9 strains had positive b-glucuronidase reactions. Thirty percent of the strains were in clade 7, 26% were in clade 3, and 22% were in clade 2, indicating that the clades 7, 3, 2 were dominant, comprising 78% of the O157 strains isolated in Chiba prefecture.
3.2. Hierarchy of O157 phylogenetic levels The 470 O157 strains were divided into three lineages by LSPA6 analysis: 242 were lineage I, 160 were lineage I/II, 67 were lineage II, and one had an atypical LSPA-6 code. The strain with an atypical code produced an Z5935 amplicon with a size intermediate between allele 1 and 2, and no amplification in yhcG. For comparison, when in silico LSPA-6 analyses were investigated on O55 strains that were thought to be ancestral strains of O157, an O55 strain (CB9165) with a 272111 LSPA-6 code was lineage II, and another O55 strain (RIM12579) with a 252111 LSPA-6 code was also lineage II. Strains in subgroup C, which contains clade 1–8, can be divided by stx phage insertion site analysis, but clade 9 strains were excluded from this analysis because clade 9 is not in subgroup C (Shaikh and Tarr, 2003; Leopold et al., 2009). The stx analysis showed that all lineage I strains had an ‘‘Occupied’’ yehV and ‘‘Occupied‘‘ wrbA (data not shown), indicating that all lineage I strains were in cluster 3 of subgroup C as described by Leopold et al. (2009). All lineage I/II and II strains had an ‘‘Occupied’’ yehV and ‘‘Intact’’ wrbA, indicating that all lineage I/II and II strains were in cluster 1 of subgroup C. Analysis of stx phage gene variation showed that the stx2c gene was not present in lineage I, cluster 3 strains. Of the 67 lineage II strains, 61 had the stx2c gene (Table 1). All clade 1, clade 2, and clade 3 strains and 8 of 11 clade 4/5 strains were lineage I, cluster 3 strains. All clade 6 strains, 80of 139 clade 7 strains, and 59 of 66 clade 8, and 2 of 11 clade 4/5 strains were lineage I/II, cluster 1 strains. The remaining clade 7, 8, and 4/5 strains were lineage II, cluster 1 strains. One of the two clade 9 strains was in lineage I/II, and the other clade 9 strain
1726
E. Yokoyama et al. / Infection, Genetics and Evolution 12 (2012) 1724–1728 0.1
Clade 2 Clade 1
Clade 3
Clade 4/5
Clade 7 Clade 9
Clade 6
Clade 8
Fig. 1. Phylogenetic network model of enterohemorrhagic Escherichia coli serotype O157:H7/H-strains in this study reconstructed from SNP data. SNP group 19 in clade 4 and SNP group 13 in clade 3 were not included in this analysis to more clearly show the relationships between the clades.
Table 1 Distribution of O157 strains and stx variants in each clade and lineage. Clade
1 2 3 4/5
6 7 8 9
Lineage
I I I I I/II II I/II I/II II I/II II I/II Atypical
Total
Number of strains
Number of strains with stx2 and stx2c variants stx2
stx2c
stx2 and stx2c
UTa
NDb
7 103 124 8 2 1 18 80 59 59 7 1 1
7 103 121 8 0 1 2 39 0 33 1 0 0
0 0 0 0 1 0 1 29 56 13 0 0 0
0 0 0 0 1 0 14 10 3 13 2 0 0
0 0 0 0 0 0 1 1 0 0 1 1 1
0 0 3 0 0 0 0 1 0 0 3 0 0
470
315
100
43
5
7
a
Unable to type. These strains did not produce amplicons by PCR for an stx2 variant, but did showed positive reactions for VT2 by RPLA. b Not detected. These strains, which had negative reactions for VT2 by RPLA, did not produce amplicons by PCR for an stx2 variant.
had an atypical LSPA-6 code and could not be assigned to a lineage (Table 1). 4. Discussion The hierarchy of O157 phylogenetic relationships in this study suggested a modification of model of O157 evolution as shown in Fig. 2. Our study demonstrated that lineage I and I/II could be included in the O157 evolutionary model reported previously (Leopold et al., 2009). The cluster 3 strains in this study were concordant with lineage I strains, and the cluster 1 strains were concordant with lineage I/II and II strains. A similar result was reported by Whitworth et al. (2010), although lineage II strains were not all in any one cluster. In this study, inclusion of lineage II in the model of O157 evolution was resolved by LSPA-6 analyses, including in silico analysis. When the lineages of the two O55 strains were analyzed by in silico LSPA-6 analysis, both strains were lineage II and had allele 2 in the yhcG locus. We previously showed that changes in the LSPA-6 code in 5 out of 6 loci were due to changes in the number of tandem repeat, changes in yhcG were due to deletion of a 79 bp sequence, not to a change in the number of tandem repeats
(Yokoyama et al., 2011b). The O55 strains in this study had the same 79 bp sequence as that in previously investigated lineage II O157 strains (data not shown). These results indicated that lineage I/II diverged from lineage II with deletion of the 79 bp sequence. This is discordant with the recent report by Eppinger et al. (2011). This discordance may be due to the difference in number of strains tested in these two studies. In the stepwise model of O157 evolution, lineage I/II would have diverged from lineage II before or within A5 group. Two strains were sorbitol-nonfermenting and positive for b-glucuronidase reactions in this study: these were clade 9 strains, one in lineage I/II and the other with an atypical LSPA-6 code. In addition, Liu et al. (2009) reported a clade 9 strain that belonged to lineage II. Since a recent report demonstrated that the revised model of O157 evolution that were sorbitol-nonfermenting and positive for b-glucuronidase were in A5 (Kyle et al., 2012), clade 9 strains of this study were classified in A5. Although this is discordant with the report by Kim et al. (2001), this discordance may be explained by the recent modification of the model of O157 evolution discussed above. Since A4 and A5 groups diverged from A3 (Kyle et al., 2012), O157 strains in A3 should be studied to indicate when lineage I/II diverged from lineage II. The organization of clades in the model of O157 evolution is complicated. Inclusion of clades in each lineage indicated that clade 4/5 included strains in several different phylogenetic groups, a lineage I/II or II ancestral group and a lineage I descendant group. The SNP loci used in this study could not differentiate clade 4 and 5 strains, so they were designated clade 4/5 strains in this study. Since clade 5 had to be excluded from SNP analysis due to mixed cultures (Whittam et al., 2010), clade 4/5 strains in this study must include clade 4 strains only. Clade 4 strains have the stx2c gene (Manning et al., 2008), so the ancestral clade 4/5 strains in this study, but not the descendant clade 4/5 strains, may be clade 4 strains as described by Manning et al. (2008). Whittam et al. (2010) suggested that one of the SNP types included in clade 4 should be separated from other clade 4 strains. We have not confirmed whether ancestral clade 4/5 was the same as that SNP type because, at present, no information has been published for differentiating that SNP type from other clades. Clade analysis in this study suggested paraphyletic evolution of lineage II and lineage I/II. Several clades (i.e., ancestral 4/5 and clade 7 and 8) were found in both lineage II and I/II in cluster 1. Previous descriptions showed that Lineage II is significantly different from Lineage I/II, and lineage I/II is more similar to lineage I than lineage II (Bono et al., 2012; Eppinger et al., 2011; Zhang et al., 2006), which may support the paraphyletic evolution
1727
E. Yokoyama et al. / Infection, Genetics and Evolution 12 (2012) 1724–1728 A4 Subgroup B
Clade 8
O55:H7
A3
Subgroup C Cluster 3
Subgroup C Cluster 1
A5 Clade 9
Subgroup C Cluster 2
Clade 6
Clade 7
Descendant clade 4/5
Ancestral clade 4/5
Clade 3
Clade 8 Lineage I
Ancestral clade 4/5
Clade 2
Lineage I/II Clade 1
Lineage II
Clade 7
Lineage has not been determined
Fig. 2. Model of O157 evolution described by Leopold et al. (2009) and Kyle et al. (2012) as modified in this study. The dotted lines denote the evolutionary route to be clarified.
hypothesis. A comparative study of whole genomes using more strains in both lineages is needed to elucidate paraphyletic evolution of lineages. In this study, the number of lineage II strains was smaller than that of lineage I strains. This may have been influenced by the source of O157 isolates, because more lineage II strains have been isolated from bovine samples than from human samples (Yang et al., 2004). In conclusion, the hierarchy of O157 phylogenetic levels, based on experiments with several phylogenetic markers, were (in descending order) lineage, subgroup, cluster and clade. Lineage I/II may have diverged from lineage II before or within the A5 group and those was paraphyletic. Clade 4/5 strains were found in several phylogenetic groups, indicating that the clade definition needs to be clarified.
Acknowledgement Part of this study was supported by a Grant from the Daido Life Welfare Foundation for Regional Health and Welfare Research in 2011.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.meegid.2012. 07.003. References Bono, J.L., Smith, P.L., Keen, J.E., Harhay, G.P., McDaneld, T.G., Mandrell, R.E., Jung, W.K., Besser, T.E., Gerner-Smidt, P., Nielaszewska, M., Karch, H., Clawson, M.L., 2012. Phylogeny of Shiga toxin-producing Escherichia coli O157 isolated from cattle and clinically ill humans. Mol. Biol. Evol 29, 2047–2062. Dowd, S.E., Williams, J.B., 2008. Comparison of Shiga-like toxin II expression between two genetically diverse lineages of Escherichia coli O157:H7. J. Food Prot. 71, 1673–1678. Eppinger, M., Mammel, M.K., Leclerc, J.E., Ravel, J., Cebula, T.A., 2011. Genomic anatomy of Escherichia coli O157:H7 outbreaks. Proc. Natl. Acad. Sci. USA 108, 20142–20147. Franz, E., van Hoek, A.H.A.M., van der Wal, F.J., de Boer, A., Zwartkruis-Nahuis, A., van der Zwaluw, K., Aarts, H.J.M., Heuvelink, A.E., 2012. Genetic features differentiating bovine, food, and human isolates of Shiga toxin-producing Escherichia coli O157 in the Netherlands. J. Clin. Microbiol. 50, 772–780.
Hirotsu, N., Murakami, N., Kashiwagi, T., Ujiie, K., Ishimaru, K., 2010. Protocol: a simple gel-free method for SNP genotyping using allele-specific primers in rice and other plant species. Plant Methods 6, 12. Hunson, D.H., Bryant, D., 2006. Application of phylogenetic networks in evolutionary studies. Mol. Biol. Evol. 23, 254–267. Kim, J., Nietfeldt, J., Ju, J., Wise, J., Fegan, N., Desmarchelier, P., Benson, A.K., 2001. Ancestral divergece, genome diversification, and phylogeographic variation in subpopulation of sorbitole-negative, b-glucuronidase-negative enterohemorrhagic Escherichia coli O157. J. Bacteriol. 183, 6885–6897. Kyle, J.L., Cummings, C.A., Parker, C.T., Quiñones, B., Vatta, P., Newton, Elizabeth, Huynh, S., Swimley, M., Degoricija, L., Barker, M., Fontanoz, S., Nguyen, K., Patel, R., Fang, R., Tebbs, R., Petrauskene, O., Furtado, M., Mandrell, R.E., 2012. Escherichia coli serotype O55:H7 diversity supports pararell acquisition of bacteriophage at Shiga toxin phage insertion sites during evolution of O157:H7 lineage. J. Bacteriol., 1885–1896. Lee, K.-I., French, L.N., Jones, G., Hara-Kudo, Y., Iyoda, S., Koyabashi, H., SugitaKonishi, Y., Tsubone, H., Kumagai, S., 2012. Variation in stress resistance patterns among stx genotypes and genetic lineages of Shiga toxin-producing Escherichia coli O157. Appl. Environ. Microbiol. 78, 3361–3368. Leopold, S.R., Magrini, V., Holt, N.J., Shaikh, N., Mardis, E.R., Cagno, J., Ogura, Y., Iguchi, A., Hayashi, T., Mellmann, A., Karch, H., Besser, T.E., Sawyer, S.A., Whittam, T.S., Tarr, P.I., 2009. A precise reconstruction of the emergence and constrained radiations of Escherichia coli O157 portrayed by backbone concatenomic analysis. Proc. Natl. Acad. Sci. USA 106, 8713–8718. Liu, K., Knabel, S.J., Dudley, E.G., 2009. Rhs genes are potential markers for multilocus sequence typing of Escherichia coli O157:H7 strains. Appl. Environ. Microbiol. 75, 5853–5862. Manning, S.D., Motiwala, A.S., Springman, A.C., Qi, W., Lacher, D.W., Ouellette, L.M., Mladonicky, J.M., Somsel, P., Rudrik, J.T., Dietrich, S.E., Zhang, W., Swaminathan, B., Alland, D., Whittam, T.S., 2008. Variation in virulence among clades of Escherichia coli O157:H7 associated with disease outbreaks. Proc. Nat. Acad. Sci. USA 105, 4868–4873. Newton, C.R., Graham, A., Heptinstall, L.E., Powell, S.J., Summers, C., Kalsheker, N., Smith, J.C., Markham, A.F., 1989. Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS). Nucleic Acids Res. 17, 2503– 2516. Riordan, J.T., Viswanath, S.B., Manning, S.D., Whittam, T.S., 2008. Genetic differentiation of Escherichia coli O157:H7 clades associated with human disease by real-time PCR. J. Clin. Microbiol. 46, 2070–2073. Sharma, R., Stanford, K., Louie, M., Munns, K., John, S.J., Zhang, Y., Gannon, V., Chui, L., Read, R., Topp, E., McAllister, T., 2009. Escherichia coli O157:H7 lineages in healthy beef and dairy cattle and clinical human cases in Alberta. Canada J. Food Prot. 72, 601–607. Shaikh, N., Tarr, P.I., 2003. Escherichia coli O157:H7 Shiga toxin-producing bacteriophages: integrations, excisions, truncations, and evolutionary implications. J. Bacteriol. 185, 3596–3605. Whittam, T., Manning, S., Alland, D., Motiwala, A. 2010. Methods of detecting and genotyping Escherichia coli O157:H7. United States Patent Application Publication. US 2010/0279294 A1. Whittam, T.S., Wilson, R.A., 1988. Genetic relationships among pathogenic Escherichia coli of serogroup O157. Infect. Immun. 56, 2467–2473. Whitworth, J., Zhang, Y., Bono, J., Pleydell, E., French, N., Besser, T., 2010. Diverse genetic markers concoedantly identified bovine origin Escherichia coli O157 genotypes underrepresented in human disease. Appl. Environ. Microbiol. 76, 361–365.
1728
E. Yokoyama et al. / Infection, Genetics and Evolution 12 (2012) 1724–1728
Yang, Z., Kovar, J., Kim, J., Nietfeldt, J., Smith, D.R., Moxley, R.A., Olson, M.E., Fey, P.D., Benson, A.K., 2004. Identification of common subpopulations of non-sorbitolfermenting, beta-glucuronidase-negative Escherichia coli O157:H7 from bovine production environments and human clinical samples. Appl. Environ. Microbiol. 70, 6846–6854. Yokoyama, E., Etoh, Y., Ichihara, S., Horikawa, K., Konishi, N., Kai, A., Matsumoto, Y., Kurosaki, M., Kasahara, H., Kurazono, T., Yoda, K., 2011a. Emergence of enterohemorrhagic Escherichia coli serovar O157 strains in
clade 8 with highly similar pulsed-field gel electrophoresis patterns. J. Food Prot. 74, 1324–1327. Yokoyama, E., Hashimoto, R., Etoh, Y., Ichihara, S., Horikawa, K., Uchimura, M., 2011b. Biased distribution of IS629 among strains in different lineages of enterohemorrhagic Escherichia coli serovar O157. Infect. Genet. Evol. 11, 78–82. Zhang, Y., Laing, C., Steele, M., Ziebell, K., Johnson, R., Benson, A.K., Taboada, E., Gannon, V.P.J., 2006. Genome evolution in major Escherichia coli O157:H7 lineages. BMC Genomics 8, 121.