ABSTRACT
The O serogrouping of pathogenic Escherichia coli is a standard method for subtyping strains for epidemiological studies and enhancing phylogenetic studies. In particular, the identification of strains of the same O serogroup is essential in outbreak investigations and surveillance. In a previous study, we analyzed the O-antigen biosynthesis gene cluster in all known E. coli O serogroups (A. Iguchi et al., DNA Res, 22:101–107, 2015, http://dx.doi.org/10.1093/dnares/dsu043). Based on those results, we have arranged 162 PCR primer pairs for the identification or classification of O serogroups. Of these, 147 pairs were used to identify 147 individual O serogroups with unique O-antigen biosynthesis genes, and the other 15 pairs were used to identify 15 groups of strains (Gp1 to Gp15). Each of these groups consisted of strains with identical or very similar O-antigen biosynthesis genes, and the groups represented a total of 35 individual O serogroups. We then used the 162 primer pairs to create 20 multiplex PCR sets. Each set contained six to nine primer pairs that amplify products of markedly different sizes. This genetic methodology (E. coli O-genotyping PCR) allowed for comprehensive, rapid, and low-cost typing. Validation of the PCR system using O-serogroup references and wild strains showed that the correct O serogroups were specifically and accurately identified for 100% (182/182) and 90.8% (522/575) of references and wild strains, respectively. The PCR-based system reported here might be a promising tool for the subtyping of E. coli strains for epidemiological studies as well as for the surveillance of pathogenic E. coli during outbreaks.
INTRODUCTION
The O-antigen polysaccharide constitutes the outermost part of the lipopolysaccharide (LPS) present in the outer membranes of Gram-negative bacteria. The chemical composition and structure of O antigens exhibit high levels of variation even in a single species, and the serotyping of strains with O antigens is used in epidemiological studies of pathogenic bacteria (1–4). Thus far, the World Health Organization Collaborating Centre for Reference and Research on Escherichia and Klebsiella, which is based at the Statens Serum Institut (SSI) in Denmark (http://www.ssi.dk/English.aspx), has recognized 184 Escherichia coli O serogroups. These are designated O1 to O187 and include three pairs of subgroups, O18ab/O18ac, O28ab/O28ac, and O112ab/O112ac, and six missing numbers, O31, O47, O67, O72, O93, and O122. The Shiga toxin-producing E. coli (STEC) strains constitute one of the most important groups of foodborne pathogens, as they can cause gastroenteritis that may be complicated by hemorrhagic colitis or hemolytic-uremic syndrome (HUS) (5). O157 is a leading STEC O serogroup associated with HUS (6), and other STEC O serogroups, including O26, O103, O111, O121, and O145, are also recognized as significant foodborne pathogens worldwide (7). Additionally, unexpected STEC O serogroups have sometimes emerged to cause sporadic cases or outbreaks. For example, STEC O104:H4 was responsible for a large foodborne disease outbreak in Europe in 2011 (8).
The O serogrouping of E. coli strains provides valuable information for identifying pathogenic clonal groups (9). In particular, the identification of strains of the same O serogroup is essential in outbreak investigations and surveillance. Many rapid and low-cost genetic methodologies for identifying E. coli O serogroups have been developed (10–15). Coimbra et al. (11) reported a broad O-typing method characterized by the restriction fragment length polymorphism pattern of amplified O-antigen biosynthesis gene clusters (O-AGCs). On the other hand, in most cases, these target sequences of O-antigen processing genes such as wzx (encoding the O-antigen flippase), wzy (encoding the O-antigen polymerase), and the wzm and wzt genes (encoding components of the ABC transporter pathway). These are highly variable orthologs and are considered to be specific to each O serogroup (12).
In our previous study (16), we analyzed the O-AGCs from all 184 known E. coli O serogroups. By comparing sequences, we revealed that among the 182 O serogroups (excluding O14 and O57, which contain no O-AGCs at the typical locus) 145 had unique O-AGCs and the other 37 shared identical or very similar O-AGCs, which were placed into 16 groups (Gp1 to Gp16). Although most wzx-wzy and wzm-wzt genes showed high levels of DNA sequence diversity (less than 70% identity with the most similar other O-AGCs or O-AGC groups), there was high DNA sequence conservation within the 16 O-AGC groups (most with ≥97% identity). An exception was O169 and O183 of Gp16, which was distinguished clearly on the basis of the wzx sequences because a fragment, including wzx, was exchanged within the Gp16 O-AGCs. Based on these results, we believe that the remarkable sequence diversity and conservation in the wzx-wzy and wzm-wzt genes is sufficient to make possible the identification of each of the known O serogroups using these sequences.
Here, we present a basic set of 162 PCR primer pairs and a multiplex PCR system (E. coli O-genotyping PCR) for the comprehensive molecular O-typing of E. coli. The system was evaluated for its specificity and usefulness using 184 O-serogroup reference strains and 579 O-serogrouped wild strains.
MATERIALS AND METHODS
Source sequences.The O-AGC sequences that were used for designing primers were obtained from the GenBank database (see Table S1 in the supplemental material).
Bacterial strains, serogrouping, and DNA preparation.The E. coli O serogroups were designated O1 to O187 and included three pairs of subgroups, O18ab/O18ac, O28ab/O28ac, and O112ab/O112ac and six missing numbers, O31, O47, O67, O72, O93, and O122. These 184 O-serogroup reference strains from SSI were used to evaluate our PCR primers in simplex and multiplex PCR assays (see Table S1 in the supplemental material). Additionally, 579 wild E. coli isolates from all but eight O serogroups (O16, O17, O18ab, O46, O60, O97, O112ac, and O170) were used for validation of the comprehensive multiplex PCR system (Table 1; see Table S4 in the supplemental material). Of these, 440 were isolated from humans, animals, and foods in Japan, and the other 139 were obtained from SSI. O serogrouping was performed by agglutination tests in microtiter plates using commercially available pooled and single antisera against all recognized E. coli O antigens (O1 to O187) (SSI Diagnostica, Hillerød, Denmark). Genomic DNAs were purified using the Wizard Genomic DNA purification kit (Promega, Madison, WI, USA) according to the manufacturer's instructions. Single template assays were performed using 10 ng/μl of template DNA, and pooled template assays were performed using mixed template DNAs (10 ng/μl each) with ≤5 reference strains. Additionally, supernatants of boiled reference strains were also prepared as DNA templates for validation. Briefly, each overnight LB broth culture was pelleted by centrifugation, the supernatant was discarded, and the cells were suspended in Tris-EDTA (TE) buffer equivalent to one-quarter volume of culture medium. The suspension was boiled for 10 min, and the centrifuged supernatant was used as a DNA template.
O-serogroup-identified wild strains used in the validation assays
Simplex PCR.Simplex PCR was performed as follows: each 30-μl reaction mixture contained 2 μl of genomic DNA, 6 μl of 5 × Kapa Taq buffer, deoxynucleoside triphosphate (dNTP) mix (final concentration, 0.3 mM each), MgCl2 (final concentration, 2.5 mM), primers (final concentration, 0.5 μM each), and 0.8 U of Kapa Taq DNA polymerase (Kapa Biosystems, Woburn, MA, USA). The thermocycling conditions were 25 cycles of 94°C for 30 s, 58°C for 30 s, and 72°C for 1 min. PCR products (2 μl) were electrophoresed in 1.5% of agarose gels in 0.5× Tris-borate-EDTA (TBE) (25 mM Tris-borate and 0.5 mM EDTA) and photographed under UV light after the gel was stained with ethidium bromide (1 mg/ml).
Multiplex PCR.Multiplex PCR was performed as follows: each 30-μl reaction mixture contained 2 μl of genomic DNA, 6 μl of 5 × Kapa Taq buffer, dNTP mix (final concentration, 0.3 mM each), MgCl2 (final concentration, 2.5 mM), 3.52 μl of primer mix (see Table S3 in the supplemental material), and 0.8 U of Kapa Taq DNA polymerase. The thermocycling conditions included an initial denaturation of 94°C for 1 min followed by 25 cycles of 94°C for 30 s, 58°C for 30 s, 72°C for 1 min, and then a final extension of 72°C for 2 min. PCR products were separated and visualized using the automated capillary electrophoresis QIAxcel system with the QIAxcel DNA Screening kit (Qiagen, Hilden, Germany). Positive products by the QIAxcel system were also confirmed with the gel electrophoresis by the same method as described for simplex PCR.
Validation with reference strains.In the simplex and multiplex PCR assays, the specificities of the primer pairs were validated using 37 pooled DNA mixtures containing DNA from all 184 reference strains, and subsequently with individual DNAs contained in PCR-positive pool(s). For validation using the boiled strain suspensions, individual template DNA samples were used.
Validation with wild strains.The appropriateness of the PCR system was assessed using purified template DNAs from O-serogroup-identified E. coli isolates. As a control for DNA integrity, E. coli-specific primers that amplify a 622-bp fragment from gryB (gyrB_F; AGTGATCATGACCGTTCTGCA and gryB_R; TTTGTCCATGTAGGCGTTCA) were used.
RESULTS
Primers for O genotyping.In this study, a total of 162 O-typing PCR primer pairs were used for identifying molecular O serogroup (O genotype) (see Table S2 in the supplemental material). Of the 162 primer pairs, 34 were taken from published primer sequences, and the other 128 were designed for this study based on a sequence set we analyzed in our previous study (16). Among the 162 primer pairs, 147 were used to identify the O genotype of each of the 147 individual O serogroups, and the other 15 were used for identifying each of 15 groups (OgGp1 to OgGp15) with identical or very similar O-AGC sequences that represented a total of 35 O serogroups (see Table S2). All primer pairs targeted unique sequences of the O-antigen processing genes wzx, wzy, wzm, or wzt except for two pairs, Og8 and Og157, which targeted unique O-AGC genes: orf469 (encoding mannosyltransferase) and rfbE (encoding perosamine synthetase), respectively (see Table S2). The PCR product sizes ranged from 132 bp (Og145) to 1,235 bp (Og9) (Fig. 1; see Table S2). Subsequently, 20 multiplex PCR sets (MP-1 to MP-20) were produced that included all 162 primer pairs (see Table S3 in the supplemental material). Each reaction set contained six to nine primer pairs whose PCR products were of different sizes and made step-like patterns on agarose gels (Fig. 2; see Table S3).
Gel images of 162 PCR products ranging from 132 bp (Og145) to 1,253 bp (Og9). The 162 primer pairs were designed to identify or classify specific E. coli O serogroups. Seven gel images were placed in series, and 100-bp DNA ladders were placed at each end.
Gel images of PCR products obtained using the 162 primer pairs, grouped as they were in the multiplex PCR sets MP-1 to MP-20. Products of primers designed for grouping into OgGp1 to OgGp15 are indicated by asterisks. M, 100-bp DNA ladder (GeneDirex). See Table S3 in the supplemental material for each product size.
Evaluation with reference strains.All 184 O-serogroup reference strains from O1 to O187 were used to evaluate our PCR primers in simplex and multiplex PCR assays. In the two assays, clear PCR products of the expected sizes were obtained only with the corresponding strain(s), and no extra products were produced in the size range of 100 to 1,500 bp (data not shown). These results confirmed the specificity and usability of each simplex and multiplex PCR set with purified and boiled template DNAs.
Validation with wild strains.We also assayed 579 O-serogrouped wild E. coli isolates from humans, animals, and foods using the full set of 20 multiplex PCRs. Each O serogroup used in this study was represented by ≥1 and ≤24 strains (Table 1). Among these strains, 522 gave O genotypes that corresponded to the O serogroups obtained by conventional E. coli serogrouping. On the other hand, 13 strains belonging to nine O serogroups gave O genotypes that did not correspond to their recorded serogroups (Table 2). Agglutination tests indicated that at least 10 of these strains reacted with the antiserum corresponding to the O genotype that we obtained. No PCR products were obtained from 39 strains belonging 20 O serogroups (Table 3). Furthermore, two PCR products were obtained from each of five strains belonging to four O serogroups (Table 4). Agglutination tests indicated that none of the secondary O genotypes associated with their serological determinants. Thus, we found that with our PCR system, some strains gave no PCR products or O genotypes that were inconsistent with their O serogroups. However, the correct O serogroups were specifically and accurately identified for the majority of strains (522 of 575, except for O14 and O57 isolates). These results validate the utility of the developed E. coli O-genotyping PCR system for O-serogroup identification.
List of strains that had no or weak association between O serogroup and O genotype
Strains whose O genotypes were not identified by the PCR system
List of strains that each had two O genotypes detected by the PCR system
DISCUSSION
This is the first report to show a comprehensive molecular O-typing platform, which was made possible by recent advances in the exploration of E. coli genomic diversity. Although serotyping is a standard method for subtyping of pathogenic E. coli isolates, agglutination reactions with specific antisera are laborious and expensive, and cross-reactions between different O serogroups often occur giving equivocal results (17). In this study, to efficiently screen O-genotypes, 20 multiplex PCR sets containing 162 primer pairs were developed. The PCR method chosen here is the most frequently and widely used molecular technique for the investigation and surveillance of pathogenic bacteria. The multiplex PCR system used in this study does not require special enzymes and solutions. The optimized and unified PCR conditions dramatically enhance work efficiency in subtyping E. coli isolates. Conventional gel electrophoresis is available as a general system for screening and identification of PCR products. Therefore, this genetic methodology allows for comprehensive, rapid, and low-cost O typing of isolates from patients and contaminated foods. The system provides usable information to evaluate the routes, sources, and prevalence of agents, leading to prompt infection control. Additionally, this system may allow large-scale epidemiological studies monitoring local, national, and international trends in pathogenic E. coli.
We confirmed that the PCR system accurately and specifically identified the O serogroups of all reference strains and the majority of wild strains. However, some strains showed inconsistent or complex results when the PCR results were compared with results from traditional O serotyping. The O genotypes from 13 strains did not correspond to their O serogroups identified by serotyping. The reason for this lack of correspondence is unclear; however, agglutination tests with a single antiserum indicated that at least 10 strains also reacted with the expected O-genotype-associated antiserum. This suggests that these strains also expressed O antigens associated with the O-genotypes, but cross-reacted more strongly with different antisera. The O genotypes of 39 strains representing 20 O serogroups were not determined because no clear PCR products were obtained. These included four strains belonging to the O14 and O57 serogroups, for which specific primers were not designed. It is known that Salmonella enterica and Citrobacter freundii share similar but sequence-diversified O-AGCs with E. coli O157, O111, and O55 but produce O antigens that are identical to those produced by E. coli (18). Plainvert et al. (19) showed that there are two types of O-AGCs in E. coli O45 strains; one type is related to intestinal pathogenic E. coli such as STEC, and the other is related to extraintestinal pathogenic E. coli. These two O-AGCs are genetically related but show low or no DNA sequence homology in orthologous genes (19). Sequence variations among the O-AGCs of some strains within the same O serogroups may explain the lack of PCR products in this study. Two PCR products were obtained from each of five strains. Our previous study showed that the E. coli O8 reference strain had two O-AGCs in series on the chromosome (16). One O-AGC was a wzm-wzt-type cluster that was identical to the O-AGC of the O8 strain IAI1 (20). The other (second) O-AGC was a wzx-wzy-type cluster that had no or low sequence identity with all reference strains. The cps and kps regions, encoding capsule K-antigen synthesis enzymes, include wzx-wzy-like and wzm-wzt-like genes, respectively (21). These homologous polysaccharide synthesis genes located inside or outside the O-AGC locus might be detected by the PCR screening. Further sequence analysis of the O-AGCs or whole genomes will be needed to explain the inconsistent and complex results that we obtained with some strains. Additional sequence data will also serve as a valuable resource for the development of a broader O-typing system covering novel O genotypes.
In conclusion, we present a basic set of PCR primers for the identification and classification of almost all known E. coli O serogroups. The PCR-based O-typing system reported here provides an accurate and reliable approach for subtyping E. coli isolates to much the same level as O serogrouping. This E. coli O-genotyping PCR system might be a promising tool for subtyping of E. coli strains for epidemiological studies as well as for the surveillance of pathogenic E. coli during outbreaks.
ACKNOWLEDGMENTS
We thank Atsuko Akiyoshi and Yuiko Kato for technical assistance.
This work was supported by Health Labor Sciences Research Grants from the Ministry of Health, Labor, and Welfare, Japan to A.I. (H25-Syokuhin-Wakate-018) and M.O. (H24-Shinkou-Ippan-012), the Adaptable and Seamless Technology Transfer Program through Target-driven R&D (AS242Z00217P) from Japan Science and Technology Agency to A.I., and a Scientific Research Grant on Priority Areas from the University of Miyazaki and the Program to Disseminate Tenure Tracking System from the Japanese Ministry of Education, Culture, Sports, Science, and Technology to A.I.
Members of Pathogenic E. coli Working Group in Japan include the following: Aomori Prefectural Institute of Public Health and Environment (Hiroko Takenuma and Osamu Fukuda), Chiba City Research Institute for Environmental Sciences and Public Health (Tomoko Kitahashi), Chiba Prefectural Institute of Public Health (Eiji Yokoyama and Shinichiro Hirai), Ehime Prefectural Institute of Public Health and Environmental Science (Keiko Semba), Environmental Hygiene Inspection Center of Higashiosaka City, Fukui Prefectural Institute of Public Health and Environmental Science (Akihiro Nagata and Fubito Ishiguro), Fukuoka City Institute for Hygiene and the Environment, Fukuoka Institute of Health and Environmental Sciences (Yoshiki Etoh and Nobuyuki Sera), Fukushima Institute of Public Health (Rie Kikuchi), Gifu Prefectural Research Institute for Health and Environmental Sciences (Noda Makiko and Shiraki Yutaka), Himeji City Environment Sanitary and Hygiene Research Institute (Shinya Kawanishi), Hiroshima Prefectural Technology Research Institute, Health and Environment Center (Yoshihiro Takeda, Hiroko Yamada, Kazumi Imai, and Kanako Masuda), Hokkaido Institute of Public Health (Toshiyuki Nagase, Keiko Ogawa, Tetsuya Ikeda, Yo Morimoto, and Shunichi Shimizu), Ishikawa Prefectural Institute of Public Health and Environmental Science (Emiko Kitagawa and Keiko Kawakami), Ishikawa Prefectural Institute of Public Health and Environmental Science, Japan Microbiological Laboratory Co., Ltd., Kagawa Prefectural Research Institute for Environmental Sciences and Public Health (Chiemi Fukuda, Youko Iwashita, Mayumi Arizuka, and Junko Uchida), Kagoshima Prefectural Institute for Environmental Research and Public Health, Kanagawa Prefectural Institute of Public Health (Ichiro Furukawa and Toshiro Kuroki), Kawasaki City Institute for Public Health (Sachiko Homma, Akiko Kubomura, and Hiroyasu Sato), Kawasaki Municipal Ida Hospital (Yuka Kojima), Kitakyusyu City Institute of Environmental Sciences, Kobe City Institute for Environment and Public Health, Kumamoto City Environmental Research Center, Kumamoto Prefectural Institute of Public Health and Environmental Science (Seiya Harada), Mie Prefecture Health and Environment Research Institute (Yuhki Nagai), Miyagi Prefectural Institute of Public Health and Environment (Takashi Hatakeyama), Miyazaki Prefectural Institute for Public Health and Environment (Kimiko Kawano, Shuji Yoshino, and Mariko Kurogi), Nagano Environmental Conservation Research Institute (Hitomi Kasahara and Maki Sekiguchi), Nagasaki Prefectural Institute for Environmental Research and Public Health (Hayato Nishimura), Nara Prefecture Institute of Health (Sumiko Tanabe), Niigata City Institute of Public Health and Environment, Niigata Prefectural Institute of Public Health and Environmental Sciences (Masao Kawase), Oita Prefectural Institute of Health and Environment (Ogata Kikuyo, Narimatsu Hiroshi, and Sasaki Mari), Okayama Prefectural Institute for Environmental Science and Public Health (Hiroshi Nakajima, Hisahiro Kawai, and Ritsuko Ohata), Okazaki City Public Health Center (Kunihiko Nakane), Okinawa Prefectural Institute of Health and Environment, Osaka City Institute of Public Health and Environmental Sciences (Hiromi Nakamura), Osaka City University (Yoshikazu Nishikawa), Osaka Prefectural Institute of Public Health (Masumi Taguchi), Saga Prefectural Center of Public Health and Sanitary (Shiyunichi Yoshitake), Sagamihara City Laboratory of Public Health, Saitama Prefectural Institute of Public Health (Takayuki Kurazono), Saitama-City Institute of Health Science and Research, Sendai City Institute of Public Health, Shiga Prefectural Institute of Public Health (Kazuhiko Ishikawa, Seiko Umehara, and Tomomi Kono), Shimane Prefectural Institute of Public Health and Environmental Science (Jun Kawase and Yuta Kawakami), Shizuoka City Institute of Environmental Sciences and Public Health, Shizuoka Institute of Environment and Hygiene, Tokushima Prefectural Public Health, Pharmaceutical and Environmental Sciences Center (Hiroko Ishida), Toyama Institute of Health (Keiko Kimata and Junko Isobe), University of Miyazaki Hospital (Akihiko Okayama and Yuji Saeki), Wakayama Prefectural Research Center of Environment and Public Health (Kayoko Nakaoka), Yamaguchi Prefectural Institute of Public Health and Environment (Kiyoshi Tominaga, Junko Yabata, and Mitsuhiro Kameyama), and Yokohama City Institute of Health (Atsuko Ogawa and Yuko Matsumoto).
FOOTNOTES
- Received 5 February 2015.
- Returned for modification 23 March 2015.
- Accepted 21 April 2015.
- Accepted manuscript posted online 29 April 2015.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/JCM.00321-15.
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.
The authors have paid a fee to allow immediate free access to this article.
