Advance Publication by J-STAGE Japanese Journal of Infectious Diseases Molecular typing of Escherichia coli O157 isolates from Romanian human cases
Mihaela Oprea, Adriana Simona Ciontea, Mădălina Militaru, Sorin Dinu, Daniela Cristea, and Codruţa-Romaniţa Usein
Received: March 23, 2018. Accepted: July 17, 2018 Published online: July 31, 2018 DOI:10.7883/yoken.JJID.2018.129
Advance Publication articles have been accepted by JJID but have not been copyedited or formatted for publication.
Title page
Molecular typing of Escherichia coli O157 isolates from Romanian human cases
Authors:
us cr ip
t
Running head: Typing of Romanian Escherichia coli O157 isolates
Mihaela Oprea1, 2, Adriana Simona Ciontea1, Mădălina Militaru1, Sorin Dinu1, Daniela Cristea1, Codruţa-Romaniţa Usein1, 3* Affiliation:
Cantacuzino National Medical - Military Institute of Research and Development, Splaiul
M an
1
Independentei 103, sector 5, 050096, Bucharest, Romania 2
The Research Institute of the University of Bucharest, Bd. M. Kogalniceanu 36-46, sector 5,
050107, Bucharest, Romania
Carol Davila University of Medicine and Pharmacy, Dionisie Lupu 37, sector 1, Bucharest,
Romania
pt ed
3
ce
The work was done at: Molecular Epidemiology Laboratory, Cantacuzino National Institute
Ac
of Medical - Military Research - Development, Splaiul Independentei 103, sector 5, 050096, Bucharest, Romania
Corresponding author: Codruţa-Romaniţa Usein Molecular Epidemiology Laboratory, Cantacuzino National Medical - Military Institute of Research and Development, Splaiul Independentei 103, sector 5, Bucharest, Romania e-mail:
[email protected]; Tel: +40213069223 Keywords: Escherichia coli O157, molecular characterisation, whole genome sequencing 1
Summary Verocytotoxin-producing Escherichia coli (VTEC) of serogroup O157 are among the most important causes of severe cases of foodborne disease and outbreaks worldwide. As little is known about the characteristic of these strains in Romania, we aimed to provide reference
us cr ip
t
information on the virulence gene content, phylogenetic background, and genetic diversity of seven autochthonous O157 strains collected during 2016 and 2017 from epidemiologically non-related cases. These strains were typed by a combination of phenotypic and molecular methods routinely used by the national reference laboratory. Additionally, four of them were also subjected to whole-genome sequencing (WGS) and public web-based tools were used to
M an
extract information on virulence gene profiles, multilocus sequence types (MLST), and SNPbased phylogenetic relatedness. Molecular typing provided evidence of the circulation of a polyclonal
population
while
distinguishing
a
cluster
of
non-sorbitol-fermenting,
glucuronidase-negative, phylogenetic group E, MLST 1804 strains, representing lineage II
pt ed
and clade 7, which harbored vtx2c, eae-gamma, and ehxA genes. A good correlation between the routine typing methods and WGS data was observed. Yet SNP-based genotyping provided a higher resolution in depicting the relationships between the O157:H7 strains than
ce
that provided by PFGE. This study should be a catalyst for improved laboratory-based
Ac
surveillance of autochthonous VTEC.
Introduction
2
Diarrheagenic Escherichia coli (DEC) strains are extensively studied due to their clinical and epidemiological importance. Classified in more categories, based on the mechanisms that lie behind the distinct syndromes they generate, DEC are known as enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), enteroinvasive E. coli (EIEC), enteroaggregative E. coli (EAEC), diffusely adherent E. coli (DAEC), and
us cr ip
t
verocytotoxin producing E. coli (VTEC) (1).
VTEC (also known as Shiga toxin-producing E. coli, STEC) are foodborne pathogens with worldwide distribution linked to illness outbreaks and severe pathology. They are defined by their capacity to produce verocytotoxins or Shiga toxins, bacteriophage-encoded cytotoxins considered the most important virulence trait that differentiate them from EPEC.
M an
Enterohemorragic E. coli (EHEC) are a subset of VTEC which are able to colonize the bowel usually by an intimate effacing adherence to the epithelial cell membrane. EHEC strains induce the formation of „attaching-and-effacing” (A/E) lesions, a histopathological feature also detected in EPEC colonization (2). Intimin is the main adherence factor demonstrated to
pt ed
play a role in the intestinal colonization, a complex process that requires mechanisms encoded by genes located on the so-called locus of enterocyte effacement (LEE) pathogenicity island (3). Commonly, EHEC are capable of producing severe bloody diarrhea
ce
and hemorrhagic colitis that can progress to complications such as hemolytic uremic
Ac
syndrome (HUS). EHEC most commonly isolated from humans worldwide are representing a limited number of serogroups although hundreds of VTEC serotypes have been identified to date (4). O157:H7 is the most common and notorious EHEC serotype, known for causing foodborne outbreaks (5). Molecular typing facilitated the characterization and comparison of E. coli O157:H7 strains recovered from different isolation sources. Relevant genomic information supported
3
their segregation into more distinct evolutionary lineages and clades which seemed to differ in host specificity and virulence characteristics (6-8). The public health importance of VTEC infections across Europe displays considerable national differences and apparently, they are quite rare in Romania (9). However, the recent HUS outbreak attributed to VTEC members of O26 serogroup registered among young
us cr ip
t
children (10) was evidence that the burden caused by VTEC infections is significantly underestimated in Romania and a more consistent laboratory-based surveillance is needed for reassessing it. In this context, we employed a molecular approach to confirm the intrinsic virulence of autochthonous E. coli selected as presumptive VTEC by O157-serotyping
serotype. Materials and methods
M an
method and to provide a much thorough genotyping of representative members of O157:H7
Bacterial strains and phenotypic characterization
The E. coli O157 strains investigated in this study were received for confirmation
pt ed
tests in the national reference laboratory for VTEC infection from Cantacuzino National Institute of Medical-Military Research-Development. The strains originated from 7 apparently sporadic cases of acute diarrhea (5 strains) and HUS (2 strains) with no
ce
epidemiological linkage, diagnosed in 2016 and 2017 in children less than 2 years old,
Ac
resided throughout Romania (Table 1). The strains were serotyped for the O157 and H7 antigens by slide agglutination with
commercially available antisera (SSI Diagnostica) and tested for sorbitol fermentation on sorbitol MacConkey agar (bioMérieux). For this study, the strains were given identification numbers (from RO_O157_1 to RO_O157_7) according to the chronological order of isolation. PCR-based characterization 4
A multiplex PCR-based protocol targeting rfbEO157 and fliCH7 with previously published primers (11) was used for the molecular serotyping of the strains. The strains were also screened for the presence of EHEC virulence markers vtx1 (verocytotoxin 1), vtx2 (verocytotoxin 2), and eae (intimin) by using a commercially available multiplex PCR-based kit (DEC Primer Mix, SSI Diagnostica). Additionally,
us cr ip
t
previously published primers (12) were used for the detection of EHEC enterohemolysin encoding gene ehxA.
Subtyping of vtx1 and vtx2 genes was performed according to a published protocol recommended by WHO Collaborating Centre for Reference and Research on Escherichia and Klebsiella from Statens Serum Institute, Denmark (13).
M an
Phylogenetic group assignment of the O157 E. coli strains was performed according to the method described by Clermont et al. (14). Pulsed-field gel electrophoresis
Pulsed-field gel electrophoresis (PFGE) was used to assess the genetic relatedness of
pt ed
the E. coli O157 strains. XbaI-digested genomic DNA was analyzed on a 1% SeaKem Gold agarose gel in 0.5 × TBE buffer using CHEF MAPPER (BioRad) (15). The run time was 18 h at 6 V cm−1, with initial and final switch times of 2.16 and 54.17 s, respectively. The gel was
ce
stained with 1 μg mL−1 ethidium bromide, visualized on the Gel Doc XR system (BioRad)
Ac
and analyzed using Bionumerics fingerprinting software version 6.6 (Applied Maths). Pulsotype designation was assigned at the 100% profile similarity level. Whole genome sequencing The genomes of E. coli strains RO_O157_2, RO_O157_3, RO_O157_5 and RO_O157_7 were sequenced using an Ion Torrent PGM (Thermo Fisher Scientific) according to 400-bp protocols for library preparation through enzymatic shearing, Ion OneTouch2 emulsion PCR, enrichment, Hi-Q View sequencing kits (Thermo Fisher 5
Scientific). The sequences obtained were subjected to de novo assembly by using the Assembler SPAdes plugin version 5.8 (16) installed on Ion Torrent Server. Analysis of WGS sequence data Whole genome sequencing (WGS) data were analyzed using the web-based tools, designed by the Center of Genomic Epidemiology (CGE), available on the public CGE server
us cr ip
t
(at https://cge.cbs.dtu.dk/services/). VirulenceFinder tool was used for the detection of virulence genes included in the CGE database (17) SerotypeFinder tool was used for WGSbased O and H typing (18). Multilocus sequence typing (MLST) tool with the configuration set to E. coli scheme 1 (adk, fumC, gyrB, icd, mdh, purA, recA) was used for determination of the multilocus sequence types (19). Call SNPs & Infer (CSI) Phylogeny tool (20) with default
M an
parameters (a minimal depth at SNP positions of 10x, a minimal distance between SNPs of 10 bp, a minimal SNP quality score of 30, a minimal Z-score of 1.96) was used to perform a single nucleotide polymorphism (SNP)-typing phylogenetic analysis with E. coli EDL 933 strain as reference sequence (NZ_CP008957.1) and construct a maximum likelihood tree
pt ed
from the SNP alignment using Mega 7 (21). Published sequences available in the NCBI GenBank and Sequence Read Archive (SRA) were used for comparison (Table 2). For a better comparison with the Romanian sequenced isolates, the reference strains were also
ce
analyzed with VirulenceFinder tool.
Ac
Additionally, intimin typing was performed in silico through a BLAST search (22) of the eae gene sequences from the WGS against the National Center for Biotechnology Information (NCBI) nucleotide repository with a threshold of 95% nucleotide sequence identity with existing intimin genes sequences. Also, a BLAST analysis for detection of the uidA gene mutations was performed in order to predict the beta-glucuronidase (GUD) activity. Besides, in order to confirm the non-sorbitol fermenting character of the strains, the annotated WGS sequences were investigated for the following supplementary markers: urease gene cluster, 6
tellurite resistance and adherence-conferring pathogenicity island, and fimbrial sfp cluster (23, 24). The lineage assignment of the four O157:H7 strains sequenced was based on the investigation of the polymorphism of the six markers proposed by Yang et al. (6) using the Integrative Genomics Viewer (IGV) software (25) and EDL933 as reference strain. For this
us cr ip
t
study, Sanger sequencing was used to confirm the insertions and deletions observed in the target loci at IGV inspection.
Clade typing relied on the identification of clade specific SNPs originally described by Manning et al. (8) targeting known informative positions (26).
The WGS reads of the four O157:H7 strains are part of the European Molecular
M an
Biology Laboratory’s (EMBL) European Nucleotide Archive (ENA) sequence database (accession number PRJEB25277). All assembled genomes were uploaded on NCBI Prokaryotic Genome Annotation Pipeline and received the following GenBank accession numbers: PUIS00000000 (RO_O157_2), PUIU00000000 (RO_O157_3), PUIT00000000
Results
pt ed
(RO_O157_5), and PUIR00000000 (RO_O157_7).
Phenotypic characterization
ce
Five O157 strains belonged to serotype O157:H7 and were sorbitol non-fermenters.
Ac
The other two O157 strains, although motile, lacked the H7 flagellar antigen. Only one of them was able to ferment the sorbitol. PCR-based serotyping, virulotyping, and phylogenetic grouping Molecular serotyping confirmed all the strains to be of serogroup O157, distinguished the fliCH7 gene in all five O157:H7 strains, and resolved the two fliCH7 gene negative strains. Of the virulence genes for whose presence the O157 E. coli strains were investigated by PCR, vtx, eae, and ehxA were detected. Overall, six PCR-defined virulotypes were 7
identified which qualified the strains as VTEC/EHEC (4 strains), EPEC (2 strains) or nonVTEC non-EPEC (1 strain) (Table. 1). The study strains were assigned to phylogenetic groups A (1 strain), B1 (1 strain), D (1 strain), and E (4 strains). The O157:H7 strains belonged to phylogroups D (1 strain) and E (4 strains) and harbored combinations of at least two of the virulence gene markers
us cr ip
t
considered for DEC diagnostic identification (i.e. eae+ehxA+, vtx2c+eae+ehxA+, vtx1a+vtx2a+). PFGE analysis
All the E. coli O157 strains yielded interpretable XbaI PFGE profiles which extended to the 68% similarity level overall. However, the 100% similarity criterion resolved six pulsotypes of which one contained 2 strains. The phylogenetic group E O157:H7 strains
(Fig. 1). WGS-based typing
M an
(RO_O157_2, RO_O157_3 and RO_O157_5) formed a discrete cluster on the dendrogram
The VTEC O157:H7 strains of phylogenetic group E were subjected to a more
pt ed
thorough analysis performed by WGS. WGS assembly metrics are presented in Table 3. The serotype detection tool confirmed the accuracy of the PCR-based molecular serotyping indicating all 4 strains as serotype O157:H7.
ce
Sequence type ST1804 was assigned by the MLST tool to RO_O157_2, RO_O157_3,
Ac
RO_O157_5, and ST 11 to RO_O157_7, respectively. Concerning the GUD encoding gene uidA, comparison of the WGS data with the
same sequence of K12 DH10B strain (GenBank access no. CP000948) revealed that the autochthonous strains displayed 4 mutations that resulted in the following amino acid changes: D31E, A64V, C253Y, and Q259R. Moreover, RO_O157_2, RO_O157_3, and RO_O157_5 strains but not RO_O157_7 had a GG insertion within uidA at nucleotide 686, resulting in a premature stop codon. 8
The WGS annotated sequences showed that all the strains carried a complete urease cluster and tellurite resistance and adherence-conferring pathogenicity island and were negative for the sfp cluster. A 100% concordance between the virulence detection tool and conventional PCR results for vtx (type/subtype), ehxA, and eae genes was observed, in silico eae subtyping
us cr ip
t
indicating a γ-type intimin. Strains RO_O157_2, RO_O157_3 and RO_O157_5 displayed almost identical WGS-based virulence profiles, the only difference being that RO_O157_2 lacked the gad gene encoding glutamate decarboxylase. RO_O157_7 strain exhibited a different profile (Table 1). All the strains analyzed carried a pO157-like plasmid which in the case of RO_O157_7 strain lacked espP and katP genes. CSI Phylogeny tool allowed SNPs
M an
identification from WGS data and further phylogenetic inference separated the four autochthonous strains from one another and from selected published genomes available in the NCBI GenBank. The topology of the SNP-based phylogenetic tree (Fig. 2) showed the clustering of the RO_O157_2, RO_O157_3, and RO_O157_5 strains and their separation
pt ed
from the ST11 RO_O157_7 strain, which resembled the topology observed on the PFGEbased tree (Fig. 1). The SNP difference matrix, calculated by CSI Phylogeny, indicated that the clustered strains were separated from one another by 30 to 75 SNPs and each of them
ce
harbored >2600 SNP differences compared to
RO_O157_7 strain. Overall, the
Ac
autochthonous strains clustered with the published strains belonging to the same STs. On the basis of the WGS data used for lineage and clade typing, RO_O157_2,
RO_O157_3, and RO_O157_5 strains were assigned to lineage II (genotype 232123) and clade 7 and RO_O157_7 to lineage I/II (genotype 211111) and clade 9.
Discussion
9
As the globalization process appears to be causing unpredictable changes in the epidemiology of infectious diseases, the local background of infectious agents warrants further attention in order to understand the risks they pose to both resident and visiting populations. The year 2016 was marked by the occurence of the first major event linked to VTEC in Romania and brought public awareness about how little was known about the
us cr ip
t
autochthonous reservoir of these pathogens. Our aim was to give an insight into the features of a set of autochthonous E. coli strains belonging to O157 serogroup which is often assumed to be the source of the most virulent VTEC (2) and provide reference information about their intrinsic virulence, phylogenetic background, and genetic diversity.
The molecular approach commonly used at national reference laboratory level reliably
M an
distinguished O157:H7 from O157:nonH7 strains and detected associations of key virulence gene markers indicative of the capacity to produce the verocytotoxins, presence of the LEE pathogenicity island, and possession of the large enterohemolysin-encoding plasmid in the O157:H7 strains. With one exception, these strains carried vtx2, the gene possessed by most
pt ed
of the O157:H7 strains circulating now in Europe as well as in other regions of the world (2729). The vtx-negative O157:H7 strain found was not a surprise, knowing that spontaneous loss of the vtx phage(s) during in vitro subcultivation was reported for both VTEC O157 (30)
ce
and non-O157 strains (31). Yet, this finding was a matter of concern from the diagnostic
Ac
point of view.Together with the evidence of O157 strains of other serotypes but H7 circulation. it indicated that presumptive DEC, selected solely on the basis of the slide agglutination with commercially available antisera, should be confirmed and further characterized in the national reference laboratory. Considering that WGS is the pinnacle for strain characterisation and epidemiological analyses, we took the opportunity to extract additional and important information about the autochthonous O157:H7 strains from WGS data. Due to budget constraints, we initially 10
selected the strains which seemed to have circulated more frequently among the Romanian population during the study period, namely the O157:H7 strains belonging to phylogenetic group E. In order to facilitate our WGS approach we made use of open sources of web-based analysis tools to compensate our limited bioinformatics skills. As a matter of fact, the bioinformatics processing and analysis of WGS data was considered by the European Centre
us cr ip
t
for Disease Prevention and Control (ECDC) as a great challenge for the strategy to harness WGS to strengthen outbreak investigations and public health surveillance (32). WGS typing allowed us not only to verify the accuracy of the routine typing approach but to expand the microbiological information that can be used for interlaboratory international comparisons and understand how the autochthonous O157:H7 strains fit among similar strains isolated
M an
elsewhere. VirulenceFinder confirmed that the investigated strains harboured vtx2c in association with intimin γ encoding eae, and ehxA genes, a virulence genotype commonly found in E. coli human strains belonging to this serotype (33, 34) and gave much more information about chromosomally and plasmidic encoded pathogenicity-associated gene
pt ed
content of the strains than the PCR-based routine typing. The WGS MLST performed with the E. coli scheme 1 that employed 7 house-keeping genes (35) showed that 3 of the Romanian O157:H7 strains studied were ST1804, a less
ce
common sequence type according to the data comprised in Escherichia coli EnteroBase (36)
Ac
where only 31 such isolates from the United Kingdom (UK), the United States (US), Spain and Italy are registered. However, the largely spread ST11 clone was also represented by the only vtx-negative O157:H7 strain detected in this study, namely RO_O157_7 strain. Unlike the ST1804 strains, in silico prediction of the GUD activity for this strain showed only the changing amino acid mutations known not to alter the bacterial capacity but not the frameshift mutation caused by G-G insertion at +686 in the uidA gene responsible for the loss of
11
GUD expression (37). This strain possessed a plasmid lacking the katP and espP. O157:H7 strains with similar characteristics were previously reported in US (38) and Japan (39). PFGE provided a better strain differentiation than MLST but when compared with WGS to evaluate the genetic relatedness of the O157:H7 strains, the SNP-based genotyping was able to separate the strains with indistinguishable PFGE profiles. When placed among
us cr ip
t
other strains belonging to the O157:H7 lineage, used as references for genomic comparisons, the autochthonous ST1804 strains formed a more homogeneous cluster with human strains of the same sequence type isolated from other regions, namely from UK and US. The VirulenceFinder results revealed that with one exception, all these ST1804 strains were vtx2c-positive and shared eae, ehxA, espA, espB, espJ, espP, etpD, iha, iss, katP, nleA, nleB,
M an
nleC, tir, and toxB genes. The exception was a British vtx2a-positive strain that possessed the eae, espA, espB, espJ, iha, iss, nleA, nleB, nleC, and tir but not the plasmid-encoded genes ehxA, katP, espP, and toxB. Of note, type III secretion system effector encoding genes espF and tccP were less commonly present among the ST1804 members, these genes being absent
pt ed
from the Romanian strains and also from the ones originated from UK. Comparison of the Romanian ST11 strain with the reference ST11 strains placed the former closer to strains which share a large plasmid lacking katP and espP genes.
ce
Based on the sequencing data and in silico analyses, we were able to integrate the
Ac
autochthonous strains into the global picture of the O157:H7 population and provide the first evidence of the presence of members of lineages II and I/II belonging to clades 7 and 9, respectively, among Romanian human O157:H7 isolates. Clade 7 was reported as one of most common clades in a study on US EHEC O157 isolates (8) with strains significantly more likely to carry vtx2c, the gene also detected in the Romanian clade 7 strains. The American study (8) as well as another one, conducted in Japan (40), showed that while strains of such clades as 6 and 8 were more frequently found in HUS patients, clade 7 strains 12
originated predominantly from less severe cases of diarrhea or even from asymptomatic individuals. Only one of the clade 7 Romanian strains was associated with a HUS patient but the limited set of strains investigated prevented us from determining the clinical relevance of the autochthonous clade 7 pathogens. It is worth noting that the clade 9 strain identified in the present study, a ST11
us cr ip
t
member, possessed an uidA gene without any base alteration that could affect its expression. Manning et al. previously reported that clade 9 included such strains with GUD positive and sorbitol-positive or negative phenotypes (8).
The low number of strains investigated and the lack of consistent patient information to better link the strain profile with disease outcome and epidemiology were perceived as
M an
main limitations of this study. However, the data it provided, resulted from an extensive molecular characterisation conducted up to WGS-level resolution, can be used as reference for future research and comparisons with similar human, food or animal isolates at national or international level. Moreover, given the public health importance of members of E. coli
pt ed
serogroup O157, accurate characterization and typing are critical for the laboratory-based surveillance and should be consistently performed at least at national reference laboratory level.
ce
Funding: This work was supported by The Research Institute of the University of
Ac
Bucharest, Romania (ICUB Fellowships for Young Researchers awarded to Mihaela Oprea, November 2016) and by Cantacuzino National Medical - Military Institute of Research and Development, Bucharest, Romania. Conflict of interest: None to declare
References
13
1. Kaper JB, Nataro JP, Mobley HL. Pathogenic Escherichia coli. Nat Rev Microbiol. 2004;2:123-40. 2. Croxen MA, Finlay BB. Molecular mechanisms of Escherichia coli pathogenicity. Nat Rev Microbiol. 2010;8:26-38. 3. Stevens MP, Frankel GM. The locus of enterocyte effacement and associated virulence
us cr ip
t
factors of enterohemorrhagic Escherichia coli. Microbiol Spectr. 2014;2:EHEC-0007-2013. 4. Werber D, Beutin L, Pichner R, et al. Shiga toxin-producing Escherichia coli serogroups in food and patients, Germany. Emerg Infect Dis. 2008;14:1803-6.
5. Pennington H. Escherichia coli O157. Lancet. 2010;376(9750):1428-35.
6. Yang Z, Kovar J, Kim J, et al Identification of common subpopulations of non-sorbitol-
M an
fermenting, beta-glucuronidase-negative Escherichia coli O157:H7 from bovine production environments and human clinical samples. Appl Environ Microbiol. 2004;70(11):6846-54. 7. Zhang Y, Laing C, Steele M, et al. Genome evolution in major Escherichia coli O157:H7 lineages. BMC Genomics. 2007;8:121.
pt ed
8. Manning SD, Motiwala AS, Springman AC, et al. Variation in virulence among clades of Escherichia coli O157:H7 associated with disease outbreaks. Proc Natl Acad Sci U S A. 2008;105(12):4868-73.
ce
9. European Food Safety Authority, European Centre for Disease Prevention and Control.
Ac
The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2016. EFSA Journal 2017;15:5077. 10. Usein CR, Ciontea AS, Militaru CM, et al. Molecular characterisation of human Shiga toxin-producing Escherichia coli O26 strains: results of an outbreak investigation, Romania, February to August 2016. EuroSurveill. 2017;22(47):17-00148.
14
11. Wang G, Clark CG, Rodgers FG. Detection in Escherichia coli of the genes encoding the major virulence factors, the genes defining the O157:H7 serotype, and components of the type 2 Shiga toxin family by multiplex PCR. J Clin Microbiol. 2002;40:3613-9. 12, Schmidt H, Beutin L, Karch H. Molecular analysis of the plasmid-encoded hemolysin of Escherichia coli O157:H7 strain EDL 933. Infect Immun. 1995;63:1055-61.
us cr ip
t
13. Scheutz F, Teel LD, Beutin L, et al. Multicenter evaluation of a sequence-based protocol for subtyping Shiga toxins and standardizing Stx nomenclature. J Clin Microbiol. 2012;50:2951-63.
14. Clermont O, Christenson JK, Denamur E, et al. The Clermont Escherichia coli phylotyping method revisited: improvement of specificity and detection of new phylo-groups.
M an
Environ Microbiol Rep. 2013;5:58-65.
15. Ribot EM, Fair MA, Gautom R, et al. Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog Dis. 2006;3:59-67.
pt ed
16. Bankevich A, Nurk S, Antipov D, et al. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol. 2012;19:455-77. 17. Joensen KG, Scheutz F, Lund O, et al. Real-time whole-genome sequencing for routine
ce
typing, surveillance, and outbreak detection of verotoxigenic Escherichia coli. J Clin
Ac
Microbiol. 2014;52:1501-10. 18. Joensen KG, Tetzschner AM, Iguchi A, et al. Rapid and easy in silico serotyping of Escherichia coli using whole genome sequencing (WGS) data. J Clin Microbiol. 2015;53:2410-26. 19. Larsen MV, Cosentino S, Rasmussen S, et al. Multilocus sequence typing of total genome sequenced bacteria. J Clin Microbiol. 2012;50:1355-61.
15
20. Kaas R, Leekitcharoenphon P, Aarestrup FM, et al. Solving the problem of comparing whole bacterial genomes across different sequencing platforms. PLoS ONE. 2014;9:e104984. 21. Kumar S, Stecher G, Tamura K. MEGA7: Molecular evolutionary genetics analysis Version 7.0 for bigger datasets. Mol Biol Evol. 2016;33:1870-4. 22. Altschul SF, Gish W, Miller W, et al. Basic local alignment search tool. J Mol Biol.
us cr ip
t
1990;215:403-10.
23. Brunder W, Karch H, Schmidt H. Complete sequence of the large virulence plasmid pSFO157 of the sorbitol-fermenting enterohemorrhagic Escherichia coli O157:H- strain 3072/96. Int J Med Microbiol. 2006;296:467-74.
24. Ferdous M, Zhou K, Mellmann A, et al. Is Shiga toxin-negative Escherichia coli
M an
O157:H7 enteropathogenic or enterohemorrhagic Escherichia coli? Comprehensive molecular analysis using whole-genome sequencing. J Clin Microbiol. 2015;53:3530-8. 25. Thorvaldsdóttir H, Robinson JT, Mesirov JP. Integrative Genomics Viewer (IGV): highperformance genomics data visualization and exploration. Brief Bioinform. 2013;14(2):178-
pt ed
92.
26. Yokoyama E, Hirai S, Hashimoto R, et al. Clade analysis of enterohemorrhagic Escherichia coli serotype O157:H7/H- strains and hierarchy of their phylogenetic
ce
relationships. Infect Genet Evol. 2012;12(8):1724-8.
Ac
27. Fierz L, Cernela N, Hauser E, et al. Characteristics of Shiga toxin-producing Escherichia coli strains isolated during 2010-2014 from human infections in Switzerland. Front Microbiol. 2017;8:1471.
28. Hedican EB, Medus C, Besser JM, et al. Characteristics of O157 versus non-O157 Shiga toxin-producing Escherichia coli infections in Minnesota, 2000-2006. Clin Infect Dis. 2009;49:358-64.
16
29. Leotta GA, Miliwebsky ES, Chinen I, et al. Characterisation of Shiga toxin-producing Escherichia coli O157 strains isolated from humans in Argentina, Australia and New Zealand. BMC Microbiol. 2008;8:46. 30. Schmidt H, Scheef J, Huppertz HI, et al. Escherichia coli O157:H7 and O157:H(-) strains that do not produce Shiga toxin: phenotypic and genetic characterization of isolates
us cr ip
t
associated with diarrhea and hemolytic-uremic syndrome. J Clin Microbiol. 1999;37:3491-6. 31. Bielaszewska M, Prager R, Köck R, et al. Shiga toxin gene loss and transfer in vitro and in vivo during enterohemorrhagic Escherichia coli O26 infection in humans. Appl Environ Microbiol. 2007;73:3144-50.
32. European Centre for Disease Prevention and Control. Expert opinion on whole genome
M an
sequencing for public health surveillance. Stockholm: ECDC; 2016.
33. Beutin L, Krause G, Zimmermann S, et al. Characterization of Shiga toxin-producing Escherichia coli strains isolated from human patients in Germany over a 3-year period. J Clin Microbiol. 2004;42:1099-108.
pt ed
34. Elhadidy M, Elkhatib WF, Elfadl EA, et al. Genetic diversity of Shiga toxin-producing Escherichia coli O157 : H7 recovered from human and food sources. Microbiology. 2015;161:112-9.
ce
35. Wirth T, Falush D, Lan R, et al. Sex and virulence in Escherichia coli: an evolutionary
Ac
perspective. Mol Microbiol 2006;60:1136-51. 36.
EnteroBase.
Escherichia/Shigella.
Available
at
Accessed March 23, 2018. 37. Monday SR, Whittam T, Feng P. Genetic and evolutionary analysis of mutations in the gusA gene that caused the absence of β-glucuronidase activity in Escherichia coli O157:H7. J Infect Dis. 2001;184:918–21.
17
38. Hayes PS, Blom K, Feng P, et al. Isolation and characterization of a β-D-glucuronidaseproducing strain of Escherichia coli O157:H7 in the United States. J Clin Microbiol. 1995;33:3347–8. 39. Nagano H, Okui T, Fujiwara O, et al. Clonal structure of Shiga toxin (Stx)-producing and
us cr ip
sporadic cases in Hokkaido, Japan. J Med Microbiol. 2002;51:405–16.
t
β-D-glucuronidase-positive Escherichia coli O157:H7 strains isolated from outbreaks and
40. Iyoda S, Manning SD, Seto K, et al. Phylogenetic clades 6 and 8 of enterohemorrhagic Escherichia coli O157:H7 with particular stx subtypes are more frequently found in isolates from hemolytic uremic syndrome patients than from asymptomatic carriers. Open Forum Infect Dis. 2014;1(2):ofu061.
M an
Figure caption
Fig. 1. PFGE dendrogram showing the genetic relatedness of the Escherichia coli O157
pt ed
human strains investigated in this study.
Fig. 2. SNP-based phylogenetic tree separated Escherichia coli O157:H7 Romanian strains from one another and from selected published genomes.
ce
The evolutionary history was inferred by using the Maximum Likelihood method, with 1000
Ac
bootstrap replicates. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. There were 5674 positions in the final dataset. Evolutionary analyses were conducted in MEGA7 (21).
18
us cr ip
t
Table 1. Characteristics of the seven Escherichia coli O157 strains investigated in this study
RO_O157_1
RO_O157_2
RO_O157_3
RO_O157_4
RO_O157_5
RO_O157_6 RO_O157_7
Year of isolation
2016
2016
2016
2016
2016
2016
2017
Patient residence Arges
Neamt
Brasov
Iasi
Bucharest
Neamt
Brasov
Specimen
Stool sample
Stool sample
Stool sample
Stool sample
Stool sample
Underlying
HUS
D
D
D
HUS
Stool sample D
Stool sample D
negative
negative
negative
negative
negative
positive
negative
Serotype
O157:H7
O157:H7
O157:H7
O157:H-
O157:H7
O157:H-
O157:H7
Sequence type
ND2)
ST1804
ST1804
ND
ST1804
ND
ST11
Phylogenetic
D
E
E
A
E
B1
E
yes
ND
yes
ND
no
disease1) Sorbitol
GG insertion in uidA gene
ce p
Ac
group
ted
fermentation
ND
M an
Strain ID
yes
vtx2c
vtx2c
-
vtx2c
eae
+
+
+
+
+
ehxA
-
+
+
-
toxB
ND
+
+
katP
ND
+
espA
ND
espB
-
t
vtx1a, vtx2a
+
+
-
+
ND
+
ND
+
+
ND
+
ND
-
+
+
ND
+
ND
+
ND
+
+
ND
+
ND
+
espF
ND
-
-
ND
-
ND
+
espJ
ND
+
+
ND
+
ND
+
espP
ND
+
+
ND
+
ND
-
etpD
ND
+
+
ND
+
ND
+
iha
ND
+
+
ND
+
ND
+
iss
ND
+
+
ND
+
ND
+
gad
ND
-
+
ND
+
ND
+
nleA
ND
+
+
ND
+
ND
+
nleB
ND
+
+
ND
+
ND
+
M an
ted
ce p
us cr ip
-
Ac
Virulence genes3)
vtx
ND
+
+
ND
+
ND
+
tccP
ND
-
-
ND
-
ND
t
+
tir
ND
+
+
ND
+
ND
+
Clade
ND
7
7
ND
7
ND
9
Lineage
ND
II
II
ND
II
ND
I/II
HUS = hemolytic uremic syndrome, D = diarrhea
2)
not determined
M an
1)
us cr ip
nleC
3) vtx, eae and ehxA genes were assessed by PCR and in silico WGS analysis and the remaining genes were assessed exclusively
Ac
ce p
ted
by in silico WGS analysis.
Serotype
beta-
sorbitol-
glucuronidas
fermenting
vtx
ST
e activity O157:H7
-
-
1, 2
11
Sakai
O157:H7
-
-
1, 2
11
Frik2000
O157:H7
-
-
1, 2
11
FDAARGOS_2
O157:H7
-
-
1, 2
Xuzhou21
O157:H7
-
-
FORC_044
O157:H7
-
-
-
E09/10
O157:N
+
M C1-057
ce p
O157:H7
Ac
343
O157:H7
-
Year
GenBank/SRA accession number
USA
1982
CP008957.1
Human
Japan
1996
NC_002695.1
Animal
USA
1991
CP015842.1
11
Human
USA
2014
CP022050.1
1, 2
11
Human
China
1999
NC_017906.1
1, 2
11
Human
South
2015
CP016755
ted
93
Country
Food
M an
EDL933
Source
us cr ip
Strain
t
Table 2. Characteristics of the reference strains used for SNP-based phylogenetic analysis1).
Korea
-
2
11
Human
Holland
2013
LDOZ00000000
+
2
11
Human
Germany
2009
LGBK00000000
-
-
11
Animal
USA
2008
LAZO00000000
+
-
1, 2
11
Human
USA
EC96038
O157:H7
+
+
1, 2
11
Human
Japan
287
O157:N
-
-
-
11
Human
+
+
-
11
Human
+
+
-
11
M 393
O157:N
E09/224
O157:N M
AETX00000000
1996
AMUI00000000
Holland
2013
LGBL00000000
Holland
2013
LGBM0000000
Human
Germany
2009
LGBP00000000
M an
M
1995
t
O157:H7
us cr ip
G5101
O157:H7
-
-
2
1804
Human
UK
2016
SRR5006331
H130660198
O157:H7
-
-
2
1804
Human
UK
2012
SRR1981404
PNUSAE01204
O157:H7
-
-
2
1804
?
USA
?
SRR6667658
O157:H7
-
2
1804
Human
UK
2012
SRR1980991
Ac
H124580172
ce p
5
ted
271122
-
t
ce p
ted
M an
us cr ip
The characteristics of the reference strains were inferred from the information linked to the corresponding databases.
Ac
1)
Table 3. Quality assembly metrics for the sequenced Escherichia coli genomes Strain ID
No. of contigs
N50
227 207 218 229
146610 146871 142561 122706
Total contigs length 5442416 5423585 5344220 5338828
No. of CDS (total) 5838 5904 5854 5780
Ac
ce
pt ed
M an
us cr ip
t
RO_O157_2 RO_O157_3 RO_O157_5 RO_O157_7
Sequence coverage 93,76x 108,7x 103,4x 99,57x
M an
RO_O157_3
ted
RO_O157_2
ce p
t
us cr ip
100
90
80
Strain ID
RO_O157_5
Ac
70
Similarity (%)
RO_O157_7
RO_O157_6
RO_O157_4
RO_O157_1
58 99
us cr ip
t
100 100
271122 RO O157/5 H130660198 RO O157/2 RO O157/3 PNUSAE012045
99
H124580172
100
343 Frik2000
100
100
M an
C1-057 FDAARGOS 293 Xuzhou21 EDL933
100
pt ed
Sakai
100
ce
FORC 044 287
Ac
100
100
100
G5101 EC96038 RO O157/7 100 98
E09/10 393 E09/224
0.05