Skip to main content
Download PDF
  • Research article
  • Open access
  • Published:

Escherichia cryptic clade I is an emerging source of human intestinal pathogens

BMC Biology volume 21, Article number: 81 (2023) Cite this article

Abstract

Background

Within the genus Escherichia, several monophyletic clades other than the traditionally defined species have been identified. Of these, cryptic clade I (C-I) appears to represent a subspecies of E. coli, but due to the difficulty in distinguishing it from E. coli sensu stricto, the population structure and virulence potential of C-I are unclear.

Results

We defined a set of true C-I strains (n = 465), including a Shiga toxin 2a (Stx2a)-producing isolate from a patient with bloody diarrhoea identified by the retrospective analyses using a C-I-specific detection system. Through genomic analysis of 804 isolates from the cryptic clades, including these C-I strains, we revealed their global population structures and the marked accumulation of virulence genes and antimicrobial resistance genes in C-I. In particular, half of the C-I strains contained hallmark virulence genes of Stx-producing E. coli (STEC) and/or enterotoxigenic E. coli (ETEC). We also found the host-specific distributions of virulence genes, which suggests bovines as the potential source of human infections caused by STEC- and STEC/ETEC hybrid-type C-I strains, as is known in STEC.

Conclusions

Our findings demonstrate the emergence of human intestinal pathogens in C-I lineage. To better understand the features of C-I strains and their infections, extensive surveillance and larger population studies of C-I strains are needed. The C-I-specific detection system developed in this study will be a powerful tool for screening and identifying C-I strains.

Background

Escherichia cryptic clade I (C-I) is one of the six cryptic clades proposed in the genus Escherichia [1,2,3]. Among these, clades III and IV have recently been defined as a separate species, E. ruysiae [4] [referred to as Er (C-III) and Er (C-IV) in this manuscript, respectively], and clade V was classified as E. marmotae [5] [Em (C-V)]. A new species, “E. whittamii”, was also recently proposed for clade II [6] [C-II (“Ew”)]. Although the taxonomic position of C-I has not yet been fully defined, several phylogenetic analyses have shown that C-I is closely related to E. coli sensu stricto (E. coli ss) and could represent a subspecies of E. coli [2, 3, 7,8,9]. Based on whole-genome sequence analysis, C-II (“Ew”), Er (C-III/IV), and Em (C-V) have been identified within discrete clusters of 95% average nucleotide identity (ANI) [10, 11], while C-I is grouped with E. coli into a 95% ANI cluster [12].

Among three common Escherichia species, namely, E. coli, E. albertii and E. fergusonii (note that reassignment of E. hermannii [13] to “Atlantibacter hermannii” was recently proposed [14]), E. coli is a genetically and phenotypically diverse group of strains, ranging from nonpathogenic commensal strains to pathogenic strains of various pathotypes that cause intestinal and extraintestinal diseases in humans and animals. Each pathotype possesses specific virulence factors introduced by mobile genetic elements (MGEs) [15]. Although E. albertii and E. fergusonii are not frequently isolated from humans [16], E. albertii has been recently recognized as a human enteric pathogen possessing a type III secretion system (T3SS) similar to that of enteropathogenic E. coli (EPEC) and major Shiga toxin (Stx)-producing E. coli (STEC) [17,18,19]. E. fergusonii causes sporadic cases of bacteraemia, urinary tract infections and diarrhoea, but its virulence determinants remain unknown [16].

While many of the strains belonging to cryptic clades or recently defined Escherichia species are often identified as E. coli in clinical laboratories due to the difficulty in accurately identifying these strains by routine diagnostic biochemical examinations C-II (“Ew”), Er (C-III/IV) and Em (C-V) strains are isolated mainly from environmental samples, such as water, wastewater and soil [2]. In addition, these strains carry no virulence genes or only minor potential virulence genes. In contrast, C-I strains are frequently isolated from the faeces of humans and other animals and sometimes from children with diarrhoea. Several strains producing Stx, heat-labile enterotoxin (LT) and/or heat-stable enterotoxin (ST), the hallmark virulence factors of STEC and enterotoxigenic E. coli (ETEC), respectively [15], have been isolated from animals [1, 2, 20, 21]. C-I strains producing ETEC enterotoxins have also been isolated from human patients [2, 22]. However, the relevance of such C-I strains in the C-I population and the virulence potential of these strains are unknown. The global population structure of C-I has also yet to be elucidated. This lack of knowledge is at least partly due to the difficulty in accurately identifying C-I strains [2]. Although a rapid PCR-based method have been recently developed to identify cryptic clades, the two-step screening strategy employed in this typing method is unsuitable for extensive screening purposes, and a possibility of false identification of C-I has also been indicated [7].

In this study, we performed an extensive database search to accurately define C-I genomes and developed a one-step PCR-based C-I detection system. This system was successfully applied in retrospective analyses to identify C-I strains among strains previously identified as E. coli, including an stx-positive C-I strain isolated from a patient suffering from bloody diarrhoea. We further performed a phylogenomic analysis of accurately defined C-I strains, along with recently defined Escherichia species and other cryptic clades, and showed, in addition to the global population structure of C-I and its phylogenetic relationship to E. coli ss, the marked accumulation of E. coli virulence genes and antimicrobial resistance (AMR) genes in C-I. In particular, half of the C-I strains in the dataset contained hallmark virulence genes of STEC and/or ETEC. These strains were classified into STEC, ETEC and STEC/ETEC hybrid types, and bovines are suggested as a potential source of human infection of STEC- and STEC/ETEC hybrid-type strains. Data showing genetic links between the accumulation of E. coli virulence genes and MGEs in the C-I lineage and the accumulation of more AMR genes in nonintestinal pathogenic strains are also presented.

Results

Identification of C-I genomes in public databases and development of C-I-specific PCR

Since cryptic clades and recently defined Escherichia species are usually identified as E. coli in routine bacterial identification [2], most of their genome sequences were assumed to be registered as E. coli sequences in public databases. The EnteroBase database includes all complete genome sequences and Illumina paired-end reads deposited in the NCBI database and original sequence data uploaded by users for several bacterial genera/species, including Escherichia spp., and raw read sequences are de novo assembled using a standardized pipeline [23]. Genotyping data for Escherichia spp. based on several tools, including ClermonTyping [24] and EzClermont [25], are also available in EnteroBase. To identify strains belonging to cryptic clades and recently defined Escherichia species in the Enterobase database, we obtained a total of 1088 genomes that were not deposited as E. albertii or E. fergusonii and were not assigned to any of the E. coli phylogroups as determined by both ClermonTyping and EzClermont. After filtering out low-quality data, 1065 genomes were used for constructing a core gene-based phylogenomic tree (Additional file 1: Fig. S1). In this tree, all strains categorized as “clade I” (n = 453), “clade II” (n = 5), “clade III” (n = 49), “clade IV” (n = 46) and “clade V” (n = 213) by ClermonTyping formed deep-branching lineages. Strains categorized as “clade III” and “clade IV” strains formed sister lineages as expected, and lineages categorized as “clade IV” and “clade V” included the type strains of Er and Em, respectively. Although strains deposited as C-VI (n = 2) or C-VIII (n = 5) were categorized as E. coli phylogroup F by ClermonTyping, these strains formed two distinct deep-branching lineages (Additional file 1: Fig. S1). Both strains deposited as C-VII (categorized as “clade II” and “unknown” in ClermonTyping) clustered with C-II (“Ew”) strains; thus, they were regarded as C-II (“Ew”) in this study. All the strains categorized as “E or clade I” (n = 133) and almost all the strains categorized as “unknown” (41 out of 43) by ClermonTyping clustered with strains categorized as E. coli phylogroups “A” and “G”, which apparently belong to the E. coli ss lineage. In EzClermont typing, many strains categorized as phylogroups “A” (157 out of 181) and “E” (202 out of 210) were grouped into the C-I lineage. Conversely, a considerable number of strains categorized as “cryptic” (127 out of 194) and “U/cryptic” (111 out of 437) were grouped into the E. coli ss or E. fergusonii lineage. EzClermont and ClermonTyping employ in silico PCR analysis using eight and fifteen primer pairs, respectively [24, 25]. The difference in primer sets was thought to be the primary cause of the variation in accuracy between these two systems for the classification of cryptic clades. Finally, we identified 453 C-I genomes in the Escherichia genomes deposited in EnteroBase. In addition, 6, 50, 46, 213, 2 and 5 genomes were identified as C-II (“Ew”), Er (C-III), Er (C-IV), Em (C-V), C-VI and C-VIII, respectively (Additional file 2: Table S1).

Using these genomes, we performed a pangenome-wide association study to search for C-I-specific genes that can be used for developing a PCR-based C-I detection system for efficient screening of C-I strains. In this analysis, we included 17 C-II genomes that were available in the NCBI database but not in EnteroBase [9] and representative strains of E. albertii, E. fergusonii and eight E. coli phylogroups (A to G) (Additional file 2: Table S1). This analysis identified one gene (1368 bp in length, of unknown function with two secreted effector protein pipB2 domains) that was conserved in nearly all the C-I strains (452 out of 453) and completely absent in the other strains analysed. The C-I-specific gene (locus_tag: EsCd1-10290c_1484 in strain 10,290) was located on putative genomic island, and no homologue gene (> 60% amino acid identity) was detected in other members of the Enterobacteriaceae family. We designed a primer pair for this gene and developed a PCR system to detect C-I strains. We confirmed that this gene is not present in any of the 1134 E. coli strains that we previously isolated from humans and animals and sequenced [26]. To evaluate the specificity of this PCR system, we first performed in silico PCR analyses of these 1134 E. coli strains and confirmed that all yielded negative results. We further performed PCR and in silico PCR analyses of seven C-I strains that were isolated from healthy bovine stools during our abovementioned study on E. coli strains [26] and defined as belonging to C-I based on their genome sequences. In this analysis, all the strains yielded a PCR product of the expected size (Additional file 1: Fig. S2). These analyses confirmed the specificity of the PCR system that we developed. Notably, the seven C-I strains were all stx2-positive; five carried stx2a, and two carried stx2g (Additional file 2: Table S1).

Retrospective screening of C-I strains in the “E. coli” strains isolated from clinical, food and animal samples

Using the PCR-based C-I detection system, we performed two series of retrospective screenings of C-I strains. In the first screening, we analysed strains isolated from human patients with intestinal diseases and identified as E. coli exhibiting minor serotypes or O-untypeable in Miyazaki (n = 41) and Osaka (n = 207) prefectures, Japan, from 1995 to 2014. This screening identified two PCR-positive strains (EC05-109 and 10290). The former was isolated from a patient with diarrhoea in Osaka in 2005 and exhibited the O1:H45 serotype. This strain carried none of the hallmark virulence genes of the five major pathotypes of intestinal pathogenic E. coli, including stx for STEC, elt and est for ETEC, eaeA for EPEC, aggR for enteroaggregative E. coli (EAEC) and ipaH for enteroinvasive E. coli (EIEC). In addition, only one colony was detected in the specimen in routine diagnostic analysis; thus, this strain was unlikely to be the causative agent of the infection. However, strain 10290 showing the O2:H28 serotype, which was isolated from a 27-year-old man who was hospitalized with severe symptoms, including bloody diarrhoea, in Miyazaki in 2002, carried not only stx2a but also esta and elt2 encoding STa and LT2, respectively. No other suspected intestinal pathogens were detected in the specimen from this patient. Furthermore, a strain isolated from the food (a variety of beef products) consumed by the patient exhibited an identical pulsed-field gel electrophoresis pattern as that of strain 10290. Thus, strain 10290 was considered the causative agent of the infection.

In the second screening, we analysed 18 strains exhibiting minor serotypes or O-untypeable strains in our laboratory E. coli collection isolated from foods, animals and the environment between 2007 and 2015 in Japan. This analysis also identified one food and two bovine C-I isolates. These three strains carried stx2a and elt2a, and one strain (strain PV14-161) additionally carried esta4 among the hallmark virulence genes of intestinal pathogenic E. coli.

Re-evaluation of the phylogenetic position of C-I in the genus Escherichia and the characteristic isolation sources and serotypes of these strains in the E. coli O:H serotyping scheme.

Following the genome sequencing of the five C-I strains identified in the PCR screening, we constructed a core gene-based phylogenomic tree of the 465 C-I strains, including the 12 C-I strains that we sequenced and the 339 strains belonging to C-II (“Ew”), Er (C-III/IV), Em (C-V), C-VI or C-VIII obtained from the databases (Fig. 1). We included one representative strain from E. albertii, E. fergusonii and each of eight E. coli phylogroups in this analysis. This analysis reconfirmed that the C-I strains formed a monophyletic sister lineage of E. coli ss. The ANI was over 98.3% within the C-I lineage and over 95.9% between the C-I strains and the eight representative E. coli strains. The C-II (“Ew”), Er (C-III/IV), Em (C-V), C-VI and C-VIII strains also formed a monophyletic lineage, respectively, with Er (C-IV) and Em (C-V) strains representing sister lineages.

Fig. 1
figure 1

Phylogenomic relationships of C-I with E. coli ss, other Escherichia cryptic clades and recently defined Escherichia species and the distribution of virulence genes in these clades and species. The ML tree of 804 isolates from cryptic clades with 10 reference strains was constructed based on 339,491 SNP sites located on 1581 core genes. E. fergusonii ATCC_35469T, E. albertii EC06-170, E. coli MG1655 (phylogroup A), E. coli SE11 (phylogroup B1), E. coli SE15 (phylogroup B2), E. coli APECO78 (phylogroup C), E. coli UMN026 (phylogroup D), E. coli CB9615 (phylogroup E), E. coli SMS-3–5 (phylogroup F) and E. coli PSUO78 (phylogroup G) were included as references. Strains belonging to cryptic clades and recently defined species are highlighted with differently coloured lines. Information about the continents and sources of isolation and the distribution of virulence genes for each strain without references is also provided. Strains for which complete genome sequences and Stx2 production levels were determined are indicated by black and red arrowheads, respectively (1: KS-P079, 2: EC42405, 3: 3057, 4: 89–3506, 5: HH-P024, 6: SI-P041, 7: 10,290, 8: KS-P062 and 9: 2013C-4282)

Full size image

More than half (59.0%) of the C-I strains were isolated in North America, while the others were isolated in various countries on the other 5 continents (Fig. 1 and Additional file 1: Fig. S3a). The majority of the C-I strains (77.3%) were isolated from livestock animals or humans (Additional file 1: Fig. S3b). In in silico serotyping, 99% of the C-I strains were serotyped as known E. coli O or H genotypes (Additional file 1: Fig. S3c). They were classified into 51 different O:H genotypes, among which the highest numbers of strains belonged to OG9:H45 (15%) or ONT:H45 (15%).

In contrast, more than half (51.5% to 95.5%) of the C-II (“Ew”), Er (C-III/IV) and Em (C-V) strains were from aquatic environments or wild animals (Fig. 1 and Additional file 1: Fig. S3b). In in silico serotyping, 38, 54 and 59% of the C-II (“Ew”), Er (C-III) and Er (C-IV) strains, respectively, were untypeable for both E. coli O and H genotypes (Additional file 1: Fig. S3c). While all the Em (C-V) strains were serotyped as known O or H genotypes, it is noteworthy that 99% were H56 with various O genotypes, with a high proportion of strains (52%) showing the ONT:H56 genotype.

Distribution of known E. coli virulence genes in C-I strains

We next analysed the distribution of the abovementioned hallmark virulence genes of the five major pathotypes of intestinal pathogenic E. coli in C-I strains (Figs. 1 and 2a and Additional file 1: Fig. S4). Although eaeA, aggR and ipaH were not detected in any of the C-I strains analysed here, many of the C-I strains (227 strains; accounting for nearly half of the strains analysed) were found to carry at least one of the stx, elt and est genes (Fig. 2a), in sharp contrast to C-II (“Ew”), Er (C-III/IV), Em (C-V), C-VI and C-VIII, in which only three and four strains of Em (C-V) were positive for eaeA and ipaH, respectively (Fig. 1). Although various subtypes of the three hallmark virulence genes were identified, the 227 C-I strains were classified into three types according to the profiles of these hallmark virulence genes: STEC (n = 118), ETEC (n = 47) and STEC/ETEC hybrid (n = 62) types.

Fig. 2
figure 2

Distribution of virulence genes in C-I, other cryptic clades and recently defined Escherichia species. a Pie chart showing the distribution of the STEC and ETEC hallmark virulence genes in C-I strains. b Grid charts showing the numbers of virulence genes other than the hallmark virulence genes identified in C-I, other cryptic clades and recently defined species. C-I strains were grouped into four pathotypes. STEC: STEC-type, ETEC: ETEC-type, Hybrid: STEC/ETEC hybrid-type, Non: nontypeable. c Distribution of virulence genes in C-I strains from humans and livestock animals (bovine, poultry and swine)

Full size image

The presence of STEC/ETEC hybrid-type C-I strains carrying stx and est was previously reported [1, 21]. In this study, STEC/ETEC hybrid-type C-I strains carrying various combinations of stx, elt and est were identified (Fig. 2a). Notably, in the stx-positive C-I strains (STEC- and STEC/ETEC hybrid types), stx2a was most frequently detected among the various subtypes of stx1 and stx2 genes (86 of 180 strains; Additional file 1: Fig. S4). In human and animal ETEC strains, various colonization factors (CFs) have been identified [27, 28]. Although, in the ETEC- and STEC/ETEC hybrid-type C-I strains (est- and/or elt-positive; n = 109), known ETEC colonization factors (CFs) were detected in only 17 strains, two types of K88-like CFs were identified in 41 C-I strains (Additional file 2: Table S1 and see Additional file 3: Supplementary text for details) [21, 27,28,29,30,31,32,33].

Various other known E. coli virulence genes were detected in the C-I strains (Fig. 1). Among the genes analysed, espP (encoding serine protease), ehxA (encoding enterohaemolysin) and ihaA (encoding adhesin) were most frequently detected (23.9 ~ 29.5% of the total C-I strains, Additional file 1: Fig. S4). espP and ehxA were identified on the virulence plasmids of STEC [34], and ihaA was identified on chromosomal integrative elements in STEC [35]. Consistent with these genomic locations, the distribution of these three genes was biased towards stx-positive strains (STEC- and STEC/ETEC hybrid-type strains) (Additional file 1: Fig. S5). In contrast, the distribution of ealA, eatA, katP and paa, which encode a pertussis toxin-like toxin, a serine protease, a catalase and an adhesin, respectively, was biased towards the est- and/or elt-positive strains (ETEC- and STEC/ETEC hybrid-type strains) (Additional file 1: Fig. S5). Despite these biased distributions, various known E. coli virulence genes were apparently accumulated in C-I strains of the STEC and STEC/ETEC hybrid types (Fig. 2b), showing a clear contrast to the strains belonging to C-II (“Ew”), Er (C-III/IV) and Em (C-V). In these lineages, cdt subtypes (encoding cytolethal distending toxins), hes (haemagglutinin), paa (adhesin), pic (serine protease) and sat (autotransporter toxin) were rather frequently detected in Er (C-IV) or Em (C-V) (Fig. 1 and Additional file 1: Fig. S4).

As described above, most of the C-I strains were isolated from humans and livestock animals (especially bovines, poultry and swine) (Additional file 1: Fig. S3b). We thus analysed the difference in the distribution of 36 virulence genes identified in the C-I strains between different hosts (Fig. 2c). Although the difference in numbers of hosts should be considered, approximately half of the genes analysed (16 genes) were much more frequently detected in bovine isolates than in isolates from the other hosts. Among the STEC and ETEC hallmark virulence genes, the stx1a, stx2a, stx2g, elt2a, esta4 and esta5 genes were detected in both human and bovine isolates but more frequently in bovine isolates. In contrast, elt1 and esta1 were specifically distributed in human isolates. estb1 was almost entirely specific to strains from swine.

Consistent with the distribution of the STEC and ETEC hallmark virulence genes in humans and livestock animals, more than 80% of the bovine isolates were classified into STEC-, ETEC- or STEC/ETEC hybrid-type strains, and STEC- and STEC/ETEC hybrid-type strains (i.e. stx-positive strains) were found in only human and bovine isolates (Fig. 3). In contrast, more than half of the swine isolates (14/23) were classified as ETEC-type strains, and only one poultry isolate (1/78) was classified as an ETEC-type strain (carrying estb1).

Fig. 3
figure 3

Distribution of STEC-, ETEC- and STEC/ETEC hybrid-type C-I strains in humans and livestock animals. “Others” includes wild animals (n = 12), companion animals (n = 4), food (n = 21), the environment (n = 28) and aquatic environments (n = 6)

Full size image

Distribution of AMR genes in clade I

Analysis of the distribution of horizontally acquired AMR genes revealed that more than 40% of the C-I strains (205 out of 465) carried at least one AMR gene, and most of these strains (175 out of 205) carried multiple AMR genes (up to 15 genes), again showing a clear difference from the C-II (“Ew”), Er (C-III/C-IV) and Em (C-V) strains (Fig. 4 and Additional file 2: Table S1). The number of C-I strains carrying AMR genes for more than three antibiotic classes was 153 (32.9%), while no such multidrug-resistant (MDR) strains were present in C-II (“Ew”) and Er (C-III), and only 4 (6.5%) and 3 (1.4%) strains of Er (C-IV) and Em (C-V), respectively, were MDR strains (Additional file 2: Table S1). In the C-I strains, a total of 79 AMR genes for 11 antibiotic classes were detected, among which strA, strB, blaTEM-105, sul2 and tetA were widely distributed in C-I and detected in 94 (20.2%), 92 (19.8%), 78 (16.8%), 103 (22.2%) and 110 strains (23.7%), respectively (Additional file 2: Table S1 and Additional file 1: Figs. S6 and S7). Although several C-I sublineages exhibited low AMR gene prevalence, closely related strains did not always have the same AMR gene profile, suggesting repeated acquisitions of AMR genes in the C-I lineage.

Fig. 4
figure 4

Prevalence of strains carrying AMR genes in C-I, other cryptic clades and recently defined Escherichia species. Bar charts indicate the percentage of strains carrying different numbers of AMR genes in each group of strains

Full size image

Among the C-I strains from humans and livestock animals, the distribution patterns of AMR genes showed notable differences between hosts (Additional file 1: Fig. S8), which may be linked to the difference in the usage of antimicrobials. In particular, genes for metallo-β-lactamase and two other carbapenemases were detected only in human isolates, and most extended-spectrum β-lactamase (ESBL) genes (9/10) were detected in human and/or poultry isolates. These strains have been isolated since 2011 (Additional file 2: Table S1).

A more striking difference in the distribution pattern of AMR genes was observed between intestinal pathogenic strains (STEC-, ETEC- or STEC/ETEC hybrid-type strains) and the other strains (Additional file 1: Fig. S9). Strains lacking any of the STEC/ETEC hallmark virulence genes (nonintestinal pathogenic strains) contained more AMR genes than intestinal pathogenic strains; the proportion of strains containing no AMR genes was significantly lower in nonintestinal pathogenic strains (38%) than in the STEC- (86.4%; P = 2.2E − 16) and STEC/ETEC hybrid-type strains (72.6%; P = 9.5E − 07). Conversely, the average number of AMR genes carried by nonintestinal pathogenic strains (3.55 genes) was significantly higher than those in STEC- (0.53 genes; P = 6.7E − 15), ETEC- (2.26 genes; P = 0.033) and STEC/ETEC hybrid-type strains (1.40 genes; P = 9.3E − 05). The abovementioned genes for carbapenemases and ESBLs were also more frequently detected in nonintestinal pathogenic strains than in the other strains (Additional file 1: Fig. S9).

MGEs carrying the STEC and ETEC hallmark virulence genes and AMR genes in C-I

Detailed analyses of MGEs are required to understand the mechanisms underlying the acquisition of virulence and AMR genes, but draft sequences do not provide enough information for such analyses. However, only four completely sequenced C-I genomes were available in the public databases (Additional file 2: Table S2). Moreover, among the STEC and ETEC hallmark virulence genes, only stx1 was identified in two (89–3506 and 2013C-4282) of these completely sequenced genomes (Additional file 2: Table S1). Among virulence plasmids, the complete sequence of an est-encoding plasmid, p7v, was the only one available [23]. Therefore, we determined the complete genome sequences of three strains, namely, a human clinical isolate (10290) and two bovine isolates (KS-P079 and HH-P024) (Additional file 2: Table S2). These strains were selected as they belonged to C-I sublineages different from those of the abovementioned four strains and carried different combinations of STEC and ETEC hallmark virulence genes, including stx2a, stx2g, esta4, esta5, elt2a and elt2c (Fig. 1 and Additional file 2: Table S1).

Similar to STEC strains, the stx genes were encoded by prophages in all five stx-positive C-I strains (Fig. 5a). The Stx phages of the C-I strains exhibited structural similarities to those of STEC strains that are available in publicly databases (see Additional file 3: Supplementary text for details and Additional file 2: Table S3).

Fig. 5
figure 5

Genomic structures of MGEs carrying major virulence genes. a Genomic structures of Stx phages and LT2 phages. The genome structures of Stx2a, Stx2g and LT2 phages of the C-I strains whose complete genome sequences were determined are drawn to scale. Phages of E. coli strains are also shown for comparison. Homologous regions are indicated by shading, and sequence identities are indicated by different colours. Integration sites and sequence lengths of each phage are also indicated. b Circular maps of the virulence plasmids in the C-I strains 10,290 and HH-P024. c Dot-plot matrices of the completed C-I virulence plasmid sequences. Virulence plasmids of ETEC strains (E24377A and H10407) and a STEC O157 strain (Sakai) are also included for comparison

Full size image

The elt2 genes in all three elt2-positive C-I strains were also encoded by prophages that were all integrated into ttcA (Fig. 5a and see Additional file 3: Supplementary text for details). A notable similarity between the C-I LT2 phages was the presence of ealAB genes encoding an ADP-ribosyltransferase toxin family, which was first identified on the LT2 phage of the ETEC strain SA53 [30]. The LT2a phages of two strains (KS-P079 and 10,290) additionally encoded Paa, which is involved in the early step of adhesion to intestinal epithelial cells [36, 37]. The genetic link of these two virulence-related genes with elt2 explains, at least partly, the high prevalence of these two genes in ETEC- and STEC/ETEC hybrid-type C-I strains (Additional file 1: Fig. S5).

A large plasmid in three est-positive C-I strains (10290, KS-P079 and HH-P024) exhibited remarkable similarity to p7v (Fig. 5b, c and see Additional file 3: Supplemental text for details). The esta4 gene, along with the genes for a K88 fimbria-like CF (Additional file 1: Fig. S10 and see Additional file 3: Supplemental text for details) and a type 6 secretion system (T6SS), was encoded by these C-I virulence plasmids (Fig. 5b). The plasmid in 10290/KS-P079/7v additionally encoded the genes for the EspP serine protease, KatP catalase, EatA serine protease, a type 2 secretion system (T2SS) and a partial enterohaemolysin (Ehx). The plasmid in HH-P024 also additionally encoded esta5 and full-length genes for Ehx. Despite these differences, these virulence plasmids of the C-I strains exhibited chimeric features of STEC and ETEC virulence plasmids as previously reported [21]; EspP, KatP, Ehx and T2SS are associated with STEC virulence plasmids, and STs, EatA and the K88 fimbria-like CF are associated with ETEC virulence plasmids. The presence of the STEC/ETEC chimeric plasmids explains the high prevalence of these virulence genes in STEC/ETEC hybrid-type C-I strains (Additional file 1: Fig. S5). However, the backbones of these C-I virulence plasmids completely differed from those of the typical ETEC strains E23487A and H10407 and the typical STEC O157 strain Sakai (Fig. 5c).

Strain 10290 contained an additional large plasmid encoding multiple AMR genes (Additional file 1: Fig. S11). All AMR genes identified in this strain were encoded by this plasmid. The sepA gene encoding Shigella extracellular protein A, which is involved in tissue invasion [38], was also found in this plasmid.

Production levels of Stx2 in stx2-positive C-I strains

A notable number of C-I strains (135 out of 465) were stx2 positive, and Stx2 production is one of the known risk factors for severe STEC infections [39,40,41]. However, it is also known that mitomycin C (MMC)-induced Stx2 production levels vary significantly among STEC strains [32, 39,40,41,42,43,44,45,46]. Therefore, we analysed the Stx2 production levels of five stx2-positive C-I strains belonging to different C-I sublineages (Fig. 1); three were stx2a positive, and two were stx2g positive. The three stx2a-positive strains, including the human clinical isolate 10,290, exhibited high levels of Stx2 production more or less comparable to that of the STEC O157:H7 strain Sakai, which caused a large outbreak in 1996 in Sakai, Japan [35] (Fig. 6). In particular, the level in strain KS-P079 was two times higher than that in strain Sakai. In contrast, of the two stx2g-positive strains, the Stx2 production level in one strain was below the detection limit, and that in the other strain was much lower than those in the stx2a-positive strains.

Fig. 6
figure 6

Stx production levels of C-I strains. Stx concentrations of five stx2a- and two stx2g-positive C-I strains after mitomycin treatment are shown. The STEC strain Sakai carrying stx2a is also included for comparison. The mean values of three measurements are presented. Error bars represent standard errors

Full size image

Discussion

C-I strains producing LT and ST have been isolated from children with diarrhoea, and thus C-I includes at least ETEC-type human pathogens [2]. However, the precise incidence and types of C-I infections remain unclear. This is partly due to the difficulty in distinguishing C-I strains from strains belonging to E. coli ss. In fact, most of the C-I genomes we identified in public databases (366 of 391 NCBI Biosample record-available genomes) were deposited as E. coli genomes. The PCR-based C-I detection system that we developed based on the genomic information of these C-I strains solved this problem, as shown in our retrospective surveys of C-I infections among the strains previously identified as E. coli from various sources. Although only two C-I strains were identified from clinical strains in this survey, it is notable that one of the two strains carried stx2a, elt2a and esta4 and apparently caused a severe intestinal infection with bloody diarrhoea, as there is no report of apparent symptomatic human infection by Stx-producing C-I.

Our genomic analysis of the global C-I strain set including the C-I strains identified in this study along with strains belonging to other cryptic Escherichia clades and recently defined Escherichia species confirmed the phylogenetic position of C-I as a sister lineage of E. coli ss at the subspecies level (> 95% ANI) (Fig. 1). Moreover, our analysis revealed several important findings on the features of the C-I lineage that were not observed in other cryptic clades and recently defined Escherichia species. Among these, the most important finding was the marked accumulation of E. coli virulence genes in C-I, in particular the presence of STEC and/or ETEC hallmark virulence genes in half of the C-I strains in the dataset (Fig. 2a). These strains were classified into three pathotypes, namely, the STEC, ETEC and STEC/ETEC hybrid types, and all the types included notable numbers of human isolates (Fig. 3). This finding, together with the abovementioned cases of human infections caused by ETEC- and STEC/ETEC hybrid-type strains, indicates that a considerable portion of C-I strains are human pathogens.

In this regard, it is noteworthy that among the various subtypes of stx detected in stx-positive C-I strains (180 of 465 strains), the most prevalent was stx2a (86 strains), followed by stx1a (62 strains) (Additional file 1: Fig. S4). STEC strains carrying stx1a and/or stx2a are often associated with human diseases [46], and stx2a-positive strains are most frequently associated with severe diseases, including haemolytic colitis and haemolytic uraemic syndrome [39,40,41]. In addition, although it is known that STEC strains produce various levels of Stx2 [32, 42,43,44,45], which is associated with the variation in the Stx2 phage genome [32], the levels of Stx2 production and the genomic structures of Stx2a phages in the stx2a-positive C-I strains analysed were comparable to those of a representative STEC O157 strain (Figs. 5 and 6). In contrast, as reported in STEC [47,48,49], the Stx2 production levels of C-I strains carrying only stx2g, which was also the stx2 subtype frequently detected in the strain set (33 strains; Additional file 1: Fig. S4), were much lower than those in the stx2a-positive strains or were below the detection limit (Fig. 6). Moreover, all the stx2g-positive strains identified possessed esta4 and/or esta5 (Additional file 2: Table S1), and stx2g-positive STEC strains are rarely isolated from humans [50, 51]. Therefore, it remains unclear to what extent stx2g is responsible for human diseases.

Regarding the pathogenicity of C-I strains, it is also noteworthy that among the three subtypes of elt (elt1, elt2a and elt2c) found in 54 elt-positive C-I strains (Additional file 1: Fig. S4 and Additional file 2: Table S1), elt1, which is predominantly associated with human disease, was detected in 22 strains, and elt2 (elt2a and elt2c), which is mainly associated with diseases in livestock animals [52], was detected in 32 strains. Moreover, all elt1-positive strains were human isolates, and almost all the elt2-positive strains (28/29) were bovine isolates (Fig. 2c). ST is also divided into two types: STa is mainly responsible for human diseases, and STb is associated with diseases in livestock animals, particularly piglets and calves [52]. The est genes detected in 97 C-I strains (esta1, esta4, esta5 and estb1) also exhibited host-specific distributions; esta1 was specifically detected in human isolates (n = 17), and esta4, esta5 and estb1 showed highly biased distributions in bovine and swine isolates (n = 61, 28 and 17, respectively) (Fig. 2c). The distribution patterns of CFs were also apparently host specific (Fig. 2c). These data suggest that C-I is also a source of LT- and/or ST-producing strains pathogenic to human and livestock animals, although the roles of esta4 and esta5 as virulence factors are unknown and disease information for these elt/est-positive livestock animal isolates is not available.

In terms of the potential virulence of C-I strains, it is also noteworthy that the other potential virulence factors of ETEC and STEC were accumulated in intestinal pathogenic C-I strains (Fig. 2b and Additional file 1: Fig. S4). Although the roles of these factors in pathogenesis have not been fully understood even in STEC and ETEC, many of them are genetically linked with MGEs that were found to be associated with ETEC and STEC. A more important finding related to potential sources of human infection caused by these intestinal pathogenic C-I strains is that most of the STEC- and STEC/ETEC hybrid-type strains were bovine isolates, accounting for 77% of the bovine isolates (Fig. 3). This finding suggests that bovines are the major reservoir of STEC- and STEC/ETEC hybrid-type C-I strains and that the source of human infections caused by these strains is similar to that of STEC [53,54,55]. In fact, beef was the most likely infection source in the human case of Stx2a-producing C-I strain infection identified in this study.

In addition to stx, elt and est, various virulence genes were more frequently detected in C-I strains isolated from bovines compare to those isolated from humans and the other animals (Fig. 2c). Previous research has indicated that among bovine commensal E. coli strains, various virulence genes are accumulated in stx-positive strains, which implies that these virulence factors are functionally associated with Stx and likely confer benefits to these E. coli strains as they adapt to the bovine intestinal environment. This phenomenon may also hold true for C-I strains.

It is also of note that although the known E. coli virulence factors were rarely identified in recently defined Escherichia species and other cryptic clades, cdt-IV was identified in many C-V strains (82 out of 213). The cdt was initially discovered in E. coli and has been categorized into five subtypes (cdt-I to cdt-V) [56]. The cdt is only present in a subset of E. coli strains, while most E. albertii strains possess the cdt belonging to the cdt-II/III/V. As E. albertii strains are also frequently isolate from the environment and wild animals, cdt may be necessary to adapt these natural niches in E. albertii and C-V strains.

The accumulation of AMR genes is another important feature of C-I. Many strains possess multiple AMR genes (Fig. 4 and Additional file 1: Figs. S6 and S7), and as many as 32.9% of the C-I strains in our dataset were genotypically MDR stains. The AMR gene-positive strains were widely distributed in the C-I lineage irrespective of isolation source (Additional file 1: Figs. S7 and S8), but nonintestinal pathogenic strains carried more AMR genes (Additional file 1: Fig. S9). These findings suggest the close association of C-I strains, particularly nonintestinal pathogenic strains, with humans and livestock animals. Among the 55 AMR genes detected, the presence of genes for ESBLs and carbapenemases has particularly strong clinical implications considering their wide spectrum. Although the prevalence of such strains was not high, their distribution was apparently biased to human and poultry isolates (Additional file 1: Fig. S8). This may also support the close association of nonintestinal pathogenic C-I strains with humans, for which third- and fourth-generation cephalosporins and carbapenems are increasingly prescribed, although the background underlying the association of these AMR genes with poultry isolates is unclear.

The present study, however, has limitations that should be considered. The primary limitation of our study is the possibility of sampling bias resulting from the employment of a substantial proportion of publicly available genome data. Specifically, in the context of C-I strains, genomic sequencing may be biased towards strains that harbour virulence and AMR genes as well as those that isolated from clinical and veterinary sources, such as humans and livestock animals. This can be attributed to extensive surveillances and large-scale genomic studies of pathogenic E. coli strains. Another limitation of this study is that only a sporadic case of clinically relevant C-I isolate was identified, and no C-I isolate that causes disease in livestock animals was detected. Consequently, the severity and prevalence of C-I infections in humans and animals remain obscure. Nevertheless, despite these limitations, through analyses using a large number of strains that were precisely classified as C-I and other cryptic clades, our study has provided invaluable insights into the global population structure and virulence potential of C-I.

Conclusions

Escherichia cryptic clade I (referred to as C-I in this article) includes human and animal intestinal pathogenic strains of STEC, ETEC or STEC/ETEC hybrid types, and the major source of human infections caused by STEC- and STEC/ETEC hybrid-type strains is most likely bovines. Multiple AMR genes have accumulated in many C-I strains, particularly nonintestinal pathogenic strains, suggesting their close association with humans and livestock animals. To better understand the features of C-I strains and their infections, extensive surveillance of C-I strains and larger population studies using strain sets well supplemented with metadata are needed. Although C-I strains are difficult to distinguish from E. coli ss strains by routine species identification methods, the PCR system highly specific to C-I developed in this study will be a powerful tool for screening and identifying C-I strains.

Methods

Database searches of the genome sequences of C-I strains

To identify strains belonging to C-I, C-II (“Ew”), Er (C-III/IV), Em (C-V) and other cryptic clades in the 170,489 Escherichia spp. genomes available in EnteroBase (last access: 2021/07/31), we excluded those deposited as E. albertii or E. fergusonii and those assigned to any of the E. coli phylogroups by both ClermonTyping and EzClermont, obtaining a total of 1088 genomes, including those of the type strains of Er and Em (strains OPT1704T and HT073016T, respectively). To eliminate low-quality assembly, CheckM was employed to assess the completeness and contamination levels [57] (Additional file 2: Table S1). Genomes with ≥ 90% completeness have been categorized as nearly complete, whereas those with > 10% contamination have been classified as having a high degree of contamination [57]. In this study, we removed 23 genomes that exhibited either < 95% completeness or > 10% contamination. We then classified the remaining 1065 genomes into Escherichia species or cryptic clades by whole-genome-based phylogenomic analysis. Raw read data were also obtained from the Enterobase database if available. Additionally, the genome sequences of a representative strain for E. fergusonii and E. albertii, alongside each of the major E. coli phylogroupes, were obtained from EnteroBase and used as references [ATCC_35469T (E. fergusonii), EC06-170 (E. albertii), MG1655 (E. coli phylogroup A), SE11 (E. coli phylogroup B1), SE15 (E. coli phylogroup B1), APEC O78 (E. coli phylogroup C), UMN026 (E. coli phylogroup D), CB9615 (E. coli phylogroup E), SMS-3–5 (E. coli phylogroup F) and PSUO78 (E. coli phylogroup G)]. The genome sequences of 17 C-II strains [9] that were available in the NCBI database but not in the EnteroBase database were also obtained from the NCBI database. Detailed information for the strains used in this study is provided in Additional file 2: Table S1.

SNP detection and phylogenomic analysis

To construct core gene-based phylogenomic trees, pangenome analyses of each strain set were performed using Roary with the options (-i 80 -cd 100 -s) [58]. SNP sites were extracted from the core gene alignment using snp-sites [59], and ML phylogenomic trees were constructed using RAxML [60] with the GTR-GAMMA model of nucleotide substitution and 500 bootstrap replicates and displayed and annotated using iTOL [61].

Development of the PCR-based C-I detection system

Through a pangenome analysis of 792 genomes of C-I, C-II (“Ew”), Er (C-III/C-IV), Em (C-V), C-VI and C-VIII strains and the reference strains of E. albertii, E. fergusonii and E. coli using Roary [58] (-i 80 -cd 100 -s) and a subsequent pangenome-wide association study using Scoary [62], we identified one gene (1368 bp in length) that was conserved in nearly all C-I strains (452/453) and completely absent in the other strains analysed and designed a C-I-specific primer pair (F: TGCGAAACATGAGACAGTCC, R: TTCGGTAAGCCCATTCTCTG) for this gene. PCR was performed using the primer pairs and KAPATaq polymerase (Kapa Biosystems, Woburn, MA, USA) with a programme of 30 cycles of 20 s at 98 °C, 30 s at 60 °C and 1 min at 72 °C.

Genome sequencing, assembly and annotation

Genomic DNAs of the C-I strains isolated in this study were purified from 1 ml of an overnight culture using the DNeasy Blood and Tissue Kit (Qiagen). Libraries were prepared using the Nextera XT DNA Sample Preparation Kit (Illumina, CA, USA) or the NEBNext Ultra DNA Library Prep Kit for Illumina (NEB) and sequenced on the Illumina MiSeq platform to generate 300-bp paired-end reads. Genome assembly of the Illumina sequence reads was performed using the Platanus_B assembler [63].

The four finished C-I genomes (HH-P024, KS-P079 and 10,290) were obtained by mixed assembly of MiSeq and MinION (Oxford Nanopore Technologies, Oxford, UK) reads using the Unicycler pipeline version 0.4.4 with the default settings [64]. Libraries for long-read sequencing were prepared using the Rapid Barcoding Kit (SQK-RBK004; Oxford Nanopore Technologies, Ltd.) and sequenced using an R9.4.1 flow cell on the Oxford Nanopore MinION platform. Annotation was carried out using the DDBJ Fast Annotation and Submission Tool (DFAST) with the default settings [65]. Comparisons of prophage, plasmid and CS locus sequences were conducted using GenomeMatcher version 3.02 [66].

In silico serotyping and identification of virulence and AMR genes

In silico serotyping was conducted by using ECTyper version 1.0.0 with the default settings [67]. The identification and variant typing of E. coli hallmark virulence genes were performed using an in-house database (Additional file 2: Table S4) by a BLASTN search with a threshold of 99% identity and 90% coverage and a read mapping-based strategy by SRST2 [68] with the options “–max-divergence 1”. Genes detected by either the BLASTN or SRST2 analyses were considered to be present. The other E. coli virulence genes were identified by BLASTN searches using an in-house database (Additional file 2: Table S5) with a threshold of 90% identity and 90% coverage. The identification of genes encoding the known ETEC colonization factors was conducted via BLASTN searches of the scaffold sequences of each strain against the database file ETEC_vir_db (https://github.com/avonm/ETEC_vir_db). The result was considered positive when all genes of a given CF were conserved (> 90% identity and > 80% coverage). AMR genes were identified using ABRicate version 1.0.1 (https://github.com/tseemann/abricate) with the default settings using the CARD (http://www.mgc.ac.cn/VFs) databases.

Stx2 production assay

Bacterial cells were inoculated into 40 ml of lysogeny broth and grown to mid-log phase at 37 °C with shaking. MMC was then added to the culture at a final concentration of 500 ng/ml. At 5 h after MMC addition, 100 µl of the culture was collected and immediately subjected to sonication using a Bioruptor (Cosmo Bio, Tokyo, Japan). The soluble fractions of each cell lysate were obtained via centrifugation at 14,000 × g at 4 °C for 10 min. The Stx2 concentration in each cell lysate was determined using a previously described sandwich ELISA [69] using RIDASCREEN Verotoxin microtiter plates (R-Biopharm GmbH, Germany) coated with capture antibodies that recognize Stx2 and monoclonal antibodies against Stx2 (LSBio, WA, USA) conjugated with horseradish peroxidase using a Peroxidase Labeling Kit–NH2 (Dojindo, Kumamoto, Japan) as detection antibodies. Note that both Stx2a and Stx2g could be quantified at a certain level of accuracy using the detection antibodies [70].

Statistical analysis

All statistical analyses were performed using R version 3.6.1 (R Project for Statistical Computing). The prop test was conducted to examine the differences in the proportion of strains containing no AMR gene between nonintestinal pathogenic strains and intestinal pathogenic strains (STEC-, ETEC- or STEC/ETEC hybrid-type strains). The t test was used to examine the differences in the average number of AMR genes between strains.

Availability of data and materials

All sequence data generated in this study are available in the DDBJ/EMBL/GenBank BioProject under accession number PRJDB7924 [71].

Abbreviations

STEC:

Stx-producing E. coli

ETEC:

Enterotoxigenic E. coli

EPEC:

Enteropathogenic E. coli

EAEC:

Enteroaggregative E. coli

EIEC:

Enteroinvasive E. coli

ANI:

Average nucleotide identity

MGE:

Mobile genetic element

T2SS:

Type 2 secretion system

T3SS:

Type III secretion system

T6SS:

Type 6 secretion system

LT:

Heat-labile enterotoxin

ST:

Heat-stable enterotoxin

AMR:

Antimicrobial resistance

CF:

Colonization factors

ESBL:

Extended-spectrum β-lactamase

References

  1. Gangiredla J, Mammel MK, Barnaba TJ, Tartera C, Gebru ST, Patel IR, et al. Draft genome sequences of Escherichia albertii, Escherichia fergusonii, and strains belonging to six cryptic lineages of Escherichia spp. Genome Announc. 2018;6(18):e00271-18.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Walk ST. The “Cryptic” Escherichia. EcoSal Plus. 2015;6(2). https://journals.asm.org/doi/10.1128/ecosalplus.ESP-0002-2015.

  3. Walk ST, Alm EW, Gordon DM, Ram JL, Toranzos GA, Tiedje JM, et al. Cryptic lineages of the genus Escherichia. Appl Environ Microbiol. 2009;75(20):6534–44.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. van der Putten BCL, Matamoros S, Mende DR, Scholl ER, Consortium C, Schultsz C. Escherichia ruysiae sp. nov., a novel Gram-stain-negative bacterium, isolated from a faecal sample of an international traveller. Int J Syst Evol Microbiol. 2021;71(2):004609.

    PubMed  PubMed Central  Google Scholar 

  5. Liu S, Jin D, Lan R, Wang Y, Meng Q, Dai H, et al. Escherichia marmotae sp. nov., isolated from faeces of Marmota himalayana. Int J Syst Evol Microbiol. 2015;65(7):2130–4.

    Article  CAS  PubMed  Google Scholar 

  6. Gilroy R, Ravi A, Getino M, Pursley I, Horton DL, Alikhan NF, et al. Extensive microbial diversity within the chicken gut microbiome revealed by metagenomics and culture. PeerJ. 2021;9: e10941.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Clermont O, Gordon DM, Brisse S, Walk ST, Denamur E. Characterization of the cryptic Escherichia lineages: rapid identification and prevalence. Environ Microbiol. 2011;13(9):2468–77.

    Article  PubMed  Google Scholar 

  8. Luo C, Walk ST, Gordon DM, Feldgarden M, Tiedje JM, Konstantinidis KT. Genome sequencing of environmental Escherichia coli expands understanding of the ecology and speciation of the model bacterial species. Proc Natl Acad Sci U S A. 2011;108(17):7200–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Shen Z, Koh XP, Yu Y, Woo CF, Tong Y, Lau SCK. Draft genome sequences of 16 strains of Escherichia Cryptic Clade II isolated from intertidal sediment in Hong Kong. Microbiol Resour Announc. 2019;8(29):e00416-19.

  10. Chun J, Oren A, Ventosa A, Christensen H, Arahal DR, da Costa MS, et al. Proposed minimal standards for the use of genome data for the taxonomy of prokaryotes. Int J Syst Evol Microbiol. 2018;68(1):461–6.

    Article  CAS  PubMed  Google Scholar 

  11. Goris J, Konstantinidis KT, Klappenbach JA, Coenye T, Vandamme P, Tiedje JM. DNA-DNA hybridization values and their relationship to whole-genome sequence similarities. Int J Syst Evol Microbiol. 2007;57(Pt 1):81–91.

    Article  CAS  PubMed  Google Scholar 

  12. Achtman M, Zhou Z, Charlesworth J, Baxter L. EnteroBase: hierarchical clustering of 100 000s of bacterial genomes into species/subspecies and populations. Philos Trans R Soc Lond B Biol Sci. 1861;2022(377):20210240.

    Google Scholar 

  13. Brenner DJ, Davis BR, Steigerwalt AG, Riddle CF, McWhorter AC, Allen SD, et al. Atypical biogroups of Escherichia coli found in clinical specimens and description of Escherichia hermannii sp. nov. J Clin Microbiol. 1982;15(4):703–13.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Hata H, Natori T, Mizuno T, Kanazawa I, Eldesouky I, Hayashi M, et al. Phylogenetics of family Enterobacteriaceae and proposal to reclassify Escherichia hermannii and Salmonella subterranea as Atlantibacter hermannii and Atlantibacter subterranea gen. nov., comb. nov. Microbiol Immunol. 2016;60(5):303–11.

    Article  CAS  PubMed  Google Scholar 

  15. Kaper JB, Nataro JP, Mobley HL. Pathogenic Escherichia coli. Nat Rev Microbiol. 2004;2(2):123–40.

    Article  CAS  PubMed  Google Scholar 

  16. Gaastra W, Kusters JG, van Duijkeren E, Lipman LJA. Escherichia fergusonii. Vet Microbiol. 2014;172(1–2):7–12.

    Article  CAS  PubMed  Google Scholar 

  17. Inglis TJJ, Merritt AJ, Bzdyl N, Lansley S, Urosevic MN. First bacteraemic human infection with Escherichia albertii. New Microbes New Infect. 2015;8:171–3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Ooka T, Seto K, Kawano K, Kobayashi H, Etoh Y, Ichihara S, et al. Clinical significance of Escherichia albertii. Emerg Infect Dis. 2012;18(3):488–92.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Ooka T, Tokuoka E, Furukawa M, Nagamura T, Ogura Y, Arisawa K, et al. Human gastroenteritis outbreak associated with Escherichia albertii. Japan Emerg Infect Dis. 2013;19(1):144–6.

    Article  PubMed  Google Scholar 

  20. Leung PH, Peiris JS, Ng WW, Robins-Browne RM, Bettelheim KA, Yam WC. A newly discovered verotoxin variant, VT2g, produced by bovine verocytotoxigenic Escherichia coli. Appl Environ Microbiol. 2003;69(12):7549–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Leonard SR, Mammel MK, Rasko DA, Lacher DW. Hybrid Shiga toxin-producing and enterotoxigenic Escherichia sp. Cryptic Lineage 1 Strain 7v Harbors a Hybrid Plasmid. Appl Environ Microbiol. 2016;82(14):4309–19.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Steinsland H, Lacher DW, Sommerfelt H, Whittam TS. Ancestral lineages of human enterotoxigenic Escherichia coli. J Clin Microbiol. 2010;48(8):2916–24.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Zhou Z, Alikhan NF, Mohamed K, Fan Y, Agama Study G, Achtman M. The EnteroBase user’s guide, with case studies on Salmonella transmissions, Yersinia pestis phylogeny, and Escherichia core genomic diversity. Genome Res. 2020;30(1):138–52.

    Article  Google Scholar 

  24. Beghain J, Bridier-Nahmias A, Le Nagard H, Denamur E, Clermont O. ClermonTyping: an easy-to-use and accurate in silico method for Escherichia genus strain phylotyping. Microb Genom. 2018;4(7):2.

    Google Scholar 

  25. Waters NR, Abram F, Brennan F, Holmes A, Pritchard L. Easy phylotyping of Escherichia coli via the EzClermont web app and command-line tool. Access Microbiol. 2020;2(9):acmi000143.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Arimizu Y, Kirino Y, Sato MP, Uno K, Sato T, Gotoh Y, et al. Large-scale genome analysis of bovine commensal Escherichia coli reveals that bovine-adapted E. coli lineages are serving as evolutionary sources of the emergence of human intestinal pathogenic strains. Genome Res. 2019;29(9):1495–505.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. von Mentzer A, Connor TR, Wieler LH, Semmler T, Iguchi A, Thomson NR, et al. Identification of enterotoxigenic Escherichia coli (ETEC) clades with long-term global distribution. Nat Genet. 2014;46(12):1321–6.

    Article  Google Scholar 

  28. Dubreuil JD, Isaacson RE, Schifferli DM. Animal Enterotoxigenic Escherichia coli. EcoSal Plus. 2016;7(1). https://journals.asm.org/doi/abs/10.1128/ecosalplus.ESP-0006-2016.

  29. Allue-Guardia A, Koenig SSK, Martinez RA, Rodriguez AL, Bosilevac JM, Feng P, et al. Pathogenomes and variations in Shiga toxin production among geographically distinct clones of Escherichia coli O113:H21. Microb Genom. 2022;8(4): 000796.

    CAS  PubMed  PubMed Central  Google Scholar 

  30. Jobling MG. The chromosomal nature of LT-II enterotoxins solved: a lambdoid prophage encodes both LT-II and one of two novel pertussis-toxin-like toxin family members in type II enterotoxigenic Escherichia coli. Pathog Dis. 2016;74(3):ftw001.

    Article  PubMed  Google Scholar 

  31. Lindsey RL, Knipe K, Rowe L, Garcia-Toledo L, Loparev V, Juieng P, et al. Complete genome sequences of two Shiga toxin-producing Escherichia coli strains from serotypes O119:H4 and O165:H25. Genome Announc. 2015;3(6):e01496-15.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Ogura Y, Mondal SI, Islam MR, Mako T, Arisawa K, Katsura K, et al. The Shiga toxin 2 production level in enterohemorrhagic Escherichia coli O157:H7 is correlated with the subtypes of toxin-encoding phage. Sci Rep. 2015;5:16663.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Wells JG, Davis BR, Wachsmuth IK, Riley LW, Remis RS, Sokolow R, et al. Laboratory investigation of hemorrhagic colitis outbreaks associated with a rare Escherichia coli serotype. J Clin Microbiol. 1983;18(3):512–20.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Makino K, Ishii K, Yasunaga T, Hattori M, Yokoyama K, Yutsudo CH, et al. Complete nucleotide sequences of 93-kb and 3.3-kb plasmids of an enterohemorrhagic Escherichia coli O157:H7 derived from Sakai outbreak. DNA Res. 1998;5(1):1–9.

    Article  CAS  PubMed  Google Scholar 

  35. Hayashi T, Makino K, Ohnishi M, Kurokawa K, Ishii K, Yokoyama K, et al. Complete genome sequence of enterohemorrhagic Escherichia coli O157:H7 and genomic comparison with a laboratory strain K-12. DNA Res. 2001;8(1):11–22.

    Article  CAS  PubMed  Google Scholar 

  36. An H, Fairbrother JM, Desautels C, Harel J. Distribution of a novel locus called Paa (porcine attaching and effacing associated) among enteric Escherichia coli. Adv Exp Med Biol. 1999;473:179–84.

    Article  CAS  PubMed  Google Scholar 

  37. Batisson I, Guimond MP, Girard F, An H, Zhu C, Oswald E, et al. Characterization of the novel factor paa involved in the early steps of the adhesion mechanism of attaching and effacing Escherichia coli. Infect Immun. 2003;71(8):4516–25.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Benjelloun-Touimi Z, Sansonetti PJ, Parsot C. SepA, the major extracellular protein of Shigella flexneri: autonomous secretion and involvement in tissue invasion. Mol Microbiol. 1995;17(1):123–35.

    Article  CAS  PubMed  Google Scholar 

  39. Bielaszewska M, Mellmann A, Bletz S, Zhang W, Kock R, Kossow A, et al. Enterohemorrhagic Escherichia coli O26:H11/H-: a new virulent clone emerges in Europe. Clin Infect Dis. 2013;56(10):1373–81.

    Article  CAS  PubMed  Google Scholar 

  40. Dallman TJ, Ashton PM, Byrne L, Perry NT, Petrovska L, Ellis R, et al. Applying phylogenomics to understand the emergence of Shiga-toxin-producing Escherichia coli O157:H7 strains causing severe human disease in the UK. Microb Genom. 2015;1(3):e000029.

    PubMed  PubMed Central  Google Scholar 

  41. Persson S, Olsen KE, Ethelberg S, Scheutz F. Subtyping method for Escherichia coli shiga toxin (verocytotoxin) 2 variants and correlations to clinical manifestations. J Clin Microbiol. 2007;45(6):2020–4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. de Sablet T, Bertin Y, Vareille M, Girardeau JP, Garrivier A, Gobert AP, et al. Differential expression of stx2 variants in Shiga toxin-producing Escherichia coli belonging to seropathotypes A and C. Microbiology (Reading). 2008;154(Pt 1):176–86.

    Article  PubMed  Google Scholar 

  43. Kawano K, Okada M, Haga T, Maeda K, Goto Y. Relationship between pathogenicity for humans and stx genotype in Shiga toxin-producing Escherichia coli serotype O157. Eur J Clin Microbiol Infect Dis. 2008;27(3):227–32.

    Article  CAS  PubMed  Google Scholar 

  44. Kulasekara BR, Jacobs M, Zhou Y, Wu Z, Sims E, Saenphimmachak C, et al. Analysis of the genome of the Escherichia coli O157:H7 2006 spinach-associated outbreak isolate indicates candidate genes that may enhance virulence. Infect Immun. 2009;77(9):3713–21.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Zangari T, Melton-Celsa AR, Panda A, Smith MA, Tatarov I, De Tolla L, et al. Enhanced virulence of the Escherichia coli O157:H7 spinach-associated outbreak strain in two animal models is associated with higher levels of Stx2 production after induction with ciprofloxacin. Infect Immun. 2014;82(12):4968–77.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Melton-Celsa AR. Shiga Toxin (Stx) Classification, structure, and function. Microbiol Spectr. 2014;2(4):EHEC-0024-2013.

    Article  Google Scholar 

  47. Bai X, Zhang J, Ambikan A, Jernberg C, Ehricht R, Scheutz F, et al. Molecular characterization and comparative genomics of clinical hybrid Shiga toxin-producing and enterotoxigenic Escherichia coli (STEC/ETEC) Strains in Sweden. Sci Rep. 2019;9(1):5619.

    Article  PubMed  PubMed Central  Google Scholar 

  48. Granobles Velandia CV, Kruger A, Parma YR, Parma AE, Lucchesi PM. Differences in Shiga toxin and phage production among stx(2g)-positive STEC strains. Front Cell Infect Microbiol. 2012;2:82.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Prager R, Fruth A, Busch U, Tietze E. Comparative analysis of virulence genes, genetic diversity, and phylogeny of Shiga toxin 2g and heat-stable enterotoxin STIa encoding Escherichia coli isolates from humans, animals, and environmental sources. Int J Med Microbiol. 2011;301(3):181–91.

    Article  CAS  PubMed  Google Scholar 

  50. De Rauw K, Jacobs S, Pierard D. Twenty-seven years of screening for Shiga toxin-producing Escherichia coli in a university hospital. Brussels, Belgium, 1987–2014. PLoS One. 2018;13(7):e0199968.

    Article  PubMed  PubMed Central  Google Scholar 

  51. Melton-Celsa AR, O'Brien AD. New therapeutic developments against Shiga toxin-producing Escherichia coli. In: Environmental Microbial Forensics. Washington, DC: American Society for Microbiology; 2015. pp. 341–358.

  52. Mainil J. Escherichia coli virulence factors. Vet Immunol Immunopathol. 2013;152(1–2):2–12.

    Article  CAS  PubMed  Google Scholar 

  53. Beutin L, Geier D, Steinruck H, Zimmermann S, Scheutz F. Prevalence and some properties of verotoxin (Shiga-like toxin)-producing Escherichia coli in seven different species of healthy domestic animals. J Clin Microbiol. 1993;31(9):2483–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Kolenda R, Burdukiewicz M, Schierack P. A systematic review and meta-analysis of the epidemiology of pathogenic Escherichia coli of calves and the role of calves as reservoirs for human pathogenic E. coli. Front Cell Infect Microbiol. 2015;5:23.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Dallman TJ, Jalava K, Verlander NQ, Gally D, Jenkins C, Godbole G, et al. Identification of domestic reservoirs and common exposures in an emerging lineage of Shiga toxin-producing Escherichia coli O157:H7 in England: a genomic epidemiological analysis. Lancet Microbe. 2022;3(8):e606–15.

    Article  CAS  PubMed  Google Scholar 

  56. Gomes TAT, Ooka T, Hernandes RT, Yamamoto D, Hayashi T. Escherichia albertii Pathogenesis. EcoSal Plus. 2020;9(1). https://journals.asm.org/doi/10.1128/ecosalplus.ESP-0015-2019.

  57. Parks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res. 2015;25(7):1043–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Page AJ, Cummins CA, Hunt M, Wong VK, Reuter S, Holden MT, et al. Roary: rapid large-scale prokaryote pan genome analysis. Bioinformatics. 2015;31(22):3691–3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Page AJ, Taylor B, Delaney AJ, Soares J, Seemann T, Keane JA, et al. SNP-sites: rapid efficient extraction of SNPs from multi-FASTA alignments. Microb Genom. 2016;2(4):e000056.

    PubMed  PubMed Central  Google Scholar 

  60. Stamatakis A. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics. 2006;22(21):2688–90.

    Article  CAS  PubMed  Google Scholar 

  61. Letunic I, Bork P. Interactive tree of life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees. Nucleic Acids Res. 2016;44(W1):W242–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  62. Brynildsrud O, Bohlin J, Scheffer L, Eldholm V. Rapid scoring of genes in microbial pan-genome-wide association studies with Scoary. Genome Biol. 2016;17(1):238.

    Article  PubMed  PubMed Central  Google Scholar 

  63. Kajitani R, Yoshimura D, Ogura Y, Gotoh Y, Hayashi T, Itoh T. Platanus_B: an accurate de novo assembler for bacterial genomes using an iterative error-removal process. DNA Res. 2020;27(3):dsaa014.

    Article  PubMed  PubMed Central  Google Scholar 

  64. Wick RR, Judd LM, Gorrie CL, Holt KE. Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol. 2017;13(6):e1005595.

    Article  PubMed  PubMed Central  Google Scholar 

  65. Tanizawa Y, Fujisawa T, Nakamura Y. DFAST: a flexible prokaryotic genome annotation pipeline for faster genome publication. Bioinformatics. 2018;34(6):1037–9.

    Article  CAS  PubMed  Google Scholar 

  66. Ohtsubo Y, Ikeda-Ohtsubo W, Nagata Y, Tsuda M. GenomeMatcher: a graphical user interface for DNA sequence comparison. BMC Bioinformatics. 2008;9:376.

    Article  PubMed  PubMed Central  Google Scholar 

  67. Bessonov K, Laing C, Robertson J, Yong I, Ziebell K, Gannon VPJ, et al. ECTyper: in silico Escherichia coli serotype and species prediction from raw and assembled whole-genome sequence data. Microb Genom. 2021;7(12):000728.

    CAS  PubMed  PubMed Central  Google Scholar 

  68. Inouye M, Dashnow H, Raven LA, Schultz MB, Pope BJ, Tomita T, et al. SRST2: Rapid genomic surveillance for public health and hospital microbiology labs. Genome Med. 2014;6(11):90.

    Article  PubMed  PubMed Central  Google Scholar 

  69. Ishijima N, Lee KI, Kuwahara T, Nakayama-Imaohji H, Yoneda S, Iguchi A, et al. Identification of a New Virulent Clade in Enterohemorrhagic Escherichia coli O26:H11/H- Sequence Type 29. Sci Rep. 2017;7:43136.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  70. Nakamura K, Tokuda C, Arimitsu H, Etoh Y, Hamasaki M, Deguchi Y, et al. Development of a homogeneous time-resolved FRET (HTRF) assay for the quantification of Shiga toxin 2 produced by E. coli. PeerJ. 2021;9:e11871.

    Article  PubMed  PubMed Central  Google Scholar 

  71. Okuno M, Arimizu Y, Miyahara S, Wakabayashi Y, Gotoh Y, Yoshino S, et al. Genomic analysis of cryptic Escherichia clade I. GenBank https://www.ncbi.nlm.nih.gov/bioproject/727999 (2021).

Download references

Acknowledgements

We thank Mai Horiguchi, Yuriko Satoh and Kie Ozaki for providing technical assistance.

Funding

This work was supported by KAKENHI (20H03486 and 20K21527 to Y.O.) from the Japan Society for the Promotion of Science (JSPS).

Author information

Author notes

  1. Miki Okuno and Yoko Arimizu contributed equally.

Authors and Affiliations

  1. Division of Microbiology, Department of Infectious Medicine, Kurume University School of Medicine, Kurume, Fukuoka, 830-0011, Japan

    Miki Okuno, Takeshi Yamamoto & Yoshitoshi Ogura

  2. Department of Bacteriology, Graduate School of Medical Sciences, Kyushu University, Fukuoka, 812-8582, Japan

    Yoko Arimizu, Yasuhiro Gotoh, Keiji Nakamura & Tetsuya Hayashi

  3. Department of Infectious Disease, National Hospital Organization Kyushu Medical Center, Fukuoka, 810-0065, Japan

    Yoko Arimizu

  4. Department of Microbiology, Miyazaki Prefectural Institute for Public Health and Environment, Miyazaki, 889-2155, Japan

    Seina Miyahara & Shuji Yoshino

  5. Division of Microbiology, Osaka Institute of Public Health, Osaka, 537-0025, Japan

    Yuki Wakabayashi & Tetsuya Harada

  6. Division of Planning, Osaka Institute of Public Health, Osaka, 537-0025, Japan

    Kazuko Seto

Authors
  1. Miki Okuno
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Yoko Arimizu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Seina Miyahara
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Yuki Wakabayashi
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Yasuhiro Gotoh
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Shuji Yoshino
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Tetsuya Harada
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Kazuko Seto
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Takeshi Yamamoto
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Keiji Nakamura
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Tetsuya Hayashi
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Yoshitoshi Ogura
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Y.O. and T.H2. designed the study. S.M., Y.W., S.Y., T.H1. and K.S. collected the samples. M.O., Y.A., Y.G., T.Y. and K.N. performed the experiments and analysed the data. M.O., T.H2. and Y.O. wrote the manuscript. The author(s) read and approved the final manuscript.

Corresponding author

Correspondence to Yoshitoshi Ogura.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1:

Figure S1. Core gene-based phylogenomic tree of the strains in EnteroBase that were suspected to belong to Escherichia cryptic clades and recently defined Escherichia species. Figure S2. Agarose gel electrophoresis analysis of the products obtained by C-I detection PCR. Figure S3. Continents, sources and serotype distributions of strains belonging to C-I, other cryptic clades and recently defined Escherichia species. Figure S4. Summary of the prevalence of virulence genes in C-I, other cryptic clades and recently defined Escherichia species. Figure S5. Distribution of AMR genes in STEC-type, ETEC-type, STEC/ETEC hybrid-type and nonintestinal pathogenic C-I strains. Figure S6. Summary of the prevalence of AMR genes in C-I. Figure S7. Distribution of AMR genes in C-I, other cryptic clades and recently defined Escherichia species. Figure S8. Summary of the prevalence of AMR genes in C-I strains from humans and livestock animals. Figure S9. Distribution of AMR genes in STEC-type, ETEC-type, STEC/ETEC hybrid-type and nonintestinal pathogenic C-I strains. Figure S10. Comparison of genes in the K88-like CF loci with those in the K88 CS locus. Figure S11. Circular map of the plasmid encoding multiple AMR genes and the sepA gene in the C-I strain 10290.

Additional file 2: Table S1.

Characteristics of strains used in this study. Table S2. Length of chromosome and plasmids in cryptic clade I strains in which the genome sequence was compldetely determined. Table S3. Alignment coverages of the completely sequenced C-I phages and plasmids encoding major virulence genes with the scaffold sequences of the C-I strains. Table S4. List of the nucleotide sequences of the major virulence genes used in this study. Table S5. List of the nucleotide sequences of E. coli virulence genes used in this study.

Additional file 3:

Supplementary Text. Colonization factors in C-I strains and structural comparisons of the Stx and LT phages and the virulence plasmids in C-I strains.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Okuno, M., Arimizu, Y., Miyahara, S. et al. Escherichia cryptic clade I is an emerging source of human intestinal pathogens. BMC Biol 21, 81 (2023). https://doi.org/10.1186/s12915-023-01584-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12915-023-01584-4

Keywords

BMC Biology

ISSN: 1741-7007

Contact us