Strain belonging to an emerging, virulent sublineage of ST131 Escherichia coli isolated in fresh spinach, suggesting that ST131 may be transmissible through agricultural products

Abstract

Food contamination with pathogenic Escherichia coli can cause severe disease. Here, we report the isolation of a multidrug resistant strain (A23EC) from fresh spinach. A23EC belongs to subclade C2 of ST131, a virulent clone of Extraintestinal Pathogenic E. coli (ExPEC). Most A23EC virulence factors are concentrated in three pathogenicity islands. These include PapGII, a fimbrial tip adhesin linked to increased virulence, and CsgA and CsgB, two adhesins known to facilitate spinach leaf colonization. A23EC also bears TnMB1860, a chromosomally-integrated transposon with the demonstrated potential to facilitate the evolution of carbapenem resistance among non-carbapenemase-producing enterobacterales. This transposon consists of two IS26-bound modular translocatable units (TUs). The first TU carries aac(6')-lb-cr, bla _OXA-1, ΔcatB3, aac(3)-lle, and tmrB, and the second one harbors bla _CXT-M-15. A23EC also bears a self-transmissible plasmid that can mediate conjugation at 20°C and that has a mosaic IncF [F(31,36):A(4,20):B1] and Col156 origin of replication. Comparing A23EC to 86 additional complete ST131 sequences, A23EC forms a monophyletic cluster with 17 other strains that share the following four genomic traits: (1) virotype E (papGII+); (2) presence of a PAI II₅₃₆-like pathogenicity island with an additional cnf1 gene; (3) presence of chromosomal TnMB1860; and (4) frequent presence of an F(31,36):A(4,20):B1 plasmid. Sequences belonging to this cluster (which we named "C2b sublineage") are highly enriched in septicemia samples and their associated genetic markers align with recent reports of an emerging, virulent sublineage of the C2 subclade, suggesting significant pathogenic potential. This is the first report of a ST131 strain belonging to subclade C2 contaminating green leafy vegetables. The detection of this uropathogenic clone in fresh food is alarming. This work suggests that ST131 continues to evolve, gaining selective advantages and new routes of transmission. This highlights the pressing need for rigorous epidemiological surveillance of ExPEC in vegetables with One Health perspective.

Keywords: ExPEC; ST131; conjugative transfer; food safety; mobile genetic elements; virulence.

Figure 1

Distribution of selected genetic elements across ST131 sublineages. Cladogram showing the phylogenetic relationships between the 87 genomes included in this study. The first column indicates the source of isolation. The second column indicates the allelic variant of FimH identified. The third column indicates the serotype. The fourth column shows lists the virotype assigned. The 5th to 8th columns indicate the virulence genes that define virotype E. The 9th column indicates the tRNA insertion site of the pathogenicity island harbouring cnf1, papGII and hly genes. The 10th shows the presence of the compound transposon TnMB1860. Columns 11-14 show the presence of ΔcatB3 and CTX-M-15 and whether they are found in a plasmid or in the chromosome). The 15th column indicates the presence of plasmid (F31,36:A4,20:B1). Next column shows presence of the Col156 replicon. The last three columns indicate the presence the other frequent F plasmid replicons (those with a frequency of occurrence > 5 plasmids) listed according to their pMLST classification. The asterisk next to the GenBank accession number indicates strains contained in the study of Biggel et al., 2022 and those with a small triangle belong to the L1 sublineage. The circle indicates strains included in the study of Shropshire et al., 2021. This graphical representation was generated using the CSI Phylogeny platform with EC598 (NZ_HG941718.1) as the reference genome.

Figure 2

Virulome of E. coli A23EC. The genome is shown, with GC content, + and – skew in the inner circles. The outer circle indicates the location of individual virulence genes, labelled in their genomic positions. The labels corresponding to virulence factors serving as molecular markers of uropathogenic strains are boxed in green. (A) Chromosomal virulome. The three PAI I pathogenicity islands identified in A23EC are shown. PAI I A23EC (similar to PAI-II-536-pheU [PAI II₅₃₆]), PAI II A23EC (similar to PAI-CFT073-pheV [PAI I_CFT073]), and PAI III A23EC (similar to PAI-CFT073-asnT [PAI II_CFT073] and corresponding to pathogenicity island [HPI]) (Llyod et al., 2009). The virulence genes harbored in each PAIs are listed in blue boxes. Virotype E-defining genes, all found within these PAIs, are highlighted in bold and pink. (B) Plasmid (pA23EC) virulome. Genes encoding virulence factors are labeled and their position is shown in purple boxes.

Figure 3

Genomic structure and genetic context of the TnMB1860 sequence in the C2b sublineage of the C2 subclade of ST131 genomes included in our study. The different genetic elements are shown as arrows. Resistance genes are shown in red; IS26 insertion sequences in yellow, ΔIS3 in purple, ISKpn11 in lilac, transposons in orange, metG in green, ΔcirA in blue and genes encoding hypothetical proteins in gray. Genes within and in the vicinity of TnMB1860 are labelled. (A) Structure of TnMB1860. The genetic structure of TnMB1860 with its two transposable units and a total of six ARGs, as described in Shropshire et al, 2021 (Shropshire et al., 2021) (B) Comparative genetic arrangement of TnMB1860 in genomes belonging to sublineage C2b. The five clusters reflect distinct genetic distribution profiles of TnMB1860. The GenBank accession numbers for the genomic sequences included in each group are shown in Table S2 . The numbers on the left indicate the cluster corresponding to the set of sequences shown on the right, with number of sequences in each cluster shown in parentheses. All sequences were chromosomal, except for the single sequence in Cluster 1. The comparison was performed with Easyfig and the BLASTn algorithm. The percent sequence identity is shown in the form of a heatmap ranging between 99 and 100%, as shown in the legend. (C) Comparative genetic arrangement of the sequences in cluster 3. In eleven of the sequences, including strain A23EC, the insertion site of TnMB1860 was close to the gene encoding methionyl-tRNA synthetase, metG. A single sequence (CP049085.1) was instead inserted in the gene encoding for the truncated colicin I receptor (cirA), as previously described by Shropshire et al., 2021 (Shropshire et al., 2021).

Figure 4

Comparative analysis of plasmid pA23EC with the other 11 F subtype FII_31/36:FIA_4/20:B1 plasmids identified within the sublineage C2b of subclade C2. From inner to outer relative to the distance map, the circles represent the following: Circle 1-2: GC content and GC Skew. Circle 3: plasmid pA23EC (highlighted in red). Circles 4-15: other plasmids belonging to subtype FII(31/36):FIA (4/20):B1 carried by the C2b sublineage of the C2 subclade of E. coli ST131; the names of the plasmids are listed in the legend, with their year of isolation in parenthesis. The % identity and % coverage are color-coded as indicated in the legend and shown in increasing coverage order. Circle 16: pA23EC gene content annotated with Prokka, with virulence, resistance, replication, conjugation, addiction systems, insertion sequences/transposons, transcription factors and hypothetical proteins in different colors, as indicated in the legend.