Host Range-Associated Clustering Based on Multilocus Variable-Number Tandem-Repeat Analysis, Phylotypes, and Virulence Genes of Atypical Enteropathogenic Escherichia coli Strains

Abstract

Atypical enteropathogenic Escherichia coli (aEPEC) strains (36 Japanese and 50 Bangladeshi) obtained from 649 poultry fecal samples were analyzed by molecular epidemiological methods. Clermont's phylogenetic typing showed that group A was more prevalent (58%, 50/86) than B1 (31%, 27/86). Intimin type β1, which is prevalent among human diarrheal patients, was predominant in both phylogroups B1 (81%, 22/27) and A (70%, 35/50). However, about 95% of B1-β1 strains belonged to virulence group I, and 77% of them were Japanese strains, while 17% (6/35) of A-β1 strains did. Multilocus variable-number tandem-repeat analysis (MLVA) distributed the strains into 52 distinct profiles, with Simpson's index of diversity (D) at 73%. When the data were combined with those of 142 previous strains from different sources, the minimum spanning tree formed five zones for porcine strains, poultry strains (excluding B1-β1), strains from healthy humans, bovine and human patient strains, and the B1-β1 poultry strains. Antimicrobial resistance to nalidixic acid was most common (74%) among the isolates. Sixty-eight percent of them demonstrated resistance to ≥3 antimicrobial agents, and most of them (91%) were from Bangladesh. The strains were assigned into two groups by hierarchical clustering. Correlation matrix analysis revealed that the virulence genes were negatively associated with antimicrobial resistance. The present study suggested that poultry, particularly Japanese poultry, could be another reservoir of aEPEC (phylogroup B1, virulence group I, and intimin type β1); however, poultry strains seem to be apart from patient strains that were closer to bovine strains. Bangladeshi aEPEC may be less virulent for humans but more resistant to antibiotics.IMPORTANCE Atypical enteropathogenic Escherichia coli (aEPEC) is a diarrheagenic type of E. coli, as it possesses the intimin gene (eae) for attachment and effacement on epithelium. Since aEPEC is ubiquitous even in developed countries, we previously used molecular epidemiological methods to discriminate aEPEC as a human pathogen. The present study assessed poultry as another source of human diarrheagenic aEPEC. Poultry could be the source of aEPEC (phylogroup B1, virulence group I, and intimin type β1) found among patient strains in Japan. However, the minimum spanning tree (MST) suggested that the strains from Japanese poultry were far from Japanese patient strains compared with the distance between bovine and patient strains. Bangladeshi avian strains seemed to be less diarrheagenic but are hazardous as a source of drug resistance genes.

Keywords: Escherichia coli; MLVA; drug resistance; molecular epidemiology; phylogeny; virulence.

FIG 1

Distribution of intimin types among phylogroups and virulence groups in Japan and Bangladesh. (a) Distribution of intimin types among different phylogroups. (b) Distribution of intimin types among different virulence groups. JP indicates the strains isolated from Japan, and BD indicates the strains isolated from Bangladesh. * indicates significant at P ≤ 0.05, and ** indicates highly significant at P ≤ 0.01.

FIG 2

Distribution of phylogroups, virulence groups, and intimin types among the aEPEC strains isolated from poultry fecal specimens in Japan and Bangladesh.

FIG 3

Population modeling using the minimum spanning tree (MST) method of 228 aEPEC strains isolated from cattle, pig, poultry, foods, healthy carriers, and patients. The MST was constructed using the highest number of single-locus variants as the priority rule with no creation of hypothetical (or missing) types. The pale brown, green, pink, blue, and ivory clouds indicate the zones A, B, C, D, and E, respectively. (a) Strains isolated from different hosts are shown in different colors. White, red, green, blue, black, gray, and yellow indicate strains of healthy carriers, patients, foods, pig, cattle, Japanese poultry, and Bangladeshi poultry, respectively. (b) Associations of phylogenetic group and MLVA are shown in different colors. Yellow circles, red circles, green circles, blue circles, purple squares, black circles, white circles, gray circles, and light green circles indicate strains of phylogenetic groups A, B1, B2, C, D, E, F, clade-1, and unknown phylogenetic group, respectively. (c) Associations of virulence group and MLVA are shown in the figure. Red closed squares, blue closed squares, black circles, and white circles indicate strains of virulence groups Ia, Ib, II, and unknown virulence group, respectively. (d) Association of intimin types and MLVA are shown in the figure. Red circles, green circles, gray circles, yellow circles, closed blue circles, closed white circles, purple squares, light blue squares, pink circles, aqua circles, white squares, and black circles indicate the intimin type β1, θ/γ2, ζ, δ/κ/β2O, ι1, ξR/β2B, νR/ε2, ε1, γ1, α1/α2/µB, η, and untypeable, respectively.

FIG 4

Heatmap and hierarchical clustering of aEPEC isolates based on virulent genes and antibiotic resistance. Green indicates the presence and red indicates the absence of genes or antibiotic resistance. The upper row of the heatmap is a color indication of the geographical location of the strains. Letters A and B denote the two clusters formed by genotyping and antibiotic resistance patterns of the isolates. The hierarchical clustering was implemented using Wald’s method and a binary distance matrix.