Sequence analysis and characterization of a transferable hybrid plasmid encoding multidrug resistance and enabling zoonotic potential for extraintestinal Escherichia coli

Abstract

ColV plasmids of extraintestinal pathogenic Escherichia coli (ExPEC) encode a variety of fitness and virulence factors and have long been associated with septicemia and avian colibacillosis. These plasmids are found significantly more often in ExPEC, including ExPEC associated with human neonatal meningitis and avian colibacillosis, than in commensal E. coli. Here we describe pAPEC-O103-ColBM, a hybrid RepFIIA/FIB plasmid harboring components of the ColV pathogenicity island and a multidrug resistance (MDR)-encoding island. This plasmid is mobilizable and confers the ability to cause septicemia in chickens, the ability to cause bacteremia resulting in meningitis in the rat model of human disease, and the ability to resist the killing effects of multiple antimicrobial agents and human serum. The results of a sequence analysis of this and other ColV plasmids supported previous findings which indicated that these plasmid types arose from a RepFIIA/FIB plasmid backbone on multiple occasions. Comparisons of pAPEC-O103-ColBM with other sequenced ColV and ColBM plasmids indicated that there is a core repertoire of virulence genes that might contribute to the ability of some ExPEC strains to cause high-level bacteremia and meningitis in a rat model. Examination of a neonatal meningitis E. coli (NMEC) population revealed that approximately 58% of the isolates examined harbored ColV-type plasmids and that 26% of these plasmids had genetic contents similar to that of pAPEC-O103-ColBM. The linkage of the ability to confer MDR and the ability contribute to multiple forms of human and animal disease on a single plasmid presents further challenges for preventing and treating ExPEC infections.

FIG. 1.

Circular map of pAPEC-O103-ColBM. The outer two circles show predicted coding regions in forward and reverse orientations, and different colors indicate different predicted functions. The next circle shows the G+C content in a 1,000-bp window with 10-bp steps. The next five circles show levels of nucleotide homology with other sequenced ColV-type plasmids. Blue indicates ≥90% homology with pAPEC-O103-ColBM, while black indicates <90% homology. The numbers on these circles indicate comparisons with pAPEC-O2-ColV (circle 1), pAPEC-O1-ColBM (circle 2), pVM01 (circle 3), pCVM29188_146 (circle 4), and pSMS-3-5_130 (circle 5). The map was created using GenVision from DNASTAR.

FIG. 2.

Linear maps of sequenced ColV and ColBM plasmids. The bar at the top indicates positions (in base pairs). The maps begin with RepFIIA and are drawn to scale. Virulence genes of interest are identified and indicated by different colors. The plasmids are ordered according to evolutionary relationships of concatenated gene sequences (hlyF, traX, finO, repA, repA1). The evolutionary history was inferred using the neighbor-joining method (54). Bootstrap consensus trees were inferred using 500 replicates. Bootstrap confidence values greater than 50% are indicated at the nodes. The tree is drawn to scale based on evolutionary distances, and dotted lines extend the branches to the plasmid maps. Distances were computed using the maximum composite likelihood method (scale bar = 0.02 base substitution per site). Phylogenetic analyses were conducted with MEGA4 (55).

FIG. 3.

Linear map of the MDR-encoding region of pAPEC-O103-ColBM. Dashed lines indicate potential evolutionary breakpoints identified by nucleotide sequence comparisons and identified inverted repeats in the sequence. Nucleotide similarity (>90%) with pSMS35_130 is indicated below the scale (in base pairs).

FIG. 4.

Growth of bacterial strains in various growth media determined using OD₆₀₀. Counts for an 18-h period are shown. Each strain was tested in quadruplicate, and each experiment was performed twice on separate occasions. The data are averages for the experiments for each time point.

FIG. 5.

Average concentrations of bacteria in the blood and CSF of rats infected with ExPEC 408 and its derivatives. The error bars indicate the average results for two experimental trials, and at least 11 rats were used for experimental group. An asterisk indicates that a value is significantly different (P < 0.05) from the value for ExPEC 408.

FIG. 6.

Percentages of animals in each experimental group in the rat model containing different concentrations of bacteria in (A) blood and (B) CSF.

FIG. 7.

Histopathology of rat brain tissue 18 h after inoculation of ExPEC wild-type strain 408 (A), ExPEC 408 cured of pAPEC-O103-ColBM (B), and ExPEC 408 complemented with pAPEC-O103-ColBM (C). In panels A and C, moderate diffuse meningitis was observed (arrows), with neutrophils and lymphocytes scattered throughout the cerebral parenchyma and mild areas of edema. There were similar inflammatory cells that were scattered throughout the cerebral parenchyma and mild areas of edema (not shown). The plasmid-cured strain (B) (the lower magnification shows a larger region) did not cause any lesions.