Pathoadaptive Mutations of Escherichia coli K1 in Experimental Neonatal Systemic Infection

Abstract

Although Escherichia coli K1 strains are benign commensals in adults, their acquisition at birth by the newborn may result in life-threatening systemic infections, most commonly sepsis and meningitis. Key features of these infections, including stable gastrointestinal (GI) colonization and age-dependent invasion of the bloodstream, can be replicated in the neonatal rat. We previously increased the capacity of a septicemia isolate of E. coli K1 to elicit systemic infection following colonization of the small intestine by serial passage through two-day-old (P2) rat pups. The passaged strain, A192PP (belonging to sequence type 95), induces lethal infection in all pups fed 2-6 x 106 CFU. Here we use whole-genome sequencing to identify mutations responsible for the threefold increase in lethality between the initial clinical isolate and the passaged derivative. Only four single nucleotide polymorphisms (SNPs), in genes (gloB, yjgV, tdcE) or promoters (thrA) involved in metabolic functions, were found: no changes were detected in genes encoding virulence determinants associated with the invasive potential of E. coli K1. The passaged strain differed in carbon source utilization in comparison to the clinical isolate, most notably its inability to metabolize glucose for growth. Deletion of each of the four genes from the E. coli A192PP chromosome altered the proteome, reduced the number of colonizing bacteria in the small intestine and increased the number of P2 survivors. This work indicates that changes in metabolic potential lead to increased colonization of the neonatal GI tract, increasing the potential for translocation across the GI epithelium into the systemic circulation.

Fig 1. E. coli A192PP has enhanced capacity to colonize the GI tract and induce systemic infection in susceptible neonatal rats.

(A) Survival of P2 rat pups fed 2–6 x 10⁶ E. coli K1 A192, A192P or A192PP (all, n = 24). Log-rank (Mantel-Cox) test; ns, nonsignificant; *, P < 0.05; **, P < 0.01). (B) Quantification of viable E. coli K1 in intestinal sections: tissues were removed 24 or 48 h after feeding, and CFU/g tissue determined by plating onto MacConkey agar. E. coli K1 colonies were identified with K1-specific phage. At 24 h, strain A192 n = 11, strain A192P n = 12, strain A192PP n = 12. At 48 hours, strain A192 n = 12, strain A192P n = 11, strain A192PP n = 11; Student's t test; *, P < 0.05; **, P < 0.01. (C) Growth of E. coli strains A192, A192P and A192PP in Mueller-Hinton (MH) broth at 37°C (200 orbits/min) (±SD, n = 3).

Fig 2. The E. coli A192 genome is comparable to other sequenced E. coli K1 genomes.

(A) The E. coli K1 A192 genome was assembled and contiguous consensus sequences ordered according to the IHE3034 genome. From the inside, the circles represent contigs coloured in alternating red/blue, GC content, BLAST analysis of E. coli completed genomes IHE3034, PMV-1, RS218, SF-048, SF-166, SF-173, SF-468, UTI89, S88, CE10 and K12, forward CDS, reverse CDS, virulence factors and unmapped contigs. (B) Phylogenetic relationships of E. coli bloodstream infection isolates; strains were grouped into four sequence types: ST69 (green) ST73, (blue) ST95 (red) and ST131 (yellow). (C) Phylogenetic relationships of E. coli bloodstream infection ST95 isolates. E. coli A192 is closely related to reference genome IHE3034 and contemporary bloodstream infection isolates.

Fig 3. E. coli A192PP has acquired mutations that alter the proteome.

(A) Genes acquiring single nucleotide polymorphisms during passage through infection-susceptible P2 rat pups. (B) Growth of E. coli A192PP strains in MH broth at 37°C (200 orbits/min) (±SD, n = 3). Total protein from (C) A192PP, (D) A192PPΔgloB, (E) A192PPΔyjgV, (F) A192PPΔtdcE and (G) A192PPΔthrA were separated on 2D gels using immobilized pH gradient strips in the range 3–10. The gels were stained with Coomassie Brilliant Blue, scanned and differences in protein abundance detected using PDQuest software. Proteins that differed at least twofold (P>0.5) are indicated. Proteins were identified by LC/MS/MS; numbered spots represent enoyl-acyl carrier protein reductase (1), transketolase (2), pyruvate dehydrogenase E1 component (3), ATP-dependent protease ATPase subunit (4), elongation factor G (5) and thiol peroxidase (6).

Fig 4. Pathoadaptive mutations of E. coli A192PP impact on GI colonization and virulence in the P2 rat.

(A) Quantification of viable E. coli K1 in the gut; PSI, MSI and DSI: proximal, mid- and distal sections of the small intestine. Tissues were removed 24 or 48 h after initiation of colonization by oral administration of 2–6 x 10⁶ E. coli A192PP and CFU/g tissue determined by viable counting. E. coli K1 colonies were detected using K1-specific phage. Student's t test; *, P < 0.05; **, P < 0.01. (B-E) Survival of pups fed 2–6 x 10⁶ E. coli A192PP or A192PP deletion mutants at P2: n = 24 for each strain in each experiment. Log-rank (Mantel-Cox) test; ns, non-significant; *, P < 0.05; **, P < 0.01.