Uropathogenic Escherichia coli Metabolite-Dependent Quiescence and Persistence May Explain Antibiotic Tolerance during Urinary Tract Infection

Abstract

In the present study, it is shown that although Escherichia coli CFT073, a human uropathogenic (UPEC) strain, grows in liquid glucose M9 minimal medium, it fails to grow on glucose M9 minimal medium agar plates seeded with ≤10(6) CFU. The cells on glucose plates appear to be in a "quiescent" state that can be prevented by various combinations of lysine, methionine, and tyrosine. Moreover, the quiescent state is characteristic of ~80% of E. coli phylogenetic group B2 multilocus sequence type 73 strains, as well as 22.5% of randomly selected UPEC strains isolated from community-acquired urinary tract infections in Denmark. In addition, E. coli CFT073 quiescence is not limited to glucose but occurs on agar plates containing a number of other sugars and acetate as sole carbon sources. It is also shown that a number of E. coli CFT073 mini-Tn5 metabolic mutants (gnd, gdhA, pykF, sdhA, and zwf) are nonquiescent on glucose M9 minimal agar plates and that quiescence requires a complete oxidative tricarboxylic acid (TCA) cycle. In addition, evidence is presented that, although E. coli CFT073 quiescence and persistence in the presence of ampicillin are alike in that both require a complete oxidative TCA cycle and each can be prevented by amino acids, E. coli CFT073 quiescence occurs in the presence or absence of a functional rpoS gene, whereas maximal persistence requires a nonfunctional rpoS. Our results suggest that interventions targeting specific central metabolic pathways may mitigate UPEC infections by interfering with quiescence and persistence. IMPORTANCE Recurrent urinary tract infections (UTIs) affect 10 to 40% of women. In up to 77% of those cases, the recurrent infections are caused by the same uropathogenic E. coli (UPEC) strain that caused the initial infection. Upon infection of urothelial transitional cells in the bladder, UPEC appear to enter a nongrowing quiescent intracellular state that is thought to serve as a reservoir responsible for recurrent UTIs. Here, we report that many UPEC strains enter a quiescent state when ≤10(6) CFU are seeded on glucose M9 minimal medium agar plates and show that mutations in several genes involved in central carbon metabolism prevent quiescence, as well as persistence, possibly identifying metabolic pathways involved in UPEC quiescence and persistence in vivo.

Keywords: E. coli persistence; E. coli quiescence; TCA cycle; carbon metabolism; urinary tract infections.

FIG 1

E. coli CFT073 quiescence on glucose plates. A 0.2% glucose plate was seeded with 10⁵ CFU of E. coli CFT073 (see Materials and Methods). (A) Sixty minutes after seeding the plate, a colony of E. coli MG1655, grown on a glucose plate, was transferred to the plate seeded with E. coli CFT073, using a toothpick; the plate was then incubated at 37°C for 24 h. (B) The same plate, incubated for 48 h. Note that E. coli CFT073 only grows around the picked E. coli MG1655. Although not shown, E. coli Nissle 1917 undergoes quiescence on glucose plates identically.

FIG 2

E. coli CFT073 nonquiescence on glycerol, ribose, and xylose plates. Glucose, glycerol, ribose, and xylose plates (0.2% each) were seeded with 10⁵ CFU of E. coli CFT073 grown overnight in liquid M9 minimal medium containing their respective sugars (0.4%). Sixty minutes after seeding the plates, a colony of E. coli MG1655, grown on a glucose plate, was transferred to a glucose plate, using a toothpick. Plates were incubated at 37°C for 24 h. (A) Glycerol; (B) ribose; (C) xylose; (D) glucose.

FIG 3

Prevention of quiescence by human urine and amino acids. Glucose (0.2%) plates were seeded with 10⁵ CFU of E. coli CFT073, and 5-µl amounts of the following mixtures were spotted onto the plates: (A) 50-fold-concentrated E. coli MG1655 culture supernatant; (B) amino acid cocktail mimicking the amino acid concentrations in the 50-fold-concentrated E. coli MG1655 culture supernatant; (C) human urine; (D) amino acid cocktail mimicking the amino acid concentrations in human urine (Table 3); (E) lysine, methionine, and tyrosine (1.0 mM each); (F) lysine and methionine (1.0 mM each); (G) lysine and tyrosine (1.0 mM each); and (H) methionine and tyrosine (1.0 mM each). Plates were incubated at 37°C for 24 h. Although not shown, the results for E. coli Nissle 1917 were essentially identical.

FIG 4

E. coli CFT073 and E. coli MG1655 persistence. Cultures were grown overnight in 0.4% glucose M9 minimal medium as described in Materials and Methods. Persister cell assays were performed as described in Materials and Methods. ▲, E. coli CFT073; ●, E. coli CFT073 plus ampicillin; ♦, E. coli MG1655; ■, E. coli MG1655 plus ampicillin. Bars representing standard errors of the means of counts from 2 independent experiments are presented for each time point. At 24 h, the approximately 1,000-fold difference between the counts of E. coli CFT073 persisters and E. coli MG1655 persisters in the presence of ampicillin is statistically significant (P = 0.0052).

FIG 5

E. coli CFT073 persistence in the presence of amino acids. Cultures were grown in 0.4% glucose M9 minimal medium as described in Materials and Methods and diluted 20-fold into 0.2% glucose M9 minimal medium either containing or lacking a mixture of the 20 standard l-amino acids, each at 100 µg/ml, and containing or lacking ampicillin (100 µg/ml). ♦, E. coli CFT073; ■, E. coli CFT073 plus ampicillin; ▲, E. coli CFT073 plus amino acids; ●, E. coli CFT073 plus amino acids plus ampicillin. Bars representing standard errors of the means of counts from 2 independent experiments are presented for each time point. At 6 h and 24 h, the approximately 100-fold differences between E. coli CFT073 persisters and E. coli MG1655 persisters in the presence of ampicillin are statistically significant (P = 0.002 and P = 0.05, respectively).

FIG 6

Diagram of E. coli central carbon metabolism. Arrows indicate the physiological directions of the reactions. Genes encoding the enzymes for each reaction are listed beside each reaction. Mini-Tn5 Km insertions in E. coli CFT073 genes that result in nonquiescence on glucose plates are circled. P, phosphate; KDPG, 2-keto-3-deoxy-6-phosphogluconate.

FIG 7

E. coli CFT073 sdhA persistence. Cultures were grown overnight in 0.4% glucose M9 minimal medium, and persister cell assays were performed as described in Materials and Methods. ♦, E. coli CFT073; ■, E. coli CFT073 plus ampicillin; ▲, E. coli CFT073 sdhA; ●, E. coli CFT073 sdhA plus ampicillin. Bars representing standard errors of the means of counts from 4 independent experiments are presented for each time point. At 24 h, the approximately 2,000-fold difference between E. coli CFT073 persisters and E. coli CFT073 sdhA persisters in the presence of ampicillin is statistically significant (P < 0.001).

FIG 8

E. coli CFT073 gdhA persistence. Cultures were grown overnight in 0.4% glucose M9 minimal medium, and persister cell assays were performed as described in Materials and Methods. ♦, E. coli CFT073; ■, E. coli CFT073 plus ampicillin; ▲, E. coli CFT073 gdhA; ●, E. coli CFT073 gdhA plus ampicillin. Bars representing standard errors of the means of counts from 3 independent experiments are presented for each time point. At 24 h, the approximately 2,000-fold difference between E. coli CFT073 persisters and E. coli CFT073 gdhA persisters in the presence of ampicillin is statistically significant (P < 0.001).

FIG 9

E. coli CFT073 sdhA persistence in the presence of fumarate. Cultures were grown overnight in 0.4% glucose M9 minimal medium containing or lacking disodium fumarate (200 µg/ml) and then diluted 20-fold into 0.2% glucose M9 minimal medium containing or lacking fumarate, and persister cell assays were performed as described in Materials and Methods. ▲, E. coli CFT073 sdhA; ●, E. coli CFT073 sdhA plus ampicillin; ♦, E. coli CFT073 sdhA plus fumarate; ■, E. coli CFT073 sdhA plus fumarate plus ampicillin. Bars representing standard errors of the means of counts from 2 independent experiments are presented for each time point. At 6 h and 24 h, the differences between E. coli CFT073 sdhA persisters in the presence and absence of fumarate are statistically significant (P = 0.013 and P = 0.046, respectively).

FIG 10

Rescue of E. coli CFT073 sdhA quiescence by fumarate. E. coli CFT073 sdhA was grown overnight in 0.4% glucose M9 minimal medium, and inocula of 10⁵ CFU were seeded on 0.2% glucose plates containing disodium fumarate (200 µg/ml) (A) or 0.2% glucose plates (B). One hour later, 5 µl of a mixture of 1.0 mM lysine, 1.0 mM methionine, and 1.0 mM tyrosine was spotted to each plate, and the plates were incubated at 37°C for 24 h.

FIG 11

Quiescence of the E. coli CFT073 original clinical isolate on glucose plates. Glucose (0.2%) plates were seeded with 10⁵ CFU of the E. coli CFT073 original clinical isolate, and 5-µl amounts of the following mixtures (containing 1.0 mM of each amino acid) were spotted onto the plates: (A) lysine, methionine, and tyrosine; (B) lysine and methionine; (C) lysine and tyrosine; (D) methionine and tyrosine. The plates were incubated at 37°C for 24 h.