Fig 1.
Three single-step mutants with resistance against 3 diverse antibiotics (upward bars) show CS to NIT (downward bars). Fold change of the MIC compared to susceptible wild-type strains. Light shading denotes an inoculum of 106 cells; the overlaid dark shading corresponds to a larger inoculum of 108 cells. Numerical data are available in S1 Data. CS, collateral sensitivity; MEC, mecillinam; MIC, minimum inhibitory concentration; NIT, nitrofurantoin; PRO, protamine sulfate; TGC, tigecycline.
Fig 2.
Uptake dynamics of radioactively labelled NIT measured as CPM in washed cell pellets of E. coli (A) and S. enterica (B). Measurements are normalized by the OD600 of the sample and expressed relative to the first measurement (n = 3 biological replicates). (C) mRNA expression of nfsA and nfsB relative to wild-type strains of E. coli and S. enterica measured with qRT-PCR (mean ± SD, n = 3 biological replicates). (D) Relative abundance of the proteins NfsA and NfsB measured with mass spectrometry (mean ± SD, n = 2 biological replicates). Student t test, ***P < 0.001, **P < 0.01, *P < 0.05. The numerical data for this figure and the full proteomic data sets are available in S1 Data. CPM, counts per minute; CS, collateral sensitivity; NIT, nitrofurantoin; OD600, optical density at 600 nm; qRT-PCR, quantitative reverse transcription PCR.
Fig 3.
Nitroreductase expression increases susceptibility to NIT.
(A) Susceptibility to NIT, as measured using Etest (bioMérieux, Marcy-l’Étoile, France) with a high inoculum of 108 cells. The nfsA and nfsB open reading frames were cloned into the pBAD vector and expressed using induction by 0.2% arabinose (mean ± SD, n = 3–5 biological replicates). Single expression constructs are labelled “A” and “B” for nfsA and nfsB, respectively. Dual expression is labelled “AB.” (B) Expression of nitroreductases reduces exponential growth rate at low concentrations of NIT. Change of growth rate relative to EV (mean ± SEM, n = 3–5 biological replicates). (C) Deletion of nitroreductases promotes survival at high concentrations of NIT, here measured for E. coli using a time-kill experiment with 24 mg/l (mean ± SEM, n = 3 biological replicates). (D) Expression of nitroreductases reduces survival to NIT, here shown for the pBAD constructs in wild-type E. coli background and 24 mg/l NIT. Dashed lines indicate 0.2% arabinose, solid lines indicate absence of arabinose (mean ± SEM, n = 3 biological replicates). The grey dashed lines in (C) and (D) indicate the limit of detection, based on Poisson estimates. Numerical data are available in S1 Data. ara, arabinose; EV, empty vector; NIT, nitrofurantoin.
Fig 4.
Change of susceptibility to NIT relative to wild-type strains for constructed single and double mutants. MIC was measured using Etest (bioMérieux, Marcy-l’Étoile, France) with a high inoculum of 108 cells (mean ± SD, n ≥ 2 biological replicates). Coloured horizontal lines connect genotypes with the same CS mutation. The spoT mutant produced a high frequency of revertants, which are indicated in brighter colour and showed suppression of CS. “Acs1” and “Atox1” are resistance cassettes for genetic engineering. “dup” indicates a forced duplication. Numerical data are available in S1 Data. Acs1, amilCP-cat-sacB cassette 1; Atox1, amilCP-toxin cassette 1; CS, collateral sensitivity; MIC, minimum inhibitory concentration; NIT, nitrofurantoin.
Fig 5.
CS due to interference with the cellular drug response.
(A) Model for elevated susceptibility to NIT in a lon mutant due to induction of the SOS response. (B) Relative accumulation of SulA protein in the E. coli lon mutant in drug-free medium (mean ± SD, n = 2 biological replicates). (C) Transcription of sulA is unchanged in the lon mutant in drug-free medium (mean ± SD, n = 3 biological replicates). (D) Partial suppression of CS at high temperature (mean ± SD, n = 5 biological replicates). (E) Relative change of MIC compared to wild-type E. coli in mutants of the SOS response. CS is completely suppressed by deletion of sulA and partially suppressed by a noninducible allele of lexA (mean ± SD, n = 3–5 biological replicates). Numerical data are available in S1 Data. CS, collateral sensitivity; MIC, minimum inhibitory concentration; NIT, nitrofurantoin; SulA, Suppressor of Lon.
Fig 6.
Temporal dynamics of CS in the lon mutant.
Delayed establishment of inhibition at low concentrations of NIT inhibition in the lon mutant. (A) Growth kinetics of the E. coli wild type and the lon mutant diverge after several hours of equal growth (mean ± SEM, n = 3 biological replicates). (B) Growth kinetics with log-transformed OD600 data for a subset of concentrations. The shaded grey regions indicate the OD600 windows that were used for the determination of exponential growth rate. (C) Dose-dependent inhibition of exponential growth rate in early (bottom) and late (top) OD600 windows (mean ± SD, n = 3 biological replicates). (D) Time point of exponential growth rate divergence between the 2 strains (mean ± SD, n = 3 biological replicates). Numerical data are available in S1 Data. CS, collateral sensitivity; NIT, nitrofurantoin; OD600, optical density at 600 nm.