The challenge of intracellular antibiotic accumulation, a function of fluoroquinolone influx versus bacterial efflux

Abstract

With the spreading of antibiotic resistance, the translocation of antibiotics through bacterial envelopes is crucial for their antibacterial activity. In Gram-negative bacteria, the interplay between membrane permeability and drug efflux pumps must be investigated as a whole. Here, we quantified the intracellular accumulation of a series of fluoroquinolones in population and in individual cells of Escherichia coli according to the expression of the AcrB efflux transporter. Computational results supported the accumulation levels measured experimentally and highlighted how fluoroquinolones side chains interact with specific residues of the distal pocket of the AcrB tight monomer during recognition and binding steps.

Fig. 1. Quinolone activities measured with MIC and DEK vary according to their capacity of membrane translocation and sensibility to efflux.

Quinolone activities measured on the efflux-deficient mutant (AG100A) were plotted on the x axis (influence of the capacity of membrane translocation on activities) against the ratio of activities measured on the efflux-proficient compared with the -deficient mutant (AG102/AG100A) plotted on the y axis (influence of the AcrAB efflux pump sensibility on activities). a Red squares, MIC values. b Blue circles, DEK (dose necessary for early killing) values. See also Supplementary Table 1. To optimize the readability of the figure, the abscissa is represented as a logarithmic scale, and the quinolone names are abbreviated as mentioned in Table 1.

Fig. 2. UV fluorescence tracks FQs in bacteria.

Spectrofluorimetry spectra and microspectrofluorimetry imaging of CIP accumulation measured in AG100A and AG102 populations (a and b, respectively) and in living individual cells (c and d). Bacterial suspensions were incubated for 5 min without and with CIP at 5 µM, and suspensions were sampled from the same incubation mixture for spectrofluorimetry and microspectrofluorimetry measurements: for spectrofluorimetry analyses, bacterial pellets were lysed, and emission spectra of bacteria incubated without (dotted lines) or with (solid lines) CIP were measured with a spectrofluorimeter (a for AG100A and b for AG102). FQ concentrations in lysates were then calculated according to calibration curves (see “Methods” and ref. ). For microspectrofluorimetry measurements in living individual cells, bacterial pellets were resuspended in buffer and analyzed by deep–ultraviolet fluorescence imaging (examples are shown in c for AG102 and in d for AG100A, scale bars indicate 10 µm). Images were then analyzed and corrected to compare the various FQs and bacterial strains (see “Methods” and ref. ) as obtained in Supplementary Fig. 3.

Fig. 3. Influx: FQs have different capacities of membrane translocation measured with SICAR_IN.100A.

SICAR_IN.100A was obtained from the accumulated concentrations (molecules/cell) in the efflux-deficient mutant AG100A for 5-min incubation with 5 µM of FQs. Data are represented by a box-and-whisker plot (the ends of the box are the upper and lower quartiles, the median is marked by a black line inside the box, and the whiskers are the two lines outside the box that extend to the highest and lowest measurements). Substituents R₁, R₇, and R₈ of the FQ structures are indicated in the table under each corresponding FQ. ANOVA and Tukey’s post hoc tests were performed to determine differences between FQs (n = 32 biologically independent samples, ω²=0.69, degree of freedom = 8). ***P < 0.001; **P < 0.01; *P < 0.05. Data normality was checked by the Shapiro–Wilk test, and homogeneity of variance by the Fligner–Killeen test.

Fig. 4. Efflux: FQs have different susceptibilities to efflux measured with SICAR_EFF.100 and SICAR_EFF.102.

SICAR_EFF.100 a and SICAR_EFF.102 b were obtained by the ratio of the accumulated concentrations in the efflux-deficient mutant AG100A to the accumulated concentrations in the wild-type strain AG100 (a) or in the efflux-proficient mutant AG102 (b), for 5-min incubation with 5 µM of FQs. See also Supplementary Fig. 3. Data are represented by a box-and-whisker plot. Substituents R₁, R₇, and R₈ of the FQ structures are indicated in the table under each corresponding FQ of panel a ANOVA and Tukey’s post hoc tests were performed to determine differences between FQs (a n = 35 biologically independent samples, ω²=0.69, degree of freedom = 8; b n = 31 biologically independent samples, ω² = 0.70, degree of freedom = 8). ***P < 0.001; **P < 0.01; *P < 0.05. Data normality was checked by the Shapiro–Wilk test and homogeneity of variance by the Fligner–Killeen test.

Fig. 5. Calculations of the free energy of binding to the DPT of AcrB allow to distinguish three classes of FQs.

For each compound, we reported the two weighted average free energies of binding obtained from the clusters populated by at least 10% of the total conformations sampled during the two performed MD simulations and computed using the Molecular Mechanics/Generalized Born Surface Area method (see Methods for details and Supplementary Table 3 for the considered values). All values are expressed in units of kilocalories per mole (kcal mol⁻¹).

Fig. 6. NOR, FLE, and ENR have different interaction profiles with AcrB DPT residues.

a Table reported the total occurrence of hydrophobic contacts and H bonds between NOR, FLE, and ENR and the residues lining the DP_T. The contacts were extracted for each compound by merging the ten MD simulations in a single trajectory, and an error of ~6% should be considered. b Representative view of the DP_T from NOR, FLE, and ENR simulations, where the residues involved in interactions with the selected FQs are represented as sticks (white for hydrophobic contacts and magenta for H bonds). Stick width is proportional to the frequency of the considered interaction. Protein residues lining the DP_T are displayed as a blue surface. The switch- and the bottom loop are represented, respectively, as a yellow and cyan tube. Exit gate residues are shown as orange spheres. For the list of residues defining DP_T, switch-, bottom loop, and exit gate see Supplementary Table 2 and ref. .