Antibiotic Susceptibility Testing of the Gram-Negative Bacteria Based on Flow Cytometry

Rapidly treating infections with adequate antibiotics is of major importance. This requires a fast and accurate determination of the antibiotic susceptibility of bacterial pathogens. The most frequently used methods are slow because they are based on the measurement of growth inhibition. Faster methods, such as PCR-based detection of determinants of antibiotic resistance, do not always provide relevant information on susceptibility, particularly that which is not genetically based. Consequently, new methods, such as the detection of changes in bacterial physiology caused by antibiotics using flow cytometry and fluorescent viability markers, are being explored. In this study, we assessed whether Alexa Fluor® 633 Hydrazide (AFH), which targets carbonyl groups, can be used for antibiotic susceptibility testing. Carbonylation of cellular macromolecules, which increases in antibiotic-treated cells, is a particularly appropriate to assess for this purpose because it is irreversible. We tested the susceptibility of clinical isolates of Gram-negative bacteria, Escherichia coli and Pseudomonas aeruginosa, to antibiotics from the three classes: β-lactams, aminoglycosides, and fluoroquinolones. In addition to AFH, we used TO-PRO®-3, which enters cells with damaged membranes and binds to DNA, and DiBAC4 (3), which enters cells with depolarized membranes. We also monitored antibiotic-induced morphological alterations of bacterial cells by analyzing light scattering signals. Although all tested dyes and light scattering signals allowed for the detection of antibiotic-sensitive cells, AFH proved to be the most suitable for the fast and reliable detection of antibiotic susceptibility.

E. coli ATCC2592 strain treated (B, D) or not (A, C) with ampicillin at 1 × MIC. The FSC/SSC areas were measured based on the pseudocolor density-plot of FSC/SSC using the "polygon selection" tool of the ImageJ 1.48v software (areas are underlined in red (C, D).

Supplementary results: Detection of the susceptibility of the Staphylococcus aureus to ciprofloxacin.
We tested our methodological approach on a Gram-positive bacterium, Staphylococcus aureus. As with P. aeruginosa, S. aureus, in particular the methicillin-resistant S. aureus (MRSA), is a pathogen frequently involved in nosocomial infections. We chose to analyze the susceptibility of the S. aureus bacterium to ciprofloxacin because fluoroquinolones are among the drugs of choice used to treat MRSA infections. However, the MRSA, which are multidrug resistant strains, are frequently resistant to ciprofloxacin (Michel and Gutmann, 1997;Weber et al., 2003). Two strains were analyzed: one sensitive (strain 8311065; MIC: 0.25 mg/L) and one resistant (strain 0807063; MIC: 32 mg/L) to ciprofloxacin. For each experiment, log-phase cultures of the sensitive strain and the resistant strain were treated, or not, with ciprofloxacin and analyzed in a 4-hour time course study. The collected FC data are reported in the Figure S8. Light scattering profiles FSC and SSC. We observed that the light scattering profile of the sensitive strain treated with 1 mg/L of ciprofloxacin became dispersed (Figure S8 A). These changes were moderate compared to the FC data obtained from E. coli. We calculated the FSC ratios (or SSC) between the mean values of the FSC (or SSC) intensity of the treated cells and the mean values of the FSC (or SSC) intensity of the untreated cells. The FSC and SSC ratios were approximately 2 when the cells were exposed for 2 hours to ciprofloxacin. These ratios decreased after 3 hours of incubation because a subpopulation of cells had shrunken, while other cells were larger than the untreated cells. We also calculated the ratios between of the area of the density dot plot of the FSC/SSC ratios of the treated cells and the untreated cells for each experiment. The differences observed between the ratios (FSC, SSC and FSC/SSC area) of the sensitive strain and those of the resistant strain were all significant (t-test; p value < 0.03), except for the difference between the ratios of SSC obtained at 4 hours of incubation (t-test, p value = 0.09) (Figure S8 C). Autofluorescence. Exposure of cells from the 2 strains of S. aureus to 1 mg/L of ciprofloxacin induced a weak cell far-red (excitation at 633 nm, emission at 660/20 nm) and blue (excitation at 488 nm, emission at 530/30 nm) autofluorescence. Less than 5% of the treated cells were more fluorescent than the untreated cells. In the conditions used in these experiments, the autofluorescence due to the treatment did not interfere with the FC measurements of the fluorescent dyes (Figure S8 D).

Changes in cells fluorescence after staining with the AFH, TO-PRO®-3 and DiBAC 4 .
After staining with AFH and TO-PRO®-3, we found a significant increase of the fluorescence intensity of the sensitive cell population treated with the ciprofloxacin compared to the untreated cell population. After staining with DiBAC 4 the increase of fluorescence intensity of the cell population after treatment was low ( Figure S9). 20% of the sensitive cells exposed to ciprofloxacin for at least 2 hours and stained with AFH were fluorescent. The percentage of fluorescents cells, which increased in a time-dependent manner, reached approximately 60% after 4 hours of incubation (Figure S8 B and D). After staining with TO-PRO®-3, 40% of the sensitive cells were fluorescent after 1 hour of incubation. However, after 3 hours of incubation, we observed a decrease in the percentage of cells stained with TO-PRO®-3. The fact that the number of cells stained with AFH still increased after 3 hours suggests that the decrease of cell stained with the TO-PRO®-3 was not due to the restoration of membrane impermeability but rather to a change in the DNA accessibility to the dye. Nevertheless, the differences between the FC results obtained for the sensitive and resistant strains at each time point in the incubation with AFH and TO-PRO®-3 were significant (t-tests, p value ≤ 0.005 and p value ≤ 0.04, respectively). Double staining with AFH and TO-PRO®-3 could allow the susceptibility of S. aureus strains to ciprofloxacin to be efficiently detected after 1 hour of incubation with the antibiotic.
As for E. coli treated with the ciprofloxacin, the staining of the sensitive S. aureus cells with DiBAC 4 (3) was weak. Fifteen to 20% of the cells were stained with DiBAC 4 (3) after 3 hours of incubation. The difference between the staining of sensitive and resistant cells with DiBAC 4 (3) was significant only after 3 hours of incubation (t-test p value ≤ 0.01). Taken together, our data showed that it is possible to assess the susceptibility of a sensitive strain of S. aureus to ciprofloxacin using FC analysis based on AFH, TO-PRO®-3 and the light scattering data.

Analysis of the susceptibility of the Staphylococcus aureus to ciprofloxacin.
(A) Light scattering profiles for FSC versus SSC of cells from the strain 8311065 (MIC = 0.25 mg/L) after 4 hours of incubation without (left) or with 1 mg/L of ciprofloxacin (right). (B) Overlay histograms of the fluorescent intensities of cells stained with AFH or TO-PRO®-3. The blue histogram represents the cells treated with ciprofloxacin and stained with dye, the red histogram represents untreated cells stained with dye, and the black histogram is the control (untreated, unstained cells). (C and D) Time course study of samples from the 8311065 (MIC = 0.25 mg/L) and 0807063 (MIC = 32 mg/L) strains treated, or not, with 1 mg/L of ciprofloxacin. (C) Left and middle: the ratios of the mean FSC (or SSC) intensities between the antibiotictreated cells and the mean FSC (or SSC) intensities of the untreated cells from the sensitive and resistant strains relative to incubation time. Right: kinetic curve of the ratios between the area of the density plots of FSC/SSC of the treated and the untreated cells from the sensitive and resistant strains relative to incubation time. (D) The percentage of fluorescent cells stained with AFH, TO-PRO®-3 and DiBAC 4 (3) relative to time.