INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial infections caused by multidrug-resistant pathogens are a major global problem, especially for nosocomial infections (6). Methicillin-resistant Staphylococcus aureus (MRSA) is one such pathogen against which options for effective antibacterial therapies are already limited (2). While certain newly developed drugs have promising activity against MRSA (1, 9), their relatively narrow spectra of activity could limit their clinical use in empirical therapy.

Recently, a series of 8-methoxy, nonfluorinated quinolones (NFQs) (Fig. 1) has been identified as potent antibacterial agents with broad-spectrum antibacterial activities (S. D. Brown, P. C. Fuchs, and A. L. Barry, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-1510, p. 210, 2000; B. Ledoussal, J. K. Almstead, S. M. Flaim, C. P. Gallagher, J. L. Gray. X. E. Hu, N. K. Kim, H. D. KcKeever, C. J. Miley, T. L. Twinem, and S. X. Zeng, Abstr. 39th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-544, p. 303, 1999; D. F. Sahm, A. Staples, I. Critchley, C. Thornsberry, K. Murfitt, and D. Mayfield, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-1509, p. 210, 2000). On the basis of in vitro potency data, these compounds are more potent against MRSA and coagulase-negative staphylococci than several fluoroquinolones and have activities comparable to that of clinafloxacin (10). An apparent advantage of the NFQs against S. aureus lies in their ability to (i) better utilize both DNA gyrase and topisomerase IV as dual targets than certain quinolones, such as ciprofloxacin and trovafloxacin, and (ii) largely circumvent existing mutations in serine and glutamate "hot spots" of the target genes, gyrA and grlA, commonly associated with quinolone resistance (10, 11). However, it is imperative to ascertain the potential for development of de novo resistance to the NFQs in these pathogens. This report describes the in vitro isolation of S. aureus mutants with reduced susceptibilities to the NFQs and other quinolones by two approaches: stepwise isolation of mutants and spiral plater-based serial passage.

View larger version (7K): [in this window] [in a new window]   FIG. 1.   Chemical structures of the NFQs used in the present study.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Materials and strains. The NFQs PGE 9262932, PGE 4175997, and PGE 9509924 and the other quinolones used in the present study were synthesized in-house. PGE 9509924 was used as a racemic mixture for the selection of mutants. Reserpine was purchased from Sigma Aldrich (St. Louis, Mo.) and was dissolved in dimethyl sulfoxide (2 mg/ml stock) prior to addition to growth medium. S. aureus Mi273 (ATCC 29213) was used as the parent strain for the selection of mutants.

Stepwise selection of mutants. S. aureus Mi273 or its mutants were resuspended (1.8 × 10⁹ to 2.3 × 10⁹ CFU) in brain heart infusion (BHI) broth from an overnight culture and plated onto BHI agar plates containing two to four times the MIC of the test compounds (for Mi273 or its mutants), and the plates were incubated at 37°C for 48 h. Control plates with no drug were incubated for 24 h. Two to three agar plates were used to have the desired inoculum exposed to each compound at each concentration. Following incubation, the bacterial colonies were counted to obtain the mutation frequencies. Selected colonies were subcultured onto BHI agar and stored as frozen cultures in 20% glycerol. These mutants were subsequently used to generate next-step mutants by the procedure described above. At each step of mutant isolation, the MICs of the NFQs and the benchmark quinolones were determined for the mutants.

Serial passage. S. aureus Mi273 was grown on petri dishes with Mueller-Hinton (MH) agar and 5% sheep blood (BBL Microbiology Systems, Cockeysville, Md.), transferred to a tube containing sterilized brucella broth, and adjusted to an optical density of 0.4 at 600 nm. Compounds were prepared at 50 µg/ml, 500 µg/ml, 5 mg/ml, and 25 mg/ml and were dispensed with a spiral plater (model 4000; Spiral Biotech, Inc.) onto MH agar-blood agar petri dishes to generate a concentration gradient. The petri dishes were inoculated in quadruplicate (referred to as replicates 1, 2, 3, and 4 in the Results section) with the broth cultures mentioned above.

MICs. The MICs of the test compounds were determined by incubating cultures (~5 × 10⁵ CFU/ml) of Mi273 and its mutants overnight (18 to 24 h) at 37°C in BHI broth in the presence of test compounds in a twofold broth microdilution series in duplicate. The MIC was recorded as the minimum concentration of the test compound required for complete inhibition of bacterial growth. MICs in the presence of reserpine (10 µg/ml) were measured by the same broth microdilution method, and all the data were generated in duplicate.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Stepwise selection of mutants. S. aureus mutants capable of growing in the presence of concentrations two times above the MIC were isolated at frequencies of 9.1 × 10⁸ and 5.7 × 10⁹ for ciprofloxacin and trovafloxacin, respectively, and <5.7 × 10¹⁰ for the NFQs, gatifloxacin, and clinafloxacin. These mutants were termed step 1 mutants (Fig. 2). One of these mutants (strain 273/T), selected with trovafloxacin, was analyzed for quinolone susceptibility and the DNA sequence of the QRDR. As shown in Table 1, strain 273/T contained the known Ser80-Phe mutation in GrlA. The MICs for this strain relative to those for the parent were twofold higher for the NFQs, gatifloxacin, and clinafloxacin and eight- and fourfold higher for ciprofloxacin and trovafloxacin, respectively (Table 2). The susceptibilities of 17 additional step 1 mutants (5 selected with trovafloxacin and 12 selected with ciprofloxacin) to the NFQs and the other quinolones tested were similar to that of strain 273/T (the MICs were within 1 twofold dilution range; data not shown).

View larger version (13K): [in this window] [in a new window]   FIG. 2.   Schematic representation of the stepwise selection of mutants via exposure to the NFQs, ciprofloxacin, trovafloxacin, gatifloxacin, and clinafloxacin. Details of the experiment are described in Materials and Methods.

Serial passage. The results of the serial passage experiments are presented in Table 3. Among all the quinolones tested, susceptibility to ciprofloxacin was reduced the most within four passages, and the final MIC for all the four strains was >128 µg/ml, as determined by the broth dilution method. All four strains isolated via exposure to trovafloxacin remained relatively susceptible to it (final MIC range, 0.5 to 2.0 µg/ml). Similar mutants isolated with gatifloxacin and clinafloxacin showed variable susceptibilities to these compounds (final MIC ranges, 1 to 32 and 1 to 16 µg/ml, respectively). Susceptibilities to the NFQs were reduced to various degrees, with final MICs being in the range of 1 to 64 µg/ml. Significant reductions in susceptibilities were seen after seven passages with subinhibitory levels of the NFQs, clinafloxacin, and gatifloxacin.

Analysis of the NFQ-resistant mutants. To elucidate the molecular basis for reduced susceptibility to the NFQs, partial DNA sequences of the target genes of some of the mutants described above were analyzed. The results are presented in Table 4. The point mutations identified include (i) those in the known hot spots in the QRDR, such as Ser84-Leu in GyrA and Glu84-Lys in GrlA (11), and (ii) those that were previously unknown, such as His103-Tyr and Ser52-Arg in GrlA, Glu477-Val in GyrB, and Glu472-Val in GrlB. The MICs of the other quinolones tested for these mutants (relative to those for the parent strain Mi273 [Table 3]) were also elevated (Table 5). The potencies of the NFQs against strains 273/932-2 and 273/924-3 (MICs, 16 to 32 µg/ml; geometric mean MIC, 28.5 µg/ml) were higher than those of ciprofloxacin (MICs, >128 µg/ml) and overall were comparable to those of trovafloxacin, gatifloxacin, and clinafloxacin (MICs, 16 to 128 µg/ml; geometric mean MIC, 32 µg/ml). In contrast, the potencies of the NFQs against strain 273/997-3 (MICs, 8 to 32 µg/ml; geometric mean MIC, 20.1 µg/ml) were overall lower than those of the other quinolones tested (MICs 2 to 16 µg/ml; geometric mean MIC, 4.8 µg/ml).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, two approaches were used to select NFQ-resistant mutants of S. aureus. First, the stepwise isolation of S. aureus mutants via exposure to drugs at concentrations above the MIC was attempted. The frequency of mutant selection was lower when S. aureus was exposed to two times the MICs of the NFQs, gatifloxacin, and clinafloxacin (<5.7 × 10¹⁰) than when it was exposed to ciprofloxacin and trovafloxacin (9.1 × 10⁸ and 5.7 × 10⁹, respectively). These data are consistent with those from previous work (10), suggesting that the former group of quinolones is better able to utilize both DNA gyrase and topoisomerase IV as dual targets in S. aureus at the whole-cell level. However, when a step 1 mutant (strain 273/T) containing the Ser80-F mutation in GrlA was exposed to all the quinolones used in the present study at four times the MICs, mutants were selected in each case.

The results, presented in Tables 1 and 2, indicate that the mutations in the QRDRs and the levels of quinolone susceptibility of the step 2 mutants were similar for both strains, irrespective of the selecting agent. Additional step 2 mutants isolated with the other two NFQs, PGE 4175997 and PGE 9509924, had similar quinolone susceptibility profiles (data not shown). Overall, the results obtained with the step 2 mutants suggest that certain common mechanisms of quinolone resistance are selected in these mutants. The step 3 mutants, strains 273/T/T/T and 273/T/T/932, contained the same mutations in their QRDRs as their step 2 parent (strain 273/T/T). However, the quinolone MICs for the step 3 mutants were overall higher than those for the step 2 mutants, suggesting the involvement of additional mechanisms of resistance. On the basis of the MICs for these mutants in the presence of reserpine, higher levels of efflux (relative to that for the parent) did not appear to be involved. Additional step 3 mutants isolated with another NFQ, PGE 4175997, showed similar quinolone susceptibility profiles (data not shown).

By the spiral plater-based serial passage technique, three S. aureus mutants with high-level NFQ resistance (MICs, ~32 µg/ml) were identified during the study. The susceptibilities of these mutants to the other quinolones tested reveal an interesting pattern. First, the potencies of NFQs against strains 273/932-2 and 273/924-3 (Table 5) were comparable to those of trovafloxacin, gatifloxacin, and clinafloxacin. Second, in strain 273/997-3, the resistance to the NFQs appeared to be more specific than the resistance to the other quinolones tested. These data suggest that the mechanism(s) of quinolone resistance in these mutants is different from that in the stepwise-selected ones, against which the level of NFQ resistance is low (MICs, 0.5 to 2 µg/ml).

DNA sequence analysis of selected segments of the target genes, including the QRDRs of these mutants (strains 273/932-2, 273/997-3, and 273/924-3; Table 4), suggests a possible link between high-level NFQ resistance and multiple point mutations. While point mutations, such as those in the Ser80/84 and Glu84/88 residues of GrlA and GyrA, are frequently encountered in quinolone-resistant MRSA isolates (11), mutations in the His103 or Ser52 residues of GrlA appear to be unique. It is interesting that, unlike the serine and glutamate hot spots, His103 and Ser52 are not well conserved across different bacterial species (3, 7, 8). In addition, these residues fall outside the conventional QRDR in S. aureus (3). Recently, several novel mutations outside the QRDR were also found to be associated with resistance to premafloxacin in S. aureus (5). However, the specific roles of individual mutations in NFQ resistance can be determined only by further research, including introduction of the unique mutations in a sensitive S. aureus strain. In addition, point mutations in the target genes may not be solely responsible for high-level NFQ resistance. The data presented in Table 6, which show an eightfold increase in the potencies of NFQs against these mutants due to the addition of reserpine (compared with a twofold increase in potencies against the parent), suggest a possible role of putative efflux mechanisms, in addition to target mutations, in NFQ resistance.

Overall, the two sets of S. aureus mutants isolated during the present study revealed two different patterns of quinolone susceptibility. First, as previously observed with the MRSA isolates (10), stepwise-selected mutants with known mutations in GyrA and GrlA were relatively susceptible to the NFQs and clinafloxacin in vitro (MICs, 1 to 2 µg/ml). Second, additional mutants that were isolated via serial passage and that had high-level NFQ resistance harbored multiple, known, and unique mutations in both targets, combined with an apparently elevated level of efflux.