Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, March 2002, p. 680-688, Vol. 46, No. 3
0066-4804/02/$04.00+0 DOI: 10.1128/AAC.46.3.680-688.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Department of Zoology, University of Oxford, Oxford, United Kingdom,1 Department of Pathology, University of Iowa College of Medicine, Iowa City, Iowa,2 Antimicrobial Development, Abbott Laboratories, Inc., Abbott Park, Illinois3
Received 17 September 2001/ Returned for modification 30 October 2001/ Accepted 21 November 2001
|
|
|---|
4 µg/ml to define resistance, the proportions of isolates from the three sampling periods for which MICs were at or above this breakpoint were 1.2, 1.6, and 1.4%, respectively. A total of 164 unique isolates (n = 58 from 1994-1995, 65 from 1997-1998, and 42 from 1999-2000) were examined for evidence of mutations in the quinolone resistance-determining regions (QRDRs) of the parC and the gyrA genes. Forty-nine isolates harbored at least one mutation in the QRDRs of one or both genes (1994-1995, n = 15; 1997-1998, n = 19; 1999-2000, n = 15). Among the 4,650 isolates of S. pneumoniae examined in the study, we estimated that 0.3% had mutations in both the parC and gyrA loci. The majority of mutations (67.3% of the mutations in 49 isolates with mutations) were amino acid substitutions in the parC locus only. Four isolates had a mutation in the gyrA locus only, and 12 isolates had mutations in both genes (8.2 and 24.5% of isolates with mutations, respectively). There was no significant difference in the number of isolates with parC and/or gyrA mutations detected during each study period. Finally, because of the magnitude of the study, we had reasonably large numbers of pneumococcal isolates with genotypically defined mechanisms of fluoroquinolone resistance and were thus able to determine the effects of specific resistance mutations on the activities of different fluoroquinolones. In general, isolates with mutations in parC only were resistant to ciprofloxacin but remained susceptible to levofloxacin, gatifloxacin, and moxifloxacin, whereas isolates with mutations in gyrA only and isolates with mutations in both parC and gyrA were resistant to all four fluoroquinolones tested. |
|
|---|
Numerous studies have demonstrated an association between the use of antimicrobials and the subsequent increase in the rate of antimicrobial resistance and/or the emergence of antimicrobial resistance (4, 11). With the increased use of fluoroquinolones in patients with respiratory tract infections in the United States during the past 5 years, it was of interest to know if rates of fluoroquinolone resistance in S. pneumoniae have increased in the United States. Such an association has recently been made in Canada (4). Three large national antimicrobial resistance surveillance studies were performed between 1994-1995 and 1999-2000, and 4,650 geographically diverse, clinical isolates of S. pneumoniae were collected. In light of the increased use of fluoroquinolones in the United States and reports of increasing fluoroquinolone resistance in Canada during the same period of time, the first objective of the present investigation was to evaluate whether changes in the in vitro activities of ciprofloxacin, levofloxacin, gatifloxacin, and moxifloxacin against S. pneumoniae had occurred over this 5-year period in the United States. Perhaps more importantly, we wished to evaluate whether fluoroquinolone resistance rates were increasing in the United States. Since a number of different classes of antimicrobials were tested during the three surveillance studies, we were also able to evaluate whether fluoroquinolone resistance had emerged in combination with resistance to other nonfluoroquinolone antimicrobials.
The second major objective of the present study was to characterize fluoroquinolone-resistant S. pneumoniae isolates at a molecular level. Fluoroquinolone activity against S. pneumoniae is the result of the inhibition of two type II topoisomerases, DNA topoisomerase IV and DNA gyrase, enzymes involved in chromosomal supercoiling and decatenation during replication. Topoisomerase IV comprises two subunits, parC and parE; likewise, DNA gyrase comprises two subunits, gyrA and gyrB. There is increasing evidence that fluoroquinolones have different primary targets, depending on the bacterial species and the fluoroquinolone. With respect to S. pneumoniae, it appears that topoisomerase IV is the primary target for ciprofloxacin but that DNA gyrase is the primary target for levofloxacin, gatifloxacin and moxifloxacin (9, 14, 16, 23, 24).
Pneumococcal resistance to the fluoroquinolones occurs primarily as a result of mutations in the quinolone resistance-determining regions (QRDRs) of the parC and gyrA genes (21, 22, 24, 26). Fluoroquinolone resistance generally occurs in two stages: a first-step mutation in a primary target reduces the binding affinity to that target and thereby decreases fluoroquinolone activity. A subsequent mutation in a secondary target results in a further decrease in in vitro activity. Using DNA sequence analysis, we endeavored to characterize the presence or absence of mutations in the QRDRs of the genes encoding the C and E subunits of topoisomerase IV and the A and B subunits of DNA gyrase in a large number of fluoroquinolone-resistant S. pneumoniae isolates. Finally, we attempted to correlate specific resistance mutations with the in vitro activities of ciprofloxacin, levofloxacin, gatifloxacin, and moxifloxacin.
|
|
|---|
Susceptibility testing. The MICs of nine agents, ciprofloxacin, levofloxacin, gatifloxacin, moxifloxacin, penicillin, erythromycin, tetracycline, chloramphenicol, and trimethoprim-sulfamethoxazole, were determined by broth microdilution as described by the National Committee for Clinical Laboratory Standards (NCCLS) (19). Antimicrobial powders were received from their respective manufacturers; erythromycin, tetracycline, and chloramphenicol, however, were obtained from Sigma-Aldrich (St. Louis, Mo.). The growth medium was Mueller-Hinton broth plus 3% lysed horse blood. Trays were incubated in ambient air at 35°C for 22 to 24 h prior to visual determination of the MICs. The final inoculum concentration was approximately 5 x 105 CFU/ml. S. pneumoniae ATCC 49619 was used as a control organism throughout the study. Stock cultures were thawed and subcultured twice onto sheep blood agar plates prior to MIC determinations.
DNA amplification and sequencing. Genomic DNA was prepared from overnight cultures of S. pneumoniae grown on 5% sheep blood agar plates incubated in 5% CO2. Two or three isolated colonies were suspended in 100 µl of distilled water and heated for 15 min at 95°C. The primers described by Pan et al. (21) were used to amplify a 382-bp region of the gyrA gene (forward primer, VGA3; reverse primer, VGA4) corresponding to amino acid residues 46 to 172 of Escherichia coli gyrA, a 457-bp region of the gyrB gene (forward primer, H4025; reverse primer, H4026) corresponding to amino acid residues 371 to 512 in E. coli, a 366-bp region of the parC gene encoding residues 35 to 157 (forward primer, MO363; reverse primer, M4271), and a 289-bp region of the parE gene encompassing residues 398 to 483 (forward primer, S6398; reverse primer, S6399). The parC gene of one isolate could not be amplified with the primers of Pan et al. (21), so the degenerate primers of Tankovic et al. (26) were used to amplify a 254-bp region of the gene. PCRs were carried out in 100-µl volumes with 4 µl of genomic DNA and each primer pair at a concentration of 0.8 µM in a buffer solution consisting of 22 mM Tris-HCl (pH 8.4), 55 mM KCl, 1.65 mM MgCl2, 220 µM dGTP, 220 µM dATP, 220 µM dTTP, 220 µM dCTP, and 22 U of recombinant Taq DNA polymerase per ml (PCR Supermix; GIBCO BRL, Grand Island, N.Y.). The amplification reactions for gyrA and parC consisted of 40 cycles of denaturation at 94°C for 15 s, annealing at 52°C for 45 s, and polymerization at 72°C for 45 s. Amplification reactions for gyrB and parE consisted of 30 cycles of denaturation at 92°C for 1 min, annealing at 48°C for 1 min, and polymerization at 72°C for 2 min. Automated DNA sequencing was performed with the ABI PRISM Big Dye Terminator Cycle Sequencing Ready Reaction kit with AmpliTaq DNA Polymerase FS according to the manufacturer's directions (Applied Biosystems Division of Perkin-Elmer, Foster City, Calif.). Purified sequencing reaction products were electrophoresed on either a model 373XL or a model 377XL DNA sequencer according to the manufacturer's directions (Applied Biosystems Division of Perkin-Elmer). DNA sequences for clinical isolates were compared with published sequences for parC and parE (GenBank accession no. Z67739) (22), gyrA (GenBank accession no. U49087) (26), and gyrB (GenBank accession no. Z67740) (22).
Reserpine efflux screen. Isolates of S. pneumoniae were subcultured onto 5% sheep blood agar plates and were grown overnight at 37°C in 5% CO2. Ciprofloxacin MICs were determined by agar dilution with Mueller-Hinton agar supplemented with 5% sheep blood with or without 10 µg of reserpine per ml and an inoculum of 104 CFU/spot (3, 19). Reserpine was freshly prepared before use and was subsequently added to the medium. After incubation for 18 to 20 h at 37°C in 5% CO2, the MIC was determined visually as the lowest antibiotic concentration that inhibited growth.
|
|
|---|
0.03 to
128 µg/ml, respectively, for ciprofloxacin; 1, 1, and
0.03 to 64 µg/ml, respectively, for levofloxacin, 0.5, 0.25, and
0.03 to 16 µg/ml, respectively, for gatifloxacin; and 0.25, 0.12, and
0.03 to 8 µg/ml, respectively, for moxifloxacin. This resulted in a rank order of in vitro activity of moxifloxacin > gatifloxacin > levofloxacin = ciprofloxacin. Interestingly, if one compares the MIC frequency distributions obtained for ciprofloxacin and levofloxacin, the cumulative percentages are quite comparable at all concentrations except 1 µg/ml, at which the cumulative percentage of strains was approximately 15% lower for ciprofloxacin. The modal MICs of each of the four fluoroquinolones examined in this study did not change in the three study periods. The susceptibility breakpoints (for susceptible, intermediate, and resistant) defined by NCCLS are as follows:
2, 4, and
8 µg/ml, respectively, for levofloxacin and
1, 2, and
4 µg/ml, respectively, for gatifloxacin and moxifloxacin (20). There are no NCCLS breakpoints for ciprofloxacin against S. pneumoniae; however, an MIC of
4 µg/ml is generally accepted as a marker for ciprofloxacin resistance in S. pneumoniae (4, 12, 13, 25). Overall, the proportion of isolates that were either intermediate or resistant to each fluoroquinolone was very low: ciprofloxacin, 1.4%; levofloxacin, 0.5%; gatifloxacin, 0.3%; and moxifloxacin, 0.3%. Fluoroquinolone resistance rates were not found to vary significantly over time. |
View this table: [in a new window] |
TABLE 1. MICs of four fluoroquinolones from three collections of S. pneumoniaea
|
4 µg/ml and a random sample of isolates for which the ciprofloxacin MICs were 1 or 2 µg/ml were selected for molecular characterization of the QRDRs in the parC and gyrA genes (Table 2).
All of the isolates from the 1994-1995 and 1997-1998 study periods selected for molecular characterization of parC and gyrA were also evaluated for evidence of parE mutations. Only three isolates had parE mutations, two mutants had parC mutations only, and one mutant had a gyrA mutation only (data not shown). A random sample of strains was also characterized to detect gyrB mutations, but none were detected. The lack of significant parE and gyrB mutations has also been reported by other investigators (1, 15, 16, 18, 22). |
View this table: [in a new window] |
TABLE 2. Selection of S. pneumoniae for molecular characterization of fluoroquinolone resistance determinants
|
In total, 164 unique patient isolates of S. pneumoniae were characterized for parC and/or gyrA mutations, and 49 of the isolates harbored one or more mutations. A summary of the specific amino acid substitutions detected in the QRDRs of the parC and the gyrA genes is listed in Table 3. Comparison of the percentages of isolates with parC and/or gyrA mutations recovered in each study period revealed no statistically significant differences. The majority (67.3%; n = 33) of all isolates with mutations had a mutation in the parC locus only, and nine different patterns of parC mutations were noted. These were generally single amino acid substitutions in the QRDR of the parC gene, although three isolates with two different amino acid substitutions in the QRDR of this gene were detected. The most common mutation was a serine79-phenylalanine substitution, detected in 17 isolates, or 34.7% of all parC mutants. The proportion of mutants with mutations in parC only (among all mutants) decreased over time: 80.0% for the 1994-1995 group, 63.2% for the 1997-1998 group, and 60.0% for the 1999-2000 group. These changes, however, were not statistically significant.
|
View this table: [in a new window] |
TABLE 3. Number of isolates with QRDR mutations, by specific mutation, study period, and total number of isolates
|
Twelve (24.5%) of the 49 strains with mutations had amino acid substitutions in the QRDRs of both the parC and the gyrA genes. Eight different patterns of mutations in the parC and gyrA genes were detected, four of which included two different amino acid substitutions in the parC gene plus a serine81-phenylalanine mutation in the gyrA gene. The most common pattern was a serine79-phenylalanine substitution in the parC gene plus a serine81-phenylalanine substitution in the gyrA gene (n = 5; 41.7% of mutants with mutations in the parC and the gyrA genes). Among all 49 mutants, the proportion of isolates with mutations in both the parC and the gyrA loci increased marginally, but not significantly, each study year: 20.0% in 1994-1995, 26.3% in 1997-1998, and 26.7% in 1999-2000.
The specific amino acid substitutions and the corresponding antimicrobial susceptibility data for each of the 49 isolates are described in detail in Table 4. The range of fluoroquinolone MICs for 33 isolates that harbored mutations in parC only was as follows: ciprofloxacin, 2 to 16 µg/ml; levofloxacin, 1 to 4 µg/ml; gatifloxacin, 0.25 to 1 µg/ml; and moxifloxacin, 0.06 to 0.5 µg/ml. Twenty-five of 33 (75.8%) mutants with mutations in parC only were resistant to ciprofloxacin, and 18.2% (n = 6) were intermediately resistant to levofloxacin. All mutants with mutations in parC only were susceptible to gatifloxacin and moxifloxacin. Thirteen mutants with mutations in parC only were resistant to between two and five nonfluoroquinolone comparator antimicrobials: penicillin, erythromycin, tetracycline, chloramphenicol, and trimethoprim-sulfamethoxazole. Three of these 13 isolates were resistant to ciprofloxacin and intermediately resistant to levofloxacin, 5 were resistant to ciprofloxacin but susceptible to levofloxacin, and 5 remained susceptible to all four fluoroquinolones.
|
View this table: [in a new window] |
TABLE 4. ParC and gyrA mutations and associated antimicrobial susceptibility data for S. pneumoniae isolates with reduced susceptibilities to fluoroquinolones
|
Twelve isolates had mutations in the QRDRs of both parC and gyrA. These included single or double amino acid substitutions in the parC gene plus a single amino acid substitution in the gyrA gene. Ten of 12 isolates were resistant to ciprofloxacin, levofloxacin, gatifloxacin, and moxifloxacin. Isolates 1901 and 1384 were susceptible to moxifloxacin, intermediately resistant to gatifloxacin, and resistant to ciprofloxacin and levofloxacin. Six of 12 mutants with mutations in parC and gyrA were also resistant to between two and five of the nonfluoroquinolone comparator antimicrobials. Four isolates were resistant to all four fluoroquinolones but were susceptible to all five nonfluoroquinolone comparator agents. The remaining two isolates were resistant to all four fluoroquinolones plus penicillin and trimethoprim-sulfamethoxazole, respectively.
Finally, in an attempt to compare the effects of mutations on the in vitro activity of each fluoroquinolone, the MIC frequency distributions and the geometric mean MICs for each group of S. pneumoniae mutants with mutations in parC only, gyrA only, and both parC and gyrA were determined (Table 5). Compared to the modal MIC for the entire population of organisms, the geometric mean MIC for the isolates with mutations in parC increased fivefold with ciprofloxacin and twofold with levofloxacin, gatifloxacin, and moxifloxacin. The geometric mean MIC for isolates with mutations in gyrA increased 8-fold with ciprofloxacin, 10-fold with levofloxacin, and 14-fold with gatifloxacin and moxifloxacin. Considerable increases in the geometric mean MICs were noted for isolates with mutations in both loci: ciprofloxacin, 40-fold; levofloxacin, 17-fold; gatifloxacin, 21-fold; and moxifloxacin, 20-fold. By the use of recognized susceptibility breakpoints, in general, mutants with mutations in parC only were resistant to ciprofloxacin but susceptible to levofloxacin, gatifloxacin, and moxifloxacin, whereas mutants with mutations in gyrA only and mutants with mutations in parC and gyrA were resistant to all four fluoroquinolones investigated in the study.
|
View this table: [in a new window] |
TABLE 5. Fluoroquinolone MICs for S. pneumoniae isolates with mutations in the QRDRs of parC and/or gyrAa
|
|
|
|---|
All S. pneumoniae isolates for which ciprofloxacin MICs were
4 µg/ml and a random sample of isolates for which the MICs were near this breakpoint for resistance (ciprofloxacin MICs, 1 or 2 µg/ml) were selected for molecular characterization. This resulted in the molecular characterization of 164 bacterial isolates. All 29 isolates for which ciprofloxacin MICs were
8 µg/ml had mutations in the QRDRs of parC and/or gyrA. Among the 36 isolates for which ciprofloxacin MICs were 4 µg/ml, 13 (36.1%) were found to have mutations. Eight of 72 isolates for which ciprofloxacin MICs were 2 µg/ml (i.e., 11.1%) had mutations, and none of 27 isolates for which MICs were 1 µg/ml were found to harbor mutations. The likelihood of mutations in both parC and gyrA increased as the ciprofloxacin MICs increased. Interestingly, we recovered one strain of S. pneumoniae for which the ciprofloxacin MIC was 16 µg/ml, but the strain lacked mutations in either parC or gyrA. This strain did not appear to have an upregulated fluoroquinolone efflux pump, at least on the basis of the results of reserpine inhibition experiments. The mechanism of fluoroquinolone resistance in this isolate remains obscure.
The observations in the present study have several implications. First, on the basis of the results of the in vitro studies, the development of fluoroquinolone resistance appears to be a stepwise process in which strains with first-step mutations go on to develop a second-step mutation (9, 24). Ciprofloxacin targets topoisomerase IV first; therefore, if selection of mutations occurs following exposure to this agent, it is likely to be in the parC gene. ParC mutants remain susceptible to moxifloxacin and gatifloxacin (since these agents target DNA gyrase first), and many also remain susceptible to levofloxacin. However, once parC mutations have occurred, the possibility for the development of a gyrA mutation with exposure to fluoroquinolones in patients with pneumococcal infections increases and the subsequent combination of mutations greatly diminishes the activities of all four fluoroquinolones. There is a need for a reliable in vitro method for the detection of these single-locus mutations in parC only in order for treatment to be judged accordingly and for purposes of surveillance of resistance.
The in vitro activities of fluoroquinolones against organisms with parC and gyrA mutations differed greatly from their activities against those with single-locus mutations. In the present study the in vitro activities of ciprofloxacin and levofloxacin against mutants with parC and gyrA mutations decreased 40- and 17-fold, respectively, compared with their activities against strains that lacked these mutations. The activities of newer fluoroquinolones, gatifloxacin and moxifloxacin, decreased 21- and 20-fold, respectively. With only two exceptions, all isolates with mutations in parC and gyrA were resistant to all four fluoroquinolones. Therefore, irrespective of the fluoroquinolone, once an organism had both a parC mutation and a gyrA mutation, the activities of all fluoroquinolones tested were greatly diminished. This would argue for the use of gatifloxacin and moxifloxacin over ciprofloxacin or levofloxacin, as the two newer agents have the greatest overall activity (by factors of four- to eightfold over those of ciprofloxacin and levofloxacin) and are less likely to select for first-step mutants (2, 9, 10, 23).
When the mutations were evaluated on the basis of the specific amino acid substitutions in either parC or gyrA, or both loci, the lack of mutational diversity in this collection of S. pneumoniae isolates was noteworthy. The most common parC mutation was serine79-phenylalanine or serine79-tyrosine, either alone (n = 18), in combination with a second parC amino acid substitution (n = 2), or in combination with a gyrA mutation (n = 9). Similarly, the most common gyrA mutation was serine81-phenylalanine or serine81-tyrosine, either alone (n = 4) or in combination with a parC mutation (n = 11). In fact, 36 of 49 mutants had one or both of these mutations, inclusive of all mutants with mutations in gyrA only and all mutants with mutations in parC and gyrA. The only strains of pneumococci with any other mutation were isolates with mutations in parC only; the mutations were one of the following: asparagine91-aspartic acid, aspartic acid83-asparagine or aspartic acid83-tyrosine, lysine137-asparagine, or aspartic acid83-asparagine plus glutamic acid135-aspartic acid. Therefore, it appears that there is a relatively low degree of mutational diversity among fluoroquinolone-resistant S. pneumoniae isolates in nature. There did not appear to be any differences in the fluoroquinolone susceptibilities of the isolates with the less commonly occurring mutations in parC only and the fluoroquinolone susceptibilities of the S. pneumoniae isolates that harbored the common parC mutations.
It should also be noted that this investigation, like many others, focused on the examination of specific regions of the C and E subunits of topoisomerase IV and the A and B subunits of DNA gyrase as a basis for the increases in the MICs. The amino acid substitutions that occur in these specific regions, referred to as QRDRs, have previously been shown to account for the vast majority of increases in fluoroquinolone MICs for S. pneumoniae. We cannot, however, rule out the potential contribution of mutations in other regions to higher MICs because we did not look for such mutations in the present study (17).
Finally, 42.9% (21 of 49) of those isolates with any mutations in the QRDRs were also resistant to between two and five nonfluoroquinolone agents. Importantly, this was true for 6 of 12 mutants with mutations in parC and gyrA. While the mechanisms of resistance to each of these antimicrobial classes are entirely different, mechanisms of antimicrobial resistance to one or more classes of antimicrobials often coexist in the same isolate. This association between fluoroquinolone resistance and resistance to multiple antimicrobial classes in S. pneumoniae has previously been reported by Whitney et al. (27). With multidrug-resistant strains, the selection of resistant isolates (or clones) of pneumococci may occur through the use of any of the antimicrobials to which they have developed resistance (6).
In conclusion, fluoroquinolone resistance has emerged but appears to have remained stable among S. pneumoniae isolates in the United States. Reports of much higher rates of fluoroquinolone resistance have been reported in other parts of the world: Spain, 5.3% (11); Hong Kong, 12.1% (13); and Northern Ireland, 15.2% (12). Therefore, it must be emphasized that the very low rates of fluoroquinolone resistance in the United States require vigilance with respect to both the appropriate use of fluoroquinolones and continued surveillance for changes in rates of fluoroquinolone resistance among S. pneumoniae.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»