Previous Article | Next Article ![]()
Applied and Environmental Microbiology, May 2005, p. 2587-2591, Vol. 71, No. 5
0099-2240/05/$08.00+0 doi:10.1128/AEM.71.5.2587-2591.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Department of Bacteriology, National University of Ireland, Galway, Ireland
Received 16 June 2004/ Accepted 11 October 2004
|
|
|---|
|
|
|---|
4 mg/liter) is rare (4). Resistance to the quinolone nalidixic acid is a marker for reduced fluoroquinolone susceptibility. The clinical significance of reduced susceptibility to fluoroquinolones is the subject of debate at present with a recently published proposal that the current National Committee for Clinical Laboratory Standards (NCCLS) interpretive criteria for ciprofloxacin (susceptible at an MIC of
1 µg/ml) be reduced significantly (1). In gram-negative bacteria the principal target of quinolone/fluoroquinolone activity is the type II topoisomerase, DNA gyrase (11). DNA gyrase is a tetramer composed of two GyrA subunits (encoded by gyrA gene) and two GyrB subunits (encoded by gyrB gene). Raised MICs to quinolones and fluoroquinolones are associated with mutations in the quinolone resistance-determining region (QRDR) of the gyrA gene. Less frequently, quinolone resistance is associated with point mutations in the type IV topoisomerase. This, too, has a tetramer structure and is encoded by the genes parC and parE. The QRDR of DNA gyrase A in S. enterica is between amino acids 67 and 122 encoded by nucleotides 185 to 361 in gyrA gene (3). The most common amino acid substitutions in the GyrA subunit associated with quinolone resistance occur at codons Ser83 and Asp87 (14).
We observed an association in Ireland between isolated nalidixic acid resistance and serovar Enteritidis (9). In 2000, 13.7% of serovar Enteritidis strains (n = 255) from Ireland were nalidixic acid resistant compared to 2.3% of serovar Typhimurium strains. Quinolone resistance was particularly associated with phage type 1 (PT1); of 26, 23 (88%) were nalidixic acid resistant compared to 6% of serovar Enteritidis of all other phage types. Isolated nalidixic acid resistance is also common in serovars Hadar, Kentucky, and Virchow.
A high incidence of nalidixic acid resistance could reflect frequent mutation/selection events. Alternatively high incidence of nalidixic acid resistance may arise from rare mutation/selection events with subsequent clonal expansion and dissemination. One would predict that clonal expansion would result in the nalidixic acid-resistant organisms being relatively homogeneous, whereas the frequent mutation/selection hypothesis should result in greater heterogeneity. We have studied a collection of nalidixic acid-susceptible and -resistant isolates of serovar Enteritidis and other salmonella serovars to determine whether there is evidence to support either hypothesis.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Phage types, nalidixic acid and ciprofloxacin MICs, pulsed-field profiles, gyrA amino acid substitutions, and mutations of 50 serovar Enteritidis isolates
|
|
View this table: [in a new window] |
TABLE 2. Mutations in QRDR of the gyrA gene in 32 representative quinolone-resistant isolates of 17 S. enterica serovars
|
Antimicrobial susceptibility testing.
MICs were determined by agar dilution method as described by the NCCLS (8). The antimicrobial agents tested were nalidixic acid, cinoxacin, flumequine, oxolinic acid, ofloxacin, and lomefloxacin (all from Sigma-Aldrich, Steinheim, Germany) and enrofloxacin (Baytril, Shawnee Mission, Kans.). It was not possible to access ciprofloxacin and marbofloxacin powders of defined potency; therefore, it was necessary to use suspensions normally used for parenteral administration (Ciproxin I.V. Flexibag [Bayer, Berkshire, United Kingdom] and marbofloxacin [Vetoquinol Ireland, Oranmore, Ireland]). E. coli ATCC 25922 and E. coli J53-2 were included as controls.
PFGE.
Pulsed-field gel electrophoresis (PFGE) was performed in accordance with the PulseNet Protocol (Centers for Disease Control and Prevention, Atlanta, Ga.) with the restriction endonuclease XbaI (Fig. 1). The isolate Salmonella serovar Braenderup H9812 was the control organism used (2).
![]() View larger version (117K): [in a new window] |
FIG. 1. Gel illustrating the seven different pulsed-field profiles obtained from 50 serovar Enteritidis isolates by using PFGE-XbaI. The label "C" indicates the PFP of the control strain serovar Braenderup H9814. A, the most commonly observed PFP; A1 to A6, the variant PFP observed.
|
Nalidixic acid-resistant laboratory mutants.
Nalidixic acid-resistant mutants were derived from serovar Enteritidis strains (3332/00, 3486/00, and 3558/00) by plating a heavy inoculum on Mueller-Hinton broth (MHB) containing a quinolone or fluoroquinolone agent at a concentration three doubling dilution steps above the organism's MIC. Suspect mutant colonies were confirmed as salmonella by colonial morphology on xylose lysine deoxycholate agar (Becton Dickinson, Oxford, United Kingdom). Nalidixic acid resistance was confirmed by NCCLS disk diffusion. Thirty-eight isolates with growth up to the nalidixic acid disk were included in further studies. The QRDR mutations present in these laboratory mutants were defined by PCR amplification, followed by restriction endonuclease digestion (5) or DNA sequencing.
Growth rate studies.
Growth rate studies were performed in triplicate with serovar Enteritidis 3332/00 and two different nalidixic acid-resistant derivative mutants: M3332/1 (Asp87Tyr) and M3332/14 (Asp87Asn). Growth studies were performed in 20 ml of MHB with an initial inoculum of 2 x 105 CFU/ml. For M3332/1 and M3332/14 growth rates were repeated in triplicate in the presence of 32 µg of nalidixic acid/ml. Viable counts were performed at times: 0, 180, 240, 300, 330, 360, 390, 420, 450, 480, 510, and 540 min. A growth curve was constructed, and the specific growth rate of each strain was determined. The nalidixic acid MICs for both mutants were assessed before and after each experiment to confirm that the phenotype of nalidixic acid resistance was stable.
|
|
|---|
PFGE.
On PFGE typing, 38 isolates were indistinguishable, and there were six minor variants (one to two band differences) among the remaining 12 isolates. For the predominant phage types studied, PT1 (90% nalidixic acid resistant) and PT4 (100% nalidixic acid susceptible), all but two isolates were indistinguishable (pulsed field profile [PFP] A). PFGE with XbaI was not sufficiently discriminatory to provide support for either clonal expansion or frequent mutation/selection as the basis for the high frequency of nalidixic acid resistance observed.
Sequence data.
The gyrA QRDR sequences obtained for quinolone-susceptible serovar Enteritidis was identical to that reported for serovar Typhimurium (6). All of the nalidixic acid-resistant PT1 (n = 18) and PT21 (n = 3) isolates showed the transversion thymine for guanine at position 259 (G259T) in the gyrA gene. The single nalidixic acid-resistant PT6a isolate had the transition G259A. The derived amino acid sequences from these mutations indicate a substitution of tyrosine (G-T) or asparagine (G-A) for aspartic acid in the amino acid sequence at position 87 in the gyrA gene (see Table 1). No mutations were detected in the parC QRDR of the 22 quinolone-resistant isolates of serovar Enteritidis.
Five distinct gyrA mutations were found among 32 isolates of serovars other than Enteritidis, as detailed in Table 2. It is noteworthy that a G259A mutation was detected in four of five serovar Hadar isolates and that C248T was detected in all seven serovar Kentucky isolates and in four of six serovar Virchow isolates. The QRDR of the parC gene was also sequenced for a nalidixic acid-resistant serovar Reading isolate and a serovar Stanley isolate that had no gyrA mutation. No mutations were identified in the parC gene of the serovar Reading isolate, whereas the serovar Stanley had a significant mutation (G170C) in the second nucleotide of codon 57, (predicted substitution Thr57Ser) and six additional silent mutations.
Thirty-eight nalidixic acid-resistant mutants were obtained from three nalidixic acid-susceptible parent serovar Enteritidis isolates. The G259T mutation present in all of the clinical nalidixic acid-resistant PT1 isolates was detected in only nine (24%) of the laboratory mutants. A total of 14 (37%) laboratory mutants had a mutation at nucleotide 248 (codon Ser83), 4 (11%) had A260G, 1 (3%) had G259A, 2 (5%) had C241T, and 2 (5%) had C356A. No mutation was identified in the QRDR of the gyrA or parC genes in 6 (16%) of laboratory-generated mutants.
Growth studies.
Growth studies indicated that both parent (3332/00) and mutants had similar specific growth rates in MHB. In the presence of nalidixic acid 32 mg/liter the mutants M3332/1 (G259T) and M3332/14 (G259A) had similar growth rates.
|
|
|---|
Therefore, we determined the specific point mutations associated with nalidixic acid resistance and compared the diversity of mutations observed with that of mutations observed in nalidixic acid-resistant isolates of other serovars and of laboratory mutants derived from susceptible Salmonella Enteritidis parent strains. The isolates studied included five phage types/lineages (1, 4, 6a, 21, and 44) with representative nalidixic acid-resistant isolates included for three (1, 6a, and 21) of the five phage types. All 18 representatives of PT1 and three of PT21 shared the same G259T mutation (Table 1), and the single PT6a isolate had a different (G259A) mutation.
To further investigate this observation, we determined the mutations associated with nalidixic acid resistance in 32 isolates representing 16 other serovars of S. enterica (Table 2). Distinct serovars represent distinct lineages in which mutations conferring resistance have clearly arisen in nature independently of mutations in serovar Enteritidis. Five distinct gyrA mutations associated with nalidixic acid resistance were observed among the 14 serovars, and a parC mutation was observed in a serovar Stanley isolate. A tendency for an association between specific mutations and particular serovars was observed when multiple isolates of a given serovar were studied, as noted with serovars Hadar, Kentucky, and Virchow. Of necessity, the isolates studied are limited in number, are subject to selection bias, and cannot be truly representative of the biology of salmonella in the environment. In order to address in some way the possibility of our results representing a chance occurrence, we performed a limited statistical assessment. If all of the mutations associated with nalidixic acid resistance described in all salmonella serovars in the present study are equally likely to occur and persist, the probability that the pattern of associations of specific mutations with specific serovars actually observed arising by repeated independent mutation alone are as follows: 1.68 x 1016 for serovar Enteritidis; 9.0 x 103 for serovar Hadar, 3.0 x 105 serovar Kentucky, and 1.0 x 103 serovar Virchow.
In laboratory-generated nalidixic acid-resistant mutants that have clearly arisen independently of each other from a well-characterized parent strain, the basis of acquired nalidixic acid resistance was more diverse. In this group, G259T was detected in only 24% of serovar Enteritidis nalidixic acid-resistant laboratory mutants, with other gyrA mutations detected in 60% of isolates and no gyrA or parC mutation detected in 16%. This finding indicates that a bias toward the repeated independent occurrence of the G259T mutation as distinct from other mutations in serovar Enteritidis is not likely to adequately explain the predominance of this specific mutation in a given lineage.
It is also conceivable that the specific G259T mutation could come to predominate through convergent evolution if the associated Asp87Tyr substitution confers an advantage over strains with other possible nalidixic acid resistance-associated substitutions. It is difficult to construct an experiment to test this hypothesis fully; however, within the limits of the approaches available to us we have demonstrated that there is no difference in growth rate in the absence or in the presence of nalidixic acid between mutants with G259T (M3332/1) and G259A (M3332/14) derived from the same parent strain. Therefore, we found no evidence of a selective advantage for a strain possessing the predominant G259T (Asp87Tyr) mutation.
We have identified no previous reports of parC mutations in naturally occurring S. enterica. Previously described parC mutations in E. coli have been at codons 80 and 84 (10). The codon 57 has not previously been associated with quinolone resistance in E. coli; however, the association observed does not confirm that the mutation is the cause of this resistance.
The predominance of specific point mutations in nalidixic acid-resistant serovar Enteritidis has previously been described. Wiuff et al. identified the amino acid substitution Ser83Phe as predominant among 15 quinolone-resistant serovar Enteritidis PT8 isolates in the Denmark (15) and suggested that this is indicative of the spread of a resistant clone of serovar Enteritidis. Another recent study from Spain demonstrated that mutations in codon 87 was predominant (91%) in the nalidixic acid-resistant S. enterica serovar Enteritidis found. The three isolates sequenced in that study possessed the Asp87Tyr (G259T) amino acid substitution, identified as predominant in our study (12). The present study expands the evidence supporting clonal expansion as a major contributor to a high incidence of nalidixic acid resistance in serovar Enteritidis. We also present some preliminary evidence of a similar phenomenon of clonal expansion of nalidixic acid resistance in serovars Hadar, Kentucky, and Virchow.
|
|
|---|
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»