Antimicrobial consumption is one of the major contributing factors facilitating the development and maintenance of bacteria exhibiting antimicrobial resistance. Plasmid-mediated quinolone resistance (PMQR) genes, such as the qnr family, can be horizontally transferred and contribute to reduced susceptibility to fluoroquinolones. We performed an observational study, investigating the copy number of PMQR after antimicrobial therapy. We enrolled 300 children resident in Ho Chi Minh City receiving antimicrobial therapy for acute respiratory tract infections (ARIs). Rectal swabs were taken on enrollment and seven days subsequently, counts for Enterobacteriaceae were performed and qnrA, qnrB and qnrS were quantified by using real-time PCR on metagenomic stool DNA. On enrollment, we found no association between age, gender or location of the participants and the prevalence of qnrA, qnrB or qnrS. Yet, all three loci demonstrated a proportional increase in the number of samples testing positive between day 0 and day 7. Furthermore, qnrB demonstrated a significant increase in copy number between paired samples (p<0.001; Wilcoxon rank-sum), associated with non-fluoroquinolone combination antimicrobial therapy. To our knowledge, this is the first study describing an association between the use of non-fluoroquinolone antimicrobials and the increasing relative prevalence and quantity of qnr genes. Our work outlines a potential mechanism for the selection and maintenance of PMQR genes and predicts a strong effect of co-selection of these resistance determinants through the use of unrelated and potentially unnecessary antimicrobial regimes.
Citation: Vien LTM, Minh NNQ, Thuong TC, Khuong HD, Nga TVT, Thompson C, et al. (2012) The Co-Selection of Fluoroquinolone Resistance Genes in the Gut Flora of Vietnamese Children. PLoS ONE 7(8): e42919. https://doi.org/10.1371/journal.pone.0042919
Editor: John R. Battista, Louisiana State University and A & M College, United States of America
Received: June 18, 2012; Accepted: July 16, 2012; Published: August 24, 2012
Copyright: © Vien et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by The Wellcome Trust of Great Britain. Stephen Baker is funded by an OAK Foundation through Oxford University, United Kingdom. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Appropriate antimicrobial usage in the management of infectious diseases is an important issue for patients and administering clinicians. The issue is particularly relevant in many low and middle income countries, where the sale of antimicrobials is often unrestricted , , . It is apparent that the sustained usage of some antimicrobial classes has, over time, reduced the effect of these drugs on once susceptible pathogens and encouraged the selection of resistant bacteria , , . The implications of antimicrobial resistant pathogenic bacteria are clear, including treatment failure, increasing treatment costs and protracted therapy . Furthermore, antimicrobial resistant organisms, or the mobile elements carrying the resistance genes, can be transmitted to other individuals and organisms, respectively, increasing the proportion of antimicrobial resistant organisms circulating in the general population , .
Plasmid-mediated quinolone resistance (PMQR) determinants, in the form of qnr genes, were first reported in 1998 in a clinical K. pneumoniae isolate from a patient with a urinary tract infection in the USA . In less than ten years, the reported prevalence of qnr genes (qnrA, qnrB, qnrS, qnrC and qnrD) has increased dramatically , reported globally across a spectrum of bacterial species, such as E. coli, Klebsiella pneumoniae , , . The phenotype conferred by Qnr determinants is generally low level resistance to fluoroquinolones, yet they appear to encourage additional fluoroquinolone resistance mechanisms, subsequently resulting in high level resistance . Our work in Ho Chi Minh City (HCMC), Vietnam, and the work of others, has shown that qnr genes are common in commensal organisms and are carried by mobile elements harbouring multiple antimicrobial resistance genes , , . Furthermore, we have suggested that qnr genes are co-selected by the use of non-quinolone antimicrobials and we have hypothesized that the use of antimicrobials, other than fluoroquinolones, plays a critical role in the selection and maintenance of qnr genes in the commensal human gut flora.
Acute respiratory tract infections (ARIs) are the most common infection in children in developed and developing countries and are the leading cause of death in children less than five years of age . A meta-analysis highlighted the global ARI disease burden, , estimating that 1.9 million (95% Confidence Interval [CI] 1.6–2.2 million) children died from ARIs in 2000, of which 70% of deaths were in Africa and Southeast Asia . Viruses are the dominant etiologic agents of ARIs, responsible for approximately 60% of all cases . However, because an etiological diagnosis is not normally established quickly, antimicrobials are typically the mainstay of empirical ARI treatment . Here, we have developed several real-time PCR assays to detect and enumerate qnr genes and quantified these loci in paired rectal swabs from children with symptoms of ARI prior to, and after seven days of, antimicrobial treatment.
Materials and Methods
Participants, enrolment and ethical approval
Children presenting to the outpatient department of Children's Hospital 1 in HCMC between April and November 2009 were eligible for enrolment in this study. The inclusion criteria were: less than 16 years of age, symptoms of ARI, not admitted to hospital, living in HCMC, prepared to come back for follow-up at day 7 and informed written consent given by a parent or a legal guardian. The age, demographics and prescribed antimicrobials were recorded at the time of enrolment (Day 0), after the consultation and a written prescription from the clinician making the medical assessment. All antimicrobials were prescribed at the discretion of the treating clinicians. Rectal swabs from all recruited children were collected on enrolment (prior to antimicrobial therapy) and on Day 7 (after antimicrobial therapy). Rectal swabs were stored in 1 ml of sterile saline, which is commonly used for transporting samples and making dilutions of Enterobacteriaceae. These solutions were processed for microbiology on the day of collection and were stored at −80°C until molecular analysis. This study was performed in accordance with the declaration of Helsinki and was granted ethical approval by Children's Hospital 1 in HCMC, Vietnam and the Oxford Tropical Research Ethics Committee (OxTREC), in the United Kingdom. Informed written consent was given by a parent or a legal guardian.
Sample processing and colony counting
Ten-fold serial dilutions (from 10−1 to 10−8) were made from all rectal swabs on the day of collection. Fifty microlitres of each dilution was inoculated onto MacConkey agar (Oxoid, UK) plates and incubated at 37°C overnight. The number of bacterial colonies on each plate was enumerated and plates containing between 30 to 300 colonies were used to calculate the CFU ml−1.
After microbiological assessment, the fecal samples were processed in batches (Day 0 and Day 7 of the same children) to ensure limited experimental variation. We compared the quality of extracted DNA using three DNA extraction methods, DNAeasy (QIAGEN, USA), QIAam DNA Stool mini kit (QIAGEN, USA) and the nucliSENS EasyMag system (bioMérieux, Mercy l'Étoile, France). The nucliSENS EasyMag system demonstrated the highest quantity, quality and reproducibility. As a result, after addition of the internal control, total nucleic acid was extracted from 200 µl of stool using the nucliSENS EasyMag system, according to the manufacturer's instructions, and was eluted in 50 µl of elution buffer.
Primer design and quantitative real-time PCR
The nucleotide sequences of all available full length alleles for qnrA, qnrB and qnrS were retrieved from NCBI , and aligned using Vector NTI (Invitrogen). Primers and probes for each of the three loci were designed using Primer Express version 2.0 (Applied Biosystems Inc.). The specificity of primers and probes was assessed using BLASTn against the NCBI database and were synthetized by Sigma-Proligo (Singapore). The primers and probes were as follows: qnrA; qnrA_RT_F; CAGTTTCGAGGATTGCAGTT, qnrA_RT_R, CCTGAACTCTATGCCAAAGC, qnrA_probe; Fam - AAGGGTGYCACTTCAGCTATGCC – Tamra; qnrB; qnrB_RT_F; CAGATTTYCGCGGCGCAAG; qnrB_RT_R; TTCCCACAGCTCRCAYTTTTC, qnr_probe; Fam - CGCACCTGGTTTTGYAGYGCMTATATCAC – Tamra; qnrS; qnrS_RT_F; TCAAGTGAGTAATCGTATGTA, qnrS_RT_R; GTCTGACTCTTTCAGTGAT, qnrS_probe; Fam – CCAGCGATTTTCAAACAACTCAC – Tamra. An internal control consisting of PhHV DNA was added to each sample. Primers and probe for Phocid Herpes Virus (PhHV), which were used as an internal control, were as previously described .
Real-time PCR was performed using HotStart Taq polymerase (QIAGEN) on a Chromo4 Real-time PCR machine (Bio-Rad, Hercules, CA, USA). Reactions were performed in 25 µl volumes, containing 2.5 µl of 10× buffer, 5 mM of MgCl2, 500 nM of dNTPs, 400 nM of each primer, 100 nM of probe, 1 U of HotStart Taq polymerase and 5 µl of DNA template. Optimized real-time PCR conditions were: one cycle of 95°C for 15 minutes, followed by 45 cycles of 30 seconds at 95°C, 30 seconds at annealing temperature for primer pair and probe and 30 seconds at 72°C. The optimized annealing temperatures were: 52°C for qnrA, 54°C for qnrB and 55°C for qnrS. A positive PCR signal was defined as any Ct (threshold cycle) value <40.
The target genes (qnrA1, qnrB1 and qnrS1) were amplified in their entirety with previously described primers , ,  and cloned into pCR2.1-TOPO and transferred into E. coli TOP10 cells using the TOPO TA Cloning kit (Invitrogen, UK), according to the manufacturer's instructions. Plasmids containing a cloned insertion were detected by blue-white colony selection on Luria-Bertani (LB) agar supplemented with 50 mg/L ampicillin and 40 mg/L X-gal. Plasmid DNA was extracted using the QIAprep Miniprep (QIAGEN), digested with XhoI restriction endonuclease (qnrA1, qnrB1 and qnrS1 do not have XhoI restriction sites), purified (QIAquick PCR Purification Kit, QIAGEN) and quantified using a NanoDrop ND-1000 Spectrophotometer (NanoDrop Technologies, Inc). Concentrations were then converted to copy number. Ten-fold serial dilutions of plasmid DNA containing the cloned target insert were used as an external standard for all quantitative real-time RT PCR experiments.
PCR Reproducibility, linearity and efficiency
The reproducibility of each assay was assessed by calculating the co-efficient of variance (CV%). The CV calculates the deviation in the Ct values of selected plasmid DNA concentrations across multiple amplifications, both in the same run and in different runs. Intra-assay reproducibility was assessed by comparing the Ct values generated in the same run of four replicates of each plasmid concentration. Inter-assay reproducibility was assessed by comparing the Ct values generated by four replicates of each plasmid concentration over a period of four days. Linearity was assessed by the Ct values of 10-fold serial dilutions of plasmid DNA containing cloned target sequences (concentrations 5×100–5×108) and was calculated by linear regression. Efficiency was again calculated from the slope of the standard curve, using the formula: Efficiency = 10(−1/slope)−1, according to the methods of Rasmussen et al .
Internal extraction and positive amplification controls
For optimizing each of the qnr real-time PCR assays, isolates that we have previously identified as being qnr-positive were used as positive controls, these were, qnrA1; Klebsiella pneumoniae K222Ca, qnrB1; Klebsiella pneumoniae K281An and qnrS1; Klebsiella pneumoniae K35N . Efficiency of nucleic acid extraction and real-time PCR amplification was monitored in all reactions by inclusion of the PhHV internal control . Aliquots of post culture supernatant containing PhHV were prepared. The precise amount of virus added to the fecal samples was assessed by PCR titration during the standard curves experiments. The final dilution and volume was selected on the basis that it reproducibly produced a Ct value within the range of 30–33, equating to approximately copies of PhHV added to each sample prior to nucleic acid extraction. Any amplification run that did not produce a Ct value between 30 and 33 for PhHV was discarded. Positive amplification controls with known Ct values for each gene and ‘no-template negative control’ were included in each amplification run.
The gene copy numbers and the CFU ml−1 of Enterobacteriaceae in each sample was calculated and logarithmically transformed for analysis using MS excel (Microsoft). The demographic data, prevalence of qnr genes and copy number of qnr genes between Day 0 and Day 7 were compared using the Fisher's exact test, McNemar's test and Wilcoxon's rank-sum test, respectively. Results were interpreted as statistically significant when p<0.05. All analyses were performed using R version 2.12.0 program (Foundation for Statistical Computing, Vienna, Austria).
Validation of quantitative real-time PCR assays for qnr genes
We developed three real-time PCR assays to quantify the copy number of qnrA, qnrB and qnrS. Plasmids containing cloned qnrA, qnrB and qnrS target sequences were used as positive controls for quantifying amplification. PCR amplicons for cloning were generated using genomic DNA extracted from bacterial isolates previously found to be positive for one of the three target loci . The real-time PCR assays were tested on nucleic acid extracted from 45 enteric organisms previously found to be qnr negative including, E. coli, Shigella spp., Salmonella spp., Citrobacter spp., Enterobacter spp., Pantoea spp., Proteus mirabilis and Klebsiella spp. None of the three loci demonstrated any unspecific amplification with any target nucleic acid from any of the tested organisms.
To assess the potential for cross hybridization between the three gene targets, each PCR assay was performed using template DNA prepared from each of the qnrA, qnrB and qnrS positive control strains. These experiments were repeated in triplicate. Each replicate for all PCR assays produced an identical result and each assay was specific for the intended target. The detection limit of each of the three assays was calculated on the basis of fifty individual Ct values for the diluted plasmid DNA samples, produced from ten replicates over five days that consistently generated a positive signal in ≥95% of reactions. These concentrations were <50 copies per reaction of the cloned target sequences of all three targets.
Reproducibility and linearity
Table 1 shows the results from a series of consecutive standard curve experiments. These data demonstrate the overall performance, intra-assay variation and inter-assay variation of the three real-time PCR assays. The intra-assay and inter-assay co-efficient of variance across the three targets ranged from 0.56–2.72% (>97% reproducibility) and 1.27–2.96% (>97% reproducibility), respectively, with target copy numbers ranging from 5×101 to 5×107 copies per reaction. Linearity was assessed by a standard curve produced from the Ct values generated from amplification of the 5×107 to the 5×101 copies per reaction (i.e. seven data points) and was calculated by linear regression. The linear regressions of the three standard curves were R2 = 0.999 for qnrA, R2 = 0.999 for qnrB and R2 = 0.998 for qnrS, indicating a strong linear correlation between Ct value and target gene concentration. The efficiencies of the amplifications were 87% for qnrA (95CI; 85–90%), 91% for qnrB (95CI; 89–92%) and 87% qnrS (95CI; 81–92%).
Baseline study population characteristics
Three hundred children were enrolled for the purposes of this study, resulting in a total of 600 rectal swabs. Bacterial colony counting and quantitative real-time PCR for all three loci were performed on all 600 samples. For subsequent data analysis we required accurate PCR amplification and a bacterial colony count on all 600 stool samples and antimicrobial usage data from each of the 300 enrolees. We rejected 37 paired samples for the purposes of our analysis. From these rejected paired samples, 19 samples had a negative internal control, 16 samples had less than 30 colonies cultured and two enrolees did not have available antimicrobial usage data. Therefore, paired data from 263 enrolled children were available for analysis.
The resulting baseline data for the detection of qnrA, qnrB and qnrS on Day 0 from 263 children is shown in Table 2. The children were resident in 18 different districts in HCMC, which we divided into three groups based on population density, an a priori variable that we believed may confound the relationship between local qnr gene exposure and circulation (Table 2). The median age of enrolees was 12 months (range: 2 months to 12 years) and the male to female ratio was 1∶0.83. On Day 0, we detected qnrA in 12 samples (4.5%), qnrB in 116 samples (44.1%) and qnrS in 196 samples (74.5%). We found no significant association between age, gender or population density of the participants with qnr gene prevalence on Day 0 using Fisher's exact test (Table 2).
Prevalence and quantification of qnr genes in stool samples from children with ARIs
We compared the proportion of rectal swabs demonstrating a PCR amplification signal for qnrA, qnrB and qnrS using paired samples taken on Day 0 and Day 7 (Table 3). All three loci demonstrated a proportional increase in the number of samples testing positive for the qnr loci. The percentage of samples testing positive increased by 4.2%, 24% and 4.2% for qnrA, qnrB and qnrS, respectively. The comparative difference between these samples was assessed by McNemar's test; only qnrB demonstrated a statistically significant increase (p<0.001).
To compare the number of Enterobacteriaceae in the intestinal flora before and after the use of antimicrobials, the CFU ml−1 of Enterobacteriaceae on MacConkey agar were calculated. We did not formally assess a coefficient of variance for the colony counts from rectal swabs. However, whilst developing this technique for the purposes of this investigation, 20 samples were tested in triplicate and the counts were found to have tolerable variation (+/−10% of the primary value). The median CFU ml−1 in paired samples demonstrated a significant increase from 1.26×107 (range; 8×102 to 9×108) on Day 0 to 3.26×107 (range 7.8×102 to 5×1011) on Day 7 (p<0.001, paired Wilcoxon rank-sum) (Table 3).
By the use of a standard curve on each independent run, we calculated the copy number of qnrA, qnrB and qnrS in the 526 stool samples. The median copy numbers of qnrA, qnrB and qnrS on Day 0 were 7.4×104 (range: 2.6×103 to 5.2×107), 4.8×104 (range: 2.6×103 to 7.8×107) and 2.8×105 (range: 2.7×103 to 7.9×108), respectively. The median copy numbers of qnrA, qnrB and qnrS on Day 7 were 2.2×104 (range: 2.7×103 to 2.8×107), 1.3×105 (range: 2.7×103 to 4.2×1010) and 1.1×106 (range: 4×103 to 2.2×109), respectively. The qnrS gene had the highest copy numbers of the three-qnr loci on both day 0 and day 7 (Figure 1). After adjusting copy number for Enterobacteriaceae CFU ml−1, qnrB demonstrated a significant increase between Day 0 and Day 7 (p<0.001, paired Wilcoxon rank-sum) (Table 3).
Box plots showing the median and interquartile ranges of qnrA, qnrB and qnrS gene copy numbers in stool samples collected from children with ARIs on enrolment (Day 0) and after antimicrobial treatment (Day 7). Statistical significance between the qnr genes was calculated using the paired Wilcoxon rank-sum test; significant variation in gene copy number between the qnr genes is denoted at the head of the figure (*), all p values<0.001.
Comparing antimicrobial usage and qnr gene copy number
We stratified the copy numbers of qnrA, qnrB and qnrS in the 263 enrolled children by their prescribed antimicrobial treatment regimen (Table 4). Twelve different antimicrobials, belonging to three different classes; β-lactam and β-lactamase inhibitor combination (amoxicillin-clavulanic acid), cephalosporins (cefaclor, cefadroxil, cefixime, ceftriaxone, cefuroxime, cefpodoxime) and macrolides (azithromycin, erythromycin, roxithromycin, spiramycin) were used to treat enrolled participants. Notably, no children were prescribed fluoroquinolones. Seven different antimicrobial regimens were prescribed, in the form of mono-therapy or in a combination of two or more antimicrobials (Table 4).
We analysed the CFU ml−1 of the Enterobacteriaceae with each treatment regimen between the two time points. Three groups (co-amoxiclav, cephalosporin only and both combined) demonstrated a significant increase in Enterobacteriaceae CFU ml−1 between the two time points (p<0.05) (Table 4). Again we adjusted the gene copy numbers from day 0 and day 7 for Enterobacteriaceae CFU ml−1 within the treatment regimes. There was a significant increase in copy number between Day 0 and Day 7 for the qnrA and qnrB loci in those treated with co-amoxiclav and co-amoxiclav combined with a cephalosporin (Table 4).
PMQR genes stimulate low level resistance to fluoroquinolones, yet their ability to facilitate other fluoroquinolone resistance mechanisms and their promiscuity threatens the long-term efficacy of the fluoroquinolones . There has been intense investigation of PMQR genes across a wide diversity of bacterial species, with some studies focusing on detecting qnr, aac(6′)-Ib-cr and qepA genes using real-time PCR or by pyrosequencing in clinical isolates , , , . Here, for the first time, we have used real-time PCR to detect and quantify qnr genes in nucleic acid extracted directly from rectal swabs of patients treated with antimicrobials.
This observational real-time PCR investigation for qnrA, qnrB and qnrS on paired stool samples from children with ARIs living in HCMC reveals several new insights into the prevalence and selection of qnr genes in this location. Notably, the prevalence of qnrB increased between the day of enrolment and seven days after antimicrobial therapy, even after stratification by the number of Enterobacteriaceae. Our data suggest that this increase is related to selection of qnrB through antimicrobial exposure, either by direct selection or enrichment of existing qnrB containing organisms in the gastrointestinal tract. The uptake and maintenance of qnrB is likely through contaminated food or constant exposure to organisms with this gene from known environmental sources , , . Moreover, co-existence of qnrB with other resistance genes, such as blaCTX-M-14 or blaCTX-M-15, on the same plasmid is a well-known phenomenon , . In our previous study of qnr genes in organisms isolated in the hospital and the community in HCMC, we found that 11%, 12% and 45% of bacterial isolates from hospitalised individuals were positive for qnrA, qnrB and qnrS, respectively, and 0.7%, 0.5%, 12% from community isolates were positive for qnrA, qnrB and qnrS, respectively . Here, we detected a substantially higher prevalence of the PMQR genes than before. Explanations for this disparity include the use of updated primers (i.e. the ability to detect newer alleles); differences in the target populations (bacterial isolates vs. metagenomic approach on stool samples); an increased sensitivity of real-time vs. conventional PCR method; and a difference of four years, which may reflect a natural increase in the circulation of organisms (and people) carrying PMQR genes.
There have been several reports related to antimicrobial usage and an increase in resistant bacteria, yet this is the first describing evidence of the co-selection of PMQR genes in this manner , . Both qnrB and qnrS demonstrated significant copy number increases between Day 0 and Day 7. Yet, none of the individuals enrolled in this study were treated with fluoroquinolones and instead were prescribed amoxicillin/clavulanic acid, macrolides or cephalosporins. This increase in copy number over the two time points, apparently via exposure to these antimicrobials, particularly amoxicillin/clavulanic acid or amoxicillin/clavulanic acid combined with a cephalosporin, substantiates our co-selection hypothesis. Indeed, additional selection is the most likely mechanism, as PMQR plasmids frequently contain other resistance genes, particularly determinants encoding resistance to β-lactams , , , , . Our currently unpublished data related to PMQR plasmids and the genetic background of qnrS1 containing mobile elements in HCMC reveals a predominant qnrS1-containing transposon type carrying the blaLAP-2 gene, which also encodes resistance to β-lactams. Additional reports have also shown an intimate association between qnrA and qnrB with ESBL genes , , , , .
It is predicted that the number and structure of the bacteria population in the human gut depends on health status, diet and antimicrobial usage , , . Here, by using MacConkey agar, we investigated potential differences in the number of Enterobacteriaceae between two time points . It is known that the diversity and the abundance of organisms in stool is great and includes a variety of Bacteroides, Clostridia, Fusobacteria and Peptostreptococci , , all of which may contribute to the overall differences observed in PMQR gene prevalence and copy number. However, we observed a significant increase in the CFU ml−1 of Enterobacteriaceae between the two time points, implying a strong selection of these organisms, potentially through antimicrobial resistance enrichment. We found a significant increase in CFU ml−1 in rectal swabs between Day 0 and Day 7 related to three antimicrobial treatment regimens: amoxicillin/clavulanic acid only, a cephalosporin only, or amoxicillin/clavulanic acid combined with a cephalosporin. However, only amoxicillin/clavulanic acid seemed to influence qnr gene copy number (before and after adjusting for CFU ml−1 of Enterobacteriaceae), implying that the increase of qnr genes is not only a consequence of Enterobacteriaceae enrichment but these genes may also be shared by other bacterial species within the intestinal tract . Our results clearly show a major shift in qnr gene copy numbers after antimicrobial therapy, yet our findings are limited by a lack of control group who did not receive antimicrobials. The results presented here warrant asking additional questions related to the effect of antimicrobials on the presence and maintenance of qnr genes and other antimicrobial resistance genes in the gut flora. We are currently longitudinally following a cohort of children in HCMC over a two-year period with and without antimicrobial treatment to address natural fluctuations of resistance genes and changes in Enterobacteriaceae.
In conclusion, our data demonstrate an increasing prevalence of qnrB and an increasing quantity of the qnrB and qnrS genes in the stools of children with ARIs between enrolment and after seven days treatment with non-fluoroquinolone antimicrobials. This is the first study describing an association between the use of non-quinolone antimicrobials and the increasing relative prevalence in qnr gene copy number in the gut flora. Our work highlights the rampant nature of PMQR genes in this locality and suggests aggressive co-selection of these resistance determinants through the use of unrelated and potentially unnecessary antimicrobial treatment regimes.
We wish to acknowledge the ongoing efforts of the microbiology laboratory of the Hospital for Tropical Diseases in HCMC, who were essential for the completion of this study. Specifically Pham Quynh Vi and Nguyen Ngoc Hung Phuc are acknowledged for the stool cultures. We also recognize the efforts of Children's Hospital 1 and wish to thank all enrollees and their parents.
Conceived and designed the experiments: SB HRvD. Performed the experiments: LTMV HDK. Analyzed the data: LTMV TVTN CT. Contributed reagents/materials/analysis tools: NNQM TCT JIC MdJ JJF CS. Wrote the paper: SB. Drafting of the article: LTMV.
- 1. Okeke IN, Laxminarayan R, Bhutta ZA, Duse AG, Jenkins P, et al. (2005) Antimicrobial resistance in developing countries. Part I: recent trends and current status. Lancet Infect Dis 5: 481–493.
- 2. Quagliarello A, Parry CM, Hien TT, Farrar J (2003) Factors associated with carriage of penicillin-resistant Streptococcus pneumoniae among Vietnamese children: a rural-urban divide. J Health Popul Nutr 21: 316–324.
- 3. Schultsz C, Vien le M, Campbell JI, Chau NV, Diep TS, et al. (2007) Changes in the nasal carriage of drug-resistant Streptococcus pneumoniae in urban and rural Vietnamese schoolchildren. Trans R Soc Trop Med Hyg 101: 484–492.
- 4. Goossens H, Ferech M, Vander Stichele R, Elseviers M (2005) Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet 365: 579–587.
- 5. Bergman M, Nyberg ST, Huovinen P, Paakkari P, Hakanen AJ (2009) Association between Antimicrobial Consumption and Resistance in Escherichia coli. Antimicrob Agents Chemother 53: 912–917.
- 6. Lautenbach E, Metlay JP, Bilker WB, Edelstein PH, Fishman NO (2005) Association between fluoroquinolone resistance and mortality in Escherichia coli and Klebsiella pneumoniae infections: the role of inadequate empirical antimicrobial therapy. Clin Infect Dis 41: 923–929.
- 7. Travers K, Barza M (2002) Morbidity of infections caused by antimicrobial-resistant bacteria. Clin Infect Dis 34 Suppl 3S131–134.
- 8. Bailey JK, Pinyon JL, Anantham S, Hall RM (2010) Commensal Escherichia coli of healthy humans: a reservoir for antibiotic-resistance determinants. Journal of medical microbiology 59: 1331–1339.
- 9. Nguyen NT, Ha V, Tran NV, Stabler R, Pham DT, et al. (2010) The sudden dominance of blaCTX-M harbouring plasmids in Shigella spp. Circulating in Southern Vietnam. PLoS Negl Trop Dis 4: e702.
- 10. Martinez-Martinez L, Pascual A, Jacoby GA (1998) Quinolone resistance from a transferable plasmid. Lancet 351: 797–799.
- 11. Zhang R, Huang YL, Cai JC, Zhou HW, Yamaguchi N, et al. (2012) High prevalence of qnr and aac(6′)-Ib-cr genes in both water-borne environmental bacteria and clinical isolates of Citrobacter freundii in China. Microbes Environ 27: 158–163.
- 12. Garcia-Fulgueiras V, Bado I, Mota MI, Robino L, Cordeiro NF, et al. (2011) Extended-spectrum beta-lactamases and plasmid-mediated quinolone resistance in enterobacterial clinical isolates in the paediatric hospital of Uruguay. J Antimicrob Chemother 66: 1725–1729.
- 13. Luo Y, Yang J, Zhang Y, Ye L, Wang L, et al. (2011) Prevalence of beta-lactamases and 16S rRNA methylase genes amongst clinical Klebsiella pneumoniae isolates carrying plasmid-mediated quinolone resistance determinants. Int J Antimicrob Agents 37: 352–355.
- 14. Garnier F, Raked N, Gassama A, Denis F, Ploy MC (2006) Genetic environment of quinolone resistance gene qnrB2 in a complex sul1-type integron in the newly described Salmonella enterica serovar Keurmassar. Antimicrob Agents Chemother 50: 3200–3202.
- 15. Poirel L, Nguyen TV, Weintraub A, Leviandier C, Nordmann P (2006) Plasmid-mediated quinolone resistance determinant qnrS in Enterobacter cloacae. Clin Microbiol Infect 12: 1021–1023.
- 16. Vien le TM, Abuoun M, Morrison V, Thomson N, Campbell JI, et al. (2011) Differential phenotypic and genotypic characteristics of qnrS1-harboring plasmids carried by hospital and community commensal enterobacteria. Antimicrob Agents Chemother 55: 1798–1802.
- 17. WHO (2002) Acute respiratory infections (Updated 2002).
- 18. Brink AJ, Cotton MF, Feldman C, Geffen L, Hendson W, et al. (2004) Guideline for the management of upper respiratory tract infections. S Afr Med J 94: 475–483.
- 19. Jacoby G, Cattoir V, Hooper D, Martinez-Martinez L, Nordmann P, et al. (2008) qnr Gene Nomenclature. Antimicrob Agents Chemother 52: 2297–2299.
- 20. van Doornum GJ, Guldemeester J, Osterhaus AD, Niesters HG (2003) Diagnosing herpesvirus infections by real-time amplification and rapid culture. J Clin Microbiol 41: 576–580.
- 21. Robicsek A, Sahm DF, Strahilevitz J, Jacoby GA, Hooper DC (2005) Broader distribution of plasmid-mediated quinolone resistance in the United States. Antimicrob Agents Chemother 49: 3001–3003.
- 22. Cattoir V, Weill FX, Poirel L, Fabre L, Soussy CJ, et al. (2007) Prevalence of qnr genes in Salmonella in France. J Antimicrob Chemother 59: 751–754.
- 23. Whichard JM, Gay K, Stevenson JE, Joyce KJ, Cooper KL, et al. (2007) Human Salmonella and concurrent decreased susceptibility to quinolones and extended-spectrum cephalosporins. Emerg Infect Dis 13: 1681–1688.
- 24. Rasmussen R (2000) Rapid Cycle Real-time PCR, Methods and Applications; Meuer, editor.
- 25. Le TM, Baker S, Le TP, Cao TT, Tran TT, et al. (2009) High prevalence of plasmid-mediated quinolone resistance determinants in commensal members of the Enterobacteriaceae in Ho Chi Minh City, Vietnam. J Med Microbiol 58: 1585–1592.
- 26. Bell JM, Turnidge JD, Andersson P (2010) aac(6′)-Ib-cr genotyping by simultaneous high-resolution melting analyses of an unlabeled probe and full-length amplicon. Antimicrobial Agents and Chemotherapy 54: 1378–1380.
- 27. Guillard T, Duval V, Moret H, Brasme L, Vernet-Garnier V, et al. (2010) Rapid detection of aac(6′)-Ib-cr quinolone resistance gene by pyrosequencing. Journal of clinical microbiology 48: 286–289.
- 28. Hidalgo-Grass C, Strahilevitz J (2010) High-resolution melt curve analysis for identification of single nucleotide mutations in the quinolone resistance gene aac(6′)-Ib-cr. Antimicrobial Agents and Chemotherapy 54: 3509–3511.
- 29. Guillard T, Moret H, Brasme L, Carlier A, Vernet-Garnier V, et al. (2011) Rapid detection of qnr and qepA plasmid-mediated quinolone resistance genes using real-time PCR. Diagnostic microbiology and infectious disease 70: 253–259.
- 30. Huang SY, Dai L, Xia LN, Du XD, Qi YH, et al. (2009) Increased prevalence of plasmid-mediated quinolone resistance determinants in chicken Escherichia coli isolates from 2001 to 2007. Foodborne Pathog Dis 6: 1203–1209.
- 31. Ma J, Zeng Z, Chen Z, Xu X, Wang X, et al. (2009) High prevalence of plasmid-mediated quinolone resistance determinants qnr, aac(6′)-Ib-cr, and qepA among ceftiofur-resistant Enterobacteriaceae isolates from companion and food-producing animals. Antimicrob Agents Chemother 53: 519–524.
- 32. Yue L, Jiang HX, Liao XP, Liu JH, Li SJ, et al. (2008) Prevalence of plasmid-mediated quinolone resistance qnr genes in poultry and swine clinical isolates of Escherichia coli. Vet Microbiol 132: 414–420.
- 33. Strahilevitz J, Jacoby GA, Hooper DC, Robicsek A (2009) Plasmid-mediated quinolone resistance: a multifaceted threat. Clin Microbiol Rev 22: 664–689.
- 34. Pomba C, da Fonseca JD, Baptista BC, Correia JD, Martinez-Martinez L (2009) Detection of the pandemic O25-ST131 human virulent Escherichia coli CTX-M-15-producing clone harboring the qnrB2 and aac(6′)-Ib-cr genes in a dog. Antimicrob Agents Chemother 53: 327–328.
- 35. Alali WQ, Scott HM, Norby B, Gebreyes W, Loneragan GH (2009) Quantification of the bla(CMY-2) in feces from beef feedlot cattle administered three different doses of ceftiofur in a longitudinal controlled field trial. Foodborne Pathog Dis 6: 917–924.
- 36. van der Veen EL, Schilder AG, Timmers TK, Rovers MM, Fluit AC, et al. (2009) Effect of long-term trimethoprim/sulfamethoxazole treatment on resistance and integron prevalence in the intestinal flora: a randomized, double-blind, placebo-controlled trial in children. J Antimicrob Chemother 63: 1011–1016.
- 37. Chmelnitsky I, Navon-Venezia S, Strahilevitz J, Carmeli Y (2008) Plasmid-mediated qnrB2 and carbapenemase gene bla(KPC-2) carried on the same plasmid in carbapenem-resistant ciprofloxacin-susceptible Enterobacter cloacae isolates. Antimicrob Agents Chemother 52: 2962–2965.
- 38. Jiang Y, Zhou Z, Qian Y, Wei Z, Yu Y, et al. (2008) Plasmid-mediated quinolone resistance determinants qnr and aac(6′)-Ib-cr in extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae in China. J Antimicrob Chemother 61: 1003–1006.
- 39. Poirel L, Cattoir V, Soares A, Soussy CJ, Nordmann P (2007) Novel Ambler class A beta-lactamase LAP-1 and its association with the plasmid-mediated quinolone resistance determinant QnrS1. Antimicrob Agents Chemother 51: 631–637.
- 40. Saito R, Kumita W, Sato K, Chida T, Okamura N, et al. (2007) Detection of plasmid-mediated quinolone resistance associated with qnrA in an Escherichia coli clinical isolate producing CTX-M-9 beta-lactamase in Japan. Int J Antimicrob Agents 29: 600–602.
- 41. Shen P, Jiang Y, Zhou Z, Zhang J, Yu Y, et al. (2008) Complete nucleotide sequence of pKP96, a 67 850 bp multiresistance plasmid encoding qnrA1, aac(6′)-Ib-cr and blaCTX-M-24 from Klebsiella pneumoniae. J Antimicrob Chemother 62: 1252–1256.
- 42. Chen SL, Hung CS, Xu J, Reigstad CS, Magrini V, et al. (2006) Identification of genes subject to positive selection in uropathogenic strains of Escherichia coli: a comparative genomics approach. Proc Natl Acad Sci U S A 103: 5977–5982.
- 43. Hu FP, Xu XG, Zhu DM, Wang MG (2008) Coexistence of qnrB4 and qnrS1 in a clinical strain of Klebsiella pneumoniae. Acta Pharmacol Sin 29: 320–324.
- 44. Kehrenberg C, Hopkins KL, Threlfall EJ, Schwarz S (2007) Complete nucleotide sequence of a small qnrS1-carrying plasmid from Salmonella enterica subsp. enterica Typhimurium DT193. J Antimicrob Chemother 60: 903–905.
- 45. Clarke JS (1974) Bacteriology of the gut and its clinical implications. West J Med 121: 390–403.
- 46. Levy J (2000) The effects of antibiotic use on gastrointestinal function. Am J Gastroenterol 95: S8–10.
- 47. Kirjavainen PV, Gibson GR (1999) Healthy gut microflora and allergy: factors influencing development of the microbiota. Ann Med 31: 288–292.
- 48. Blood RM, Curtis GD (1995) Media for ‘total’ Enterobacteriaceae, coliforms and Escherichia coli. Int J Food Microbiol 26: 93–115.
- 49. Madigan MT, Martinko JM, Parker J (2003) Brock biology of microorganisms. New Jersey: Prentice Hall.
- 50. Moore WE, Holdeman LV (1974) Human fecal flora: the normal flora of 20 Japanese-Hawaiians. Appl Microbiol 27: 961–979.