| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
Laboratoire de Recherche en Bactériologie, Institut Pasteur de Nouvelle-Calédonie, Nouvelle-Calédonie
aAddress correspondence to this author at: Institut Pasteur of New Caledonia, BP 61 98845 Noumea cedex, New Caledonia. Fax 00-687-27-33-90; e-mail fmerien{at}pasteur.nc
Abstract
Background: For many years, the pathogenic bacterium Neisseria gonorrhoeae, the etiologic agent of gonorrhea, was generally susceptible to penicillin, until the emergence of resistant strains. Well-characterized genetic variations in the penicillin resistance–determining region correlate with decreased susceptibility to penicillin. At least 5 genes (penA, penB, mtrR, ponA, and penC) are involved in the chromosomally mediated resistance to this antibiotic. To date, no development of multiplex PCR assays targeting a range of gonococcal genes and variations as a means of predicting antibiotic resistance has been reported.
Methods: The aim of this study was to develop a duplex assay using DNA from isolated strains. We describe the development and evaluation on the LightCycler platform of a real-time duplex PCR assay (hybridization probe format) for rapid and specific detection of ponA and penA variations, predicting penicillin susceptibilities.
Results: The real-time duplex PCR assay successfully detected variations in ponA and penA genes by use of distinct melting temperatures from a total of 120 Neisseria gonorrhoeae isolates. Moreover, the variation profiles obtained with the real-time PCR and the melting analysis showed good correlation with the pattern of penicillin susceptibility generated with classical antibiograms. Nucleotide sequencing data were in complete agreement with multiplex assay results.
Conclusions: The presented assay is suitable for the detection of chromosomally mediated resistant strains of Neisseria gonorrhoeae in genotyping studies and could be valuable in the effective antimicrobial strategy to gonococci.
In many countries worldwide, the temporary decrease in the incidence of gonorrhea since the 1980s was accompanied by an increase in the frequency of strains of chromosomally mediated resistant Neisseria gonorrhoeae (CMRNG).1 For example, for the past decade in Japan, there has been a dramatic increase in the prevalence of gonococcal isolates exhibiting chromosomally mediated resistance to penicillin, tetracycline, fluoroquinolones, and oral cephems (1)(2)(3). Widespread prescription of these drugs, often coupled with misuse, has led to the development of multidrug-resistant strains that pose a real threat to human health. Interestingly, in New Caledonia (South Pacific) all N. gonorrhoeae (NG) isolates were susceptible to penicillin until the emergence in late 2004 of strains with minimum inhibitory concentrations (MICs) of 
0.125 mg/L. In the gonococci, the chromosomally mediated expression of multiple resistance genes and the production of the penicillinase plasmid TEM-1 ß lactamase are 2 independent mechanisms involved in a susceptible-to-resistant strain switch. Penicillin-binding proteins (PBPs) 1 and 2 of NG are the major targets of ß-lactam antibiotics. Resistance to penicillin in non–ß-lactamase-producing strains of NG is mediated in part by the production of altered forms of PBP 2 that have a decreased affinity for penicillin. This reduction is largely, although not exclusively, due to the insertion of an aspartic acid residue (Asp-345A) into the amino acid sequence of the PBP 2 penA gene (4). This alteration led to a mosaic structure in the transpeptidase region of this gene and was associated with a decreased susceptibility to penicillin in all NG strains with MIC 
0.03 mg/L (5). In the same way, a single–amino acid variation in the ponA gene that encodes PBP 1 is due to a single base change and was reported in all CMRNG strains for which MICs of penicillin were 1 mg/L (6).
Recent advances in PCR have made it possible to use clinical samples to characterize both the pathogenic agent and its susceptibility to antimicrobial drugs (7)(8)(9)(10). A recently validated real-time PCR assay (10) has been used for rapid detection of a ponA variation associated with a decreased susceptibility to penicillin. This method confirmed the phenotype, but the variation in the ponA gene was not the sole alteration involved. Because the resistance mechanisms are based on stepwise accumulation of point variations correlating with increased MICs (11)(12)(13), an improvement was imperative to differentiate NG strains with intermediate susceptibility to penicillin (0.06 mg/L < MIC < 1 mg/L). In this study, we used both conventional bacteriologic methods and a novel real-time PCR assay to identify all gonococci isolated in New Caledonia since July 2003 and to determine whether phenotypes obtained by E-test correlate with genotypes identified by melting curve analysis after amplification of specific gene sequences.
We collected a total of 120 NG strains in the bacteriology laboratory of the Institut Pasteur in New Caledonia. The identity of the organism was confirmed by specific characteristics after culture on chocolate agar supplemented with PolyVitex plus colistin, vancomycin, amphotericin B, and trimethoprim antibiotics (bioMérieux SA). MICs of penicillin were determined by E-test (AB Biodisk) on chocolate agar PolyVitex. To check the presence of ß-lactamase, all clinical isolates were subjected to cefinase test (bioMérieux SA, France); only 1 strain showed a positive result, indicating the presence of plasmid-mediated penicillin resistance. The real-time monoplex PCR (10) allowed rapid screening of clinical samples for penicillin-resistant NG strains. Two hybridization probes were used to distinguish the wild-type ponA gene from the variant gene (Leu421
Pro). In our novel duplex PCR, 2 additional hybridization probes and primers were used to enable real-time PCR detection of the aspartic acid residue (Asp-345A) in the NG penA gene. Template DNA was purified from positive cultures with a QIAamp DNA Minikit (Qiagen) according to the manufacturer’s instructions. Primers, as well as probes, were designed with the LightCycler Probe Design software version 2.0 (Roche Diagnostics) and were synthesized by Proligo Singapore Pte. Ltd. Primers and probes targeting the ponA gene were previously described (10). Oligonucleotides targeting penA included the following: penA-F (5'-GAC CTT GGC CTA TGA AGA G-3'), penA-R (5'-GGT TAC GGA AAC CAT CAG ATT-3'), sensor probe penA-P1 (5'-CGC GAC GAT ACC CAT GTT TAC C-fluorescein-3'), and anchor probe penA-P2 (5'-LC-Red 640-TCT TTG GAT GTG CGC GGC ATT ATG CA-phosphate-3'). Reactions were performed in 20-µL capillaries (2 µL of sample DNA and 18 µL of master mix) as previously described (10), with final concentrations of probes of 0.2 µM. As in our original assay (10), we used asymmetric primer concentrations (forward primers, 0.5 µmol/L; reverse primers, 1 µmol/L) as recommended by Barratt and Mackay (14) to improve real-time signal. Controls in every test assay included blanks and wild-type strains, strains with one variation, and strains with both variations, as determined by sequencing. Fluorescence was measured during the annealing step as a ratio of the emitted fluorescence in channel 640 (penA insert) or in channel 705 (ponA variation) to emitted fluorescence in channel 530. Color compensation to eliminate any fluorescent signal of LC-Red 640 in channel 705 was performed according to manufacturer’s instructions. Analytical sensitivity was determined with serial 10-fold dilutions of gonococcal DNA; amplification was observed in a corresponding interval of 5 x 109 to 5 x 102 genome copies per reaction according to Overbergh’s formula (15). To investigate the specificity of our test, we also tested 2 strains of Neisseria lactamica [Collection Institut Pasteur (CIP) 73.27, CIP 107857] and 1 Neisseria cinerea strain (CIP 73.16T) obtained from the CIP in France. To further investigate the specificity of our assay, the DNA of non-Neisseria species, including Mycoplasma hominis, Ureaplasma urealyticum, Trichomonas vaginalis, Candida albicans, and other gram-positive and gram-negative bacteria (Gardnerella vaginalis, Streptococcus agalactiae, Streptococcus faecalis, Staphylococcus aureus, and Escherichia coli) were also purified and subjected to our real-time assay. No amplification was detected for these strains, a result that confirmed the specificity of our duplex assay.
Oligonucleotide hybridization and melting analysis with fluorescein and LC-Red 640 probes generated 2 distinct melting temperature (Tm) curves (Fig. 1A
), permitting differentiation between wild-type and insert Asp-345A in penA genes. Differentiation of the single-base variation in the ponA gene was also possible with fluorescein and LC-Red 705 probes (Fig. 1B
). Variations in Tm (1.0 °C) may occur because amounts of DNA from culture may be much larger than those obtained from patients samples (data not shown), but such variations did not affect the Tm shift or the shape of the curves in our study. Of 120 NG strains, 59 showed MICs for penicillin of 0.008–0.06 mg/L (PenS), 43 were classified as intermediate (PenI) with MICs of 0.06–0.75 mg/L, and 18 as resistant (PenR) with MICs 1 mg/L (Table 1
).
|
|
It is noteworthy that 36 of 38 PenS with MIC 
0.016 mg/L peaked at a mean (SD) Tm of 60.0 (0.5) °C, specific for the wild-type penA genotype, whereas 2 peaked at 65.0 (0.5) °C, indicating the presence of the Asp-345A insert. The altered penA Tm was observed for 3 of 9 strains with MIC = 0.032 mg/L, and the Asp-345A+ Tm for 9 of 12 strains with MIC = 0.047 mg/L. All of these NG PenS were wild-type ponA [Tm 61.0 (0.5) °C].
Residue Asp-345A was found in 37 of 43 PenI. Except for 1 strain, all wild-type penA among the PenI strains showed an MIC to penicillin of 0.064 mg/L. We considered an MIC of 0.094 mg/L as a cutoff point because all NG strains (except 1) with MIC >0.064 mg/L had penA genes with Tms indicating the presence of the Asp-345A insertion. Complete concordance was demonstrated except for 1 discrepancy, which suggested a possible contribution of other chromosomally mediated resistance mechanisms. Interestingly, 11 of the PenI strains also showed a variant ponA genotype, indicating that this alteration was present in fewer than expected penicillin phenotypes. All PenR strains included in this study showed the presence of both penA and ponA variant genotypes. The biological significance of these findings is related to specific mechanisms involved in chromosomally mediated antibiotic resistance to penicillin in NG. The 5 genes involved in resistance to penicillin interact in a specific manner. Briefly, in transformation and homologous recombination experiments (4), genes are transferred from a resistant strain to a susceptible strain in a specific order, the altered ponA and penC genes being the last to be reported in the sequence of events leading to resistance. Thus, no apparent increase in resistance occurred in strains with ponA variations unless penA, mtrR, penB, and penC resistance determinants were also present. These specific relationships could explain why the phenotypical absence of resistance in strains with low penicillin MICs, despite the presence of the variant ponA gene. In addition to insertion, simultaneous variations in penA can also contribute to resistance (16). The presence of strains harboring the insertion in penA but lacking these previous point variations may also explain the detection of low MICs of penicillin.
Finally, DNAs of 10 NG strains with different genotypes differentiated by their Tm were amplified, purified, and sent to the Auckland University DNA sequencing facility (Auckland, New Zealand). Nucleotide sequencing data were in complete agreement with multiplex assay results. All experiments were performed in triplicate, and our results were fully reproducible. They were confirmed by conventional sequencing and correlated with phenotypes obtained by E-test. Our protocol was shown to be a good method for CMRNG detection and genotyping studies. Moreover, our method could be a valuable molecular tool in an effective antimicrobial strategy to control gonococci.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  F. Vernel-Pauillac, S. Nandi, R. A. Nicholas, and C. Goarant Genotyping as a Tool for Antibiotic Resistance Surveillance of Neisseria gonorrhoeae in New Caledonia: Evidence of a Novel Genotype Associated with Reduced Penicillin Susceptibility Antimicrob. Agents Chemother., September 1, 2008; 52(9): 3293 - 3300. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
