J Antimicrob Chemother 2016; 71: 353 – 356
doi:10.1093/jac/dkv366 Advance Access publication 3 November 2015
A real-time PCR assay for direct characterization of the
Neisseria gonorrhoeae GyrA 91 locus associated with
ciprofloxacin susceptibility
Cameron Buckley1,2, Ella Trembizki1,2, Basil Donovan3, Marcus Chen4,5, Kevin Freeman6,
Rebecca Guy3, Ratan Kundu7, Monica M. Lahra7, David G. Regan3, Helen Smith8 and David M. Whiley1,2,9*
on behalf of the GRAND study investigators†
1
Queensland Paediatric Infectious Diseases Laboratory, Queensland Children’s Medical Research Institute, Brisbane, Queensland 4029,
Australia; 2University of Queensland Child Health Research Centre, Brisbane, Queensland 4029, Australia; 3Kirby Institute, UNSW Australia,
Sydney, New South Wales 2052, Australia; 4Melbourne Sexual Health Centre, Alfred Health, Carlton, Melbourne, Victoria 3053, Australia;
5
Central Clinical School, Monash University, Melbourne, Victoria 3181, Australia; 6Microbiology Laboratory, Pathology Department, Royal
Darwin Hospital, Darwin, Northern Territory, Australia; 7WHO Collaborating Centre for STD, Microbiology Department, South Eastern Area
Laboratory Services, Prince of Wales Hospital, Sydney, New South Wales 2031, Australia; 8Public Health Microbiology, Communicable
Disease, Queensland Health Forensic and Scientific Services, Archerfield, Brisbane, Queensland, Australia; 9UQ Centre for Clinical Research
(UQCCR), The University of Queensland, Brisbane, Queensland 4029, Australia
*Corresponding author. E-mail: [email protected]
†Other GRAND study investigators are listed in the Acknowledgements section.
Received 30 July 2015; returned 5 September 2015; revised 5 October 2015; accepted 8 October 2015
Objectives: The objective of this study was to develop a real-time PCR method for specific detection of the gonococcal GyrA amino acid 91 locus directly in clinical samples so as to predict Neisseria gonorrhoeae ciprofloxacin
susceptibility.
Methods: The real-time PCR assay, GyrA91-PCR, was designed using two probes, one for detection of the WT S91
sequence and the other for detection of the S91F alteration. The performance of the assay was initially assessed
using characterized N. gonorrhoeae isolates (n ¼ 70), a panel of commensal Neisseria and Moraxella species
(n ¼ 55 isolates) and clinical samples providing negative results by a commercial N. gonorrhoeae nucleic acid
amplification test (NAAT) method (n ¼ 171). The GyrA91-PCR was then applied directly to N. gonorrhoeae
NAAT-positive clinical samples (n ¼ 210) from the year 2014 for which corresponding N. gonorrhoeae isolates
with susceptibility results were also available.
Results: The GyrA91-PCR accurately characterized the GyrA 91 locus of all 70 N. gonorrhoeae isolates
(sensitivity ¼ 100%, 95% CI ¼ 94.9% – 100%), whereas all non-gonococcal isolates and N. gonorrhoeae NAATnegative clinical samples gave negative results by the GyrA91-PCR (specificity ¼ 100%, 95% CI ¼ 98.4% –
100%). When applied to the 210 N. gonorrhoeae NAAT-positive clinical samples, the GyrA91-PCR successfully
characterized 195 samples (92.9%, 95% CI ¼ 88.5% –95.9%). When compared with the corresponding bacterial
culture results, positivity by the GyrA91-PCR WT probe correctly predicted N. gonorrhoeae susceptibility to ciprofloxacin in 161 of 162 (99.4%, 95% CI¼ 96.6% –99.9%) samples.
Conclusions: The use of a PCR assay for detection of mutation in gyrA applied directly to clinical samples can
predict ciprofloxacin susceptibility in N. gonorrhoeae.
Introduction
Concerns have continued to escalate over the rise and potential
impact of antimicrobial-resistant Neisseria gonorrhoeae. It is
now listed as an urgent antimicrobial resistance threat by the
US CDC.1 Gonococcal resistance has now been observed to all
first-line antimicrobials recommended for treatment of gonorrhoea, including the extended-spectrum cephalosporins cefixime
and ceftriaxone,2 although ceftriaxone resistance has been sporadic. therefore a need to consider new potential treatment strategies for gonorrhoea.
One proposed strategy is the recycling of previously effective
antibiotics via the use of molecular methods to predict antibiotic
susceptibility and inform individualized treatment.3 Of the various
treatments previously used to treat gonorrhoea, ciprofloxacin
appears to be the most suitable candidate for individualized
# The Author 2015. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: [email protected]
353
Buckley et al.
treatment for a number of reasons: (i) it can be given to the
patient as a single observed oral treatment, is well tolerated
with few side effects and has excellent pharmacokinetics at
both urogenital and extra-genital sites; (ii) the genetics of ciprofloxacin resistance are relatively simple and thus amenable to
assay design; and (iii) in Australia 60% of all N. gonorrhoeae isolates are ciprofloxacin susceptible.4 Thus, despite being unsuitable
for empirical treatment on the basis of being well above the WHO
5% resistance threshold,5 up to 60% of infections in Australia
could potentially be treated with ciprofloxacin if suitable susceptibility testing strategies were available. The aim of this study was
to develop a real-time PCR method to predict N. gonorrhoeae
ciprofloxacin susceptibility directly in clinical samples.
N. gonorrhoeae and non-gonococcal isolates
The performance of the assay was initially assessed using N. gonorrhoeae
isolates (n¼70) for which the GyrA 91 locus had previously been characterized using the Sequenom iPLEX mass-array platform as part of an
ongoing national study of gonococcal resistance in our country.11 The isolates represented the most common genotypes found within Australia in
the year 2012 and comprised 42 WT (S91) strains that were susceptible
to ciprofloxacin and 28 ciprofloxacin-resistant strains with the S91F alteration. To assess specificity, a panel of commensal Neisseria and Moraxella
species (n ¼55 isolates; Table 2) was also tested. The detection limit of the
GyrA91-PCR assay was compared with that of a previously described
N. gonorrhoeae diagnostic assay targeting the gonococcal porA pseudogene12 by testing 10-fold dilutions of N. gonorrhoeae isolate DNA.
N. gonorrhoeae nucleic acid amplification test
(NAAT)-negative clinical samples
Methods
Assay design, reaction mixture and cycling conditions
The GyrA91-PCR assay targeted amino acid 91 of the gonococcal GyrA
protein. The decision to only target this amino acid was made on the
basis of previous studies showing that alterations at this position are highly
associated with ciprofloxacin resistance and that, while other alterations
in GyrA and ParC are also implicated in N. gonorrhoeae ciprofloxacin resistance, these typically only occur in parallel with GyrA 91 alterations.6 – 9 In
developing the GyrA91-PCR assay, we sought to limit the potential for
cross-reaction with commensal Neisseria strains so that the method
could be also used on extra-genital sites, including the pharynx, where
commensal Neisseria species are ubiquitous and may otherwise undermine assay specificity. The reaction mixture of the GyrA91-PCR assay consisted of 10.0 mL of QuantiTect Probe PCR Master Mix (Qiagen, Australia),
10.0 mM each of forward and reverse primers (Table 1), 0.2 mM each of
probe 1 and probe 2, and 2.0 mL of nucleic acid extract, made up to a
final reaction volume of 20 mL using DNase-free water. The primers were
designed based on available sequence information in the GenBank database to enable specific detection of N. gonorrhoeae gyrA. The forward primer included one non-template base (in lower case; Table 1), which was
included to limit a potential 3′ dimer as well as to further limit the potential
for cross-reaction with commensal Neisseria species. We have previously
used this non-template base approach to enhance the specificity of assays
targeting the gonococcal 23S rRNA genes.10 For the probes, probe 1 targeted the WT serine (S91) amino acid, whereas probe 2 was designed to
detect the phenylalanine (S91F) alteration. Amplification and detection
were achieved on the Rotorgene Q real-time PCR instrument using the following two-step cycling conditions: an initial hold at 958C for 15 min, followed by 55 cycles at 958C for 15 s and finally 608C for 60 s, with
fluorescent signal read on both the green and red detection channels.
Positivity for either probe was then discriminated on the basis of relative
fluorescent signal in each channel using the Rotorgene Q allelic discrimination software tool.
Specificity was further assessed by testing samples that had given
negative results for N. gonorrhoeae by the Cobas 4800 CT/NG method
(Roche Diagnostics, Australia). These samples (n ¼ 171) were submitted
to Pathology Queensland for routine gonorrhoea testing, and comprised
27 anorectal swabs, 49 cervical swabs, 24 vaginal swabs, 34 pharyngeal
swabs, 33 urine specimens and 4 samples where the anatomical site was
not specified.
N. gonorrhoeae NAAT-positive clinical samples
The GyrA91-PCR assay was applied to 210 N. gonorrhoeae NAAT-positive
clinical samples from Pathology Queensland (n ¼ 70) and Royal Darwin
Hospital Pathology (n¼140) from the year 2014. The samples were stored
at 2208C following DNA extraction, prior to screening in early 2015. These
samples were selected on the basis that N. gonorrhoeae had been isolated
by bacterial culture and that phenotypic-based ciprofloxacin susceptibility
results were available (Table 2). These samples comprised 14 anorectal
swabs, 7 pharyngeal swabs, 46 penile swabs, 26 cervical swabs, 27 vaginal
swabs, 3 genital swabs (site unspecified), 3 joint fluids, 1 aspirate (site
unspecified) and 83 urine samples.
Results and discussion
When tested by the GyrA91-PCR, all 70 N. gonorrhoeae isolates
gave results that were consistent with their recognized characteristics (sensitivity ¼ 100%, 95% CI ¼ 94.9% – 100%; Table 2).
All non-gonococcal isolates and N. gonorrhoeae NAAT-negative
clinical samples gave negative results by the GyrA91-PCR
(specificity¼ 100%, 95% CI ¼ 98.4% –100%; Table 2). The detection limit of the GyrA91-PCR was equivalent to the previously
Table 1. Primers and probes targeting the gonococcal gyrA gene
Primers and probes
GyrA91-F
GyrA91-R
GyrA91-Probe1
GyrA91-Probe2
a
Sequence (5′ to 3′ )
Target sequencea
Source
GCGACGGCCTAAAGCCaGTGb
GTCTGCCAGCATTTCATGTGAG
FAM-CGGCGATTCCGCAGT-BHQplus-1
Quasar 670-CGGCGATTTCGCAGTT-BHQplus-2
137– 156
414– 435
264– 278
264– 279
Integrated DNA Technologies, Australia
Integrated DNA Technologies, Australia
Biosearch Technologies, USA
Biosearch Technologies, USA
Nucleotide positions are based on GenBank accession number U08817.
The lower-case adenine base within the GyrA91-F primer indicates the non-template base.
b
354
JAC
A PCR for gonococcal ciprofloxacin susceptibility
Table 2. Summary of results
GyrA91-PCR; WT probe 1 [Ct value/range;
mean value (cycles)]
GyrA91-PCR; S91F probe 2 [Ct value/range;
mean value (cycles)]
N. gonorrhoeae isolates (n¼70)
ciprofloxacin susceptible; WT S91 (n¼42)
ciprofloxacin less susceptible; S91F (n¼1)
ciprofloxacin resistant; S91F (n ¼27)
positive (17– 37; 23)
negative
negative
negative
positive (20)
positive (18 – 31; 21)
Commensal Neisseria and Moraxella isolates (n¼55)
N. flavescens (n¼1)
N. mucosa (n¼1)
N. cinerea (n¼4)
N. lactamica (n ¼16)
N. polysacchareae (n ¼4)
N. sicca (n ¼4)
N. subflava (n ¼14)
N. weaveri (n ¼1)
N. elongata (n ¼1)
M. catarrhalis (n¼7)
M. osloensis (n¼2)
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
N. gonorrhoeae NAAT-negative clinical samples (n¼171)
anorectal swab (n¼27)
cervical swab (n¼49)
pharyngeal swab (n¼34)
vaginal swab (n¼24)
urine (n¼33)
unspecified (n ¼4)
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
Isolates and clinical samples
N. gonorrhoeae NAAT-positive clinical samples from which ciprofloxacin-susceptible gonococci were isolated (n¼175)
anorectal swab (n¼11)
positive (29– 42; 34)
cervical swab (n¼19)
positive (33– 41; 36)
unspecified aspirate (n¼1)
positive (36)
joint fluid (n¼3)
positive (33– 37; 35)
penile swab (n ¼34)
positive (29– 44; 34)
pharyngeal swab (n¼3)
positive (34– 39; 37)
urine (n¼65)
positive (25– 43; 34)
unspecified genital swab (n¼3)
positive (34– 39; 36)
vaginal swab (n¼22)
positive (23– 41; 35)
cervical swab (n¼1)
negative
penile swab (n ¼3)
negative
pharyngeal swab (n¼1)
negative
urine (n¼6)
negative
vaginal swab (n¼3)
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
negative
N. gonorrhoeae NAAT-positive clinical samples from which ciprofloxacin-resistant gonococci were isolated (n¼34)
anorectal swab (n¼2)
negative
cervical swab (n¼5)
negative
penile swab (n ¼9)
negative
pharyngeal swab (n¼3)
negative
urine (n¼12)
negative
vaginal swab (n¼2)
negative
cervical swab (n¼1)
negative
positive (32 – 35; 33)
positive (34 – 40; 37)
positive (30 – 43; 37)
positive (39 – 45; 42)
positive (26 – 36; 32)
positive (31 – 32; 31)
negative
Discrepant sample: an N. gonorrhoeae NAAT-positive clinical sample from which both ciprofloxacin-susceptible and ciprofloxacin-resistant gonococci
were isolated (n¼1)
anorectal swab (n¼1)
positive (32)
negative
355
Buckley et al.
described PCR targeting the gonococcal porA pseudogene,12 with
both assays detecting to the same 10-fold dilution.
When applied to the 210 N. gonorrhoeae NAAT-positive clinical
samples, the GyrA91-PCR successfully characterized 195 (92.9%,
95% CI¼ 88.5% –95.9%) samples with 15 (7.1%) providing negative results (Table 2). For 194 of the 195 characterized samples,
the results of the GyrA91-PCR assay were in agreement with
those of the isolate-based susceptibility testing; 161 samples
from which ciprofloxacin-susceptible gonococci were isolated
were positive by the WT probe 1, whereas 33 samples with
ciprofloxacin-resistant isolates were positive by the S91F probe 2
(Table 2). The final sample (an anorectal swab; Table 2) gave a discrepant result; a ciprofloxacin-resistant strain was initially isolated
from this sample, but it was indicated by the GyrA91-PCR to harbour a susceptible strain, providing positive results by the WT
probe 1 only. Overall, positivity by the GyrA91-PCR WT probe 1 correctly predicted N. gonorrhoeae susceptibility to ciprofloxacin in
161 of 162 (99.4%, 95% CI ¼ 96.6% –99.9%) of these samples.
The above discrepant sample was further investigated.
Subsequent sub-culturing of the isolate from this sample identified the presence of a ciprofloxacin-susceptible isolate in addition
to the resistant isolate. Both isolates from this discrepant sample
were then tested by the GyrA91-PCR and no inconsistencies were
observed, i.e. the ciprofloxacin-susceptible isolate was positive by
probe 1 and the resistant isolate positive by probe 2. The discrepancy observed for this anorectal sample therefore appears to be
due to the sample harbouring a mixture of both ciprofloxacinsusceptible and -resistant strains and, presumably due to the susceptible strain being at higher DNA load, only the susceptible
strain was selectively amplified by the GyrA91-PCR. The basis of
the GyrA91-PCR-negative results for 15 samples (Table 2) was
not further investigated, but may reflect degradation of DNA
after long-term storage.
In summary, the results provide further evidence that the GyrA
91 locus can be targeted to predict N. gonorrhoeae ciprofloxacin
susceptibility directly in clinical samples. The GyrA91-PCR assay
successfully characterized 92.9% of N. gonorrhoeae NAAT-positive
clinical samples tested and provided accurate (99.4%) prediction
of ciprofloxacin susceptibility. Nevertheless, it should be noted
that the targeted nature of the GyrA91-PCR method (i.e. characterizing the GyrA 91 locus only) does make the assay prone to
error if ciprofloxacin resistance arises via mutations in other loci.
This highlights the need for local validation and, ideally, the
ongoing testing of local isolates to verify assay performance.
Overall, the GyrA91-PCR assay represents a promising step
towards individualized treatment of gonococcal infection using
ciprofloxacin and may be useful in regions, such as Australia,
where significant proportions of strains remain susceptible to
this drug.
Acknowledgements
This study was conducted as part of the Gonorrhoea Resistance
Assessment by Nucleic Acid Detection (GRAND) study.
We thank Dr Graeme Nimmo and Dr Cheryl Bletchly (Pathology
Queensland, Queensland), Dr Robert Baird (Royal Darwin Hospital,
Northern Territory), Vicki Hicks (Queensland Health Forensic and Scientific
Services, Brisbane, Queensland) and Dr Tiffany Hogan (Prince of Wales
Hospital, New South Wales) for their assistance with this study, including
the provision of samples and isolates.
356
Other GRAND study investigators
John Kaldor and Handan Wand from the Kirby Institute, UNSW Australia,
James Ward from the South Australian Health and Medical Research
Institute, Christopher Fairley from the Melbourne Sexual Health Centre,
Nathan Ryder from the Newcastle Sexual Health Service and Jiunn-Yih
Su from the Sexual Health and Blood Borne Virus Unit, Northern Territory.
Funding
This work was funded by the National Health and Medical Research Council
(APP1025517), the Children’s Health Foundation Queensland and the reference work of the National Neisseria Network, Australia, which is funded
by the Australian Government Department of Health.
Transparency declarations
None to declare.
References
1 U.S. Department of Health and Human Services, CDC. Antibiotic
Resistance Threats in the United States, 2013. Atlanta, GA, USA: U.S.
Department of Health and Human Services, CDC, 2013. http://www.cdc.
gov/drugresistance/pdf/ar-threats-2013-508.pdf.
2 Unemo M, Shafer WM. Antimicrobial resistance in Neisseria gonorrhoeae
in the 21st century: past, evolution, and future. Clin Microbiol Rev 2014; 27:
587–613.
3 Buono SA, Watson TD, Borenstein LA et al. Stemming the tide of
drug-resistant Neisseria gonorrhoeae: the need for an individualized
approach to treatment. J Antimicrob Chemother 2015; 70: 374– 81.
4 Lahra MM for the Australian Gonococcal Surveillance Programme.
Australian Gonococcal Surveillance Programme, 1 April to 30 June 2014.
Commun Dis Intell Q Rep 2014; 38: E390– 2.
5 Tapsall J. Antimicrobial Resistance in Neisseria gonorrhoeae. 2001. http://
www.who.int/drugresistance/Antimicrobial_resistance_in_Neisseria_
gonorrhoeae.pdf.
6 Tanaka M, Takahashi K, Saika T et al. Development of fluoroquinolone
resistance and mutations involving GyrA and ParC proteins among
Neisseria gonorrhoeae isolates in Japan. J Urol 1998; 159: 2215–9.
7 Magooa MP, Muller EE, Gumede L et al. Determination of Neisseria gonorrhoeae susceptibility to ciprofloxacin in clinical specimens from men using
a real-time PCR assay. Int J Antimicrob Agents 2013; 42: 63– 7.
8 Kulkarni S, Bala M, Sane S et al. Mutations in the gyrA and parC genes
of quinolone-resistant Neisseria gonorrhoeae isolates in India. Int J
Antimicrob Agents 2012; 40: 549–53.
9 Siedner MJ, Pandori M, Castro L et al. Real-time PCR assay for detection
of quinolone-resistant Neisseria gonorrhoeae in urine samples. J Clin
Microbiol 2007; 45: 1250– 4.
10 Trembizki E, Buckley C, Donovan B et al. Direct real-time PCR-based detection of Neisseria gonorrhoeae 23S rRNA mutations associated with azithromycin resistance. J Antimicrob Chemother 2015; 70: 3244–9.
11 Whiley D, Trembizki E, Regan D et al. Molecular epidemiology of
Neisseria gonorrhoeae in Australia. In: Abstracts of the Australian Society
for Microbiology Annual Scientific Meeting, Melbourne, 2014. Abstract 64.
Australian Society for Microbiology, Melbourne, Victoria, Australia.
12 Goire N, Nissen MD, LeCornec GM et al. A duplex Neisseria gonorrhoeae
real-time polymerase chain reaction assay targeting the gonococcal porA
pseudogene and multicopy opa genes. Diagn Microbiol Infect Dis
2008; 61: 6 –12.