Journal of Antimicrobial Chemotherapy Advance Access published September 11, 2012

J Antimicrob Chemother
doi:10.1093/jac/dks340

Real-time PCR for detection of blaOXA-48 genes from stools

Thierry Naas*, Garance Cotellon, Ayla Ergani and Patrice Nordmann
Service de Bacteriologie-Virologie, INSERM U914: Emerging Resistance to Antibiotics, LabEx LERMIT, Hopital de Bicetre,
94275 Le Kremlin-Bicetre, Assistance Publique-Hopitaux de Paris, Faculte de Medecine Paris-Sud, France
*Corresponding author. Tel: +33-1-45-21-29-86; Fax: +33-1-45-21-63-40; E-mail. thierry.naas@bct.aphp.fr

Received 22 May 2012; returned 10 July 2012; revised 25 July 2012; accepted 27 July 2012

Methods: An in-house real-time quantitative PCR (qPCR) assay using TaqMan chemistry to detect blaOXA-48-like
genes was compared with bacterial culturing on ChromID ESBL and SUPERCARBA media of spiked stool samples
with several species producing OXA-48 variants.
Results: qPCR amplification using plasmid DNA was linear over 10 log dilutions (r 2 0.998 and slope 23.14),
with an amplification efficiency of 1.10, and the detection limit of the assay was reproducibly estimated at
10 plasmid molecules/PCR. No cross-reaction was detected with DNA extracted from several multidrugresistant bacteria harbouring other b-lactam resistance genes. The blaOXA-48 qPCR assay was capable of detecting
1050 cfu of OXA-48 producers/100 mg of faeces. ChromID ESBL was capable of detecting OXA-48 producers
(1101 to 3102 cfu/100 mg of faeces), as long as the isolates exhibited a high level of resistance to cephalosporins due to an associated extended-spectrum b-lactamase. The SUPERCARBA screening medium was
capable of detecting all the OXA-48-like producers (1 3101 cfu/100 mg of faeces), except those producing
OXA-163, a variant lacking carbapenem-hydrolysing activity.
Conclusions: The qPCR is likely to shorten the time for blaOXA-48 detection from 48 to 4 h and will be a valuable
tool for outbreak follow-up in order to rapidly isolate colonized patients and assign them to cohorts.
Keywords: carbapenemases, carriers, qPCR

Introduction
OXA-48-like carbapenemases are increasingly identified in many
European countries, such as France, Germany and the UK,
through transfer of hospitalized patients from endemic areas,
but also from likely autochthonous diffusion.1,2 The main reservoir of OXA-48-like producers is likely to be countries located
on the southern to eastern borders of the Mediterranean Sea
(Morocco to Turkey) and the Indian subcontinent.1,2 Several outbreaks have been described that have often been the consequence of difficulties in detection, especially with isolates with
low MICs of carbapenems that may remain in the susceptibility
range.1,2
Several OXA-48 variants have been described: OXA-48,
OXA-162, OXA-181, OXA-204 and OXA-232 hydrolyse carbapenems and hydrolyse broad-spectrum cephalosporins very
weakly, while OXA-163 identified from Argentinean isolates
lacks carbapenem-hydrolysing activity, but hydrolyses expandedspectrum cephalosporins.2 The level of resistance to carbapenems

provided by OXA-48-like-carbapenemase producers may vary significantly, making their detection difficult when based just on high
levels of carbapenem resistance. Therefore, detection of OXA-48
producers is particularly challenging in a routine laboratory, especially in the course of managing an outbreak, when rapid identification and cohorting of patients is required. Several media have
been proposed for the detection of carbapenemase producers.
The ChromID ESBL medium has good sensitivity, its main disadvantage being its lack of detection of OXA-48-like producers that
are susceptible to cefpodoxime in the absence of co-production
of an extended-spectrum b-lactamase (ESBL).3 The ChromAgar
KPC medium has lower sensitivity due to lack of detection of bacteria with low MICs of carbapenems.3 A Drigalski agar-based
culture medium containing a carbapenem, cloxacillin and zinc
sulphate (SUPERCARBA medium) has recently been described
to identify accurately all carbapenemase-producing, including
OXA-48-producing, Enterobacteriaceae.4
Recently, the development of real-time quantitative PCR
(qPCR) for detecting carbapemenases directly from rectal

# The Author 2012. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: journals.permissions@oup.com

1 of 4

Downloaded from http://jac.oxfordjournals.org/ by guest on December 7, 2015

Objectives: Outbreaks of OXA-48-like carbapenemase producers are increasingly reported in many European
countries and are often the result of difficulties in detection, especially for isolates with MICs of carbapenems
that remain in the susceptibility range.

Naas et al.

swabs or stool samples with comparable, if not superior, detection limits compared with culture has shown its usefulness
due to a faster turnaround time of results (4 versus 48 h).5,6
Here, we describe the development of a qPCR assay combined
with a fully automated extraction system for the detection of
blaOXA-48-like genes from culture or from spiked stool samples.

Materials and methods

[bracketed
5 -FAM-TCC(+A)GA(+G)CA(+C)AA(+C)TACG-TAMRA-3 ]
nucleotides with+signs are locked nucleic acid (LNA) nucleotides]
were designed in-house. The concentrations of the primers and probe
used in this study that gave the best detection limits were 400 nM for
blaOXA-48 primers and 200 nM for the blaOXA-48 probe. The Quantitect
Virus Kit (Qiagen) was used as recommended by the manufacturer.
One PCR experiment (including mix preparation and amplification)
TM
takes ,2 h. Simplexa from Focus Diagnostics (Eurobio, Courtaboeuf,
France) was used as an internal control, for extraction and amplification.6

Bacterial strains and plasmids

Serial dilutions and spiking experiments

Bacterial suspensions of strains (see Table 2) with an optical density at
600 nm of 0.5 were serially diluted in PBS (bioMerieux); nine 10-fold dilutions were made. To quantify the viable bacteria in each dilution step, a
blood agar plate (bioMerieux) was inoculated with 100 mL of a suspension and incubated overnight at 378C; the number of colonies that
grew was counted the following day.
Spiked Escherichia coli cultures were prepared by adding 100 mL of
each dilution in PBS was added to 900 mL of fresh overnight E. coli
Ec1.6 Spiked faecal samples were made by adding 100 mL of each dilution
in PBS to 900 mL of a faecal suspension that was obtained by suspending
6 g of freshly pooled faeces from three healthy volunteers in 60 mL of
distilled water, as previously described.6 A faecal suspension without
the addition of a bacterial strain was used as a negative control. Aliquots
of 200 mL of spiked stool samples were subjected to DNA extraction. Each
faecal suspension was plated onto the ChromID ESBL and SUPERCARBA
media. Viable bacteria were counted after 24 h at 378C and growth on
selective media was compared with qPCR results. All experiments were
performed in triplicate.

DNA extraction from bacterial cultures and from stool

samples
Fresh overnight bacterial cultures were used for DNA extraction with a
QIAamp DNA Minikit (Qiagen, Les Ulis, France) according to the protocol
suggested by the manufacturer. Extracted bacterial DNA was eluted from
the columns in 200 mL of elution buffer and stored at 2208C.
Spiked E. coli Ec1 cultures and spiked stool samples were extracted
with the EasyMag automated DNA extraction system (bioMerieux),
which enables simultaneous DNA extractions from 24 samples in ,2 h.

OXA-48 qPCR assay

The qPCR Rotor-Gene 6000 amplification/detection system (Qiagen)
was used for the amplification and detection of the blaOXA-48 amplicon
(100 bp) by TaqMan technology. The forward PCR primer (OXA-48For; 5 -GTAGCAAAGGAATGGCAA-3 ) and the reverse PCR primer
(OXA-48-Rev; 5 -CCTTGCTGCTTATTGTCA-3 ) specific for the detection of
blaOXA-48-like genes and the blaOXA-48-specific probe [OXA-48-Probe;

2 of 4

Results
The analytical sensitivity of the blaOXA-48 qPCR assay was determined after serially diluting plasmid pOXA-48. The amplification
was linear over 10 log dilutions (r 2 0.998 and slope 23.14),
and the amplification efficiency was 1.10. The detection limit
of the assay was reproducibly estimated at 10 plasmid molecules/PCR.
None of the OXA-48-negative bacterial pathogens was positive in the blaOXA-48 qPCR assay (Table 1). Of the 14 isolates negative for the blaOXA-48 gene, 3 were imipenem susceptible and 11
were intermediate or resistant, by means of another carbapenemase gene. Of the 11 isolates positive for the blaOXA-48-like genes,
5 were susceptible to imipenem (MIC ,1 mg/L) and 6 were intermediate or resistant (MIC .4 mg/L). All were identified using this
assay, with Ct values of 19 20. Overall, there was 100% concordance between genotype of bacteria and the blaOXA-48 qPCR
assay results (Table 1), suggesting excellent specificities of the
primers and probe. The assay identified OXA-48, OXA-181 and
OXA-163 equally well.
The blaOXA-48 qPCR assay on extracted DNA from serially
diluted OXA-48 producers into E. coli Ec1 and subsequent automated DNA extraction was linear over 9 log dilutions
(r 2 0.994 and slope 22.92), and the amplification efficiency
was 1.09. The limit of the linear range of the qPCR was reproducibly 2101 cfu. The qPCR allowed detection of 1 5 cfu;
however, the Ct values were out of the linear range of the qPCR.
The blaOXA-48 qPCR assay was capable of detecting 10 50 cfu/
100 mg of faeces. Below 10 cfu (,1 equivalent cfu/PCR), results
were not reproducible. The analytical sensitivity of the assay did
not change upon the use of different blaOXA-48-like-positive strains
with different carbapenem susceptibility patterns and different
associated b-lactamases: Citrobacter freundii, Enterobacter
cloacae, E. coli and Klebsiella pneumoniae blaOXA-48--positive
strains (Table 2).
The qPCR results were compared with those obtained with
two selective culture media: the chromogenic ChromID ESBL
(bioMerieux) and SUPERCARBA. With serially diluted OXA-48 producers co-producing an ESBL or with OXA-163, which can hydrolyse expanded-spectrum cephalosporins, the lowest limit of
detection ranged from 1101 to 3102 cfu/mL for ChromID
ESBL (Table 2). However, for isolates lacking an ESBL the
detection limit was .105 cfu/mL (Table 2). ChromID ESBL,
which contains cefpodoxime as a selector and is routinely used
for screening ESBL producers,8 was capable of detecting
carbapenemase-producing Enterobacteriaceae of the OXA-48
type, as long as a high level of resistance to cephalosporins
was present due to an associated ESBL.3 For SUPERCARBA
medium, the lowest limit of detection of OXA-48 producers
ranged between 1101 and 3101 cfu/mL (Table 2). The

Downloaded from http://jac.oxfordjournals.org/ by guest on December 7, 2015

Eleven well-characterized enterobacterial OXA-48-like producers (7

OXA-48, 2 OXA-181 and 2 OXA-163 producers), along with 14
non-OXA-48-like producers used as negative control strains, were used
for direct detection in culture or for spiking experiments, in order to validate the novel detection assay (see Tables 1 and 2). A recombinant
plasmid, pOXA-48, consisting of a pCR-II-TOPO backbone (Invitrogen,
Saint Aubin, France) carrying an 1 kb insert expressing the OXA-48 carbapenemase, was used for qPCR optimization. The concentration of the
plasmid solution was 1011 plasmid molecules/mL. MICs were determined
using the Etest (bioMerieux, Marcy lEtoile, France) and interpreted
according to CLSI.7

JAC

qPCR detection of OXA-48 producers in stools

Table 1. Bacterial isolates used in this study

MIC (mg/L)
Isolate

b-Lactamases expressed

IPM

MEM

ETP

OXA-48 qPCR

Reference

K. pneumoniae
E. cloacae
K. pneumoniae UCL
K. pneumoniae HPA-1
E. cloacae
E. coli
E. coli
Providencia rettgeri
C. freundii
E. cloacae
K. pneumoniae
K. pneumoniae A28006
E. coli
E. cloacae HGM
K. pneumoniae A
E. coli A
E. cloacae A
K. pneumoniae
E. coli DIH-1
K. pneumoniae
E. cloacae
E. coli
K. pneumoniae 6852
E. coli Ec1
K. pneumoniae KPS1

OXA-163, TEM-1
OXA-48, SHV-5
OXA-48
OXA-48, CTX-M-15, SHV-1
OXA-163, TEM-1
OXA-48, TEM-150
OXA-48, CTX-M-24
OXA-48, TEM-101
OXA-181
OXA-48
OXA-181
KPC-2, CTX-M-2, TEM-1, SHV-1
KPC-2, CTX-M-9, TEM-1
KPC-2
NDM-1, CTX-M-15, TEM-1, SHV-1, CMY-2
NDM-1, CTX-M-15, TEM-1
NDM-1
VIM-19, TEM-1
VIM-19, CTX-M-3, TEM-1
VIM-1, SHV-5
IMP-4
IMP-1, TEM-1
IMP-1
WT
OKP-A

0.5
0.5
1
1.5
0.5
24
24
.32
.32
0.5
0.5
16
24
12
.32
6
8
8
8
.32
.32
0.5
1
0.06
0.12

0.25
0.5
1
12
0.5
32
32
.32
.32
0.5
0.5
32
16
12
.32
4
6
4
4
.32
.32
0.5
8
0.016
0.12

0.38
3
4
32
2
24
24
.32
.32
2
2
24
32
12
.32
4
6
16
16
.32
.32
3
2
0.006
0.032

+
+
+
+
+
+
+
+
+
+
+
2
2
2
2
2
2
2
2
2
2
2
2
2
2

3
3
3
4
3
3
3
3
3
4
4
4
3
3
4
4
4
4
4
4
4
4
4
6
6

IPM, imipenem; MEM, meropenem; ETP, ertapenem.

Table 2. Spiking of faecal samples with OXA-48-like-producing enterobacterial isolates

Isolate
1
2
3
4
5
6
7
8
9
10
11
16

Bacteria

OXA/NDM

K. pneumoniae
E. cloacae
K. pneumoniae UCL
K. pneumoniae HPA-1
E. cloacae
E. coli
E. coli
P. rettgeri
C. freundii
E. cloacae
K. pneumoniae
E. coli A

OXA-163b
OXA-48
OXA-48
OXA-48
OXA-163b
OXA-48
OXA-48
OXA-48
OXA-181
OXA-48
OXA-181
NDM-1

ESBL

SHV-5
CTX-M-15
TEM-150
CTX-M-24
TEM-101

IPM MIC
(mg/L)a

Lowest limit of detection

in stool (cfu/mL of
stoola), ChromID ESBL

0.5
0.5
1
1.5
0.5
24
24
.32
.32
0.5
0.5
6

3102
2101
NG
1101
1102
2101
2101
1102
NG
NG
NG
2101

Lowest limit of detection in Lowest limit of detection

of OXA-48 qPCR
stool (cfu/mL of stoola),
SUPERCARBA
(cfu/mL of stoola)
NG
3101
1101
1101
NG
2101
2101
1102
2101
2101
1101
1101

2101
5101
3101
3101
101
101
2101
2102
4101
101
3101
c

IPM, imipenem; NG, no growth with 105 cfu/mL of stool.

a
cfu/mL of stool. One millilitre of stool contains 100 mg of stool. Values were calculated using the standard curve made with serially diluted OXA-48producing E. coli isolates diluted in 108 E. coli WT.
b
OXA-163, unlike OXA-48 or OXA-181, can hydrolyse expanded-spectrum cephalosporins, but not carbapenems.
c
No amplification.

3 of 4

Downloaded from http://jac.oxfordjournals.org/ by guest on December 7, 2015

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

Bacteria

Naas et al.

SUPERCARBA screening medium, which may detect not only

KPC b-lactamases and metallo-b-lactamases but also OXA-48
producers,4 was capable of detecting all the carbapenemhydrolysing variants of OXA-48-like, even those with low MICs.
OXA-163 was not detected on this medium, due to lack of
carbapenem-hydrolysing activity.

screening for carriage of OXA-48 producers should be based on

selective culture media, while the molecular techniques for
screening directly from stools should reveal their usefulness in
an outbreak situation, for preventing further spread of OXA-48
producers.

Discussion

Acknowledgements

The prevention of spread of carbapenemase producers relies on

early detection of carriers. Screened patients should be kept in
strict isolation prior to obtaining the screening results, meaning
that these screening techniques have to be sensitive, specific
and rapid. Since the reservoir of carbapenemase producers
remains the intestinal flora, stools and rectal swabs are adequate samples for performing this screening. ChromID ESBL is
a reliable culture medium for screening most carbapenemase
producers, except many OXA-48 producers, directly from faecal
samples.3 However, this medium is not capable of distinguishing
between ESBL and carbapenemase producers. The SUPERCARBA
medium represents a significant improvement compared with
the available screening media to detect carbapenemase producers, particularly for the detection of OXA-48 producers that do
not co-express an ESBL. In addition, this medium is useful for
specifically selecting carbapenemase producers in stools that
also contain a large amount of ESBL producers, the growth of
which will be inhibited.4 This property is particularly relevant,
since high rates of ESBL carriage are now reported worldwide.9
Alternative strategies, such as qPCR, have proven their usefulness for detecting KPC and for the management of epidemic
situations.8 Hindiyeh et al. 5 described the development and verification of a qPCR assay for detection of blaKPC genes directly in
perianal swabs. Direct detection of blaKPC by PCR shortened the
time to identification of patients colonized or infected with
carbapenem-resistant organisms and was more sensitive than
culture. Similarly, blaNDM-1 qPCR showed excellent sensitivity
and specificity, with detection limits lower than those of the
culture techniques, and with a significant gain in time to
results (4 versus 48 h).6 For OXA-48 qPCR, a similar time to
results was observed: 2 h for automated extraction and ,2 h
for qPCR. The OXA-48 qPCR thus yielded results in ,4 h, versus
2 3 days using conventional methods. This qPCR assay is
capable of detecting OXA-48, OXA-181 and OXA-163, an
OXA-48 variant capable of hydrolysing third-generation cephalosporins but not carbapenems.2 The newly described variants,
OXA-162, OXA-204 and OXA-232, should also be detected
using this qPCR assay, since the nucleotide changes between
the variants are located outside the primer and probe binding
sequences.2 We found good sensitivity and specificity, and excellent agreement with culture screening techniques. Routine

We thank Gaelle Cuzon for technical support.

This work was funded by grants from INSERM, by the Ministe`re de lEducation Nationale et de la Recherche and by the European Communitys
Seventh Framework Programme FP7/2007-2013 under grant agreement
no. 241742.

Transparency declarations
None to declare.

References
1 Nordmann P, Naas T, Poirel L. Global spread of carbapenemaseproducing Enterobacteriaceae. Emerg Infect Dis 2011; 17: 1791 8.
2 Poirel L, Potron A, Nordmann P. OXA-48-like carbapenemases: the
phantom menace. J Antimicrob Chemother 2012; 67: 1597 606.
3 Carrer A, Fortineau N, Nordmann P. Use of ChromID ESBL medium for
detecting carbapenemase-producing Enterobacteriaceae. J Clin Microbiol
2010; 48: 1913 4.
4 Nordmann P, Girlich D, Poirel L. Detection of carbapenemase producers
in Enterobacteriaceae by use of a novel screening medium. J Clin
Microbiol 2012; 50: 27616.
5 Hindiyeh M, Smollen G, Grossman Z et al. Rapid detection of blaKPC
carbapenemase genes by real-time PCR. J Clin Microbiol 2008; 46:
2879 83.
6 Naas T, Ergani A, Carrer A et al. Real-time PCR for detection of NDM-1
carbapenemase genes from spiked stool samples. Antimicrob Agents
Chemother 2011; 55: 4038 43.
7 Clinical and Laboratory Standards Institute. Performance Standards
for Antimicrobial Susceptibility Testing: Twenty-second Informational
Supplement M100-S22. CLSI, Wayne, PA, USA, 2012.
8 Reglier-Poupet H, Naas T, Carrer A et al. Performance of chromID ESBL,
a chromogenic medium for detection of Enterobacteriaceae producing
extended-spectrum b-lactamases. J Med Microbiol 2008; 57: 310 5.
9 Rogers BA, Sidjabat HE, Paterson DL. Escherichia coli O25b-ST131: a
pandemic, multiresistant, community-associated strain. J Antimicrob
Chemother 2011; 66: 1 14.

Downloaded from http://jac.oxfordjournals.org/ by guest on December 7, 2015

4 of 4

Funding