Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, May 2009, p. 1808-1816, Vol. 53, No. 5
0066-4804/09/$08.00+0 doi:10.1128/AAC.00451-08
Copyright © 2009, American Society for Microbiology. All Rights Reserved.

Aurélia Delauné,1,
Emmanuelle Espié,2
Karin Nygard,3
Maria Pardos,1
Lucette Polomack,1
Françoise Guesnier,1
Marc Galimand,4
Jørgen Lassen,3 and
François-Xavier Weill1*
Institut Pasteur, Centre National de Référence des Salmonella, Laboratoire des Bactéries Pathogènes Entériques, Paris, France,1 Institut de Veille Sanitaire, Saint-Maurice, France,2 Norwegian Institute of Public Health, Division of Infectious Disease Control, Oslo, Norway,3 Institut Pasteur, Unité des Agents Antibactériens, Paris, France4
Received 4 April 2008/ Returned for modification 30 May 2008/ Accepted 19 February 2009
|
|
|---|
|
|
|---|
The number of records of ESC-resistant human isolates of S. enterica serotype Concord has been increasing in France since 2004 and in Norway since 2001. Concord is a very rare serotype with the antigenic formula 6,7:l,v:1,2. It was identified for the first time in 1944 from four cultures, three of which were isolated from fatal infections in chicks in the United States and one from the stools of a patient affected in a small outbreak of food poisoning in England (14). In the early 1980s, Concord was isolated in chicken-breeding farms in Saudi Arabia (2). Two sporadic ESBL-producing isolates of serotype Concord were reported in recent studies. One, containing blaSHV-12, was isolated in Holland in 2001 (21). The other, a CTX-M-15-producing isolate, was isolated in Ireland in 2005 in a patient originally from Ethiopia (28).
We studied the French and a selection of the Norwegian S. enterica serotype Concord isolates (i) to determine the genomic diversity of these isolates by standardized pulsed-field gel electrophoresis (PFGE), (ii) to determine the genetic basis for antibiotic resistance, and (iii) to briefly describe the epidemiology of infections with the organisms.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Antimicrobial resistance patterns, bla genes, class 1 integrons, and XbaI PFGE types of S. enterica serotype Concord and monophasic 6,7:l,v:- isolates from France, 1996 to 2006, and of selected S. enterica serotype Concord isolates from Norway, 2001 to 2005
|
The S. enterica serotype Concord reference strain 156K (isolated from a chick in 1944 in California), obtained from the World Health Organization Collaborative Centre for Reference and Research on Salmonella, Institut Pasteur, was used for molecular typing methods. S. enterica serotype Braenderup H9812 was used as a molecular size marker for PFGE.
Epidemiological investigations. Demographical, clinical, and epidemiological data were collected for each case identified between 2004 and 2006 in France by contacting physicians by mail. Eight associations certified by the French Government for international adoption from Ethiopia were contacted by e-mail.
In Norway, it is recommended, but not required, that all adopted infants be screened for intestinal parasites and Salmonella infection. However, the extent to which this recommendation is followed is unclear. Clinicians and microbiological laboratories report all laboratory-confirmed Salmonella infections to the Norwegian Surveillance System for Communicable Diseases. Demographical, clinical, and epidemiological information is collected on the clinical notification forms. The Norwegian association responsible for adoption from Ethiopia was informed.
Antimicrobial susceptibility.
Antibiotic susceptibility was determined by the disk diffusion method with 32 antimicrobial drugs (Bio-Rad, Marnes la Coquette, France), as previously described (35). The MICs of ceftriaxone (CRO) and ceftazidime (CAZ) were determined by Etest (AB Biodisk, Solna, Sweden). The ESBL phenotype was detected by using the ESBL detection Etest strips (AB Biodisk) and the double-disk synergy method (25). Isolates were categorized as susceptible, intermediate, or resistant according to Antibiogram Committee of the French Society for Microbiology cutoff values (http://www.sfm.asso.fr/nouv/general.php?pa=2). The cutoff values used for CRO and CAZ are slightly different from those determined by the Clinical and Laboratory Standards Institute (CLSI); susceptible strains were thus defined by MICs of
4 µg/ml (CLSI,
8 µg/ml) and resistant strains by MIC of >32 µg/ml (CLSI,
64 µg/ml for CRO and
32 µg/ml for CAZ).
PCR amplification of antimicrobial resistance genes and sequence analysis. Total DNA was extracted using the InstaGene matrix kit (Bio-Rad) according to the manufacturer's recommendations. The resistance genes blaTEM, blaSHV, blaOXA-1 group, and blaCTX-M and class 1 integron gene cassettes were amplified by PCR as described previously (15, 35).
Sequencing was performed at Genome Express (Meylan, France) or at the Plateforme de Génotypage des Pathogènes et Santé Publique, PF8 (Institut Pasteur). The nucleotide sequences and the deduced protein sequences were analyzed with EditSeq and Megalign software (Dnastar, Madison, WI). The BLASTN program of NCBI was used for database searches (http://www.ncbi.nlm.nih.gov/BLAST/).
PFGE typing. The genetic diversity of 68 clinical Salmonella isolates of serotype Concord or the monophasic variant 6,7:l,v:- (60 from France and 8 from Norway) and serotype Concord reference strain 156K was assessed by PFGE of genomic DNA digested with XbaI (Roche, Mannheim, Germany), as described previously (34). The running conditions and the molecular size marker were as described in the standardized PulseNet protocol (24). BioNumerics 4.0 (Applied Maths, Sint-Martens-Latem, Belgium) was used for image normalization and construction of similarity matrices. Bands were assigned manually. Clustering was carried out by the unweighted pair-group method with arithmetic averages based on the Dice similarity index, using a 1% optimization parameter and 1% band position tolerance. Each profile that differed by one or more extra bands of >100 kb in size was assigned a type (e.g., type X10). Profiles that differed only by the position of a high-molecular-weight band (400- to 600-kb region) were assigned to subtypes (e.g., subtypes X10a, X10b, and X10c).
Resistance transfer determination. We carried out a resistance transfer experiment using a subset of 12 (8 French and 4 Norwegian) ESBL-producing isolates on liquid and solid media, using either Escherichia coli K-12 BM14 resistant to sodium azide or E. coli C1a resistant to nalidixic acid (NAL) as the recipient strain (35). Transconjugants were selected on Drigalski agar (Bio-Rad) supplemented with cefotaxime (CTX) (4 µg/ml) and sodium azide (500 µg/ml) or NAL (64 µg/ml). Three E. coli transconjugants were arbitrarily selected for each experiment.
Plasmid DNA from four S. enterica serotype Concord isolates that was unsuccessfully transferred by conjugation was used to transform electrocompetent E. coli DH10B by standard electroporation techniques with a MicroPulser electroporation apparatus (Bio-Rad). Transformants were selected on Mueller-Hinton agar containing CTX (4 µg/ml).
Plasmid analysis. Plasmids were characterized for a subset of 19 Salmonella isolates (and four E. coli transconjugants) according to the year of isolation, antimicrobial resistance phenotype, and PFGE profile. Two plasmid-profiling methods were used: alkaline lysis and S1 nuclease analysis.
Plasmid DNA extracted by alkaline lysis (33) was analyzed by electrophoresis in 0.8% agarose gels. The molecular sizes of plasmids were estimated by comparison with plasmids of known sizes: pIP173 (125.8 kb), RP4 (56 kb), and Tp116 (210 kb), mixed with supercoiled DNA ladder (Invitrogen, Groningen, The Netherlands).
We used S1 nuclease treatment and PFGE to accurately determine the molecular sizes of large bacterial plasmids. Proteinase K-treated gel plugs prepared for PFGE analysis were cut into 1-mm slices and digested with 1 U of S1 nuclease (Roche) as described previously (3). DNA fragments separated by PFGE were transferred onto a nylon membrane (Hybond N+; Amersham) and hybridized with blaCTX-M, blaTEM, IncA/C, and IncY probes (8, 15). Hybridization, labeling, and detection were performed according to the manufacturers' recommendations, using either an enhanced-chemiluminescence nonradioactive kit (GE Healthcare, United Kingdom) or a radioactive kit (MegaprimeTM DNA-labeling system; GE Healthcare).
PCR-based replicon-typing analysis was performed as described by Carattoli et al. (8). The 18 primer pairs targeting FIA, FIB, FIC, HI1, HI2, I1-I
, L/M, N, P, W, T, A/C, K, B/O, X, Y and FII replicons were used in separate PCRs.
Chromosomal localization of the blaCTX-M-15 gene by PFGE-I-CeuI. To determine the chromosomal location of the blaCTX-M-15 gene, plugs of 11 isolates were prepared as described above and digested with the I-CeuI endonuclease (New England Biolabs). The digested fragments were separated using the Chef DRIII system as described previously (27). The sizes of I-CeuI restriction fragments were determined using known I-CeuI fragment sizes of chromosomal DNA from S. enterica serotype Typhimurium LT2 (27) and XbaI-restricted fragments from the chromosomal DNA of S. enterica serotype Braenderup H9812. The I-CeuI restriction fragments were subjected to Southern hybridization with PCR-generated probes for blaCTX-M, blaTEM, and16S rRNA gene probes (15, 20).
Analysis of the genetic environment of the blaCTX-M-15 gene by cloning. We used cloning techniques to determine the DNA sequences flanking the blaCTX-M-15 gene in isolate 05-0004 (a representative Concord isolate with a chromosomal blaCTX-M-15 gene, as determined by PFGE I-CeuI and S1 nuclease experiments). DNA prepared using the Promega Wizard kit was partially digested with Sau3aI, purified with Qiaquick PCR purification kit columns (Qiagen, Courtaboeuf, France), and ligated into dephosphorylated, BamHI-restricted phagemid pBK-CMV (Roche, Meylan, France). Recombinant plasmids were introduced into E. coli DH10B by electroporation (Gene Pulser II; Bio-Rad). Antibiotic-resistant colonies were selected on Luria-Bertani agar containing kanamycin (KAN) (30 µg/ml) and CRO (4 µg/ml). Recombinant plasmid DNA was recovered using Qiaprep spin miniprep columns (Qiagen).
The following primers were used for PCR mapping: CP604Av, 5'-GGGTATTTACGAGATGGCAGC-3' (nucleotides [nt] 157212 to 157232 of the plasmid pSN254 of S. enterica serotype Newport SL254; GenBank accession number CP000604); CP604Am, 5'-GCATTTCTCGGGTGACTTCCT-3' (nt 12446 to 12466; accession number CP000604); and CP60412000, 5'-GTTGGCAATCTCCTGGGTGAT-3' (nt 12000 to 12020; accession number CP000604).
Nucleotide sequence accession number. The nucleotide sequence of the class 1 integron gene cassette qacH-aadA1 was assigned GenBank accession number EU200458.
|
|
|---|
![]() View larger version (27K): [in a new window] |
FIG. 1. (A) Numbers of S. enterica serotype Concord isolates producing or not producing an ESBL in France (A) and in Norway (B) between 1996 and 2006.
|
In Norway, the single serotype Concord isolate collected between 1996 and 2000 (acquired after travel to Eritrea) was susceptible to all antibiotics tested. Between 2000 and 2006, 27 Concord isolates were registered. All but one isolate were from children adopted from Ethiopia. Twenty-three of the 27 Concord isolates were MDR with resistance to ESC (not shown and Table 1). The first ESBL-positive isolate was registered in 2000 in an adopted child from Ethiopia.
We evaluated the prevalence of carriers using data obtained from the systematic initial screening of children arriving in France (clinical examination, stool bacteriology and parasitology, and viral serologies). To this end, a French-certified association contacted all 187 families that adopted children from Ethiopia through the association between January 2005 and August 2006. One hundred and four families participated, and three carriers (2.9%) were found. Most of the adopted infants were from the same orphanage in Addis Ababa. We learned that there were two or three children per bed and that ESC were used extensively for treating life-threatening respiratory infections. The strain may have then been disseminated due to the promiscuity between infants and perhaps through the hands of the staff, thus explaining the long-lasting circulation of the strain. Overcrowding and understaffing are risk factors commonly observed in neonatal units in developing countries, for example, in Tunisia in 1988 (serotype Wien producing SHV-2) (19) and in 2002 (serotype Livingstone producing CTX-M-27) (6). However, the source of this strain in Ethiopia has yet to be investigated. As it was very difficult to prevent transmission in Ethiopia, we focused on preventing dissemination of the strain in France by informing all certified associations about the risk of transmission of resistant Salmonella and the adoptive families about the importance of basic hygiene practice. In 2007, the number of ESBL-producing Salmonella isolates in adopted children from Ethiopia was decreasing in France, with only 13 isolates, and no secondary cases, detected at the FRNC-Salm. In Norway, the adoption agency changed transit home in 2005, and in 2006, no new cases were reported.
Antimicrobial resistance phenotypes and genes and molecular typing. The antimicrobial resistance phenotypes of the 60 French isolates are described in Table 1. Thirty-six MDR serotype Concord and seven MDR monophasic isolates resistant to ESC contained the blaCTX-M-15 ESBL gene. The MICs of CRO, CAZ, and cefepime (FEP) were >256 µg/ml. Of the 43 ESBL-producing isolates, 38 were positive for the penicillinase blaTEM-1 gene and none were positive for the penicillinase blaOXA-1 group gene. All but three MDR strains harbored a class 1 integron containing a novel qacH-aadA1 gene cassette. One isolate harbored an additional class 1 integron containing a dfrA7 gene cassette. Three CTX-M-15-producing isolates with decreased susceptibility to ciprofloxacin (0.25 to 0.5 µg/ml) were found to contain the plasmid-mediated quinolone resistance gene qnrA1. One of these isolates, 05-3728, was studied previously (11).
The resistance profiles of the eight selected Norwegian isolates are described in Table 1. All of the isolates produced an ESBL, with the exception of one, which was positive for the penicillinase blaTEM-1 gene. We identified blaSHV-12 in five ESBL isolates collected before 2004 and blaCTX-M-15 in the remaining two ESBL isolates collected from 2004 to 2005. The MICs of CRO, CAZ, and FEP were >256 µg/ml for the CTX-M-15-producing isolates. The SHV-12-producing isolates had CAZ MICs of >256 µg/ml, CRO MICs of 64 µg/ml, and FEP MICs of 8 to 16 µg/ml. Several different class 1 integron gene cassettes were identified (Table 1).
The clonal relatedness of S. enterica serotype Concord and monophasic 6,7:l,v:- isolates was assessed by standard PFGE analysis of XbaI-digested chromosomal DNA (Fig. 2 and Table 1). Twenty different PFGE profiles were obtained from 68 clinical isolates and the reference strain. All ESBL-producing isolates clustered together (Dice correlation coefficient, 72%). SHV-12-producing isolates displayed the same profile, X10 (divided into three subtypes), distinct from the eight obtained from CTX-M-15-producing isolates. Among the CTX-M-15-producing isolates, X12 was predominant (n = 20). This profile was close to other CTX-M-15 isolate profiles, X11 and X13 to X16. However, two profiles, X17 and X18, differed more significantly. Profile X17 was observed in the three isolates containing the qnrA1 gene, and profile X18 was observed in four isolates displaying additional resistance to KAN. The two other clusters contained 15 pansusceptible isolates. The profiles obtained from the seven monophasic isolates were X11c (n = 1), X12 (n = 2), X15 (n = 3), and X16 (n = 1), suggesting that they were derived from the Concord serotype.
![]() View larger version (112K): [in a new window] |
FIG. 2. Representative XbaI-PFGE profiles of S. enterica serotype Concord isolates studied. A dendrogram was generated with Bionumerics software. The PFGE profile, number of isolates, and resistance type are indicated.
|
Transferability and location of the ESBL genes. Conjugation experiments were carried out on liquid and solid media with a subset of 12 ESBL-producing isolates selected based on the year and country of isolation, the antibiotic resistance phenotype, and the PFGE type. ESC transfer was successful only for the two Norwegian SHV-12-producing isolates tested and two French CTX-M-15-producing isolates (Table 2).
|
View this table: [in a new window] |
TABLE 2. Cefotaxime resistance transfer experiments
|
A large plasmid of approximately 300 kb that did not hybridize to the blaCTX-M probe, but which did hybridize to a blaTEM probe, was present in all CTX-M-15-producing Salmonella isolates, with the exception of isolate 04-347 and the two monophasic isolates tested. This plasmid was also absent in the single susceptible isolate tested (data not shown). Although the Norwegian isolate 07-699 did not undergo transfer of ESC resistance to E. coli, it contained at least two plasmids, one of approximately 250 kb that hybridized to the blaCTX-M-15 probe and one of approximately 140 kb that hybridized to an IncA/C probe (data not shown).
Two French isolates resistant to KAN, 06-8636 and 06-8404, which did not undergo transfer of ESC resistance, were found to contain a large plasmid of approximately 320 kb that did hybridize to the blaCTX-M-15 probe (data not shown).
PCR typing targeting IncA/C and IncY replicons was performed in all other Salmonella isolates. An IncA/C replicon was found only in isolate 07-699, whereas the IncY replicon was found in all isolates except 06-8404 and 06-8636, 07-699, and the susceptible isolates tested. S1 nuclease and Southern blotting showed that the IncY plasmid in Salmonella isolates (except that of 04-347) was approximately 100 kb in size (data not shown). In a previous study, two different replicons were associated with blaCTX-M-15-positive plasmids from Salmonella and E. coli isolates collected between 2001 and 2003 from the United Kingdom, Honduras, and Pakistan (22). Nine of 22 plasmids were positive for the IncI1 replicon, and 12 of 22 plasmids were positive for IncFII (in some strains associated with IncFIA and/or IncFIB replicons in fusion plasmids).
The fact that the ESC resistance transfer experiments were unsuccessful for the majority of the isolates, with no hybridization of their plasmid DNAs to a blaCTX-M-15 gene probe, suggested a chromosomal location of the blaCTX-M-15 gene in these isolates. This hypothesis was confirmed by I-CeuI PFGE, which clearly showed that the blaCTX-M-15 gene was located in a chromosomal fragment of approximately 900 kb (Fig. 3). This
900-kb fragment hybridizing to the 16S rRNA gene probe was absent in isolates 05-5343 and 04-347 (carrying blaCTX-M-15 on a plasmid), 06-8636 (an isolate additionally resistant to KAN), and 07-699 (a Norwegian isolate) and in susceptible isolates (Fig. 3 and data not shown). These isolates that did not have the
900-kb fragment displayed increased intensity of the band at
700 kb, suggesting that two DNA fragments comigrated at this position, as observed for 779-kb and 738-kb fragments of S. enterica serotype Typhimurium strain LT2 (27). Because the endonuclease I-CeuI cleaves only rrn operons and because the number (n = 7) and locations of the rrn genes are highly conserved in Salmonella (27), this
900-kb chromosomal fragment was interpreted to derive from one of the two DNA chromosomal fragments contained in the
700-kb band by insertion of the blaCTX-M-15 gene and additional DNA. The blaTEM gene was not detected on chromosomal fragments (data not shown). Chromosomal location of ESBL genes in enterobacteria is extremely rare, with only two reports showing chromosomal integration of blaCTX-M-9 or blaCTX-M-15 in E. coli (12, 16).
![]() View larger version (60K): [in a new window] |
FIG. 3. Chromosomal location of the blaCTX-M-15 gene. (A) PFGE separation of I-CeuI-digested DNAs from seven S. enterica serotype Concord and one S. enterica serotype 6,7:l,v:- isolates. (B and C) Southern hybridization with a blaCTX-M-specific internal probe (B) and a 16S rRNA gene probe (C). M, XbaI-digested DNA from S. enterica serotype Braenderup strain H9812; lane 1, isolate 04-347 (Concord, ESBL); lane 2, isolate 05-5343 (Concord, ESBL, qnrA1); lane 3, isolate 04-7498 (Concord, ESBL); lane 4, isolate 05-705 (Concord, ESBL); lane 5, isolate 05-5547 (Concord, ESBL); lane 6, isolate 05-9115 (6,7:l,v:-, ESBL); lane 7, isolate 04-9447 (Concord, ESBL); lane 8, isolate 04-2781 (Concord, pansusceptible).
|
![]() View larger version (13K): [in a new window] |
FIG. 4. Schematic representations of the various genetic environments of the blaCTX-M-15 gene. (A) The plasmid blaCTX-M-15 gene from S. enterica serotype Concord 04-347. (B) Recombinant clone p05-0004-C1 with the chromosomally located blaCTX-M-15 gene (9-kb insert) from S. enterica serotype Concord 05-0004. Open reading frames and genes surrounding the blaCTX-M-15 gene are shown as boxes, with an arrow indicating the orientation of each coding sequence and the gene name shown under the corresponding box. The putative IRR motif of ISEcp1 is indicated by a horizontal bar. The primers used for PCR mapping are shown with arrows above the boxes. The white boxes denote homology with sequences from MDR IncA/C plasmid pSN254 from S. enterica serotype Newport (GenBank accession number CP000604) (37). Coordinates of the pSN254 sequences are indicated. The nucleotides underlined in panel A correspond to the 5-bp directed repeats.
|
We found that DNA sequences from the pSN254 IncA/C2 plasmid were in the vicinity of the chromosomal blaCTX-M-15. One of our earliest Concord isolates (04-347), isolated in 2004, contained a blaCTX-M-15-carrying fusion plasmid with replicon elements of types IncY (resident plasmid of MDR Concord) and IncA/C2. The blaCTX-M-15 gene was inserted into this IncA/C2 plasmid at position 12411 by ISEcp1. ISEcp1 was complete (1,656 nt), and 5-bp direct repeats flanked it upstream and downstream from an IRR-like sequence. In isolate 07-699, the blaCTX-M-15 gene was also found in an IncA/C2 plasmid, but in a different position from that identified in isolate 04-347. ESC resistance in isolates 07-699 and 06-8636 did not appear to be transferable, despite the presence of large plasmids; this may have been due to the absence in these isolates of a helper IncY plasmid, which provides functions necessary for mobilization or allows the generation of an autotransferable fusion plasmid (as observed for isolate 04-347). Failure of the transfer of ESC resistance by electroporation might have been due to the loss of integrity of these large plasmid DNAs (250 to 320 kb) after a standard lysis alkaline extraction.
The sequences downstream of the chromosomally located blaCTX-M-15 gene were identical to those of the IncA/C2-IncY plasmid-located blaCTX-M-15 gene of isolate 04-347 (corresponding to nucleotide positions 12400 to 16000 of plasmid pSN254). However, there were rearrangements in the upstream region. The insertion of the IS26 element in ISEcp1 might have been the first event on the original IncA/C2 plasmid (three intact IS26s are present in pSN254). Genetic rearrangements (deletion and inversion) due to IS26 mobilization might have resulted in a structure in which the two distant described parts of pSN254 have been brought in proximity. We can also hypothesize that the whole, or part of, the IncA/C2-IncY fusion plasmid contained in isolate 04-347 was integrated into the chromosome of other Concord-derived isolates collected later; this probably occurred through IS26-mediated cointegration. Naas et al. (29) demonstrated Tn1- or IS26-mediated integration of a plasmid-borne inhibitor-resistant TEM, blaTEM-21, from Proteus mirabilis into E. coli chromosomal DNA. In our study, sequencing the ends of the inserted DNA by genome walking will be necessary in order to elucidate the precise mechanism of integration (homologous recombination, IS- or transposon-mediated transposition, or cointegration).
Published ahead of print on 9 March 2009. ![]()
L.F. and A.D. contributed equally to this work. ![]()
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»