Salmonella enterica serovar Typhimurium Virulence-Resistance Plasmids Derived from the pSLT Carrying Nonconventional Class 1 Integrons with dfrA12 Gene in Their Variable Region and sul3 in the 3′ Conserved Segment

Bacterial strains

In the investigation performed by Amar et al., ¹ 66 S. Typhimurium isolates showed a SGI1-like resistance phenotype, although SGI1 was not present. From this series 14 isolates which shared a core resistance pattern to ampicillin, tetracyclines, sulfamethoxazole, and full or intermediate resistance to streptomycin (MIC≥32 mg/L, resistant; MIC=16 mg/L, intermediate) were selected for the present study. Ten of these isolates displayed the SGI1 characteristic pentaresistance (AMP-CHL-STR-SMX-TET) phenotype. Additional resistance to gentamicin, kanamycin, nitrofurantoin and/or trimethoprim was also present in some of the isolates. They were collected between 2004 and 2006 and were provided by following project partner institutes: Istituto Superiore di Sanità (ISS), Rome, Italy (five human isolates) and Animal Health and Veterinary Laboratories Agency (AHVLA), Weybridge, United Kingdom (nine porcine isolates). The strain designations used in this manuscript correspond to BfR lab codes, which are made up by the year of strain receipt followed by a serial processing number. S. Typhimurium SUO5 (containing pUO-StVR2) ²² was included as a control in the experiments involving analysis of ‘bla_OXA-1-In8-like’-positive isolates.

Phenotypic analysis

To confirm the features of the isolates received by BfR, they were subjected to the following analyses. Serotyping was performed according to the White-Kauffmann-Le Minor scheme, ¹⁸ and phage typing following Anderson et al. ² Isolates were tested by broth microdilution, according to the guidelines of the CLSI (M07-A8), ¹² using custom-defined microtitre plates (TREK Diagnostic Sytems, East Grinstead, UK) with 14 antimicrobial substances: ampicillin (AMP), cefotaxime (FOT), ceftazidime (TAZ), chloramphenicol (CHL), ciprofloxacin (CIP), colistin (COL), florfenicol (FLO), gentamicin (GEN), kanamycin (KAN), nalidixic acid (NAL), streptomycin (STR), sulfamethoxazole (SMX), tetracycline (TET), and trimethoprim (TMP). ⁸ The evaluation of the results was performed according to the European Community Decision 2007/407/EC using the epidemiological cut-off values (ECOFFs) set by the European Committee on Antimicrobial Susceptibility Testing ¹⁴ (www.eucast.org).

All isolates were tested by disk diffusion method (M07-A8) ¹² to determine their susceptibility to amoxicillin-clavulanic acid (AMC), nitrofurantoin (NIT), spectinomycin (SPE), and sulfamethoxazole/trimethoprim (SXT) (Oxoid Ltd., London, England). The results were interpreted using CLSI clinical breakpoints (M100-S21; for these antimicrobials no ECOFFs for disc diffusion were available for Salmonella spp.). ¹³

Detection of resistance and virulence determinants

Depending on their respective resistance phenotype all isolates were tested by PCR amplifications to assess the presence/absence of 30 resistance genes, the class 1 integrase gene, class 1, and class 2 integrons and 15 pSLT genes (spvC, rck, pefA, traT, repA2, traX-finO, oriT, samA, pefB, pefD, parA, parB, ccdAB, pefI/ORF7, and srgA). Primers and amplification conditions were described previously.^{6,10,20,25,44} Mutations in the gyrA gene in the nalidixic acid-resistant isolates and in integrons inserted gene cassettes were identified by DNA sequencing of the amplicons (Qiagen, Hilden, Germany). ²¹ The sequences obtained were analyzed using the BLAST program (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Plasmid analysis and Southern blot/hybridization

All isolates were screened for the presence of plasmids by alkaline lysis using the Kado and Liu method. ³⁰ Bacterial plasmid DNA was separated on 0.8% vertical agarose gels. Escherichia coli reference plasmids R27 (169 kb), R1 (94 kb), RP4 (55 kb), and ColE1 (6 kb) were used as size standards. Plasmid sizes were determined by the S1-nuclease (Takara, Saint-Germain-en-Laye, France) digestion method followed by pulsed field gel electrophoresis (PFGE) as previously described.^5,22 For size determination, the Lambda Ladder PFG Marker (New England Biolabs, Schwalbach, Germany) was used. The incompatibility (Inc) groups of plasmids were determined by PCR-based replicon typing.^9,16 DNA obtained either from PFGE- or from alkaline lysis-generated plasmid gels was transferred by Southern blot onto nylon membranes and hybridized with different DNA probes specific for bla_TEM-1, bla_OXA-1, aadA1, strA, catA, cmlA, aphA1, tet(A), tet(B) and dfrA12 and also pSLT genes (spvC, and IncFII), as previously described. ¹⁹ The transfer potential of plasmid pMVN-VR42 (Table 2) was investigated by conjugation experiments using S. Typhimurium strain 08-02874 as donor and a sodium azide-resistant E. coli J53 derivative as recipient. Two parallel test series were conducted both at room temperature and 37°C, one in Luria-Bertani (LB) broth and the other one on solid LB (Oxoid) and M9 ³⁸ media on Millipore Membrane Filter (25 mm, 0,45 μm; Neolab, Heidelberg, Germany). Transconjugants were selected on eosin–methylene blue agar containing sodium azide (100 mg/L) and ampicillin (100 mg/L) or chloramphenicol (30 mg/L). Triparental mating was used to investigate the mobility of the plasmid pMVN-VR42. Overnight cultures of donor strain 08-02874, recipient E. coli J53 and a KAN resistant E. coli “helper” carrying plasmid pRK2013 ¹⁵ were grown in LB broth. From each culture, 100 μl were pooled to a final volume of 300 μl, pelleted by centrifugation and resuspended in 100 μl bi-distilled water. The resuspension was streaked onto a LB agar plate and incubated for 4–6 hours at 37°C. Growth surface was washed in 2 ml LB broth, which was then plated on eosin–methylene blue agar containing the selective antimicrobials indicated above. For restriction fragment length polymorphisms (RFLP) analyses pMVN-VR42 and pSLT were extracted from S. Typhimurium 08-02874 and S. Typhimurium LT2, respectively, using the Qiagen Plasmid Maxi Kit (Qiagen). The 100 kb CTX-M-1 IncI1 plasmid pSal60 from S. Paratyphi B (dT+) 09-01801 (unrelated to pSLT; characterised within the “ESBLs” EU-SafefoodERA project 08176) was used as negative control in RFLP-hybridization experiments (data not shown). Each plasmid was independently digested with the restriction endonucleases PstI, SalI, SphI, HindIII, ClaI, and EcoRI under the conditions recommended by the manufacturer (Fermentas, Leon-Rot, Germany). Restriction endonuclease-generated fragments were separated on 0.6% horizontal agarose gels by conventional electrophoresis, transferred by Southern blot onto a nylon membrane and hybridized with the entire pSLT plasmid. ²⁷ To prepare the probe, the pSLT plasmid of S. Typhimurium LT2 was extracted from a 0.45% agarose gel, purified by electroelution and labelled with the DIG DNA labelling kit (Roche, Mannheim, Germany) according to Herrero et al. ²⁷

Macrorestriction analysis and multilocus sequence typing

All isolates were subjected to macrorestriction analysis of genomic DNA using the restriction endonuclease XbaI (Roche Diagnostics, Mannheim, Germany). A CHEF-DRIII SYS220/240 system (Bio-Rad Laboratories, Munich, Germany) was used to separate the fragments by PFGE. Agarose gel plugs were prepared and PFGE was performed using the PulseNet standardized protocol (www.pulsenet-europe.org). The resulting profiles were analyzed by recording the presence or absence of fragments larger than 33 kb. Profiles in which there were differences in two or more bands were designated as XT1, XT2, etc.; patterns with only one band difference were sub-designated using letters (e.g., XT1a, XT1b, XT1c).

For further typing, all isolates were analysed by multilocus sequence typing (MLST) according to Kidgell et al. ³¹ and to the recommendations of the MLST database hosted by the University College Cork, Ireland (http://mlst.ucc.ie/mlst/dbs/Senterica). For each isolate seven housekeeping genes aroC, dnaN, hemD, hisD, purE, sucA, and thrA were amplified by PCR and sequenced (Qiagen). The DNA sequences obtained were processed by the software SeqMan Pro (DNAStar, Inc., Madison, WI) and the alleles and sequence types subsequently determined using the MLST database website.

Identification of bla-positive isolates and integron detection

All 14 S. Typhimurium isolates were analysed by PCR for the presence of bla genes and class 1 integrons. Twelve isolates generated amplicons for bla_TEM-1 (503 bp), while the other two tested positive for bla_OXA-1 (708 bp). The gene bla_PSE-1 was not identified in any of the isolates. The results of the PCR-screening for further resistance genes are summarized in Table 1.

The 14 isolates were screened for class 1 integrons by the classical PCR analysis using the 5′CS/3′CS primers (Table 1). ³² The two bla_OXA-1-positive isolates generated a PCR product of 2,000 bp containing the bla_OXA-1 and aadA1 genes. From the 12 isolates positive for the bla_TEM-1 gene, only one generated a class 1 integron PCR-product (1,600 bp), which carried dfrA1-aadA1 gene cassettes. Ten bla_TEM-1 isolates, negative for this PCR-amplification, were positive for the class 1 integrase-coding gene intI1 and the antimicrobial resistance-conferring genes dfrA12, aadA2, cmlA1 and sul3. The presence of a nonconventional (no amplification with the cited primers) class 1 integron described in S. enterica isolates from Portugal was therefore, suspected. ³ This integron contains an unusual 3′ conserved segment with the sulfamethoxazole resistance conferring gene sul3. The presence of this integron was confirmed by PCR using primer sets for intI1/dfrA12 (1,021 bp), dfrA12/aadA2 (1,697 bp), and cmlA1/aadA1 (1,577 bp). The size of the expected amplicons was determined by means of the nucleotide sequence of the sul3-carrying integron deposited by Antunes et al. in the EMBL database (accession number EF051037). ³ The remaining bla_TEM-1 positive isolate did not possess any class 1 integron.

Characterization of bla_OXA-1-In8-like positive isolates

The two S. Typhimurium DT120 isolates from cases of human infection in Italy (08-02878 and 08-02879), and which were ‘bla_OXA-1-In8-like’-positive, showed resistance phenotype [AMP-AMC]-CHL-[STR(r/i)-SPE(r/i)]-TET-SMX-NIT, as displayed by the control strain SUO5, which carries the pUO-StVR2 plasmid. The isolates were positive for the spvC virulence gene typical for pSLT. Subsequent plasmid analysis revealed that both isolates from Italy lacked pSLT but contained an IncFII plasmid of ∼140 kb (the same Inc group as pSLT) (Table 1). These IncFII plasmids also hybridized with the spvC and the bla_OXA-1probes. These hybrid plasmids also carried the ∼140 kb pUO-StVR2 genes catA1, aadA1, tet(B), and sul1, and were therefore, assigned to the S. Typhimurium pUO-StVR2 group. The screening for other selected typical pSLT virulence genes showed that both isolates were, like control strain SUO5, positive for rck, pefI, traT, traX, oriT, samA, parA, parB, ccdAB, and srgA genes. No amplification of the pefA, pefB, pefD genes of the pef (plasmid-encoded fimbriae) operon, the repA2 of the IncFIB replication region, and the rsk locus (regulation of serum killing), were observed, suggesting the absence of these genes as reported in pUO-StVR2 (accession no. AM991977) (Table 2).

Both bla_OXA-1-In8-like-positive isolates were subjected to XbaI-based PFGE analysis. By this method, they were assigned to two closely-related XbaI PFGE profiles designated XT1 and XT1a (XT1a possessed an additional band of ∼230 kb, data not shown). The pUO-StVR2 control strain SUO5 also showed the XT1 PFGE profile. The results are summarized in Table 2.

Characterization of bla_TEM-1-dfrA12 positive isolates

Two of the 12 bla_TEM-1-positive isolates, ascribed to MLST type ST34, only harboured plasmids <30 kb and were negative for the spvC gene. The remaining 10 isolates, all of them ascribed to ST19, carried the spvC gene and an integron associated to sul3. They also contained one IncFII plasmid ≥110 kb (Fig. 1). The ‘classical’ 94 kb pSLT plasmid was not detected in any of these isolates. All IncFII plasmids hybridized with probes for the genes spvC, bla_TEM-1, and dfrA12. These 10 isolates that contained VR plasmids showed the common core-resistance pheno-/genotype AMP-CHL-[STR-SPE]-SMX-TMP-SXT/bla_TEM-1-cmlA1-[aadA1-aadA2]-sul3-dfrA12. In some isolates additional resistances to KAN/aphA1 (in two isolates), TET/tet(A) (eight), and NIT (three) were detected. In these isolates, five different virulence genotypes were identified. All isolates were positive for the genes spvC, rck, pefI, traT and traX. With the exception of one isolate, all the pef genes were present.

OPEN IN VIEWER

FIG. 1.

(A) XbaI-pulsed field gel electrophoresis (PFGE) profiles (X) of representative isolates displaying the following PFGE patterns: XT2, XT3, XT4, XT5, XT6, XT7, XT8, XT1, XT1b, and XT1c (Table 2). Profile XT1a differed from profile XT1 in the presence of one additional band of ∼230 kb. Lanes M contained XbaI-digested DNA of Salmonella Braenderup H9812, which was used as a size standard. T, S. Typhimurium; (A) profile of LT2; (B) profile of S. Typhimurium DT104 isolate 07-00531 for comparison reasons. (B) Plasmid size determination by S1-PFGE in bla_TEM-1-positive isolates. Lanes M1 and M2 contained Lambda Ladder PFG Marker and XbaI-digested DNA of S. Braenderup H9812, respectively, used as size standards.

PFGE analysis revealed related but slightly different profiles for each isolate. One isolate was assigned to the XT1 profile (also detected in the bla_OXA-1-In8-like-positive isolates) and two exhibited the closely-related patterns XT1b and XT1c. The results are shown in Table 2 and Fig. 1.

Characterization of selected VR plasmids in bla_TEM-1-dfrA12 positive isolates

The three isolates (08-02873, 08-02874, and 08-02881) that showed an identical or highly similar XbaI-PFGE profiles (XT1, XT1b, XT1c) to those of the pUO-StVR2-carrying isolates (XT1 and XT1a) were subjected to further characterization. By plasmid DNA hybridization, the genes aadA1, cmlA, and strA were found to be located on the VR-plasmids. The tet(A) gene, present in the three isolates was located on smaller plasmids. The aphA1 gene, harboured only by the isolate 08-02873, was also located on a small plasmid. The IncFII plasmids carried by these strains were designated pMVN-VR41, pMVN-VR42, and pMVN-VR49, respectively (Tables 1 and 2).

The ∼160 kb plasmid pMVN-VR42 was selected for further characterization. The plasmid pMVN-VR42 could not be transferred by conjugation nor mobilized. Restriction analysis and hybridization of pMVN-VR42 and pSLT revealed individual profiles (Fig. 2). Nevertheless, many restriction enzyme-generated fragments of pMVN-VR42 had counterparts in pSLT, thereby suggesting a close relationship between the two plasmids. The CTX-M-1 IncI pSal60 plasmid, used as a negative control, generated restriction patterns which neither corresponded to those of pSLT nor hybridized with the pST90 probe (Supplementary Fig. S1, Supplementary data; Supplementary Data are available online at www.liebertpub.com/mdr).

OPEN IN VIEWER

FIG. 2.

Relation between pMVN-VR42 and pSLT. (A) Restriction analysis of pMVN-VR42 and pSLT. (B) Hybridization of the profiles obtained with the entire pSLT plasmid. λ, lambda phage DNA digested with PstI, used as a size standard. Lane 1, pMVN-VR42; lane 2, pSLT.