To force the loss of the ESBL-plasmid, a heat technique was performed [
15]. Single colonies of seven wild-type ESBL-
E. coli strains (VB977549, IMT19205, IMT27685, IMT16316, VB964041.2, IMT21183, IMT23463 of successful and pandemic sequence types [STs] ST131 and ST648 and different hosts including humans, companion and wild animals (Table
1)) were picked and inoculated in 5 ml BHI broth. BHI tubes were incubated at 45°C for 24 hours. Ten microliters of the overnight culture were spread on CHROMagar
TM plates. Plates were incubated at 37°C overnight. Replicate CHROMagar
TM plates containing an identical numbered grid on the backside of the plate were then prepared. One contained cefotaxime (4 μg/ml cefotaxime) and the other was prepared without supplementation of antibiotics. Twenty single colonies of each strain were randomly picked from the overnight incubated CHROMagar® plate and single colonies were placed on their identical grid locations in the agar of the replicate plates. This was to ensure that colonies from the two different plates could be assigned to the previously selected, single colony. ESBL-plasmid-“cured” single clones should not grow on cefotaxime-containing plates. These visually “cured” single clones were picked from the corresponding CHROMagar
TM plate without cefotaxime according to the grid and their phenotypic resistance against cefotaxime and other antimicrobial classes (Table
2) was screened using agar disc diffusion according to the CLSI method [
16]. They were further investigated using plasmid-profile-analysis to prove the loss of the plasmid [
17]. Clonal identity of the wild-type and the ESBL-plasmid-“cured”-variant was tested via
Xba I-pulsed-field gel electrophoresis (PFGE) [
18] and following comparative bioinformatic analyses. First, the number of orthologous genes in a pairwise comparison of the genome of the wild-type strain and the corresponding plasmid-“cured”-variant was checked using the OrthoMCL pipeline [
19]. In a second approach the phylogenetic distances of all strains were tested. The set of genes, which is present in each of all strains, the Maximum Common Genome (MCG), was therefore calculated, the allelic variants of the MCG from the strains was then extracted and a multiple alignment was built (Semmler, personal communication). Verified ESBL-plasmid-free strains were henceforward named PCV (plasmid-“cured” variant: PCV977549, PCV19205, PCV27685, PCV16316, PCV964041.2, PCV21183 and PCV23463). Presence of genes for TA systems on plasmids was investigated evaluating sequence data using bioinformatic methods. Both wild-type strains as well as PCVs were sequenced by an Illumina HiSeq 2000 sequencer. The resulting reads for the PCVs were used for a
de novo assembly (CLC Genomics Workbench 6.5, CLC Bio, Denmark). The contigs were then used as reference sequences for a reference mapping of the reads from the wild-type strains. All reads from the wild-type strains, which could not be mapped to the PCV sequence are supposed to represent the extracted plasmids and were used for another
de novo assembly, which resulted in the contigs of the plasmid sequences. Using BLAST for the plasmid and the PCV contigs, the genes for TA systems could be localized. Other plasmid-“curing” methods including treatment of bacteria with acridine orange and ethidium bromide [
14] were additionally tested and modified, however, as they turned out not to be successful, data are not included in this manuscript.
Table 1
Origin and genotypical characteristics of the ESBL-wild-type strains
VB977549 | Dog (C. lupus familiaris) | Urinary tract infection | 131 | CTX-M-14, CTX-M-15 | FIA/FIB |
pemI/K, vagC/D, hok/sok
| blaTEM-1, blaOXA-1, tet(A), tet(R), aadA, aac(6’)-ib-cr |
IMT19205 | Brown rat (R. norvegicus) | Feces | 131 | CTX-M-9a, CTX-M-14, CTX-M-15 | FIA/FIB |
hok/sok
| blaTEM-1, tet(A), sul2, strA, aac(3)-IV, aac(6’)-Ib-cr. |
IMT27685 | Raven (C. corax) | Feces | 131 | CTX-M-15 | not typed |
pemI/pemK, vagC/D
| blaOXA-1, tet(A), sul1, strA, strB aac(6’)-Ib-cr |
IMT16316 | Blackbird (T. merula) | Feces | 648 | CTX-M-14, CTX-M-15 | FIA/FIB |
pemI/K, vagC/D, srnB/C
| tet(A), tet(R), sul1, sul2, strA, strB, aadA, aac(3)-II mph(A), mrx, mphR, dhfrVII, |
VB964041.2 | Horse (E. ferus caballus) | Soft tissue/wound infection | 648 | CTX-M-15 | FIA/FIB |
pemI/K, vagC/D, srnB/C
| tet(A), tet(R), sul1, sul2, strA, strB, aadA, mph(A), mphR, dhfrVII |
IMT21183 | Human (H. sapiens) | Urinary tract infection | 648 | CTX-M-14, CTX-M-15 | FIA |
vagC/D, srnB/C
| tet(A), tet(R), sul1, sul2, strA, strB, aadA, aac(3)-II, mph(A), mphR, dhfrVII |
IMT23463 | Monk vulture (A. monachus) | Feces | 648 | CTX-M-9 | FIB |
PemI/pemK, srnB/C, hok/sok
| blaTEM-1, blaOXA-1, tet(A),sul2, strA, strB, aac(6’)-Ib-cr |
Table 2
Results of agar disc diffusion testing
VB977549 | R | S | R | S | R | R | S |
PCV977549 | S | S | R | S | S | S | S |
IMT19205 | R | S | R | R | I | S | S |
PCV19205 | S | S | S | R | S | S | S |
IMT27685 | R | R | R | S | R | R | R |
PCV27685 | S | S | R | S | S | S | S |
IMT16316 | R | S | R | R | R | R | R |
PCV16316 | S | S | R | S | S | S | S |
VB964041.2 | R | S | R | R | R | R | R |
PCV964041.2 | S | S | R | S | S | S | S |
IMT21183 | R | S | R | R | R | R | R |
PCV21183 | S | S | R | S | S | S | S |
IMT23463 | R | S | R | S | R | R | R |
PCV23463 | S | S | R | S | R | R | R |