Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Development of amoxicillin resistance in Escherichia coli after exposure to remnants of a non-related phagemid-containing E. coli: an exploratory study

Development of amoxicillin resistance in Escherichia coli after exposure to remnants of a... Objective: To determine the effect of exposure to remnants of a phagemid-containing E. coli, killed by treatment with a propanol-based hand rub, on antimicrobial resistance in E. coli isolates. Methods: An in vitro model was developed in which a clinical E. coli isolate (EUR1) was exposed to remnants of an E. coli K-12 strain containing a phagemid (pBS-E12) strain treated with Sterillium®. A series of 200 experiments was performed using this in vitro model. As a control, a series of 400 experiments was performed where the EUR1 was exposed either to the remnants of an E. coli K-12 strain (not containing a phagemid) (E12) treated with Sterillium® (n = 200) or to dried Sterillium® only (n = 200). The number of experiments that showed growth of an amoxicillin- resistant EUR1 isolate was evaluated in all three groups. An additional 48 experiments were performed in which a different clinical E. coli isolate (EUR2) was exposed to remnants of the pBS-E12 treated with Sterillium®. Whole- genome sequencing and phenotypic testing for AmpC beta-lactamase production was performed to investigate the mechanism behind this resistance development. Results: In 22 (11.0%) of 200 experiments in which the EUR1 isolate was exposed to remnants of a pBS-E12 an amoxicillin-resistant mutant isolate was obtained, as opposed to only 2 (1.0%) of 200 experiments involving the exposure of the EUR1 to Sterillium® only (risk difference: 10.0%; 95% CI 5.4–14.6%)) and 1 (0.5%) of 200 experiments involving the exposure of the EUR1 isolate to the remnants of the phagemid-free E12 (risk difference: 10.5%; 95% CI 6.1–14.9%). In 1 (2.1%) of the 48 experiments in which the EUR2 isolate was exposed to remnants of a pBS-E12 an amoxicillin-resistant mutant isolate was obtained. The development of resistance in all experiments was due to mutations in the promoter/attenuator region of the chromosomal AmpC beta-lactamase (cAmpC) gene leading to cAmpC hyperproduction. (Continued on next page) * Correspondence: joep.stohr@gmail.com Department of Infection Control, Amphia Hospital, Breda, the Netherlands Laboratory for Medical Microbiology and Immunology, Elisabeth-TweeSteden Hospital, Tilburg, the Netherlands Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 2 of 10 (Continued from previous page) Conclusion: Exposure of an E. coli isolate to another phagemid-containing E. coli that was treated with propanol- based hand rub increased the development of amoxicillin resistance. Although phagemids are cloning vectors that are not present in clinical isolates, this finding may have implications for hand disinfection practices in healthcare facilities. Keywords: Sterillium, AmpC, Antimicrobial resistance, Disinfection Introduction AMR through mechanisms other than transformation The last decades we have seen a dramatic worldwide [22, 23]. In this exploratory study, we used an in vitro increase in antimicrobial resistance (AMR) among model to compare the rate at which amoxicillin-resistant Gram-negative bacteria. One of the most remarkable mutants developed for an E. coli isolate exposed to phenomena is the rapid increase of plasmid-mediated amoxicillin and remnants of a pBleuscript KS(−) phage- beta-lactam resistance in Escherichia coli [1–3]. Inter- mid containing E. coli K-12 strain treated with national infection control guidelines recommend several Sterillium®, amoxicillin and remnants of a phagemid-free measures to control the spread of AMR, among which K-12 strain treated with Sterillium® or amoxicillin and propanol-based disinfection of hands is vital [4, 5]. dried Sterillium® only. These disinfection methods rapidly and effectively de- crease the number of viable bacteria on hands and Method thereby limiting the spread of resistant bacteria and Isolate selection healthcare-related infections [6–9]. However, intact bac- An E. coli -K12 (JM83, ATCC® 35607™) harbouring a II- terial DNA could potentially persist after propanol-based pBleuscript KS (−) phagemid (ATCC® 87047™) (copy bacterial cell lysis and mechanical cleaning and serve as number: 300–500) containing an amoxicillin-resistance a source of resistance determinants for other bacteria gene (bla ) (pBS-E12), was chosen to be exposed TEM-116 that reach the disinfected area [10, 11]. Uptake of plas- to the propanol-based hand rub and two E. coli isolates mid and chromosomal bacterial DNA from the environ- (EUR1 and EUR2) obtained from routine clinical ment through natural transformation has already been cultures (EUR1: peritoneal fluid culture, EUR2: urine described as a method of resistance acquisition in culture) of epidemiologically unrelated patients were se- streptococci, Helicobacter spp. and various other bacteria lected to be exposed to remnants of the pBS-E12, [12–16]. Recent studies have also shown that E. coli is Sterillium® and amoxicillin (for mutant selection). Strain able to take up DNA in various environments [16–21]. characteristics of the used strains are depicted in Supple- However, it remains unclear to what extent the uptake mentary table S1. Features of the high copy number of resistance plasmids via natural transformation con- pBleuscript phagemid are depicted in Supplementary tributes to the development of AMR in E. coli. More- table S2. The selection of the EUR1 and EUR2 was based over, it is unknown whether exposure to environmental on antimicrobial susceptibility pattern, i.e. susceptible to DNA remainders (plasmids) could facilitate this form of amoxicillin and resistant to trimethoprim based on AMR acquisition. Therefore, we developed an in vitro EUCAST clinical breakpoints version 9.0 [24]. The pBS- model in which a clinical E. coli isolate (EUR1) was ex- E12 has a trimethoprim MIC below the clinical break- posed to a pBleuscript KS(−) phagemid (encoding an point. Antimicrobial susceptibility testing was performed amoxicillin resistance gene)-containing E. coli K-12 using the VITEK 2® system (bioMérieux, Marcy l’Etoile, strain which had been treated with alcohol (Sterillium®). France). The minimal inhibitory concentration (MIC) In this experiment, we did not observe plasmid transfer for amoxicillin was additionally tested with ETEST® (results on file), yet we did observe the development of (bioMérieux, Marcy l’Etoile, France). The MICs of the beta-lactam resistance in the EUR1 isolate. This observa- EUR1, EUR2, and pBS-E12 for the various tested anti- tion led to the hypothesis that exposure to remnants of microbial agents are shown in Table 1. phagemid-containing E. coli (killed by treatment with Sterillium®) could lead to the development of AMR in E. Exposure to propanol-based hand rub (Sterillium®) and coli isolates that came in contact with the remnants of resistance induction the killed E. coli K12 strain, through increased chromo- Sterile glass surfaces (25mmx18mm) were inoculated somal mutations. Although not previously reported for with 10 μL of a 0.5 McFarland suspension of the pBS- E. coli, studies have already shown that in Salmonella E12 isolate (1.5 × 10 colony forming units (CFU)) in spp. and Pseudomonas spp. the presence of external brain-heart infusion broth (BHI). The inoculation of the DNA could lead to an increase in the development of glasses was performed in a Kojair biosafety cabinet class Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 3 of 10 Table 1 Minimal inhibitory concentration for various antibiotics Isolate MIC (mg/L) D68C AmpC & ESBL Detection set conclusion E-test® Vitek® amox amcl pita cfrx cfxt cftz cftx AmpC hyperproduction¥ pBS-E12 > = 256 > = 32 > = 128 8 <=4 <=1 <=1 No EUR1 6 <=2 <=4 4 <=4 <=1 <=1 Yes EUR1M1 > = 256 > = 32 8 32 > = 64 4 <=1 Yes EUR1M2 > = 256 > = 32 16 32 > = 64 4 <=1 Yes EUR1M3 > = 256 > = 32 <=4 16 16 <=1 <=1 Yes EUR1M4 > = 256 > = 32 8 32 32 <=1 <=1 Yes EUR1M5 > = 256 > = 32 8 > = 64 > = 64 4 2 Yes EUR1M6 256 > = 32 <=4 16 > = 64 <=1 <=1 Yes EUR1M7 > = 256 > = 32 8 32 > = 64 4 <=1 Yes EUR1M8 > = 256 > = 32 <=4 16 16 <=1 <=1 Yes EUR1M9 256 > = 32 <=4 16 16 <=1 <=1 Yes EUR1M10 128 > = 32 <=4 16 32 <=1 <=1 Yes EUR1M11 128 > = 32 <=4 32 16 <=0.25 <=0.25 Yes EUR1M12 > = 256 > = 32 16 32 16 2 1 Yes EUR1M13 > = 256 > = 32 16 32 32 2 1 Yes EUR1M14 > = 256 > = 32 16 32 16 1 <=0.25 Yes EUR1M15 196 > = 32 <=4 16 16 0.5 <=0.25 Yes EUR1M16 196 > = 32 <=4 16 16 0.5 <=0.25 Yes EUR1M17 256 > = 32 <=4 32 32 2 0.5 Yes EUR1M18 > = 256 > = 32 16 32 16 2 1 Yes EUR1M19 > = 256 > = 32 8 32 16 2 <=0.25 Yes EUR1M20 64 16 <=4 4 <=4 0.5 <=0.25 Yes EUR1M21 > = 256 > = 32 16 32 16 4 1 Yes EUR1M22 256 > = 32 16 32 16 4 1 Yes EUR1S1 > = 256 > = 32 <=4 16 32 <=1 <=1 Yes EUR1S2 > = 256 > = 32 8 32 > = 64 2 <=1 Yes EUR1E1 > = 256 > = 32 <=4 16 32 <=1 <=1 Yes EUR2 1 <=2 <=4 2 <=4 <=1 <=1 No EUR2M1 > = 256 > = 32 8 8 <=4 <=1 <=1 Yes MICs were measured with ETEST® (bioMérieux, Marcy l’Etoile, France) (amoxicillin) or the VITEK 2® system (bioMérieux, Marcy l’Etoile, France). ¥Interpretation of D68C AmpC & ESBL Detection set according to manufacturers instruction. Zone diameters as measured in the D68C AmpC & ESBL Detection set are depicted in Supplementary table S3. amox: amoxicillin; amcl: amoxicillin-clavulanic acid; pita: piperacillin-tazobactam; cfrx: cefuroxime; cfxt: cefoxitin; cftz: ceftazidime; cftx: cefotaxime; mero: meropenem; imip: imipenem;M: amoxicillin-resistant mutant after exposure to remnants of the pBS-E12 and Sterillium®S: amoxicillin-resistant mutant after exposure to Sterillium® onlyE: amoxicillin-resistant mutant after exposure to E12 and Sterillium® II Silver Line. The inoculated glasses were left to dry at PBS-E12 and Sterillium®-containing glasses were left to ambient air temperature for 10 +/− 1min. Subsequently, dry at ambient room temperature for 10 +/− 1 min. the glasses were inoculated with 30 μL of Sterillium®, Subsequently, the glasses were inoculated with 10 μLof an which was spread across the entire glass surface using 0.5 McFarland suspension of the EUR1 isolate (1.5 × 10 sterile plastic sticks. The Sterillium® used contained per CFU) in BHI, which was spread across the entire glass 100 g of solution: Propan-2-ol 45.0 g, propan-1-ol 30.0 g surface using sterile plastic sticks. The re-inoculated and mecetronium etilsulfate 0.2 g. The ratio of the volume glasses were left to dry at ambient air temperature for 10 − 5 of inoculated Sterillium® to glass surface area (3 × 10 L: +/− 1 min. Subsequently, the glasses were placed in a con- 0.00045 m ) was chosen to reflect the ratio of the volume tainer with 4 mL of BHI, containing 1 mg/L amoxicillin, of Sterillium® used in hand disinfection to the average vortexed for 30 s at 2800 rotations per minute and incu- − 3 2 hand surface area (3 × 10 L: 0.045 m )[25, 26]. The bated at 35 to 37C°. The duration the isolates were Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 4 of 10 subjected to a subinhibitory amoxicillin concentration is Experiments with the EUR1 and EUR2 isolates were meant to represent the duration of subinhibitory amoxicil- performed on separate days. For each experiment colonies lin plasma-concentrations in patients starting amoxicillin were picked of the EUR1 and EUR2 isolate and the pBS- treatment [27]. After 1 h of incubation, an additional 1 mL E12 strain in stationary growth phase after incubation at of BHI broth containing 320 mg/L amoxicillin was added 35 to 37C° for 18 to 24 h on Muller Hinton agar. The to the glasses carrying containers, resulting in 5 mL of pBS-E12 strain was inoculated on Muller Hinton agar BHI broth containing 64.8 mg/L amoxicillin in the glasses plates containing an amoxicillin 10 μg disk. Only colonies carrying containers. The containers were then incubated of the pBS-E12 growing in the direct surrounding of the at 35 to 37C° for 72 h and visually inspected for bacterial amoxicillin disk were used in the experiments. The experi- growth at 8, 24, 48 and 72 h. At visible growth or after 72 mental setup is illustrated in Fig. 1. h of incubation in absence of visible growth, 1 μLofthe suspension was plated on a Muller Hinton agar containing Control experiments 8 mg/L trimethoprim and 64 mg/L amoxicillin (MH-TA). A series of 200 control experiments were performed in The inoculated MH-TA plates were incubated for 18 to which the glasses were only inoculated with Sterillium® 24 h at 35 to 37C°. Colonies growing on the MH-TA and the EUR1 and a series of 200 experiments were underwent species identification performed with VITEK performed in which the glasses were inoculated with the MS (bioMérieux, Marcy l’Etoile, France), using the VITEK E. coli K-12 strain (JM109, ATCC® 53323™) not contain- MS Knowledge Base v.2.0, and antimicrobial susceptibility ing the pBleuscript KS (−) high copy number phagemid testing performed with the VITEK 2® system (bioMérieux, (E12) (Supplementary Table S1), Sterillium® and the Marcy l’Etoile, France), using EUCAST susceptibility EUR1 isolate. Experiments were performed on separate breakpoints version 9.1 [24]. The MIC of all isolates to days. For each experiment colonies were picked of the amoxicillin was additionally tested using ETEST® (bioMér- EUR1 isolate and E12 strain after incubation at 35 to ieux, Marcy l’Etoile, France). Grown E. coli colonies with a 37C° for 18 to 24 h on Muller Hinton agar. The setup of MIC of > 8 mg/L for trimethoprim, measured in the the control experiments is illustrated in Fig. 2. VITEK 2® system (bioMérieux, Marcy l’Etoile, France), and with a MIC of > 64 mg/L for amoxicillin, measured Disinfection experiments with ETEST® (bioMérieux, Marcy l’Etoile, France), were To determine if the pBS-E12 was still culturable after considered mutants. A series of 200 experiments were exposure to Sterillium®, we performed a series of 100 performed using the EUR1 isolate. Additionally, a series of experiments in which the glasses were inoculated with 48 experiments were performed using the EUR2 isolate. the pBS-E12 followed by Sterillium® only. The inoculated Fig. 1 An illustration of the experiments in which the EUR1 or EUR2 isolates were exposed to amoxicillin and the pBS-E12 treated with Sterillium® Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 5 of 10 Fig. 2 An illustration of the experiments in which the EUR1 isolate was exposed to amoxicillin and the E12 treated with Sterillium®, or to amoxicillin and Sterillium® glasses were placed in a container with 5 mL of BHI, Acquired resistance genes were called when at least 60% vortexed for 30 s at 2800 rotations per minute and incu- of the length of the best matching gene in the ResFinder bated for 72 h at 35 to 37C°, visually inspecting growth database was covered with a sequence identity of at least at 24, 48 and 72 h. In case there was no visible growth 90%. If an unknown or known chromosomal mutation after 72 h of incubation, 1 μL of broth was plated on was detected in genomic regions implicated in beta- Muller Hinton agar. lactam resistance using PointFinder, the region was extracted from the assembled genome using Biopython Whole-genome shotgun sequencing (WGS) and de novo v.1.73. Subsequently, the extracted genomic regions were assembly aligned using Vector NTI Advance 11 software (Ther- The wildtype isolates (EUR1, EUR2), the pBS-E12 and moFisher Scientific, Waltham, USA) to the correspond- E12 strains, and a selection of the mutant isolates were ing region of either the EUR1 or EUR2 isolate and the sequenced on an Illumina MiSeq (Illumina, San Diego, corresponding region of the E. coli K-12 strain MG1655 United States) and assembled with CLC Genomics (GenBank database accession number NC000913.3), the Workbench v.11 or v.12 (Qiagen, Hilden, Germany). Se- pBS-E12 strain, and the E12 strain. Moreover, to verify lection criteria were: in the first thirteen experiments in that the observed amoxicillin resistance was not the which growth of a mutant isolate was detected, an iso- result of the acquisition of the amoxicillin resistance late was sent for sequencing. The following quality con- gene from the pBleuscript(−) phagemid, all assembled trol criteria were used: coverage ≥20; number of genomes were screened for the presence of this gene scaffolds ≤1000; N50 ≥ 15.000 bases and maximum scaf- using ABRicate v.0.8.13 with the same coverage and fold length ≥ 50.000 bases. identity thresholds as those used in ResFinder. Addition- ally, all mutated strains and the pBS-E12, EUR1, EUR2 Mechanism of resistance isolates were phenotypically screened for AmpC produc- The presence of genomic resistance determinants tion using the D68C AmpC & ESBL Detection set conferring amoxicillin and trimethoprim resistance in (Mastdiscs, Mastgroup Ltd., Bootle United Kingdom) the assembled genomes of the mutated E. coli was iden- and interpreted according to manufacturer’s instruction. tified with the online bioinformatic tools ResFinder v3.1 and PointFinder v3.1 (Center for Genomic Epidemi- Whole-genome multilocus sequence typing ology, Technical University of Denmark, Lingby, Whole-genome multilocus sequence typing (wgMLST) Denmark) [28, 29]. Genomes were screened for known (core and accessory genome) was performed of both and unknown chromosomal resistance mutations. wildtype isolates and all sequenced mutant isolates using Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 6 of 10 Ridom SeqSphere+, version 5.1.0. (Ridom, Münster, shows the various mutations at given positions of the Germany). Species-specific typing schemes were used as promoter/attenuator region of the cAmpC gene for described by Kluytmans-Van den Bergh et al. [30]. All- every sequenced isolate in this study. Known and un- to-all pairwise genetic difference was calculated between known mutations in the promoter/attenuator region of the isolates by counting the total number of allele differ- the chromosomal AmpC beta-lactamase (cAmpC) gene ences in the wgMLST typing scheme and by dividing the were detected in the genomes of all mutant E. coli total number of allele differences in the wgMLST typing isolates that were sequenced (EUR1M1-EUR1M10, scheme by the total number of shared alleles in the EUR1S1-EUR1S2, EUR2M1), but not in the EUR1 wgMLST typing scheme, ignoring pairwise missing isolate. Two unknown mutations were detected in the values. promoter/attenuator region of the EUR2 isolate (Table 2). Alignment against the corresponding region of an E. Statistics coli K-12 strain revealed that one of these mutations was Risk differences were estimated using a generalised lin- in the alternate − 10 box promoter box (Table 2). ear model with binomial distribution, an identity link However, these mutations did not increase the MIC for and robust error estimation (SPSS version 25). amoxicillin in the EUR2 isolate. In every mutant (EUR1M1-EUR1M10, EUR1S1-EUR1S2, EUR2M1) Accession numbers alignment of the cAmpC promoter/attenuator region All generated raw reads were submitted to the European revealed mutations, when compared to the same Nucleotide Archive (ENA) of the European Bioinformat- genomic region of the corresponding wildtype isolates, ics Institute (EBI) under the study accession number: in regions implicated in cAmpC hyperproduction as PRJEB34354. described by Tracz et al. (Table 2)[31]. Moreover, the different promoter/attenuator regions present in the mu- Results tant EUR1 isolates (Table 2) suggests de novo mutation Resistance induction experiment rather than the selection of a previously present mutated In 22 (11.0%) of 200 experiments in which the EUR1 subpopulation. No mutations in other regions implicated strain was exposed to the remnants of the pBS-E12 and in beta-lactam resistance were detected, nor were ac- Sterillium® an amoxicillin- and trimethoprim-resistant E. quired beta-lactam resistance genes detected in any of coli isolate (EUR1M1 - EUR1M22) was obtained, as the mutated (EUR1M1-EUR1M10, EUR1S1-EUR1S2, opposed to only 2 (1.0%) (EUR1S1, EUR1S2) of 200 EUR2M1) or wildtype isolates (EUR1, EUR2). The experiments involving the exposure of the EUR1 strain amoxicillin-resistance gene of the pBleuscript KS(−) pha- to Sterillium® only (risk difference 10.0%; 95% CI 5.4– gemid was not detected in any of the sequenced isolates 14.6%), and 1 (0.5%) (EUR1E1) of 200 experiments except the pBS-E12. In the pBS-E12 and E12 strain involving the exposure of the EUR1 strain to the rem- alignment of the cAmpC promotor/attenuator did not nants of the E12 and Sterillium® (risk difference 10.5%; reveal mutations, when compared to the same genomic 95% CI 6.1–14.9%). In the experiments performed using region of the E. coli K-12 strain MG1655. A dfrA1 gene the EUR2, 1 mutant isolate (EUR2M1) was grown in 48 conferring trimethoprim resistance was detected in the experiments. Amoxicillin-resistant strains did not only EUR1, EUR2 and all sequenced mutant isolates but not show increased MICs for amoxicillin but also for in the pBS-E12 and E12 strains. amoxicillin-clavulanic acid, piperacillin-tazobactam or Phenotypic tests for AmpC beta-lactamase production cephalosporins (Table 1). No increases in the MIC were were negative for the two wildtype isolates and the pBS- observed for carbapenems and non-beta-lactam antibi- E12 isolate. However, for all mutants, phenotypic testing otics tested. One out of 100 disinfection experiments showed AmpC beta-lactamase production (Table 1; showed growth of the pBS-E12, indicating that in only a Supplementary Table S3). minimal number of experiments the amoxicillin concen- tration during the experiments was influenced by viable Whole-genome multilocus sequence typing beta-lactamase-producing pBS-E12. The number of allele differences between the mutated EUR1M1-EUR1M10 isolates and wildtype EUR1 isolate Mechanisms of amoxicillin resistance ranged from 21 (0.59%) to 69 (1.96%) (median: 44.5 A selection of the isolates was sequenced, i.e. 10 mutant (1.27%)). In all but one (EUR1M9) of the mutant EUR1 E. coli isolates from the experiment in which EUR1 was isolates, the difference between the mutant and corre- exposed to the remnants of the pBS-E12, 2 mutants sponding control exceeded the threshold for genetic dis- from the control experiments, 1 mutant from the experi- tance between related and unrelated isolates as defined ments with the EUR2 isolate, the wildtype isolates EUR1 by Klutymans-van den Bergh et al. (Fig. 3)[30]. Both and EUR2, and the pBS-E12 and E12 strains. Table 2 mutated isolates from the control experiment (EUR1S1 Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 7 of 10 Table 2 Mutations in cAmpC promoter/attenuator region of the wildtype and mutant isolates E.coli K-12 Base in wildtype cAmpC promoter/attenuator region (wildtype) −32 − 23.1 − 21.1 −19 − 18 −16.1 − 11 + 58 + 81 T –– TG – CC G Isolate Position of mutation in cAmpC promoter/attenuator region Date experiments −32 −23.1 −21.1 −19 −18 −16.1 −11 + 58 + 81 performed pBS-E12 E12 EUR1 (wildtype) A EUR1M1 A A 10th May EUR1M2 G C A A 10th May EUR1M3 G C A A 11th May EUR1M4 G C A A 12th May EUR1M5 G C A A 12th May EUR1M6 T A 13th May EUR1M7 G C A A 25th May EUR1M8 G C A A 26th May EUR1M9 G C A A 1st June EUR1M10 T A 1st June EUR1S1 T A 23th May EUR1S2 A A 17th August EUR2 (wildtype) A T EUR2M1 A A T 12th May Position numbering of the cAmpC promoter/attenuator region as defined by Mulvey et al. [31]. In position number n.x: decimal number x refers to an insertion at position n. Positions −32 and − 11 are part of the wild-type promoter boxes. Position − 19 is part of the alternate promoter box. Positions − 21.1 and − 23.1 are part of the spacer region of both the wild-type and alternate promoter. Position − 16.1 is part of the spacer region of the wild-type promoter. ¥ strain used in experiments on 1st of June. ¤ all experiments were performed in 2016 Fig. 3 A Neighbour-joining tree representing the percentage of wgMLST allele differences between all sequenced isolates (Green: EUR1 wildtype/ mutant isolates; Red: EUR2 wildtype/mutant isolates) Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 8 of 10 and EUR1S2) also showed a high number of allele differ- assays with other antibiotics (e.g. streptomycin and nali- ences when compared to the EUR1 wildtype (EUR1S1: dixic acid) need to be performed, to further assess the n = 34 (0.96%), EUR1S2: n = 36 (1.02%)). The high num- mutagenic potential of the remnants of the pBS-E12 and ber of allele differences between the various mutated Sterillium® in the EUR1 isolate. However, the remnants EUR1 isolates suggests de novo mutation acquisition ra- of the pBS-E12 could also facilitate the development of ther than the selection of a previously present mutated AMR by enabling the EUR1 and EUR2 isolates to survive subpopulation (Fig. 3). In the EUR2M1, when compared longer under high amoxicillin concentrations, possibly to its wildtype, no allele differences were observed. The through providing nutrients or through efflux pump ac- number of allele differences between the EUR1 and tivation, thereby increasing the chance of the occurrence EUR2 isolate was 3184 (97.73%)(Fig. 3). of a mutation leading to cAmpC hyperproduction. A small degree of heterogeneity in the cAmpC promotor Discussion region of the pBS-E12 population that lead to cAmpC An in vitro model was developed to simulate propanol- hyperproduction cannot be completely ruled out. There- based hand disinfection as is common practice in health- fore, it still remains possible that the cAmpC promotor care facilities nowadays. Bacteria are effectively killed, region of this heterogeneous subpopulation was horizon- but their remnants remain present after disinfection. tally transferred to the wildtype EUR1 and EUR2 Subsequently susceptible bacteria are exposed to the isolates. remaining debris, including DNA. Experiments per- Several studies have shown that environmental formed using this model showed a significant increase in stressors other than antibiotics can influence the the development of amoxicillin resistance in E. coli iso- development of AMR [35–38]. However, the association lates after exposure to the remnants of the pBS-E12 between exposure to remnants of phagemid-carrying E. treated with a propanol-based hand rub as compared to coli isolates and increased development of AMR has not exposure to remnants of the E-12 treated with a been investigated to date. Interestingly, this increase in propanol-based hand rub and as compared to exposure the development of AMR was only related to exposure to only a propanol-based hand rub. AmpC beta- to remnants of phagemid-containing E. coli isolates. lactamase hyperproduction, due to mutations in the Other studies have already shown that extracellular AmpC promoter/attenuator region, was responsible for DNA can lead to the development of AMR in other the development of the observed AMR. The number of bacteria [22, 23]. However, this is the first study relat- resistant mutants that developed in the isolates exposed ing exposure to external DNA to the development of to the remnants of the pBS-E12 also exceeded the num- beta-lactam resistance through chromosomal muta- ber of pBS-E12 growing in the disinfection experiments, tions in E. coli. which indicates that amoxicillin degradation by the pBS- Despite the increased development of AMR was only E12 beta-lactamase was not responsible for the observed related to exposure to remnants of the pBS-E12, it re- increase in AMR. All mutated EUR1 isolates showed a mains unknown if and what specific compounds of high number of allele differences when compared to the Sterillium® contribute to the observed increased develop- wildtype isolate, suggesting genome-wide mutations. ment of AMR. Contrary to other alcohol-based hand Interestingly this genome-wide mutational pattern was rubs, Sterillium® contains mecetronium etilsulfate which not detected in the EUR2 isolate. Indicating that the potentially has a lasting antimicrobial effect [39, 40]. genome-wide mutational pattern might not only be Further studies are needed to evaluate if this increased (control) experiment but also isolate related. development of AMR also occurs with other alcohol- The exact underlying mechanism leading to the AMR based hand rubs not containing such compounds. conferring mutations in the promoter/attenuator region Furthermore, the current study only simulates hand of the cAmpC remains unknown. Since the amoxicillin disinfection procedures so it remains unknown to what resistance gene of the pBluescript phagemid was not de- extent this phenomenon also could apply to environ- tected in the EUR1 and EUR2 mutant isolates, it seems mental disinfection. more likely that the remnants of the pBS-E12 increased Although cAmpC hyperproduction has been impli- the mutation rate in the EUR1 and EUR2 isolate cated in beta-lactam resistance in E. coli isolates follow- facilitating AMR development through cAmpC hyper- ing exposure to stepwise increasing concentrations of production. Previous studies have implicated several amoxicillin over several days [41, 42]. cAmpC hyperpro- mechanisms that may increase the mutation rate and re- duction in this study occurred after only a short time of sult in AMR, such as the general stress response and an exposure to amoxicillin concentrations lower than the increased constitutive mutation frequency [32–35]. Fur- MIC of the exposed E. coli isolate. Even in our control ther studies are needed to evaluate the role of these experiments, in which the E. coli isolates were only ex- mechanisms in our experiments. Moreover, mutagenic posed to Sterillium®, cAmpC hyperproduction developed Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 9 of 10 after only short sub-inhibitory amoxicillin concentra- Supplementary information Supplementary information accompanies this paper at https://doi.org/10. tions. Moreover, in the isolates EUR1M2-EUR1M5 and 1186/s13756-020-00708-7. EUR1M7–9 three mutations were detected in the cAmpC promotor region. Each of these mutations Additional file 1: Table S1. Characteristics of used strains. Table S2. individually could increase cAmpC production in E. coli Features of pBleuscript KS(-) phagmid ATCC® 87047TM. Table S3. Inhibition zone diameters of D68C AmpC & ESBL detection set for the [31, 43]. Perhaps, consecutive mutations, not a single pBS-E12, EUR1, EUR2 and mutant isolates. mutational event during the short time of sub-inhibitory amoxicillin concentration, lead to a step-wise increase in Abbreviations amoxicillin MIC in these mutant isolates. In the study AMR: Antimicrobial resistance; pBS-E12: E. coli -K12 containing a pBleuscript by Kohanski et al. [44] exposure to sub-lethal levels of KS (−) phagemid encoding for an amoxicillin resistance gene (bla ); TEM-116 EUR1: Clinical E. coli isolate 1; EUR2: Clinical E. coli isolate 2; MIC: Minimal amoxicillin also resulted in MIC increases to antimicro- inhibitory concentration; BHI: Brain-heart infusion broth; MH-TA: Muller bials other than from the beta-lactam group. We did not Hinton agar containing 8 mg/L trimethoprim and 64 mg/L amoxicillin; observe MIC increases or known mutations in WGS: Whole genome shotgun sequencing; ENA: European Nucleotide Archive; EBI: European Bioinformatics Institute; cAmpC: chromosomal AmpC resistance-associated genes of antibiotics other than the beta-lactamase beta-lactams. Contrary to other studies investigating the develop- Acknowledgments We kindly thank the Avans university of applied sciences Breda for providing ment of AMR in E. coli [41, 44], we used two clinical E. the JM83, JM109 isolates and pBluescript KS(-) phagemid. coli isolates for amoxicillin resistance induction. More- over, our in vitro model simulates propanol-based hand Authors’ contributions JS performed the experiments, analysed and interpreted the data, and disinfection procedures which are very common in drafted the manuscript. MK analysed and interpreted the data, and revised clinical practice [4, 5, 45]. Also, the duration of exposure the manuscript. CV performed the experiments, and revised the manuscript. to sub-inhibitory amoxicillin concentrations closely re- JR analysed and interpreted the data, and revised the manuscript. JK analysed and interpreted the data, and revised the manuscript. All authors flects the duration of sub-inhibitory amoxicillin plasma read and approved the final manuscript. concentrations in patients at the start of amoxicillin treatment [27]. Funding This study has some limitations. The pBluescript-KS None (−) phagemid is a cloning vector not present in clinical Availability of data and materials isolates. Whether remnants of E. coli isolates containing The datasets supporting the conclusions of this article is included within the wildtype phagemids also increase the development of article and its additional file. AMR needs further investigation. Moreover, since the Ethics approval and consent to participate pBluescript KS(−) phagemid contains both a bacterio- Not applicable. phage origin of replication and a plasmid origin of Consent for publication replication future studies are required to assess whether Not applicable. the observed effect is bacteriophage or plasmid related. Also, only a limited number of isolates were used in this Competing interests The authors declare that they have no competing interests. study. The extent to which this AMR induction is possible in other E. coli isolates or other species remains Author details unknown. Moreover, it remains unknown whether and Department of Infection Control, Amphia Hospital, Breda, the Netherlands. Laboratory for Medical Microbiology and Immunology, to what extent this in vitro phenomenon plays a role in Elisabeth-TweeSteden Hospital, Tilburg, the Netherlands. Amphia Academy the development of AMR in vivo. Infectious Disease Foundation, Amphia Hospital, Breda, the Netherlands. To the best of our knowledge, this is the first study Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands. Department of showing development of amoxicillin resistance in an E. Medical Microbiology and Infection Prevention, University of Groningen, coli isolate after exposure to a phagemid-containing E. University Medical Center Groningen, Groningen, the Netherlands. coli treated with a propanol-based hand rub. Received: 29 October 2019 Accepted: 11 March 2020 Conclusion References This exploratory study showed the development of 1. Bonnet R. Growing group of extended spectrum: the CTX-M enzymes. Antimicrob Agent Chemother. 2004;48(1):1–14. amoxicillin resistance in an E. coli isolate after exposure 2. Nordmann P, Naas T, Poirel L. Global spread of Carbapenemase producing to an unrelated phagemid-containing E. coli treated with Enterobacteriaceae. Emerg Infect Dis. 2011;17(10):1791–8. a propanol-based hand rub. Although phagemids are 3. Mathers AJ, Peirano G, Pitout JDD. The role of epidemic resistance plasmids and international high- risk clones in the spread of multidrug-resistant cloning vectors that are not present in clinical isolates, Enterobacteriaceae. Clin Microbiol Rev. 2015;28(3):565–91. this finding may have implications for hand disinfection 4. Tacconelli E, Cataldo MA, Dancer SJ, De Angelis G, Falcone M, Frank U, et al. practices in healthcare facilities. ESCMID guidelines for the management of the infection control measures Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 10 of 10 to reduce transmission of multidrug-resistant gram-negative bacteria in 28. Zankari E, Allesøe R, Joensen KG, Cavaco LM, Lund O, Aarestrup FM. hospitalized patients. Clin Microbiol Infect. 2014;20(s1):1–55. PointFinder: a novel web tool for WGS-based detection of antimicrobial 5. Boyce J, Pittet D. Centers for Disease Control and Prevention. Guideline for resistance associated with chromosomal point mutations in bacterial hand hygiene in health-care settings: recommendations of the healthcare pathogens. J Antimicrob Chemother. 2017;72(10):2764–8. infection control practices advisory committee and the HICPAC/SHEA/APIC/ 29. Zankari E, Hasman H, Cosentino S, Vestergaard M, Rasmussen S, Lund O, IDSA hand hygiene task force. MMWR Recomm Rep. 2002;51:1–45. et al. Identification of acquired antimicrobial resistance genes. J Antimicrob 6. Kampf G, Ostermeyer C, Heeg P, Paulson D. Evaluation of two methods of Chemother. 2012;67(11):2640–4. determining the efficacies of two alcohol-based hand rubs for surgical hand 30. Kluytmans-van den Bergh MFQ, Rossen JWA, Bruijning-Verhagen PCJ, Bonten MJM, Friedrich AW, Vandenbroucke-Grauls CMJE, et al. Whole- antisepsis. Appl Environ Microbiol. 2006;72(6):3856–61. genome multilocus sequence typing of extended-Spectrum-Beta-lactamase- 7. Kampf G, Kapella M. Suitability of Sterillium gel™ for surgical hand producing Enterobacteriaceae. J Clin Microbiol. 2016;54(12):2919–27. disinfection. J Hosp Infect. 2003;54(3):222–5. 31. Tracz DM, Boyd DA, Hizon R, Bryce E, McGeer A, Ofner-Agostini M, et al. 8. Kampf G, Ostermeyer C, Heeg P. Surgical hand disinfection with a propanol- ampC gene expression in promoter mutants of cefoxitin-resistant based hand rub: equivalence of shorter application times. J Hosp Infect. Escherichia coli clinical isolates. FEMS Microbiol Lett. 2007;270(2):265–71. 2005;59(4):304–10. 32. Firmino N, Goldsmith J-M, Morin A. Ampicillin resistance is increased in 9. Kramer A, Rudolph P, Kampf G, Pittet D. Limited efficacy of alcohol-based Escherichia coli K12 relA and spoT mutants but sub-inhibitory pretreatment hand gels. Lancet. 2002;359(9316):1489–90. does not induce adaptive resistance. J Exp Microbiol Immunol. 2013; 10. Mao D, Luo Y, Mathieu J, Wang Q, Feng L, Mu Q, et al. Persistence of 17(April):34–9. extracellular DNA in river sediment facilitates antibiotic resistance gene 33. Baharoglu Z, Mazel D. Vibrio cholerae triggers SOS and mutagenesis in propagation. Environ Sci Technol. 2014;48(1):71–8. response to a wide range of antibiotics: a route towards multiresistance. 11. Nielsen KM, Smalla K, Van Elsas JD. Natural transformation of Acinetobacter Antimicrob Agents Chemother. 2011;55(5):2438–41. sp. strain BD413 with cell lysates of Acinetobacter sp., Pseudomonas 34. Eliopoulos GM, Blazquez J. Hypermutation as a factor contributing to the fluorescens, and Burkholderia cepacia in soil microcosms. Appl Environ acquisition of antimicrobial resistance. Clin Infect Dis. 2003;37(9):1201–9. Microbiol. 2000;66(1):206–12. 35. Petrosino JF, Galhardo RS, Morales LD, Rosenberg SM. Stress-induced β- 12. Kwon DH, Dore MP, Kim JJ, Kato M, Lee M, Wu JY, et al. High-level β-lactam lactam antibiotic resistance mutation and sequences of stationary-phase resistance associated with acquired multidrug resistance in Helicobacter mutations in the Escherichia coli chromosome. J Bacteriol. 2009;191(19): pylori. Antimicrob Agents Chemother. 2003;47(7):2169–78. 5881–9. 13. Bae J, Oh E, Jeon B. Enhanced transmission of antibiotic resistance in 36. Fruci M, Poole K. Bacterial stress responses as determinants of antimicrobial Campylobacter jejuni biofilms by natural transformation. Antimicrob Agents resistance. In: Stress and environmental regulation of gene expression and Chemother. 2014;58(12):7573–5. adaptation in Bacteria. Hoboken: Wiley; 2016. p. 115–36. 14. Trzciński K, Thompson CM, Lipsitch M. Single-step capsular transformation 37. Nishino K, Yamasaki S, Hayashi-Nishino M, Yamaguchi A. Effect of NlpE and acquisition of penicillin resistance in Streptococcus pneumoniae.J overproduction on multidrug resistance in Escherichia coli. Antimicrob Bacteriol. 2004;186(11):3447–52. Agents Chemother. 2010;54(5):2239–43. 15. Johnston C, Martin B, Fichant G, Polard P, Claverys J-P. Bacterial 38. Raivio TL, Leblanc SKD, Price NL. The Escherichia coli Cpx envelope stress transformation: distribution, shared mechanisms and divergent control. Nat response regulates genes of diverse function that impact antibiotic Rev Microbiol. 2014;12(3):181–96. resistance and membrane integrity. J Bacteriol. 2013;195(12):2755–67. 16. Domingues S, Harms K, Fricke WF, Johnsen PJ, da Silva GJ, Nielsen KM. 39. Kampf G. Lack of antimicrobial efficacy of mecetronium etilsulfate in Natural Transformation Facilitates Transfer of Transposons, Integrons and propanol-based hand rubs for surgical hand disinfection. J Hosp Infect. Gene Cassettes between Bacterial Species. PLoS Pathog. 2012;8(8):e1002837. 2017;96(2):189–91. 17. Sun D, Zhang Y, Mei Y, Jiang H, Xie Z, Liu H, et al. Escherichia coli is naturally 40. Kampf G, Kramer A, Suchomel M. Lack of sustained efficacy for alcohol- transformable in a novel transformation system. FEMS Microbiol Lett. 2006; based surgical hand rubs containing ‘residual active ingredients’ according 265(2):249–55. to EN 12791. J Hosp Infect. 2017;95(2):163–8. 18. Baur B, Hanselmann K, Schlimme W, Jenni B. Genetic transformation in 41. van der Horst MA, Schuurmans JM, Smid MC, Koenders BB, ter Kuile BH. De freshwater: Escherichia coli is able to develop natural competence. Appl novo Acquisition of Resistance to three antibiotics by Escherichia coli. Environ Microbiol. 1996;62(10):3673–8. Microb Drug Resist. 2011;17(2):141–7. 19. Maeda S, Ito M, Ando T, Ishimoto Y, Fujisawa Y, Takahashi H, et al. 42. Händel N, Schuurmans JM, Feng Y, Brul S, ter Kuile BH. Interaction between Horizontal transfer of nonconjugative plasmids ina colony biofilm of mutations and regulation of gene expression during development of De novo Escherichia coli. FEMS Microbiol Lett. 2006;255(1):115–20. antibiotic resistance. Antimicrob Agents Chemother. 2014;58(8):4371–9. 20. Maeda S, Sawamura A, Matsuda A. Transformation of colonial Escherichia 43. Mulvey MR, Bryce E, Boyd DA, Ofner-Agostini M, Land AM, Simor AE, et al. coli on solid media. FEMS Microbiol Lett. 2004;236(1):61–4. Molecular characterization of cefoxitin-resistant Escherichia coli from 21. Etchuuya R, Ito M, Kitano S, Shigi F, Sobue R, Maeda S. Cell-to-cell Canadian hospitals. Antimicrob Agents Chemother. 2005;49(1):358–65. transformation in Escherichia coli: a novel type of natural transformation 44. Kohanski MA, DePristo MA, Collins JJ. Sublethal antibiotic treatment leads to involving cell-derived DNA and a putative promoting pheromone. PLoS multidrug resistance via radical-induced mutagenesis. Mol Cell. 2010;37(3): One. 2011;6(1):e16355. 311–20. 22. Johnson L, Horsman SR, Charron-Mazenod L, Turnbull AL, Mulcahy H, 45. Rutala WA, Weber DJ. Disinfection and sterilization in health care facilities: Surette MG, et al. Extracellular DNA-induced antimicrobial peptide resistance what clinicians need to know. Clin Infect Dis. 2004;39(5):702–9. in Salmonella enterica serovar Typhimurium. BMC Microbiol. 2013;13(1):1 Available from: BMC Microbiology. 23. Lewenza S. Extracellular DNA-induced antimicrobial peptide resistance Publisher’sNote mechanisms in Pseudomonas aeruginosa. Front Microbiol. 2013;4(FEB):1–6. Springer Nature remains neutral with regard to jurisdictional claims in 24. The European Committee on Antimicrobial Susceptibility Testing. published maps and institutional affiliations. Breakpoint tables for interpretation of MICs and zone diameters. Version 8.1, 2018. http://www.eucast.org. 25. Lee J, Choi J, Kim H. Determination of hand surface area by sex and body shape using alginate. J Physiol Anthropol. 2007;26(4):475–83. 26. Girard R, Aupee M, Erb M, Bettinger A, Jouve A. Hand rub dose needed for a single disinfection varies according to product: a bias in benchmarking using indirect hand hygiene indicator. J Epidemiol Glob Health. 2012;2(4):193–8. 27. de Velde F, de Winter BCM, Koch BCP, van Gelder T, Mouton JW. Non-linear absorption pharmacokinetics of amoxicillin: consequences for dosing regimens and clinical breakpoints. J Antimicrob Chemother. 2016;71(10): 2909–17. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Antimicrobial Resistance & Infection Control Springer Journals

Development of amoxicillin resistance in Escherichia coli after exposure to remnants of a non-related phagemid-containing E. coli: an exploratory study

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/development-of-amoxicillin-resistance-in-escherichia-coli-after-qABOKKLPXk

References (53)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2020
eISSN
2047-2994
DOI
10.1186/s13756-020-00708-7
Publisher site
See Article on Publisher Site

Abstract

Objective: To determine the effect of exposure to remnants of a phagemid-containing E. coli, killed by treatment with a propanol-based hand rub, on antimicrobial resistance in E. coli isolates. Methods: An in vitro model was developed in which a clinical E. coli isolate (EUR1) was exposed to remnants of an E. coli K-12 strain containing a phagemid (pBS-E12) strain treated with Sterillium®. A series of 200 experiments was performed using this in vitro model. As a control, a series of 400 experiments was performed where the EUR1 was exposed either to the remnants of an E. coli K-12 strain (not containing a phagemid) (E12) treated with Sterillium® (n = 200) or to dried Sterillium® only (n = 200). The number of experiments that showed growth of an amoxicillin- resistant EUR1 isolate was evaluated in all three groups. An additional 48 experiments were performed in which a different clinical E. coli isolate (EUR2) was exposed to remnants of the pBS-E12 treated with Sterillium®. Whole- genome sequencing and phenotypic testing for AmpC beta-lactamase production was performed to investigate the mechanism behind this resistance development. Results: In 22 (11.0%) of 200 experiments in which the EUR1 isolate was exposed to remnants of a pBS-E12 an amoxicillin-resistant mutant isolate was obtained, as opposed to only 2 (1.0%) of 200 experiments involving the exposure of the EUR1 to Sterillium® only (risk difference: 10.0%; 95% CI 5.4–14.6%)) and 1 (0.5%) of 200 experiments involving the exposure of the EUR1 isolate to the remnants of the phagemid-free E12 (risk difference: 10.5%; 95% CI 6.1–14.9%). In 1 (2.1%) of the 48 experiments in which the EUR2 isolate was exposed to remnants of a pBS-E12 an amoxicillin-resistant mutant isolate was obtained. The development of resistance in all experiments was due to mutations in the promoter/attenuator region of the chromosomal AmpC beta-lactamase (cAmpC) gene leading to cAmpC hyperproduction. (Continued on next page) * Correspondence: joep.stohr@gmail.com Department of Infection Control, Amphia Hospital, Breda, the Netherlands Laboratory for Medical Microbiology and Immunology, Elisabeth-TweeSteden Hospital, Tilburg, the Netherlands Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 2 of 10 (Continued from previous page) Conclusion: Exposure of an E. coli isolate to another phagemid-containing E. coli that was treated with propanol- based hand rub increased the development of amoxicillin resistance. Although phagemids are cloning vectors that are not present in clinical isolates, this finding may have implications for hand disinfection practices in healthcare facilities. Keywords: Sterillium, AmpC, Antimicrobial resistance, Disinfection Introduction AMR through mechanisms other than transformation The last decades we have seen a dramatic worldwide [22, 23]. In this exploratory study, we used an in vitro increase in antimicrobial resistance (AMR) among model to compare the rate at which amoxicillin-resistant Gram-negative bacteria. One of the most remarkable mutants developed for an E. coli isolate exposed to phenomena is the rapid increase of plasmid-mediated amoxicillin and remnants of a pBleuscript KS(−) phage- beta-lactam resistance in Escherichia coli [1–3]. Inter- mid containing E. coli K-12 strain treated with national infection control guidelines recommend several Sterillium®, amoxicillin and remnants of a phagemid-free measures to control the spread of AMR, among which K-12 strain treated with Sterillium® or amoxicillin and propanol-based disinfection of hands is vital [4, 5]. dried Sterillium® only. These disinfection methods rapidly and effectively de- crease the number of viable bacteria on hands and Method thereby limiting the spread of resistant bacteria and Isolate selection healthcare-related infections [6–9]. However, intact bac- An E. coli -K12 (JM83, ATCC® 35607™) harbouring a II- terial DNA could potentially persist after propanol-based pBleuscript KS (−) phagemid (ATCC® 87047™) (copy bacterial cell lysis and mechanical cleaning and serve as number: 300–500) containing an amoxicillin-resistance a source of resistance determinants for other bacteria gene (bla ) (pBS-E12), was chosen to be exposed TEM-116 that reach the disinfected area [10, 11]. Uptake of plas- to the propanol-based hand rub and two E. coli isolates mid and chromosomal bacterial DNA from the environ- (EUR1 and EUR2) obtained from routine clinical ment through natural transformation has already been cultures (EUR1: peritoneal fluid culture, EUR2: urine described as a method of resistance acquisition in culture) of epidemiologically unrelated patients were se- streptococci, Helicobacter spp. and various other bacteria lected to be exposed to remnants of the pBS-E12, [12–16]. Recent studies have also shown that E. coli is Sterillium® and amoxicillin (for mutant selection). Strain able to take up DNA in various environments [16–21]. characteristics of the used strains are depicted in Supple- However, it remains unclear to what extent the uptake mentary table S1. Features of the high copy number of resistance plasmids via natural transformation con- pBleuscript phagemid are depicted in Supplementary tributes to the development of AMR in E. coli. More- table S2. The selection of the EUR1 and EUR2 was based over, it is unknown whether exposure to environmental on antimicrobial susceptibility pattern, i.e. susceptible to DNA remainders (plasmids) could facilitate this form of amoxicillin and resistant to trimethoprim based on AMR acquisition. Therefore, we developed an in vitro EUCAST clinical breakpoints version 9.0 [24]. The pBS- model in which a clinical E. coli isolate (EUR1) was ex- E12 has a trimethoprim MIC below the clinical break- posed to a pBleuscript KS(−) phagemid (encoding an point. Antimicrobial susceptibility testing was performed amoxicillin resistance gene)-containing E. coli K-12 using the VITEK 2® system (bioMérieux, Marcy l’Etoile, strain which had been treated with alcohol (Sterillium®). France). The minimal inhibitory concentration (MIC) In this experiment, we did not observe plasmid transfer for amoxicillin was additionally tested with ETEST® (results on file), yet we did observe the development of (bioMérieux, Marcy l’Etoile, France). The MICs of the beta-lactam resistance in the EUR1 isolate. This observa- EUR1, EUR2, and pBS-E12 for the various tested anti- tion led to the hypothesis that exposure to remnants of microbial agents are shown in Table 1. phagemid-containing E. coli (killed by treatment with Sterillium®) could lead to the development of AMR in E. Exposure to propanol-based hand rub (Sterillium®) and coli isolates that came in contact with the remnants of resistance induction the killed E. coli K12 strain, through increased chromo- Sterile glass surfaces (25mmx18mm) were inoculated somal mutations. Although not previously reported for with 10 μL of a 0.5 McFarland suspension of the pBS- E. coli, studies have already shown that in Salmonella E12 isolate (1.5 × 10 colony forming units (CFU)) in spp. and Pseudomonas spp. the presence of external brain-heart infusion broth (BHI). The inoculation of the DNA could lead to an increase in the development of glasses was performed in a Kojair biosafety cabinet class Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 3 of 10 Table 1 Minimal inhibitory concentration for various antibiotics Isolate MIC (mg/L) D68C AmpC & ESBL Detection set conclusion E-test® Vitek® amox amcl pita cfrx cfxt cftz cftx AmpC hyperproduction¥ pBS-E12 > = 256 > = 32 > = 128 8 <=4 <=1 <=1 No EUR1 6 <=2 <=4 4 <=4 <=1 <=1 Yes EUR1M1 > = 256 > = 32 8 32 > = 64 4 <=1 Yes EUR1M2 > = 256 > = 32 16 32 > = 64 4 <=1 Yes EUR1M3 > = 256 > = 32 <=4 16 16 <=1 <=1 Yes EUR1M4 > = 256 > = 32 8 32 32 <=1 <=1 Yes EUR1M5 > = 256 > = 32 8 > = 64 > = 64 4 2 Yes EUR1M6 256 > = 32 <=4 16 > = 64 <=1 <=1 Yes EUR1M7 > = 256 > = 32 8 32 > = 64 4 <=1 Yes EUR1M8 > = 256 > = 32 <=4 16 16 <=1 <=1 Yes EUR1M9 256 > = 32 <=4 16 16 <=1 <=1 Yes EUR1M10 128 > = 32 <=4 16 32 <=1 <=1 Yes EUR1M11 128 > = 32 <=4 32 16 <=0.25 <=0.25 Yes EUR1M12 > = 256 > = 32 16 32 16 2 1 Yes EUR1M13 > = 256 > = 32 16 32 32 2 1 Yes EUR1M14 > = 256 > = 32 16 32 16 1 <=0.25 Yes EUR1M15 196 > = 32 <=4 16 16 0.5 <=0.25 Yes EUR1M16 196 > = 32 <=4 16 16 0.5 <=0.25 Yes EUR1M17 256 > = 32 <=4 32 32 2 0.5 Yes EUR1M18 > = 256 > = 32 16 32 16 2 1 Yes EUR1M19 > = 256 > = 32 8 32 16 2 <=0.25 Yes EUR1M20 64 16 <=4 4 <=4 0.5 <=0.25 Yes EUR1M21 > = 256 > = 32 16 32 16 4 1 Yes EUR1M22 256 > = 32 16 32 16 4 1 Yes EUR1S1 > = 256 > = 32 <=4 16 32 <=1 <=1 Yes EUR1S2 > = 256 > = 32 8 32 > = 64 2 <=1 Yes EUR1E1 > = 256 > = 32 <=4 16 32 <=1 <=1 Yes EUR2 1 <=2 <=4 2 <=4 <=1 <=1 No EUR2M1 > = 256 > = 32 8 8 <=4 <=1 <=1 Yes MICs were measured with ETEST® (bioMérieux, Marcy l’Etoile, France) (amoxicillin) or the VITEK 2® system (bioMérieux, Marcy l’Etoile, France). ¥Interpretation of D68C AmpC & ESBL Detection set according to manufacturers instruction. Zone diameters as measured in the D68C AmpC & ESBL Detection set are depicted in Supplementary table S3. amox: amoxicillin; amcl: amoxicillin-clavulanic acid; pita: piperacillin-tazobactam; cfrx: cefuroxime; cfxt: cefoxitin; cftz: ceftazidime; cftx: cefotaxime; mero: meropenem; imip: imipenem;M: amoxicillin-resistant mutant after exposure to remnants of the pBS-E12 and Sterillium®S: amoxicillin-resistant mutant after exposure to Sterillium® onlyE: amoxicillin-resistant mutant after exposure to E12 and Sterillium® II Silver Line. The inoculated glasses were left to dry at PBS-E12 and Sterillium®-containing glasses were left to ambient air temperature for 10 +/− 1min. Subsequently, dry at ambient room temperature for 10 +/− 1 min. the glasses were inoculated with 30 μL of Sterillium®, Subsequently, the glasses were inoculated with 10 μLof an which was spread across the entire glass surface using 0.5 McFarland suspension of the EUR1 isolate (1.5 × 10 sterile plastic sticks. The Sterillium® used contained per CFU) in BHI, which was spread across the entire glass 100 g of solution: Propan-2-ol 45.0 g, propan-1-ol 30.0 g surface using sterile plastic sticks. The re-inoculated and mecetronium etilsulfate 0.2 g. The ratio of the volume glasses were left to dry at ambient air temperature for 10 − 5 of inoculated Sterillium® to glass surface area (3 × 10 L: +/− 1 min. Subsequently, the glasses were placed in a con- 0.00045 m ) was chosen to reflect the ratio of the volume tainer with 4 mL of BHI, containing 1 mg/L amoxicillin, of Sterillium® used in hand disinfection to the average vortexed for 30 s at 2800 rotations per minute and incu- − 3 2 hand surface area (3 × 10 L: 0.045 m )[25, 26]. The bated at 35 to 37C°. The duration the isolates were Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 4 of 10 subjected to a subinhibitory amoxicillin concentration is Experiments with the EUR1 and EUR2 isolates were meant to represent the duration of subinhibitory amoxicil- performed on separate days. For each experiment colonies lin plasma-concentrations in patients starting amoxicillin were picked of the EUR1 and EUR2 isolate and the pBS- treatment [27]. After 1 h of incubation, an additional 1 mL E12 strain in stationary growth phase after incubation at of BHI broth containing 320 mg/L amoxicillin was added 35 to 37C° for 18 to 24 h on Muller Hinton agar. The to the glasses carrying containers, resulting in 5 mL of pBS-E12 strain was inoculated on Muller Hinton agar BHI broth containing 64.8 mg/L amoxicillin in the glasses plates containing an amoxicillin 10 μg disk. Only colonies carrying containers. The containers were then incubated of the pBS-E12 growing in the direct surrounding of the at 35 to 37C° for 72 h and visually inspected for bacterial amoxicillin disk were used in the experiments. The experi- growth at 8, 24, 48 and 72 h. At visible growth or after 72 mental setup is illustrated in Fig. 1. h of incubation in absence of visible growth, 1 μLofthe suspension was plated on a Muller Hinton agar containing Control experiments 8 mg/L trimethoprim and 64 mg/L amoxicillin (MH-TA). A series of 200 control experiments were performed in The inoculated MH-TA plates were incubated for 18 to which the glasses were only inoculated with Sterillium® 24 h at 35 to 37C°. Colonies growing on the MH-TA and the EUR1 and a series of 200 experiments were underwent species identification performed with VITEK performed in which the glasses were inoculated with the MS (bioMérieux, Marcy l’Etoile, France), using the VITEK E. coli K-12 strain (JM109, ATCC® 53323™) not contain- MS Knowledge Base v.2.0, and antimicrobial susceptibility ing the pBleuscript KS (−) high copy number phagemid testing performed with the VITEK 2® system (bioMérieux, (E12) (Supplementary Table S1), Sterillium® and the Marcy l’Etoile, France), using EUCAST susceptibility EUR1 isolate. Experiments were performed on separate breakpoints version 9.1 [24]. The MIC of all isolates to days. For each experiment colonies were picked of the amoxicillin was additionally tested using ETEST® (bioMér- EUR1 isolate and E12 strain after incubation at 35 to ieux, Marcy l’Etoile, France). Grown E. coli colonies with a 37C° for 18 to 24 h on Muller Hinton agar. The setup of MIC of > 8 mg/L for trimethoprim, measured in the the control experiments is illustrated in Fig. 2. VITEK 2® system (bioMérieux, Marcy l’Etoile, France), and with a MIC of > 64 mg/L for amoxicillin, measured Disinfection experiments with ETEST® (bioMérieux, Marcy l’Etoile, France), were To determine if the pBS-E12 was still culturable after considered mutants. A series of 200 experiments were exposure to Sterillium®, we performed a series of 100 performed using the EUR1 isolate. Additionally, a series of experiments in which the glasses were inoculated with 48 experiments were performed using the EUR2 isolate. the pBS-E12 followed by Sterillium® only. The inoculated Fig. 1 An illustration of the experiments in which the EUR1 or EUR2 isolates were exposed to amoxicillin and the pBS-E12 treated with Sterillium® Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 5 of 10 Fig. 2 An illustration of the experiments in which the EUR1 isolate was exposed to amoxicillin and the E12 treated with Sterillium®, or to amoxicillin and Sterillium® glasses were placed in a container with 5 mL of BHI, Acquired resistance genes were called when at least 60% vortexed for 30 s at 2800 rotations per minute and incu- of the length of the best matching gene in the ResFinder bated for 72 h at 35 to 37C°, visually inspecting growth database was covered with a sequence identity of at least at 24, 48 and 72 h. In case there was no visible growth 90%. If an unknown or known chromosomal mutation after 72 h of incubation, 1 μL of broth was plated on was detected in genomic regions implicated in beta- Muller Hinton agar. lactam resistance using PointFinder, the region was extracted from the assembled genome using Biopython Whole-genome shotgun sequencing (WGS) and de novo v.1.73. Subsequently, the extracted genomic regions were assembly aligned using Vector NTI Advance 11 software (Ther- The wildtype isolates (EUR1, EUR2), the pBS-E12 and moFisher Scientific, Waltham, USA) to the correspond- E12 strains, and a selection of the mutant isolates were ing region of either the EUR1 or EUR2 isolate and the sequenced on an Illumina MiSeq (Illumina, San Diego, corresponding region of the E. coli K-12 strain MG1655 United States) and assembled with CLC Genomics (GenBank database accession number NC000913.3), the Workbench v.11 or v.12 (Qiagen, Hilden, Germany). Se- pBS-E12 strain, and the E12 strain. Moreover, to verify lection criteria were: in the first thirteen experiments in that the observed amoxicillin resistance was not the which growth of a mutant isolate was detected, an iso- result of the acquisition of the amoxicillin resistance late was sent for sequencing. The following quality con- gene from the pBleuscript(−) phagemid, all assembled trol criteria were used: coverage ≥20; number of genomes were screened for the presence of this gene scaffolds ≤1000; N50 ≥ 15.000 bases and maximum scaf- using ABRicate v.0.8.13 with the same coverage and fold length ≥ 50.000 bases. identity thresholds as those used in ResFinder. Addition- ally, all mutated strains and the pBS-E12, EUR1, EUR2 Mechanism of resistance isolates were phenotypically screened for AmpC produc- The presence of genomic resistance determinants tion using the D68C AmpC & ESBL Detection set conferring amoxicillin and trimethoprim resistance in (Mastdiscs, Mastgroup Ltd., Bootle United Kingdom) the assembled genomes of the mutated E. coli was iden- and interpreted according to manufacturer’s instruction. tified with the online bioinformatic tools ResFinder v3.1 and PointFinder v3.1 (Center for Genomic Epidemi- Whole-genome multilocus sequence typing ology, Technical University of Denmark, Lingby, Whole-genome multilocus sequence typing (wgMLST) Denmark) [28, 29]. Genomes were screened for known (core and accessory genome) was performed of both and unknown chromosomal resistance mutations. wildtype isolates and all sequenced mutant isolates using Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 6 of 10 Ridom SeqSphere+, version 5.1.0. (Ridom, Münster, shows the various mutations at given positions of the Germany). Species-specific typing schemes were used as promoter/attenuator region of the cAmpC gene for described by Kluytmans-Van den Bergh et al. [30]. All- every sequenced isolate in this study. Known and un- to-all pairwise genetic difference was calculated between known mutations in the promoter/attenuator region of the isolates by counting the total number of allele differ- the chromosomal AmpC beta-lactamase (cAmpC) gene ences in the wgMLST typing scheme and by dividing the were detected in the genomes of all mutant E. coli total number of allele differences in the wgMLST typing isolates that were sequenced (EUR1M1-EUR1M10, scheme by the total number of shared alleles in the EUR1S1-EUR1S2, EUR2M1), but not in the EUR1 wgMLST typing scheme, ignoring pairwise missing isolate. Two unknown mutations were detected in the values. promoter/attenuator region of the EUR2 isolate (Table 2). Alignment against the corresponding region of an E. Statistics coli K-12 strain revealed that one of these mutations was Risk differences were estimated using a generalised lin- in the alternate − 10 box promoter box (Table 2). ear model with binomial distribution, an identity link However, these mutations did not increase the MIC for and robust error estimation (SPSS version 25). amoxicillin in the EUR2 isolate. In every mutant (EUR1M1-EUR1M10, EUR1S1-EUR1S2, EUR2M1) Accession numbers alignment of the cAmpC promoter/attenuator region All generated raw reads were submitted to the European revealed mutations, when compared to the same Nucleotide Archive (ENA) of the European Bioinformat- genomic region of the corresponding wildtype isolates, ics Institute (EBI) under the study accession number: in regions implicated in cAmpC hyperproduction as PRJEB34354. described by Tracz et al. (Table 2)[31]. Moreover, the different promoter/attenuator regions present in the mu- Results tant EUR1 isolates (Table 2) suggests de novo mutation Resistance induction experiment rather than the selection of a previously present mutated In 22 (11.0%) of 200 experiments in which the EUR1 subpopulation. No mutations in other regions implicated strain was exposed to the remnants of the pBS-E12 and in beta-lactam resistance were detected, nor were ac- Sterillium® an amoxicillin- and trimethoprim-resistant E. quired beta-lactam resistance genes detected in any of coli isolate (EUR1M1 - EUR1M22) was obtained, as the mutated (EUR1M1-EUR1M10, EUR1S1-EUR1S2, opposed to only 2 (1.0%) (EUR1S1, EUR1S2) of 200 EUR2M1) or wildtype isolates (EUR1, EUR2). The experiments involving the exposure of the EUR1 strain amoxicillin-resistance gene of the pBleuscript KS(−) pha- to Sterillium® only (risk difference 10.0%; 95% CI 5.4– gemid was not detected in any of the sequenced isolates 14.6%), and 1 (0.5%) (EUR1E1) of 200 experiments except the pBS-E12. In the pBS-E12 and E12 strain involving the exposure of the EUR1 strain to the rem- alignment of the cAmpC promotor/attenuator did not nants of the E12 and Sterillium® (risk difference 10.5%; reveal mutations, when compared to the same genomic 95% CI 6.1–14.9%). In the experiments performed using region of the E. coli K-12 strain MG1655. A dfrA1 gene the EUR2, 1 mutant isolate (EUR2M1) was grown in 48 conferring trimethoprim resistance was detected in the experiments. Amoxicillin-resistant strains did not only EUR1, EUR2 and all sequenced mutant isolates but not show increased MICs for amoxicillin but also for in the pBS-E12 and E12 strains. amoxicillin-clavulanic acid, piperacillin-tazobactam or Phenotypic tests for AmpC beta-lactamase production cephalosporins (Table 1). No increases in the MIC were were negative for the two wildtype isolates and the pBS- observed for carbapenems and non-beta-lactam antibi- E12 isolate. However, for all mutants, phenotypic testing otics tested. One out of 100 disinfection experiments showed AmpC beta-lactamase production (Table 1; showed growth of the pBS-E12, indicating that in only a Supplementary Table S3). minimal number of experiments the amoxicillin concen- tration during the experiments was influenced by viable Whole-genome multilocus sequence typing beta-lactamase-producing pBS-E12. The number of allele differences between the mutated EUR1M1-EUR1M10 isolates and wildtype EUR1 isolate Mechanisms of amoxicillin resistance ranged from 21 (0.59%) to 69 (1.96%) (median: 44.5 A selection of the isolates was sequenced, i.e. 10 mutant (1.27%)). In all but one (EUR1M9) of the mutant EUR1 E. coli isolates from the experiment in which EUR1 was isolates, the difference between the mutant and corre- exposed to the remnants of the pBS-E12, 2 mutants sponding control exceeded the threshold for genetic dis- from the control experiments, 1 mutant from the experi- tance between related and unrelated isolates as defined ments with the EUR2 isolate, the wildtype isolates EUR1 by Klutymans-van den Bergh et al. (Fig. 3)[30]. Both and EUR2, and the pBS-E12 and E12 strains. Table 2 mutated isolates from the control experiment (EUR1S1 Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 7 of 10 Table 2 Mutations in cAmpC promoter/attenuator region of the wildtype and mutant isolates E.coli K-12 Base in wildtype cAmpC promoter/attenuator region (wildtype) −32 − 23.1 − 21.1 −19 − 18 −16.1 − 11 + 58 + 81 T –– TG – CC G Isolate Position of mutation in cAmpC promoter/attenuator region Date experiments −32 −23.1 −21.1 −19 −18 −16.1 −11 + 58 + 81 performed pBS-E12 E12 EUR1 (wildtype) A EUR1M1 A A 10th May EUR1M2 G C A A 10th May EUR1M3 G C A A 11th May EUR1M4 G C A A 12th May EUR1M5 G C A A 12th May EUR1M6 T A 13th May EUR1M7 G C A A 25th May EUR1M8 G C A A 26th May EUR1M9 G C A A 1st June EUR1M10 T A 1st June EUR1S1 T A 23th May EUR1S2 A A 17th August EUR2 (wildtype) A T EUR2M1 A A T 12th May Position numbering of the cAmpC promoter/attenuator region as defined by Mulvey et al. [31]. In position number n.x: decimal number x refers to an insertion at position n. Positions −32 and − 11 are part of the wild-type promoter boxes. Position − 19 is part of the alternate promoter box. Positions − 21.1 and − 23.1 are part of the spacer region of both the wild-type and alternate promoter. Position − 16.1 is part of the spacer region of the wild-type promoter. ¥ strain used in experiments on 1st of June. ¤ all experiments were performed in 2016 Fig. 3 A Neighbour-joining tree representing the percentage of wgMLST allele differences between all sequenced isolates (Green: EUR1 wildtype/ mutant isolates; Red: EUR2 wildtype/mutant isolates) Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 8 of 10 and EUR1S2) also showed a high number of allele differ- assays with other antibiotics (e.g. streptomycin and nali- ences when compared to the EUR1 wildtype (EUR1S1: dixic acid) need to be performed, to further assess the n = 34 (0.96%), EUR1S2: n = 36 (1.02%)). The high num- mutagenic potential of the remnants of the pBS-E12 and ber of allele differences between the various mutated Sterillium® in the EUR1 isolate. However, the remnants EUR1 isolates suggests de novo mutation acquisition ra- of the pBS-E12 could also facilitate the development of ther than the selection of a previously present mutated AMR by enabling the EUR1 and EUR2 isolates to survive subpopulation (Fig. 3). In the EUR2M1, when compared longer under high amoxicillin concentrations, possibly to its wildtype, no allele differences were observed. The through providing nutrients or through efflux pump ac- number of allele differences between the EUR1 and tivation, thereby increasing the chance of the occurrence EUR2 isolate was 3184 (97.73%)(Fig. 3). of a mutation leading to cAmpC hyperproduction. A small degree of heterogeneity in the cAmpC promotor Discussion region of the pBS-E12 population that lead to cAmpC An in vitro model was developed to simulate propanol- hyperproduction cannot be completely ruled out. There- based hand disinfection as is common practice in health- fore, it still remains possible that the cAmpC promotor care facilities nowadays. Bacteria are effectively killed, region of this heterogeneous subpopulation was horizon- but their remnants remain present after disinfection. tally transferred to the wildtype EUR1 and EUR2 Subsequently susceptible bacteria are exposed to the isolates. remaining debris, including DNA. Experiments per- Several studies have shown that environmental formed using this model showed a significant increase in stressors other than antibiotics can influence the the development of amoxicillin resistance in E. coli iso- development of AMR [35–38]. However, the association lates after exposure to the remnants of the pBS-E12 between exposure to remnants of phagemid-carrying E. treated with a propanol-based hand rub as compared to coli isolates and increased development of AMR has not exposure to remnants of the E-12 treated with a been investigated to date. Interestingly, this increase in propanol-based hand rub and as compared to exposure the development of AMR was only related to exposure to only a propanol-based hand rub. AmpC beta- to remnants of phagemid-containing E. coli isolates. lactamase hyperproduction, due to mutations in the Other studies have already shown that extracellular AmpC promoter/attenuator region, was responsible for DNA can lead to the development of AMR in other the development of the observed AMR. The number of bacteria [22, 23]. However, this is the first study relat- resistant mutants that developed in the isolates exposed ing exposure to external DNA to the development of to the remnants of the pBS-E12 also exceeded the num- beta-lactam resistance through chromosomal muta- ber of pBS-E12 growing in the disinfection experiments, tions in E. coli. which indicates that amoxicillin degradation by the pBS- Despite the increased development of AMR was only E12 beta-lactamase was not responsible for the observed related to exposure to remnants of the pBS-E12, it re- increase in AMR. All mutated EUR1 isolates showed a mains unknown if and what specific compounds of high number of allele differences when compared to the Sterillium® contribute to the observed increased develop- wildtype isolate, suggesting genome-wide mutations. ment of AMR. Contrary to other alcohol-based hand Interestingly this genome-wide mutational pattern was rubs, Sterillium® contains mecetronium etilsulfate which not detected in the EUR2 isolate. Indicating that the potentially has a lasting antimicrobial effect [39, 40]. genome-wide mutational pattern might not only be Further studies are needed to evaluate if this increased (control) experiment but also isolate related. development of AMR also occurs with other alcohol- The exact underlying mechanism leading to the AMR based hand rubs not containing such compounds. conferring mutations in the promoter/attenuator region Furthermore, the current study only simulates hand of the cAmpC remains unknown. Since the amoxicillin disinfection procedures so it remains unknown to what resistance gene of the pBluescript phagemid was not de- extent this phenomenon also could apply to environ- tected in the EUR1 and EUR2 mutant isolates, it seems mental disinfection. more likely that the remnants of the pBS-E12 increased Although cAmpC hyperproduction has been impli- the mutation rate in the EUR1 and EUR2 isolate cated in beta-lactam resistance in E. coli isolates follow- facilitating AMR development through cAmpC hyper- ing exposure to stepwise increasing concentrations of production. Previous studies have implicated several amoxicillin over several days [41, 42]. cAmpC hyperpro- mechanisms that may increase the mutation rate and re- duction in this study occurred after only a short time of sult in AMR, such as the general stress response and an exposure to amoxicillin concentrations lower than the increased constitutive mutation frequency [32–35]. Fur- MIC of the exposed E. coli isolate. Even in our control ther studies are needed to evaluate the role of these experiments, in which the E. coli isolates were only ex- mechanisms in our experiments. Moreover, mutagenic posed to Sterillium®, cAmpC hyperproduction developed Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 9 of 10 after only short sub-inhibitory amoxicillin concentra- Supplementary information Supplementary information accompanies this paper at https://doi.org/10. tions. Moreover, in the isolates EUR1M2-EUR1M5 and 1186/s13756-020-00708-7. EUR1M7–9 three mutations were detected in the cAmpC promotor region. Each of these mutations Additional file 1: Table S1. Characteristics of used strains. Table S2. individually could increase cAmpC production in E. coli Features of pBleuscript KS(-) phagmid ATCC® 87047TM. Table S3. Inhibition zone diameters of D68C AmpC & ESBL detection set for the [31, 43]. Perhaps, consecutive mutations, not a single pBS-E12, EUR1, EUR2 and mutant isolates. mutational event during the short time of sub-inhibitory amoxicillin concentration, lead to a step-wise increase in Abbreviations amoxicillin MIC in these mutant isolates. In the study AMR: Antimicrobial resistance; pBS-E12: E. coli -K12 containing a pBleuscript by Kohanski et al. [44] exposure to sub-lethal levels of KS (−) phagemid encoding for an amoxicillin resistance gene (bla ); TEM-116 EUR1: Clinical E. coli isolate 1; EUR2: Clinical E. coli isolate 2; MIC: Minimal amoxicillin also resulted in MIC increases to antimicro- inhibitory concentration; BHI: Brain-heart infusion broth; MH-TA: Muller bials other than from the beta-lactam group. We did not Hinton agar containing 8 mg/L trimethoprim and 64 mg/L amoxicillin; observe MIC increases or known mutations in WGS: Whole genome shotgun sequencing; ENA: European Nucleotide Archive; EBI: European Bioinformatics Institute; cAmpC: chromosomal AmpC resistance-associated genes of antibiotics other than the beta-lactamase beta-lactams. Contrary to other studies investigating the develop- Acknowledgments We kindly thank the Avans university of applied sciences Breda for providing ment of AMR in E. coli [41, 44], we used two clinical E. the JM83, JM109 isolates and pBluescript KS(-) phagemid. coli isolates for amoxicillin resistance induction. More- over, our in vitro model simulates propanol-based hand Authors’ contributions JS performed the experiments, analysed and interpreted the data, and disinfection procedures which are very common in drafted the manuscript. MK analysed and interpreted the data, and revised clinical practice [4, 5, 45]. Also, the duration of exposure the manuscript. CV performed the experiments, and revised the manuscript. to sub-inhibitory amoxicillin concentrations closely re- JR analysed and interpreted the data, and revised the manuscript. JK analysed and interpreted the data, and revised the manuscript. All authors flects the duration of sub-inhibitory amoxicillin plasma read and approved the final manuscript. concentrations in patients at the start of amoxicillin treatment [27]. Funding This study has some limitations. The pBluescript-KS None (−) phagemid is a cloning vector not present in clinical Availability of data and materials isolates. Whether remnants of E. coli isolates containing The datasets supporting the conclusions of this article is included within the wildtype phagemids also increase the development of article and its additional file. AMR needs further investigation. Moreover, since the Ethics approval and consent to participate pBluescript KS(−) phagemid contains both a bacterio- Not applicable. phage origin of replication and a plasmid origin of Consent for publication replication future studies are required to assess whether Not applicable. the observed effect is bacteriophage or plasmid related. Also, only a limited number of isolates were used in this Competing interests The authors declare that they have no competing interests. study. The extent to which this AMR induction is possible in other E. coli isolates or other species remains Author details unknown. Moreover, it remains unknown whether and Department of Infection Control, Amphia Hospital, Breda, the Netherlands. Laboratory for Medical Microbiology and Immunology, to what extent this in vitro phenomenon plays a role in Elisabeth-TweeSteden Hospital, Tilburg, the Netherlands. Amphia Academy the development of AMR in vivo. Infectious Disease Foundation, Amphia Hospital, Breda, the Netherlands. To the best of our knowledge, this is the first study Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands. Department of showing development of amoxicillin resistance in an E. Medical Microbiology and Infection Prevention, University of Groningen, coli isolate after exposure to a phagemid-containing E. University Medical Center Groningen, Groningen, the Netherlands. coli treated with a propanol-based hand rub. Received: 29 October 2019 Accepted: 11 March 2020 Conclusion References This exploratory study showed the development of 1. Bonnet R. Growing group of extended spectrum: the CTX-M enzymes. Antimicrob Agent Chemother. 2004;48(1):1–14. amoxicillin resistance in an E. coli isolate after exposure 2. Nordmann P, Naas T, Poirel L. Global spread of Carbapenemase producing to an unrelated phagemid-containing E. coli treated with Enterobacteriaceae. Emerg Infect Dis. 2011;17(10):1791–8. a propanol-based hand rub. Although phagemids are 3. Mathers AJ, Peirano G, Pitout JDD. The role of epidemic resistance plasmids and international high- risk clones in the spread of multidrug-resistant cloning vectors that are not present in clinical isolates, Enterobacteriaceae. Clin Microbiol Rev. 2015;28(3):565–91. this finding may have implications for hand disinfection 4. Tacconelli E, Cataldo MA, Dancer SJ, De Angelis G, Falcone M, Frank U, et al. practices in healthcare facilities. ESCMID guidelines for the management of the infection control measures Stohr et al. Antimicrobial Resistance and Infection Control (2020) 9:48 Page 10 of 10 to reduce transmission of multidrug-resistant gram-negative bacteria in 28. Zankari E, Allesøe R, Joensen KG, Cavaco LM, Lund O, Aarestrup FM. hospitalized patients. Clin Microbiol Infect. 2014;20(s1):1–55. PointFinder: a novel web tool for WGS-based detection of antimicrobial 5. Boyce J, Pittet D. Centers for Disease Control and Prevention. Guideline for resistance associated with chromosomal point mutations in bacterial hand hygiene in health-care settings: recommendations of the healthcare pathogens. J Antimicrob Chemother. 2017;72(10):2764–8. infection control practices advisory committee and the HICPAC/SHEA/APIC/ 29. Zankari E, Hasman H, Cosentino S, Vestergaard M, Rasmussen S, Lund O, IDSA hand hygiene task force. MMWR Recomm Rep. 2002;51:1–45. et al. Identification of acquired antimicrobial resistance genes. J Antimicrob 6. Kampf G, Ostermeyer C, Heeg P, Paulson D. Evaluation of two methods of Chemother. 2012;67(11):2640–4. determining the efficacies of two alcohol-based hand rubs for surgical hand 30. Kluytmans-van den Bergh MFQ, Rossen JWA, Bruijning-Verhagen PCJ, Bonten MJM, Friedrich AW, Vandenbroucke-Grauls CMJE, et al. Whole- antisepsis. Appl Environ Microbiol. 2006;72(6):3856–61. genome multilocus sequence typing of extended-Spectrum-Beta-lactamase- 7. Kampf G, Kapella M. Suitability of Sterillium gel™ for surgical hand producing Enterobacteriaceae. J Clin Microbiol. 2016;54(12):2919–27. disinfection. J Hosp Infect. 2003;54(3):222–5. 31. Tracz DM, Boyd DA, Hizon R, Bryce E, McGeer A, Ofner-Agostini M, et al. 8. Kampf G, Ostermeyer C, Heeg P. Surgical hand disinfection with a propanol- ampC gene expression in promoter mutants of cefoxitin-resistant based hand rub: equivalence of shorter application times. J Hosp Infect. Escherichia coli clinical isolates. FEMS Microbiol Lett. 2007;270(2):265–71. 2005;59(4):304–10. 32. Firmino N, Goldsmith J-M, Morin A. Ampicillin resistance is increased in 9. Kramer A, Rudolph P, Kampf G, Pittet D. Limited efficacy of alcohol-based Escherichia coli K12 relA and spoT mutants but sub-inhibitory pretreatment hand gels. Lancet. 2002;359(9316):1489–90. does not induce adaptive resistance. J Exp Microbiol Immunol. 2013; 10. Mao D, Luo Y, Mathieu J, Wang Q, Feng L, Mu Q, et al. Persistence of 17(April):34–9. extracellular DNA in river sediment facilitates antibiotic resistance gene 33. Baharoglu Z, Mazel D. Vibrio cholerae triggers SOS and mutagenesis in propagation. Environ Sci Technol. 2014;48(1):71–8. response to a wide range of antibiotics: a route towards multiresistance. 11. Nielsen KM, Smalla K, Van Elsas JD. Natural transformation of Acinetobacter Antimicrob Agents Chemother. 2011;55(5):2438–41. sp. strain BD413 with cell lysates of Acinetobacter sp., Pseudomonas 34. Eliopoulos GM, Blazquez J. Hypermutation as a factor contributing to the fluorescens, and Burkholderia cepacia in soil microcosms. Appl Environ acquisition of antimicrobial resistance. Clin Infect Dis. 2003;37(9):1201–9. Microbiol. 2000;66(1):206–12. 35. Petrosino JF, Galhardo RS, Morales LD, Rosenberg SM. Stress-induced β- 12. Kwon DH, Dore MP, Kim JJ, Kato M, Lee M, Wu JY, et al. High-level β-lactam lactam antibiotic resistance mutation and sequences of stationary-phase resistance associated with acquired multidrug resistance in Helicobacter mutations in the Escherichia coli chromosome. J Bacteriol. 2009;191(19): pylori. Antimicrob Agents Chemother. 2003;47(7):2169–78. 5881–9. 13. Bae J, Oh E, Jeon B. Enhanced transmission of antibiotic resistance in 36. Fruci M, Poole K. Bacterial stress responses as determinants of antimicrobial Campylobacter jejuni biofilms by natural transformation. Antimicrob Agents resistance. In: Stress and environmental regulation of gene expression and Chemother. 2014;58(12):7573–5. adaptation in Bacteria. Hoboken: Wiley; 2016. p. 115–36. 14. Trzciński K, Thompson CM, Lipsitch M. Single-step capsular transformation 37. Nishino K, Yamasaki S, Hayashi-Nishino M, Yamaguchi A. Effect of NlpE and acquisition of penicillin resistance in Streptococcus pneumoniae.J overproduction on multidrug resistance in Escherichia coli. Antimicrob Bacteriol. 2004;186(11):3447–52. Agents Chemother. 2010;54(5):2239–43. 15. Johnston C, Martin B, Fichant G, Polard P, Claverys J-P. Bacterial 38. Raivio TL, Leblanc SKD, Price NL. The Escherichia coli Cpx envelope stress transformation: distribution, shared mechanisms and divergent control. Nat response regulates genes of diverse function that impact antibiotic Rev Microbiol. 2014;12(3):181–96. resistance and membrane integrity. J Bacteriol. 2013;195(12):2755–67. 16. Domingues S, Harms K, Fricke WF, Johnsen PJ, da Silva GJ, Nielsen KM. 39. Kampf G. Lack of antimicrobial efficacy of mecetronium etilsulfate in Natural Transformation Facilitates Transfer of Transposons, Integrons and propanol-based hand rubs for surgical hand disinfection. J Hosp Infect. Gene Cassettes between Bacterial Species. PLoS Pathog. 2012;8(8):e1002837. 2017;96(2):189–91. 17. Sun D, Zhang Y, Mei Y, Jiang H, Xie Z, Liu H, et al. Escherichia coli is naturally 40. Kampf G, Kramer A, Suchomel M. Lack of sustained efficacy for alcohol- transformable in a novel transformation system. FEMS Microbiol Lett. 2006; based surgical hand rubs containing ‘residual active ingredients’ according 265(2):249–55. to EN 12791. J Hosp Infect. 2017;95(2):163–8. 18. Baur B, Hanselmann K, Schlimme W, Jenni B. Genetic transformation in 41. van der Horst MA, Schuurmans JM, Smid MC, Koenders BB, ter Kuile BH. De freshwater: Escherichia coli is able to develop natural competence. Appl novo Acquisition of Resistance to three antibiotics by Escherichia coli. Environ Microbiol. 1996;62(10):3673–8. Microb Drug Resist. 2011;17(2):141–7. 19. Maeda S, Ito M, Ando T, Ishimoto Y, Fujisawa Y, Takahashi H, et al. 42. Händel N, Schuurmans JM, Feng Y, Brul S, ter Kuile BH. Interaction between Horizontal transfer of nonconjugative plasmids ina colony biofilm of mutations and regulation of gene expression during development of De novo Escherichia coli. FEMS Microbiol Lett. 2006;255(1):115–20. antibiotic resistance. Antimicrob Agents Chemother. 2014;58(8):4371–9. 20. Maeda S, Sawamura A, Matsuda A. Transformation of colonial Escherichia 43. Mulvey MR, Bryce E, Boyd DA, Ofner-Agostini M, Land AM, Simor AE, et al. coli on solid media. FEMS Microbiol Lett. 2004;236(1):61–4. Molecular characterization of cefoxitin-resistant Escherichia coli from 21. Etchuuya R, Ito M, Kitano S, Shigi F, Sobue R, Maeda S. Cell-to-cell Canadian hospitals. Antimicrob Agents Chemother. 2005;49(1):358–65. transformation in Escherichia coli: a novel type of natural transformation 44. Kohanski MA, DePristo MA, Collins JJ. Sublethal antibiotic treatment leads to involving cell-derived DNA and a putative promoting pheromone. PLoS multidrug resistance via radical-induced mutagenesis. Mol Cell. 2010;37(3): One. 2011;6(1):e16355. 311–20. 22. Johnson L, Horsman SR, Charron-Mazenod L, Turnbull AL, Mulcahy H, 45. Rutala WA, Weber DJ. Disinfection and sterilization in health care facilities: Surette MG, et al. Extracellular DNA-induced antimicrobial peptide resistance what clinicians need to know. Clin Infect Dis. 2004;39(5):702–9. in Salmonella enterica serovar Typhimurium. BMC Microbiol. 2013;13(1):1 Available from: BMC Microbiology. 23. Lewenza S. Extracellular DNA-induced antimicrobial peptide resistance Publisher’sNote mechanisms in Pseudomonas aeruginosa. Front Microbiol. 2013;4(FEB):1–6. Springer Nature remains neutral with regard to jurisdictional claims in 24. The European Committee on Antimicrobial Susceptibility Testing. published maps and institutional affiliations. Breakpoint tables for interpretation of MICs and zone diameters. Version 8.1, 2018. http://www.eucast.org. 25. Lee J, Choi J, Kim H. Determination of hand surface area by sex and body shape using alginate. J Physiol Anthropol. 2007;26(4):475–83. 26. Girard R, Aupee M, Erb M, Bettinger A, Jouve A. Hand rub dose needed for a single disinfection varies according to product: a bias in benchmarking using indirect hand hygiene indicator. J Epidemiol Glob Health. 2012;2(4):193–8. 27. de Velde F, de Winter BCM, Koch BCP, van Gelder T, Mouton JW. Non-linear absorption pharmacokinetics of amoxicillin: consequences for dosing regimens and clinical breakpoints. J Antimicrob Chemother. 2016;71(10): 2909–17.

Journal

Antimicrobial Resistance & Infection Control – Springer Journals

Published: Mar 16, 2020

There are no references for this article.