A 10-year microbiological study of Pseudomonas aeruginosa strains revealed the circulation of populations resistant to both carbapenems and quaternary ammonium compounds

P. aeruginosa bacterial strains

Reference strains

Four well-described and genome-available reference strains were used in the present study, ATCC27853 and ATCC15442, obtained from the American Type Culture Collection (ATCC), and PAO1 and PA14, from the collection of Institut Pasteur (Paris, France). Strain ATCC15442 is recommended for disinfectant susceptibility testing44, strain ATCC27853 is the Pseudomonas spp. reference for antibiotic susceptibility testing45, PAO1 is the reference genome for the P. aeruginosa species46 and strain PA14 is a highly virulent isolate representing the most P. aeruginosa common clonal group worldwide contrary to PAO147.

Hospital strains and study panel

Data on all P. aeruginosa strains from Caen University Hospital, a 1,410-bed teaching hospital in Normandy, France, were extracted from the laboratory management system (TD-Synergy, Montbonnot-Saint-Martin, France) for human strains (only the first strain isolated from each patient, without distinguishing between strains sought for carriage or diagnosis), and from annual reports of the hospital hygiene ward for hospital environment strain48. In total during the period of 1 January 2011 to 31 December 2020, 13,049 P. aeruginosa strains were isolated; 6,661 strains (51.0%) came from patients (P strains) and 6,388 (49.0%) from the hospital water environment (H strains) (Fig. 1 and SD 5). All were stored at room temperature in an agar medium, or at − 80 °C on brain–heart infusion medium (BHI; bioMérieux, Marcy-l’Étoile, France) with 15% glycerol (VWR, Radnor, Pennsylvania, USA). Among them, 180 strains were selected retrospectively based on the year and hospital ward isolation and the antibiotype to constitute the “study panel” (Fig. 1 and SD 2 Part 1). Of these, 124 were from patients (without distinguishing between strains sought for carriage or diagnosis) over the 2011 to 2020 period, and 56 were from the hospital environment over the 2016 to 2020 period (before 2016, H strains could not be included because they were not conserved). All were subcultured on tryptic soy broth (TS; Bio-Rad, Hercules, California, USA) and then stored on 15% glycerinated BHI at − 80 °C. The species identification was determined by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (Microflex; Bruker Daltonik, Bremen, Germany). Then, from the study panel, 77 strains were selected based on the resistance profile to represent the variability of antibiotic and disinfectant resistance patterns, the origin of the sample, and the sampling year to constitute the panel for genomic characterization (SD 2 Part 2). Thirty-seven of them were from patients, and the other 40 were from environmental samples. Among them, 13 strains (6 from patients and 7 from the hospital environment, with various resistomes) were selected to determine the overexpression of efflux pumps and cephalosporinase (AmpC), and it constituted the “representative short panel” (SD 2 Part 3).

In parallel to this study, 100 strains of P. aeruginosa (10 per year) were randomly selected from the 6661 P strains of Caen UH, to estimate the frequency of nonsusceptibility to DDAC (data not shown).

Strains and patient data (SD 2 part 1)

Available data were extracted for all 180 strains from the laboratory software (TD-Synergy TECHNIDATA, Montbonnot-Saint-Martin, France), including the year, origin and type of sampling, hospital ward, and carbapenemase-expressing status. In addition, patient information was retrospectively collected, including the patient gender, age, presence of comorbidities, immune status, whether it was a P. aeruginosa infection or colonization, whether or not death occurred (P. aeruginosa-related or not), and antipseudomonal antibiotic therapy administered during hospitalization (before and after P. aeruginosa identification). According to the European Committee on Antimicrobial Susceptibility Testing (EUCAST), the following drugs were considered active against P. aeruginosa for medical use: antipseudomonal penicillins (ticarcillin, ticarcillin-clavulanic acid, piperacillin, and piperacillin-tazobactam), carbapenems (doripenem, imipenem, imipenem-relebactam, meropenem, and meropenem-vaborbactam), monobactams (aztreonam), 3rd- and 4th-generation cephalosporins (cefepime, cefiderocol, ceftazidime, ceftazidime-avibactam, and ceftolozane-tazobactam), aminoglycosides (amikacin, gentamicin, netilmicin, and tobramycin), fluoroquinolones (levofloxacin, ciprofloxacin), fosfomycin and polymyxins (colistin and polymyxin B)49. The P strain P. aeruginosa acquisition type was also determined: community (i.e., the acquisition occurred before or during the first 48 h of hospitalization) or hospital acquired. The Charlson comorbidity index score was calculated using the available tool at www.rdplf.org/calculated50.

Antimicrobial susceptibility testing (AST)

Antibiotic susceptibility testing

Among the 6661 P strains, ASTs were available for 4375 P. aeruginosa strains against 16 antipseudomonal drugs categorized into 8 classes: penicillin (ticarcillin, ticarcillin-clavulanic acid, piperacillin, piperacillin-tazobactam), carbapenems (imipenem, meropenem), monobactams (aztreonam), cephalosporins (cefepime, ceftazidime, ceftolozane-tazobactam), phosphonic acid (fosfomycin), polymyxins (colistin), aminoglycosides (amikacin, tobramycin, gentamicin) and fluoroquinolones (ciprofloxacin, levofloxacin). Disk diffusion method was performed for all except for colistin and fosfomycin, tested by broth microdilution by VITEK® 2 (bioMérieux, Marcy-l’Étoile, France) or by the Sensititre™ Vizion™ System (Thermo Fisher Scientific, Waltham, Massachusetts, USA). Interpretation of values according to the guidelines of the Comité de l’antibiogramme de la Société Française de Microbiologie (CASFM) each year, from 2011 to 2020. The CASFM guidelines differ from the EUCAST guidelines until 2013 (inoculum, incubation time, and temperature), then the hospital CASFM will follow the EUCAST guidelines. The data were gathered to estimate the percentage of resistance (R) for each antibiotic and classification in multidrug-resistant (MDR), or extensively drug-resistant (XDR) profiles.

For the 180-strain study panel, antibiotic susceptibility testing was performed again for 16 antipseudomonal antibiotics (Bio-Rad, Hercules, California, USA) in association or not with a β-lactamase inhibitor that are distributed in 7 distinct classes: the same as described above, except for polymyxins which were excluded. MIC for colistin was not performed. Antibiotic susceptibility testing was performed on Mueller–Hinton agar (Becton Dickinson, Franklin Lakes, New Jersey, USA) following the EUCAST guidelines for Pseudomonas spp. disk diffusion method45. Following the 2021 edition of the EUCAST breakpoint tables for interpretation, breakpoints were set49. Strains showing resistance (R) in at least three antibiotic classes were considered MDR. Those with remaining susceptibility (“Susceptible, standard dose”, S and “Susceptible, increased exposure”, I) in one or two antibiotic categories were considered XDR. Pan resistance could be defined as resistance to all eight antibiotic classes51.

Quaternary ammonium susceptibility testing

Susceptibility to the detergent/disinfectant DDAC was evaluated for the 180-strain study panel by determining the minimum inhibitory concentration (MIC) using the reference liquid microdilution method52 in cation-adjusted Mueller–Hinton broth with N-Tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid (Thermo Fisher Scientific, Waltham, Massachusetts, USA). The DDAC was tested in the concentration range from 0.5 to 1024 mg/L (successive two-fold dilutions), three times for each strain. By relying on the concentration of DDAC in the disinfectant solution according to the manufacturer’s instructions, the threshold for decreased susceptibility (DS) to DDAC in this study was set at MIC > 62.9 mg/L, equivalent to 0.00629%.

Furthermore, for the 100 randomly selected P. aeruginosa strains between 2011 and 2020 the MIC of DDAC was determined under the same technical conditions37.

Whole-genome sequencing and bioinformatic analysis

For the reference strains, ATCC27853 and PA14 genomes were obtained from the European Nucleotide Archive (ENA) database with accession numbers CP015117 and ASWV01000001, respectively, while the ATCC15442 and PAO1 sequences were obtained from GenBank with accession numbers GCF_000504485.1 and GCA_000006765.1, respectively.

The 77 strains were sequenced by the “Plateforme de Microbiologie Mutualisée P2M” (Institut Pasteur, Paris, France). A MagNA Pure 96 instrument (Roche Diagnostics, Meylan, France) was used for DNA extraction, a Nextera XT library kit (Illumina Inc., San Diego, USA) was used for NGS library construction, and a NextSeq500 (Illumina Inc., San Diego, USA) was used for sequencing. FastQC V0.11.053 and MultiQC V1.954 software were used for quality control checks on raw sequence data. The paired-end reads were preprocessed (filtered and trimmed) using fqCleanER (gitlab.pasteur.fr/GIPhy/fqCleanER), with a minimal read size of 70 bp and a Phred quality score of 28. De novo assembly was performed using SPAdes 3.1255 with a 50X minimum average sequencing depth, and Quast software V5.056 was used for final assembly quality checks.

Species identifications based on the sequences have been validated with Ribosomal Multilocus Sequence Typing (rMLST), available at pubmlst.org/57. In silico, P. aeruginosa strain serotyping was performed using PAst1.0 software, available at cge.cbs.dtu.dk/services/PAst-1.0/58,59. Then, multilocus sequence typing (MLST) was performed on the sequence variation of 7 housekeeping genes using MLST2.0 software (version 2.0.4, 2019/05/08; database version 2021/10/04). This software uses the MLST allele sequence and profile obtained from PubMLST.org58,60. Finally, the core genome MLST was determined based on the P. aeruginosa MLST scheme targeting 3,867 loci (available on cgmlst.org at www.cgmlst.org/ncs/schema/16115339/locus/, database version 2021/05/26)32. Chewbacca software version 2.8.561 was used for the cgMLST scheme conversion and allele calling. Finally, a neighbor-joining tree based on cgMLST was visualized with iTOL v6.3.362. The nucleotide sequences were submitted to AMRFinderPlus analysis (version 3.10.30, database version 2022–05-26.1)63 with a minimal identity of 80% and minimal coverage of 50% to identify antimicrobial resistance genes and known resistance-associated point mutations. All 77 assembled genomes were deposited in the BioProject PRJNA884650.

Analysis of the overexpression of efflux pumps and cephalosporinase activity

Imipenem, meropenem, and DDAC MICs values were determined in the presence/absence of the efflux pump inhibitor PaβN (at 25, 50 and 100 mg/L) and the cephalosporinase (AmpC) inhibitor cloxacillin (at 250 mg/L) separately. Susceptibility testing was performed by broth microdilution method52, as previously described in the study, for the 13 strains of the representative short panel and 2 reference strains (PAO1 and PA14). In addition, the expression of the mexA, mexB, oprM, mexE, mexF, and oprN genes was determined by quantitative real-time PCR. Total RNA was extracted from bacterial cells to the late exponential phase using the Direct-Zol RNA miniprep kit (Zymo Research, Irvine, California, USA), and the residual chromosomal DNA was removed using a Turbo DNA-free kit (Life Technologies, Carlsbad, California, USA). A NanoDrop One spectrophotometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA) was used for quantification, and cDNA was synthesized from total RNA using a QuantiTect reverse transcription kit (Qiagen, Hilden, Germany) following the manufacturer’s guidelines. The transcript levels were determined by the ΔΔCT method (CT is the threshold cycle) using the expression of the housekeeping gyrB gene as a reference transcript. The level of transcription measured for the PAO1 reference strain was used to determine the expression ratios for the strains of interest. The sequences of the primers associated with each gene are listed in SD 6.

Statistical analysis

All statistical tests were performed with GraphPad Prism version 9.0.0 for macOS (GraphPad Software, San Diego, California USA). Independence tests of populations using Fisher’s exact test were performed successively to determine if the DS to DDAC was linked to nonsusceptibility of strains (MDR or XDR phenotypes) and/or if the DS to DDAC was linked to carbapenem resistance. A chi-square test was performed to compare the distribution of DS to DDAC according to the strain origin. Carbapenem resistance average annual expression frequency was obtained by averaging annual expression frequency obtained for each year between 2011 and 2020, and 95% confidence intervals were calculated by the Wilson/Brown method.

Ethics statement

This study has been conducted in compliance with the Helsinki Declaration (ethical principles for medical research involving human subjects) and in accordance with the guidelines of research board of our teaching hospital, Caen, France. Ethic committee of CHU Caen Normandie reviewed and approved the study number ID 3784. It was a non-interventional study: specimens used in this study were part of the routine patient management without any additional sampling. Furthermore informed consent was obtained from all subjects and/or their legal guardian(s).

Informed consent

Informed consent was obtained from all subjects and/or their legal guardian(s).

Read more here: Source link