Genetic basis for in vivo piperacillin-tazobactam resistance

doi:10.21203/rs.3.rs-2752509/v1

Download PDF

Article

Genetic basis for in vivo piperacillin-tazobactam resistance

https://doi.org/10.21203/rs.3.rs-2752509/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Piperacillin-tazobactam resistance (P/T-R) is increasingly reported among Escherichia coli isolates. Although in vitro experiments have suggested that blaTEM gene plays a key role in the P/T-R acquisition, no clinical in vivo study has yet confirmed the role of blaTEM or other genes. Therefore, we aimed to identify the mechanisms underlying P/T-R by following up patients with E. coli intraabdominal infections (IAI) who experienced P/T treatment failure. We found a higher copy number of blaTEM gene in P/T-R isolates, generated by three different genetic events: (1)IS26-mediated duplication of the blaTEM gene, (2) generation of a small multicopy plasmid (ColE-like) carrying blaTEM, and (3) adaptive evolution via reduction of plasmid size, leading to a higher plasmid copy number. Moreover, two P/T-R strains showed reduced expression of OmpC. Thus, P/T treatment may lead to the development of resistance in patients with IAI by E. coli, through three blaTEM-dependent mechanisms and downregulation of OmpC.

Biological sciences/Microbiology/Antimicrobials/Antimicrobial resistance

Health sciences/Medical research/Genetics research

The ß-lactamases are the main cause of ß-lactam resistance in Gram-negative bacteria¹, whose mechanisms of action are based on the hydrolysis of the ß-lactam ring. Combinations of ß-lactams/ß-lactamase inhibitors (BL/BLI) were introduced to address ß-lactamase activity^1,2. Some of the most commonly used combinations in clinical practice include ampicillin/sulbactam (A/S), amoxicillin/clavulanic acid (A/C), and piperacillin/tazobactam (P/T), as well as the recently introduced ceftazidime/avibactam (CZA) and ceftolozane/tazobactam (C/T).

P/T is a broad-spectrum BL/BLI commonly used for the empiric treatment of severe infections, including intra-abdominal infection (IAI) and febrile neutropenia¹. The use of P/T has increased in recent years, as well as the number of P/T-resistant Escherichia coli isolates, limiting its therapeutic use^3–6. TEM enzymes have been described to lead the P/T resistance acquisition in E. coli, either through an overproduction of TEM-1 or an evolution of inhibitor-resistant TEM variants (IRT), or an evolution of TEM variants with increased hydrolytic capabilities^3,4,6−9. Other ß-lactamases such as OXA-1 or AmpC, as well as loss of porins are also frequent causes of P/T resistance^10–14.

In E. coli, BL/BLI resistance is a gradual and unidirectional process that extends first from A/S, then A/C and finally P/T. We have previously described a new concept on antimicrobial resistance, called extended-spectrum BL/BLI resistance (ESRI), which is defined as the acquisition of high-level BL/BLI resistance (resistance to A/S, A/C and P/T) from a low- (resistance to A/S with susceptibility to A/C and P/T) or moderate-level (resistance to A/S and A/C and susceptibility to P/T) resistance phenotype, in which the presence of TEM ß-lactamase plays an important role^7,15. Exposure to subinhibitory concentrations of P/T has been shown to trigger ESRI isolates in vitro by two different mechanisms: the first based on duplication of bla_TEM through the action of Insertion Sequence 26 (IS26); and the other based on transposition of the bla_TEM gene into a multicopy plasmid (such as ColE1-like plasmids)⁷.

Here, we have identified four pairs of E. coli isolates, recovered from four different patients over two different times, prior and after P/T-treatment, which showed the acquisition of P/T resistance after its administration. Our aim was to determine whether P/T treatment in patients could select resistance to P/T in E. coli, and to characterize the bacterial resistance mechanisms leading to this BL/BLI resistance development.

Study design, patients and ethics

This is a prospective observational cohort study from February 1st, 2020 to December 31th, 2021. Adult patients (≥ 18 years) with IAI and bloodstream infection (BSI) caused by E. coli at the Virgen del Rocío University Hospital (Seville, Spain) were included in the study. The role of E. coli isolated from intraabdominal and blood samples, as colonization or infection, was evaluated according to previously defined criteria¹⁶. The study was approved by the Ethics Committee of Virgen Macarena and Virgen del Rocío University Hospitals (approval number C.I. 0088-N-20). Written informed consent was signed by all patients before the inclusion in the study.

Bacterial isolates

Four isolates couples (AEC-11/AEC-51, AEC-15/AEC-24, AEC-99/AEC-106 and PT3/PT4) from four patients were isolated in the pre- and post-therapy with P/T dosed 4 g/0.5 g intravenously every 8 hours in extended infusion. E. coli isolates were identified by MALDI-TOF (Bruker Daltonik, GmbH, Germany). In addition, E. coli MG1655 and its OmpC mutant (ΔompC) strains were used¹⁷. Finally, E. coli ATCC 25922 strain was used as a standard control in all the experiments.

Resistance profile

The A/S, A/C and P/T antimicrobial susceptibility profiles of the four isolates couples were initially tested by broth microdilution using the MicroScan WalkAway NMDR-1 panels. P/T MICs for these isolates and for the MG1655 and ΔompC strains were subsequently confirmed by broth microdilution reference method and interpreted according to European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines¹⁸.

Antimicrobial selection pressure

The diluted bacterial inoculum (10⁵ CFU/mL) of the different isolates and strains were incubated with sub-inhibitory concentrations of P/T (Merck, Spain) corresponding to dilutions 1-fold below the MICs at 37 ºC for 24 h. Positive bacterial growth was re-adjusted to 10⁵ cfu/mL for P/T MIC determinations and for further incubation with a 2-fold increased concentration of P/T concentration of 256/4 mg/L or a P/T concentration that did not allow bacterial growth.

Removal of antimicrobial selection pressure

In order to evaluate the stability of the resistance, daily passages of the P/T-resistant isolates (AEC-51, AEC-24, AEC-106 and PT4) on sheep blood agar plates (ThermoFisher, Spain) were made for 15 days and the P/T MICs were determined on days 5, 10 and 15 for each isolate.

Whole-genome sequencing and bioinformatic analysis

Whole-genome sequencing (WGS) was performed with two different techniques: short-read Illumina sequencing and long-read Oxford Nanopore technology to obtain a concensus sequence. Total DNA was extracted from colonies using (i) the Invitrogen PureLink™ Genomic DNA Mini Kit following the manufacturer’s instructions (ThemoFisher, France) and (ii) the Magcore® extractor system H16 with a Magcore Genomic DNA Large Volume Whole Blood Kit (RBC Bioscience Corp., Taiwan). A DNA library was prepared either (i) using a NEB Ultra II FS DNA library prep kit for Illumina (NEB, France) and then run on the MiSeq sequencer (Illumina, France) to generate paired-end 150 pb reads and (ii) by using a R9.4 flow cell and the 1D native barcoding genomic DNA kit (SQK-LSK109) on a MinION sequencer (Oxford Nanopore Technologies, UK). De novo assembly of Illumina reads was performed using CLC Genomics Workbench 21.0 (QIAGEN, France). Basecalling were performed in a GPU cluster with four Tesla v100 using guppy 5.0.16 in high accuracy mode to obtain the Nanopore reads. Then and in order to perform a genomic analysis, nanopore long reads were assembled using flye 2.9. The acquired antimicrobial resistance genes were identified using ResFinder v.4.1 (Center for Genomic Epidemiology, CGE, https://cge.food.dtu.dk/services/ResFinder/). The genome was annotated using the Rapid Annotations using Subsystems Technology (RAST) server tool⁷. The ISfinder database (https://isfinder.biotoul.fr/) was used to identify insertion sequence (IS) elements. The plasmid content was analyzed using PlasmidFinder (https://cge.cbs.dtu.dk/services/PlasmidFinder/) and manually searched by homology. Additionally, MLST tool was used to determine the clonality of the different pair of isolates from each patient⁷. The bla_TEM gene copy number was estimated using the coverage ratio of the bla_TEM− 1b gene (or its genetic context) and the mean of three chromosomally encoded genes (replication initiator protein dnaA, the ß-subunit of RNA polymerase rpoB and the malate dehydrogenase mdh). To obtain the consensus sequence, reads were filtered with fastp and assembled with flye (version 2.9). Later, the assembly was corrected with medaka (version 1.7.2). Finally, the consensus sequence produced by medaka was corrected by pilon (version 1.20.1).

BLAST analysis and sequence comparison using BRIG

Consensus sequences of the plasmid harboring bla_TEM gene were blasted by couple (P/T-susceptible vs P/T-resistant) using the BLAST Ring Image Generator (BRIG) software (version 0.95-dist). The blastn algorithm was used to generate a circular BLAST between the different plasmids.

Quantification of ß-lactamase activity

The ß-lactamase activity was determined using the ß-lactamase Activity Kit (BioVision Inc, Milpitas, CA, USA)⁷. Briefly, the bacterial culture pellets were weighed and resuspended in ß-lactamase assay buffer. Samples were sonicated and centrifuged. Five microliters of each sample were incubated with nitrocefin, which is a chromogenic cephalosporin substrate frequently used to find beta-lactamase enzymes. The optical density (OD) at 490 nm was immediately measured in kinetic mode by the CLARIOstar reader (BMG LABTECH, France).

RNA extraction and RT-qPCR

Bacterial RNA was purified using the RNeasy Mini Kit (Qiagen, Netherlands) from overnight cultures. Contaminating DNA was removed from the RNA samples and the reverse transcription step were carried out using the QuantiTect Reverse Transcription kit (Qiagen, Spain), according to the manufacturer’s instructions. The primer3 v.0.4.0 software (http://bioinfo.ut.ee/primer3-0.4.0/) was used to design the primers. We used specific primers to amplify the bla_TEM and ompC genes and the housekeeping rrsG gene^7,19 (Table S1). The quantitative real-time PCR assay was carried out with SYBR Premix Ex Taq (Takara, Japan) in a MxPro 3005p system (Stratagene, CA, USA). Gene expression was relatively quantified using the Comparative CT (∆∆Ct) Method (Applied Biosystems Guide, USA).

Growth rate measurements

Growth rate was measured as the maximum increase in OD over time during exponential growth. OD measurements were performed in 96-well plates containing 200 µL Luria Bertani (LB) or Mueller Hinton (MH) (Merck, Spain) broth/well by the Clariostart plate reader. OD at 600 nm was measured with 20 min intervals for 12 h and incubated at 37ºC with shaking for 5 sec every measure point. The doubling time (DT) was determined by the exponential curve fitting function to generate an equation in the form y = Ae^Bx, where A and B are numbers and x is the time between doubling of y (A_600nm). The equation simplifies to DT = ln2/B. Colonies were pre-cultured during 2–3 h with shaking at 37ºC to reach the exponential phase before the inoculation of the final plate.

In vitro competition experiments

The in vitro competition index (CI) was determined from growth cultures, consisting of AEC-11 or AEC-51, and a mixed inoculum of equivalent CFUs from AEC-11 and AEC-51 as previously described²⁰. Briefly, bacteria with a titre of 5.5 log CFU/mL were grown in LB broth. At 0, 4, 8 and 24 h, 10 µL aliquots were taken and AEC-11 and AEC-51 CFUs were determined by plating serial log dilutions on LB agar (Sigma, Spain) or LB agar plus P/T (64/4 µg/mL). The AEC-51 in competition with AEC-11 was determined after plating the mixed inoculum on LB agar plus P/T to select only the AEC-51 colonies. The AEC-11 in competition with AEC-51 was determined after plating the mixed inoculum on LB agar. The obtained log CFU/mL was subtracted from the log CFU/mL of AEC-51 in competition. The CI was defined as the number of AEC-51 CFUs recovered per number of AEC-11 CFUs recovered at 4, 8 and 24 h, divided by the number of AEC-51 CFUs inoculated per number of AEC-11 CFUs inoculated at 4, 8 and 24 h, respectively. The same protocol was used with the rest of the paired isolates (AEC-15 vs AEC-24, AEC-99 vs AEC-106 and PT3 vs PT4).

Analysis of outer membrane proteins by SDS-PAGE

Bacterial cells were grown in LB to the logarithmic phase and lysed by sonication. Outer membrane proteins (OMPs) were extracted with sodium lauroylsarcosinate (Merck, Spain) and recovered by ultracentrifugation as described previously²⁰. The OMP profiles were determined by SDS-PAGE using 10% SDS gels and 6 µg of OMP, followed by Simply Blue SafeStain gel (ThermoFisher, Spain). For peptide mass fingerprinting, a Simply Blue SafeStain band representing an unidentified protein with molecular weight 40 kDa was excised from the SDS-PAGE gel and sent for MALDI-TOF-TOF (MS-MS/MS) analysis. Data obtained from peptide MS fingerprinting were matched against the National Centre for Biotechnology Information (NCBI) database (http://www.ncbi.nlm.nih.gov) using the Mascot program.

Data availability

Whole-genome sequences were deposited in GenBank under the BioProject accession number PRJNA949846.

Demographics and clinical data analysis.

Demographics and clinical features of the four patients included in this study are summarized in Fig. 1 and Table 1.

Table 1

Demographics, clinical, epidemiological and microbiological data of the patients and the *Escherichia coli* isolates.
Patient	Sex	Age	Source	Dose regimen	Treatment duration (days)	Isolates	MLST	Resistance phenotype	Resistance genotype
#1	Male	63	Pancreatic abscess	P/T dosed 4 g/0.5 g IV (extended-infusion 4 h)^a	12	AEC-11	162	AMP, TIC, PIP, CIP, LEV, TMP-SXT	bla_TEM−1, tet(B), sul2, dfrA17, aadA5, aph(6)-ld, aph(3”)-lb
#1	Male	63	Pancreatic abscess	P/T dosed 4 g/0.5 g IV (extended-infusion 4 h)^a	12	AEC-51	162	AMP, AMC, TIC, TCC, PIP, P/T, CIP, LEV, TMP-SXT	bla_TEM−1, tet(B), sul2, dfrA17, aadA5, aph(6)-ld, aph(3”)-lb
#2	Female	66	Post-surgical intraabdominal abscess	P/T dosed 4 g/0.5 g IV (extended-infusion 4 h)^a	18	AEC-15	48	AMP, TIC, PIP	bla_TEM−1, tet(A)
#2	Female	66	Post-surgical intraabdominal abscess	P/T dosed 4 g/0.5 g IV (extended-infusion 4 h)^a	18	AEC-24	48	AMP, AMC, TIC, TCC, PIP, P/T	bla_TEM−1, tet(A)
#3	Male	72	Post-surgical intraabdominal abscess	P/T dosed 4 g/0.5 g IV (extended-infusion 4 h)^a	14	AEC-99	73	AMP, AMC, TIC, PIP, TMP-SXT	bla_TEM−1, tet(B), sul2, dfrA8, aph(6)-ld, aph(3”)-lb
#3	Male	72	Post-surgical intraabdominal abscess	P/T dosed 4 g/0.5 g IV (extended-infusion 4 h)^a	14	AEC-106	73	AMP, AMC, TIC, TCC, PIP, P/T, TMP-SXT	bla_TEM−1, sul2, dfrA8, aph(6)-ld, aph(3”)-lb
#4	Male	63	Blood (Perianal abscess)	P/T dosed 4 g/0.5 g IV (extended-infusion ever4 h)^a	8	PT3	88	AMP, AMC, TIC, PIP, TMP-SXT	bla_TEM−1, tet(A), sul2, dfrA5, aph(6)-ld, aph(3”)-lb
#4	Male	63	Blood (Perianal abscess)	P/T dosed 4 g/0.5 g IV (extended-infusion ever4 h)^a	8	PT4	88	AMP, AMC, TIC, TCC, PIP, P/T, TMP-SXT	bla_TEM−1, tet(A), sul2, dfrA5, aph(6)-ld, aph(3”)-lb
AMP, ampicillin; AMC, amoxicillin/clavulanic acid; TIC, ticarcillin; TCC, ticarcillin/clavulanic acid; PIP, piperacillin; P/T, piperacillin/tazobactam; CIP, ciprofloxacin; LEV, levofloxacin; TMP-SXT, trimethoprim/sulfamethoxazole. ^a extended infusion every 8 hours
The sources are the same for the first and second isolates of each patient.

Patient #1 was an 63-year-old male who had been diagnosed with a neuroendocrine tumor of the pancreas and was undergoing surgery. He was carrying an abdominal drainage catheter due to an infection of the surgical site. He was admitted due to an increase in flow through the drainage, and a diagnosis of intra-abdominal abscess was made. Samples were taken and multisusceptible E. coli (AEC-11) was isolated. Treatment with P/T dosed 4 g/0.5 g every 8 hours iv. in extended infusion was started. After 12 days, treatment was simplified to A/C, and he was discharged from the hospital. Treatment with A/C was continued at home. However, 73 days later, he was readmitted for removal of the drainage catheter and sampling, where P/T-resistant E. coli (AEC-51) was isolated.

Patient #2 was an 66-year-old woman diagnosed with colorectal adenocarcinoma who underwent emergency surgery due to intestinal obstruction. On the 8th day after admission, she was diagnosed with a postsurgical intra-abdominal abscess which was drained, and samples were taken. Multisusceptible E. coli (AEC-15) was isolated, and she started treatment with P/T on the 3rd day since admission, dosed 4 g/0.5 g every 8 hours iv. in extended infusion for 19 days. Seventeen days after the start of treatment, a new sample was taken from the abscess, where P/T-resistant E. coli (AEC-24) was isolated.

Patient #3 was a 72-year-old man diagnosed with ampulloma who was admitted with symptoms suggestive of cholangitis. On the 13th day of admission, a post-surgical intra-abdominal abscess was drained, and E. coli P/T-susceptible (AEC-99) was recovered. He began treatment with P/T dosed 4 g/0.5 g every 8 hours iv. in extended infusion on the 1st day of admission, continuing for 14 days. Ten days after the start of treatment, a new sample was taken due to fever onset, and E. coli P/T-resistant (AEC-106) was recovered.

Patient #4 was an 63-year-old man diagnosed with myelodysplastic syndrome with excess blasts. He was admitted for febrile neutropenia with an abdomino-pelvic abscess focus. Blood cultures were taken on admission, and multisusceptible E. coli (PT3) was isolated. Treatment was started with P/T, dosed 4 g/0.5 g every 8 hours iv. in extended infusion, which was maintained for 8 days. On the 7th day, a new blood culture was taken, and P/T-resistant E. coli (PT4) was isolated.

MLST analysis

The MLST analysis performed with the assembled Whole Genome sequences revealed the same ST type for each pair of isolates from the same patient, being ST162, ST48, ST73, and ST88 recovered from the patient #1, #2, #3, and #4, respectively (Table 1).

Analysis of P/T resistance in E. coli isolates

For each patient, a pair of E. coli isolates with the same ST were recovered from each single patient (AEC-11 and AEC-51 from patient #1; AEC-15 and AEC-24 from patient #2; AEC-99 and AEC-106 from patient #3, PT3 and PT4 from patient #4) from two separate infection episodes, suggesting within-patient evolution to P/T resistance. All these isolates were carbapenem susceptible and none produced an extended-spectrum ß-lactamase. MIC for P/T of the second isolate was 32-fold higher than that of the first isolate (before intravenous therapy with P/T) in all four cases (Fig. 2A).

Since E. coli harboring TEM acquired resistance to P/T after incrementally increasing the selection pressure (i.e. concentration) of P/T^5,7, we hypothesized that P/T-susceptible isolates (AEC-11, AEC-15, AEC-99 and PT3) became resistant to P/T after exposure to P/T. In the in vitro selection pressure, the new P/T MIC of these isolates increased 32-fold with respect to the inicial MIC; these isolates thus became resistant to P/T (Table S2).

Of note, the P/T resistance observed for AEC-51, AEC-24, AEC-106 and PT4 remained stable when the antibiotic pressure was removed from the medium. Indeed, after growth in medium free of P/T for 15 days, none had returned to their initial phenotype, except for one isolate (PT4) that decreased its P/T MIC from 256 to 32 mg/L (which is still considered as resistant by EUCAST) (Fig. 2B). This latter is in line with our previous in vitro study, in which different E. coli isolates acquiring P/T resistance by selective pressure kept their resistance to P/T in medium free of P/T for 15 days⁷.

Genomic analysis of the region surrounding the bla _TEM-1 gene in E. coli

To better understand and describe the underlying mechanisms responsible for the significative increase in P/T MICs, the four pairs of isolates were sequenced by Oxford Nanopore technologies (long read) and corrected using Illumina short reads, allowing complete assembly of the plasmids and of the genomes. Analysis of the different pairs of isolates revealed the following results:

(i) in the AEC-11 isolate the mean chromosomal coverage (MCC) was 12.06x, and the bla_TEM-1 coverage was 9.07x, representing a single copy per cell. Meanwhile, in AEC-51 isolate, the MCC was 12.21x, and the bla_TEM-1 coverage was 233.75x representing 25 copies per cell. Analysis of the coverage of the region surrounding the bla_TEM-1 gene in AEC-51 isolate revealed five copies of IS26 bracketing the bla_TEM-1 gene, generated through a series of ̴ 15 kb DNA fragment translocation events by IS26. The repeated fragment is composed by resistance genes (bla_TEM-1, tet(B), dfrA, and ant(3’)-la), and MGEs (IS4, IS26, and IntI) (Fig. 3A). Resistance to antibiotics by increase of resistance gene copy numbers has already been described among Enterobacterales in vitro and in vivo^7,21;

(ii) in the AEC-15 isolate the MCC was 51.44x, and the bla_TEM-1 gene coverage was 4.63x, representing a single copy per cell. Meanwhile, in AEC-24 isolate, the MCC was 48.14x, and the bla_TEM-1 coverage was 108.67x representing 25 copies per cell. Similarly, in the AEC-99 isolate the MCC was 5.39x, and the bla_TEM-1 coverage was 2.51x, representing a single copy per cell. Meanwhile, in AEC-106 isolate, the MCC was 2.56x, and the bla_TEM-1 coverage was 45.17x representing 35 copies per cell. Genetic context analysis revealed that plasmids of AEC-24 and AEC-106 isolates underwent similar modifications with slight differences. While a single copy of bla_TEM-1 gene and no single nucleotide polymorphisms (SNPs) were identified in the two couples of E. coli, a consistent deletion of approximately 60 kb and 30 kb was observed in the pAEC-24 and the pAEC-106 plasmids respectively (Fig. 3C and 3D). In pAEC-24, the consistenly deleted region encoded SOS system inhibitor genes, plasmid partition system, methyltransferases genes, and the main conjugation machinery. Interestingly, deletions was surrounded by IS26, suggesting that intramolecular transposition and recombination lead by the insertion sequence allowed for the observed loop-out dynamics (Fig. 3C).

On the other hand, pAEC-106 lost regions encoding transporter system, iron related genes, and two resistance determinant. In this case, deletions was surrounded by IS1, IS4, and IS26, which could be the responsible of the reorganization of the plasmid backbone (Fig. 4D). ISs and mainly IS26 is often found on bacterial plasmid associated with antibiotic resistance genes, and believed to be involved in the plasmid reorganization through intramolecular replicative transposition events^22–25. Hence, we showed that the reported adaptative evolution mechanism (deletions of costly plasmid genes) in pAEC-24 and pAEC-106 is also used as a resistance mechanism by increasing plasmid copy number in order to overcome antibiotics, even when this change entails an increased fitness cost.

(iii) in the PT3 isolate the MCC was 26.24x, and the bla_TEM−1 coverage was 21.66x, representing a single copy per cell. In PT4 isolate, the MCC was 28.75x, and the bla_TEM−1 coverage was 517.89x representing 18 copies per cell. Genetic context analysis of PT4 isolate reveals that two copies of bla_TEM−1 gene localized in the pPT4-IncFIB plasmid (150 kb), which showed 100% identity with the pPT3-IncFIB plasmid (150 kb), and in the pSmPT4 plasmid (8 kb), which was generated through the action of the IS26 and harbour resistance genes such as bla_TEM−1, dfrA, aph(6’)-lc and aph(3’)-l, and two origin of replication (repA and repC) (Fig. 3B). The reduced size of the pSmPT4 plasmid is commonly associated to a increased copy number, leading to a major expression and production of TEM-1 with the consequent elevated MIC of P/T. Similar adaptative evolution in conditions selecting for plasmid carriage was observed in E. coli after an gradual pressure by P/T in vitro⁷. The latter study described the transposition event of a transposon (Tn2) from a plasmid IncI1 to a ColE1-like plasmid, in contradiction to our study, where the pSmPT4 was the product of the circularization of a plasmidic fragment induced by the action of the IS26⁷.

Impact of the P/T resistance on bacterial fitness

Previous studies on plasmid evolution have described that the deletions of large DNA fragments increased the plasmid stability in the host through reducing the fitness cost of the original plasmid^26,27. In order to test the effect on the fitness of the P/T-susceptible and -resitant bacteria, growth rates were determined. Surprisingly, growth rate assays revealed increased doubling times (DT) for the resistant isolates AEC-24 (DT = 149 min) and AEC-106 (DT = 171 min) in comparison with their paired isolates AEC-15 (DT = 95.5 min) and AEC-99 (DT = 104.5 min), respectively (Fig. 4A). This increase in doubling time is probably related to their reduced fitness. For the two other pair of isolates (AEC-11/AEC-51 and PT3/PT4), the growth rates were identical, suggesting no fitness cost in those type of resistance mechanisms (Fig. 4A).

To confirm this observation, in vitro competition experiments between the four couples of isolates have been performed. The competition index between AEC-51 and AEC-11, and PT4 and PT3 did not show significant decrease of AEC-51 and PT4 growth from 1 to 0.55 and 0.65 at 4 h of culture, respectively. Meanwhile, the competition index between AEC-24 and AEC-15, and AEC-105 and AEC-99 showed a significant decrease of AEC-24 and AEC-105 growth from 1 to 0.29 and 0.007, respectively, at 4 h of culture. Finally, at 24 h of culture all the resistant isolates have reduced their competiton index siginificantly, from 1 to 0.006–0.099 (Fig. 4B).

These data seem to indicate that the different genetic events involved in P/T resistance have different cost for the bacteria.

TEM-1 is involved in the acquired resistant to P/T

Since isolates of E. coli carrying TEM enzymes could acquire P/T resistance by expressing higher β-lactamase activity⁷, we suggest that the increase in P/T MIC values of the study isolates would be due to higher β-lactamase activities after hyperproduction of TEM enzymes. Indeed, Fig. 5A shows that AEC-51, AEC-24, AEC-106 and PT4 isolates resistant to P/T showed higher β-lactamase activity than their paired susceptible E. coli isolates. Noteworthy, in the absence of selective pressure, these isolates maintained higher β-lactamase activity, which may explain the long-term stability of P/T resistance in the absence of this antibiotic observed in this and in another study⁷. In addition, we analysed the transcription of bla_TEM in these isolates, and found that the PT-R isolates contained higher levels of transcripts of bla_TEM (Fig. 5B). Of note, other ß-lactamases genes were not detected in these isolates.

In order to verify whether differential transcription of bla_TEM−1 gene was due to different promoter sequences, upstream regions were analyzed. In all cases, the promoter upstream the bla_TEM−1 gene was Pa/Pb, which has been shown to be a strong promoter responsible of TEM-1 hyperproduction, leading to resistance to ß-lactamase inhibitors such as clavulanate and sulbactam²⁸ and also tazobactam^6,8. No mutation was evidenced in the Pa/Pb region of any P/T resistant isolates as compared to the P/T susceptible ones.

Role of OmpC in the P/T resistance

Additional comparative analysis of the WGS revealed mutations in OmpC and OmpR, a regulator of OmpC expression²⁹, in the P/T-resistant AEC-51 and AEC-24 isolates as compared with P/T-susceptible AEC-11 and AEC-15, respectively. The mutation in OmpC consist in an chromosomal-insertion of adenosine duplication at position 51 of the gene (c.51 dupA) generating an frame shift change, and the mutation of OmpR consist in single nucleotide variation (c.136C > T) in the position 136 of the ompC gene leading to an amino acid substitution (Arg46Cys). Transcription analysis of ompC gene revealed that ompC gene was less transcribed in P/T-resistant AEC-51 and AEC-24 isolates (Fig. 6A). Furthermore, SDS-PAGE analysis of OMPs showed a reduction of the expression of a protein identified as OmpC by MALDI-TOF-TOF (MS-MS/MS) (Fig. 6B). Notably, after exposure of MG1655 wt and ΔompC strains with 2-fold increased P/T concentration, the MIC of ΔompC strain was increased progressively from 8 to 32-folds as compared with the MG1655 wt strain (Fig. 6C).

OmpC constitute the main OMP in E. coli that is necessary for drug transport across cellular membranes³⁰. A reduction in the OmpC expression has been associated with the ß-lactam reduced susceptibility/resistance^31–33. The importance of the OmpC in the P/T resistance is less clear³⁴. Additional studies are need to decipher how P/T regulates the expression of OmpC.

Long read sequencing has shown to be crucial in the elucidation of the antimicrobial resistance mechanisms, mainly those involving numerous IS elements and repeat sequences. Here, we have found three ways in the acquisition of resistance to P/T: (i) IS26-based duplication, as previously shown⁶. We found modifications of a pre-existing plasmid by transposition, rather than acquisition of foreign DNA; (ii) generation of a smaller plasmid (ColE-like) harboring the bla_TEM−1 gene as previously reported in in vitro studies⁷, and; (iii) adaptative evolution through the reduction of the plasmid size and fitness cost, leading to a higher plasmid copy number. To the best our knowledge, this is the first study describing IS-mediated deletions of important traits of the plasmid as a driving force of bacterial resistance by increasing the plasmid copy number.

This kind of genetic events (recombination, transposition or conjugation) occurs frequently, but the genetic events leading to antibiotic resistance, such as, higher production levels of TEM, seems to be responsible in clinical failures of P/T treatments. The sub-inhibitory conditions could be created in certain circunstances when the antibiotic is not able to reach enough concentrations or time above the MICs in the site of infection (i.e. intraabdominal abscess), which could create an environment appropriate for the induction of the antimicrobial resistance or the selection of resistant subpopulations. In this sense, Murao et al. have determined the pharmacokinetic parameters of P/T administred in bolus in patients and showed significant differences in the P/T concentrations in plasma and peritoneal fluid, being lower in the latter case³⁵. Therefore, this underpines the need for determining the pharmacokinetics and pharmacodynamics parameters of the P/T extendly perfused in patients, not only in the blood, but also in other fluids such as peritoneal, where the antibiotic might diffuse with difficulty. Thus in light of our results, one should consider different dosing regimens for different kind of infections based on-site specific PK/PD target attaitments. Hence, it is crucial to promote the development and implementation of new rapid molecular tests, which allow us to detect these types of “resistance development markers” in a short time^15,36, as well as, knowing precisely the PK/PD to P/T in other sites of infections, helping the clinicians to optimize the antimicrobial treatment in order to reduce the probabilities of therapeutical failures in patients with severe E. coli infections.

In this study, we demonstrated that treatment of E. coli IAI with P/T can lead to the development of resistance to this antibiotic. Three different mechanisms of resistance to P/T were found, in all of which the bla_TEM−1 gene played a central role and OmpC expression was reduced. Therefore, this underscores the great need to target specific dosing regimens based not only on the resistance phenotype of the bacteria or species but also on markers that may increase the likelihood of resistance development in certain environments, such as the expression of the bla_TEM gene.

FUNDING

This work was funded by the Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía, Industria y Competitividad (grant PI19/01009), by Plan Nacional de I+D+i 2013 to 2016, and by the Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía, Industria y Competitividad, Spanish Network for Research in Infectious Diseases (grant RD16/0016/0009), cofinanced by the European Development Regional Fund “A way to make Europe”/”Investing in your future”. A.R.V, R.A-M, J.A.L. and J.M.C. were supported (grant CB21/13/00006) by CIBERINFEC - Consorcio Centro de Investigación Biomédica en Red, Instituto de Salud Carlos III, Ministerio de Ciencia e Innovación and Unión Europea – NextGenerationEU.

L.G.B. was partly financed by a grant from the Spanish Society of Infectious Diseases and Clinical Microbiology (SEIMC) and is supported by the Subprograma Rio Hortega, Instituto Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Ciencia, Innovación y Universidades, Spain (CM22/00196). J.M.O.R. is supported by the Subprograma Sara Borrell, Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Ciencia, Innovación y Universidades, Spain (CD21/00098). A.R.V. is supported by the Subprograma Juan Rodés, Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Ciencia, Innovación y Universidades, Spain (JR20/00023), and Y.S. has received a Miguel Servet Tipo II contract from the Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía y Competitividad, Spain (grant CPII20/00018).

Acknowledgements

We like to thank Andrea Vila Domínguez for technical help and Sanna Koskiniemi for the gift of the MG1655 and its OmpC mutant strains.

Bush K, Bradford PA. ß-lactams and ß-lactamase inhibitors: an overview. Cold Spring Harb. Perspect. Med. 2016; 6: a025247.
Drawz SM, Bonomo RA. Three decades of ß-lactamase inhibitors. Clin. Microbiol. Rev. 2010; 23: 160-201.
Shubert C, Slaughter J, Creely D, van Belkum A, Gayral JP, Dunne WM, Zambardi G, Shortridge D. Population analysis of Escherichia coli isolates with discordant resistance levels by piperacillin-tazobactam broth microdilution and agar dilution testing. Antimicrob. Agents Chemother. 2014; 58(3):1779-81.
Schechter LM, Creely DP, Garner CD, Shortridge D, Nguyen H, Chen L, Hanson BM, Sodergren E, Weinstock GM, Dunne WM Jr., van Belkum A, Leopold SR. Extensive gene amplification as a mechanism for piperacillin-tazobactam resistace in Escherichia coli. mBio 2018; 9: e00583-18.
Hansen KH, Andreasen MR, Pedersen MS, Westh H, Jelsbak L, Schønning K. Resistance to piperacillin/tazobactam in Escherichia coli resulting from extensive IS-26-associated gene amplification of blaTEM-1. J. Antimicrob. Chemother. 2019; 74: 3179-83.
Hubbard ATM, Mason J, Roberts P, Parry CM, Corless C, van Aartsen J, Howard A, Bulgasim I, Fraser AJ, Adams ER, Roberts AP, Edwards T. Piperacillin/tazobactam resistance in a clinical isolate of Escherichia coli due to IS26-mediated amplification of bla_TEM-1b. Nat. Comm. 2020;11:4915.
Rodriguez-Villodres A, Gil-Marqués ML, Alvarez-Marín R, Bonnin RA, Pachón-Ibáñez ME, Aguilar-Guisado M, Naas T, Aznar J, Pachón J, Lepe JA, Smani Y. Extended-spectrum resistant to ß-lactams/ß-lactamase inhibitors (ESRI) evolved from low-level resistant Escherichia coli. J. Antimicrob. Chemother. 2020;75: 77-85.
Zhou K, Tao Y, Han L, Ni Y, Sun J. Piperacillin-tazobactam (TZP) Resistance in Escherichia coli due to Hyperproduction of TEM-1 ß-lactamase mediated by the Promoter Pa/Pb. Front. Microbiol. 2019;10: 833.
Chiaibi, EB, Sirot D, Paul G, Labia R. Inhibitor-resistant TEM ß-lactamases: phenotypic, genetic and biochemical characteristics. J. Antimicrob. Chemother. 1999; 43: 447-58.
Henderson A, Paterson DL, Chatfield MD, Tambyah PA, Lye DC, De PP, Lin RTP, Chew KL, Yin M, Lee TH, Yilmaz M, Cakmak R, Alenazi TH, Arabi YM, Falcone M, Bassetti M, Righi E, Rogers BA, Kanj SS, Bhally H, Iredell J, Mendelson M, Boyles TH, Looke DFM, Runnegar NJ, Miyakis S, Walls G, Khamis MAI, Zikri A, Crowe A, Ingram PR, Daneman N, Griffin P, Athan E, Roberts L, Beatson SA, Peleg AY, Cottrell K, Bauer MJ, Tan E, Chaw K, Nimmo GR, Harris-Brown T, Harris PNA; MERINO Trial Investigators and the Australasian Society for Infectious Disease Clinical Research Network (ASID-CRN). Association between minimum inhibitory concentration, beta-lactamase genes and mortality for patients treated with piperacillin/tazobactam or meropenem from the MERINO Study. Clin. Infect. Dis. 2021; 73(11): e3842-e3850.
Livermore DM, Day M, Cleary P, Hopkins KL, toleman MA, Wareham DW, Wiuff C, Doumith M, Woodford N. OXA-1 ß-lactamase and non-susceptibilty to peinicillin/ß-lactamase inhibitor combinations among ESBL-producing Escherichia coli. J. Antimicrob. Chemother. 2019; 74(2): 326-33.
Bethel CR, Hujer AM, Helfand MS, Bonomo RA. Exploring the effectiveness of tazobactam against ceftazidime resistant Escherichia coli: insights from the comparison between susceptibility testing and beta-lactamase inhibition. FEMS Microbiol. Lett. 2004; 234(1): 99-103.
Vimont S, Aubert D, Mazoit JX, Poirel L, Nordmann P. Broad-spectrum ß-lactams for treating experimental peritonitis in mice due to Escherichia coli producing plasmid-encoded cephalosporinases. J. Antimicrob. Chemother. 2007;60(5): 1045-50.
Pagès JM, James CE, Winterhalter M. The porin and the permeating antibiotic: a selective diffusion barrier in Gram-negative bacteria. Nat. Rev. Microbiol. 2009; 6: 893-903.
Rodríguez-Villodres Á, Gutiérrez Linares A, Gálvez-Benitez L, Pachón J, Lepe JA, Smani Y. Semirapid detection of piperacillin/tazobactam resistance and extended-spectrum resistance to β-Lactams/β-Lactamase inhibitors in clinical isolates of Escherichia coli. Microbiol. Spectr. 2021; 9(2): e0080121.
Solomkin JS, Mazuski JE, Baron EJ, Sawyer RG, Nathens AB, DiPiro JT, Buchman T, Patchen Dellinger E, Jernigan J, Gorbach S, Chow AW, Bartlett J. Guidelines for the selection of anti-infective agents for complicated intra-abdominal infections. Clin. Infect. Dis. 2003; 37: 997-1005.
Virtanen P, Wäneskog M, Koskiniemi N. Class II contact-dependent growth inhibition (CDI) systems allow for broad-range cross-species toxin delivery within the Enterobacteriaceae family. Mol. Microbiol . 2019; 111(4): 1109-25.
The European Committee on Antimicrobial Susceptibility Testing. Breakpoint Tables for Interpretation of MICs and Zone Diameters. EUCAST, 2021. Version 11.0, 2021. http://www.eucast.org.
Vinson HM, Gautam A, Olet S, Gibbs PS, Barigye R. Molecular analysis of porin gene transcription in heterogenotypic multidrug-resistant Escherichia coli isolates from scouring calves. J. Antimicrob. Chemother. 2010; 65(9): 1926-35.
Smani Y, Dominguez-Herrera J, Pachón J. Association of the outer membrane protein Omp33 with fitness and virulence of Acinetobacter baumannii. J Infect Dis 2013; 208: 1561-70.
Simner PJ, Mostafa HH, Bergman Y, Ante M, Tekle T, Adebayo A, Beisken S, Dzintars K, Tamma PD. Progressive development of cefiderocol resistance in Escherichia coli during therapy is associated with an increase in blandm-5 copy number and gene expression. Clin. Infect. Dis. 2022;7 5(1): 47-54.
Cullik A, Pfeifer Y, Prager R, von Baum H, Witte W. A novel IS26 structure surrounds blaCTX-M genes in different plasmids from German clinical Escherichia coli isolates. J. Med. Microbiol. 2010; 59(Pt 5): 580-7.
Curiao T, Cantón R, Garcillán-Barcia MP, de la Cruz F, Baquero F, Coque TM. Association of composite IS26-sul3 elements with highly transmissible IncI1 plasmids in extended-spectrum-beta-lactamase-producing Escherichia coli clones from humans. Antimicrob. Agents Chemother. 2011; 55(5): 2451-7.
Partridge SR, Zong Z, Iredell JR. Recombination in IS26 and Tn2 in the evolution of multiresistance regions carrying blaCTX-M-15 on conjugative IncF plasmids from Escherichia coli. Antimicrob. Agents Chemother. 2011; 55(11): 4971-8.
He S, Hickman AB, Varani AM, Siguier P, Chandler M, Dekker JP, Dyda F. Insertion sequence IS26 reorganizes plasmids in clinically isolated multidrug-resistant bacteria by replicative transposition. mBio 2015; 6(3): e00762.
Fricke WF, Welch TJ, McDermott PF, Mammel MK, LeClerc JE, White DG, Cebula TA, Ravel J. Comparative genomics of the IncA/C multidrug resistance plasmid family. J. Bacteriol. 2009; 191(15): 4750-7.
Porse A, Schønning K, Munck C, Sommer MO. Survival and Evolution of a Large Multidrug Resistance Plasmid in New Clinical Bacterial Hosts. Mol. Biol. Evol. 2016; 33(11): 2860-73.
Lartigue MF, Leflon-Guibout V, Poirel L, Nordmann P, Nicolas-Chanoine MH. Promoters P3, Pa/Pb, P4 and P5 Upstream from bla_TEM genes and their relationship to ß-lactam resistance. Antimicrob. Agents Chemother. 2002; 46: 4035-7.
Foo YH, Gao Y, Zhang H, Kenney LJ. Cytoplasmic sensing by the inner membrane histidine kinase EnvZ. Prog Biophys Mol Biol 2015; 118: 119-29.
Bafna JA, Sans-Serramitjana E, Acosta-Gutiérrez S, Bodrenko IV, Hörömpöli D, Berscheid A, Brötz-Oesterhelt H, Winterhalter M, Ceccarelli M. Kanamycin uptake into Escherichia coli is facilitated by OmpF and OmpC porin channels located in the outer membrane. ACS Infect. Dis. 2020; 6(7): 1855-65.
Yi-Fang L, Jing-Jou Y, Huan-Yao L, Ching-Hao T, Ming-Cheng W, Chin-Chung T, Jiunn-Jong W. Loss of outer membrane protein c in Escherichia coli contributes to both antibiotic resistance and escaping antibody-dependent bactericidal activity. Infect. Immun. 2012; 80(5): 1815-22.
Chetri S, Singha M, Bhowmik D, Nath K, Dhar Chanda D, Chakravarty A, Bhattacharjee A. Transcriptional response of OmpC and OmpF in Escherichia coli against differential gradient of carbapenem stress. BMC Res. Notes 2019; 12: 128.
Larkin PMK, Realegeno S, Ward KW, Garner OB, Yang S. an unusual carbapenem resistant Escherichia coli carrying plasmid-mediated ampc and mutated ompC in a patient with recurrent urinary tract infections. IDCases 2020; 20: e00781.
Mediavilla JR, Schneider Z, Nwaigwe C, Chavda K, Chen L, Satlin M, Jenkins SG, Nicolau DP, Kreiswirth BN. Molecular characterization of cephalosporin/carbapenem/monobactam susceptible but piperacillin-tazobactam (TZP) resistant E. coli, abstr 1181 Abstr ID Week 2015, San Diego, CA.
Murao N, Ohge H, Ikawa K, Watadani Y, Uegami S, Shigemoto N, Shimada N, Yano R, Kajihara T, Uemura K, Murakami Y, Morikawa N, Sueda T. Pharmacokinetics of piperacillintazobactam in plasma, peritoneal fluid and peritoneum of surgery patients, and dosing considerations based on site-specific pharmacodynamic target attainment. Int. J. Antimicrob. Agents 2017; 50: 393-8.
Rodríguez Villodres A, Gálvez Benítez L, Arroyo MJ, Méndez G, Mancera L, Vila Domínguez A, Lepe Jímenez JA, Smani Y. Ultrasensitive and rapid identification of ESRI developer- and piperacillin/tazobactam-resistant Escherichia coli by the MALDIpiptaz test. Emerg. Microbes Infect. 2022; 11(1): 2034-44.

There is NO Competing Interest.

SUPPLEMENTARYTABLES.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Genetic basis for in vivo piperacillin-tazobactam resistance

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Study design, patients and ethics

Bacterial isolates

Resistance profile

Antimicrobial selection pressure

Removal of antimicrobial selection pressure

Whole-genome sequencing and bioinformatic analysis

BLAST analysis and sequence comparison using BRIG

Quantification of ß-lactamase activity

RNA extraction and RT-qPCR

Growth rate measurements

Analysis of outer membrane proteins by SDS-PAGE

Data availability

Results And Discussion

MLST analysis

Impact of the P/T resistance on bacterial fitness

TEM-1 is involved in the acquired resistant to P/T

Role of OmpC in the P/T resistance

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1