
==== Front
J Antimicrob Chemother
J Antimicrob Chemother
jac
Journal of Antimicrobial Chemotherapy
0305-7453
1460-2091
Oxford University Press UK

38953288
10.1093/jac/dkae213
dkae213
Original Research
AcademicSubjects/MED00740
AcademicSubjects/MED00290
AcademicSubjects/MED00230
Extended period of selection for antimicrobial resistance due to recirculation of persistent antimicrobials in broilers
Swinkels Aram F Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands

Berendsen Bjorn J A Wageningen Food Safety Research, Wageningen University & Research, Wageningen, The Netherlands

https://orcid.org/0000-0002-0599-701X
Fischer Egil A J Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands

https://orcid.org/0000-0002-0758-5190
Zomer Aldert L Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
WHO Collaborating Centre for Reference and Research on Campylobacter and Antimicrobial Resistance from a One Health Perspective/WOAH Reference Laboratory for Campylobacteriosis, Utrecht, The Netherlands

Wagenaar Jaap A Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands
WHO Collaborating Centre for Reference and Research on Campylobacter and Antimicrobial Resistance from a One Health Perspective/WOAH Reference Laboratory for Campylobacteriosis, Utrecht, The Netherlands
Wageningen Bioveterinary Research, Wageningen University & Research, Lelystad, The Netherlands

Corresponding author. E-mail: j.wagenaar@uu.nl
9 2024
02 7 2024
02 7 2024
79 9 21862193
07 3 2024
31 5 2024
© The Author(s) 2024. Published by Oxford University Press on behalf of British Society for Antimicrobial Chemotherapy.
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com.

Abstract

Objectives

Antimicrobials can select for antimicrobial-resistant bacteria. After treatment the active compound is excreted through urine and faeces. As some antimicrobials are chemically stable, recirculation of subinhibitory concentrations of antimicrobials may occur due to coprophagic behaviour of animals such as chickens.

Methods

The persistence of three antimicrobials over time and their potential effects on antimicrobial resistance were determined in four groups of broilers. Groups were left untreated (control) or were treated with amoxicillin (unstable), doxycycline or enrofloxacin (stable). Antimicrobials were extracted from the faecal samples and were measured by LC-MS/MS. We determined the resistome genotypically using shotgun metagenomics and phenotypically by using Escherichia coli as indicator microorganism.

Results

Up to 37 days after treatment, doxycycline and enrofloxacin had concentrations in faeces equal to or higher than the minimal selective concentration (MSC), in contrast to the amoxicillin treatment. The amoxicillin treatment showed a significant difference (P ≤ 0.01 and P ≤ 0.0001) in the genotypic resistance only directly after treatment. On the other hand, the doxycycline treatment showed approximately 52% increase in phenotypic resistance and a significant difference (P ≤ 0.05 and P ≤ 0.0001) in genotypic resistance throughout the trial. Furthermore, enrofloxacin treatment resulted in a complete non-WT E. coli population but the quantity of resistance genes was similar to the control group, likely because resistance is mediated by point mutations.

Conclusions

Based on our findings, we suggest that persistence of antimicrobials should be taken into consideration in the assessment of priority classification of antimicrobials in livestock.

Netherlands Centre for One Health
==== Body
pmcIntroduction

Antimicrobial compounds are crucial and lifesaving medicines, which can be applied to treat infections with a bacterial pathogen and preventatively for surgery or organ transplantations.1–3 Unfortunately antimicrobial resistance (AMR) has become a growing problem during recent decades.4 The rising development of AMR has been considered as one of the most serious health threats for humans and animals by organizations such as the WHO, the Food and Agriculture Organization of the UN (FAO) and the World Organisation for Animal Health (WOAH).3,5,6 The excessive use of antimicrobials in livestock is contributing to the emergence of resistant bacteria.7 Measures to reduce antimicrobial usage in livestock are important to reduce the exposure to antimicrobial concentrations that select for resistant bacteria.

In terms of reducing worldwide antimicrobial usage, antimicrobial stewardship programmes have been established.8,9 These programmes consist of guidelines to reduce the use of antimicrobials and to stimulate prudent usage of antimicrobials in livestock.10–12 For example, the EMA has classified antimicrobial compounds into four different categories; A (avoid), B (restrict), C (caution) and D (prudence).13 This categorization of antimicrobial compounds has been established to reduce potential consequences for public health as a result of increased AMR.

Based on the classification of the EMA, antimicrobials are selected and applied by veterinarians to limit the selection of AMR. However, in this categorization, only direct selection during treatment is considered. Studies have shown that there is an extensive difference in half-life properties between antimicrobial compounds, ranging from hours to over a year.14–16 This implies that, for specific antimicrobials, selection for resistant bacteria can be maintained long after the application and withdrawal time.16,17 Especially in poultry, the persistence of antimicrobial compounds is of concern due to coprophagic behaviour, which induces recirculation of the antimicrobial compounds, thereby re-exposing the gut microbiome to residual concentrations that may be above the minimal selective concentration (MSC) for a longer period, resulting in prolonged selective pressure.17–20 Since persistence is not included in the assessment of the classification of antimicrobials it could be a missed opportunity for reducing AMR. To elaborate, to our knowledge, no studies have been conducted that show a relationship between persistence of antimicrobial residues and resistance levels.

The current study measured the temporal antimicrobial residue concentration in faecal droppings and caecal material after treatment of broilers. Alongside this, we investigated the prevalence of AMR in Escherichia coli isolates and the resistance genes in resistome data after cessation of the antimicrobial treatment. An animal trial was conducted, in which groups of broilers were treated with amoxicillin (unstable), and doxycycline and enrofloxacin (stable). Antimicrobials were extracted from the faecal samples and analysed by LC-MS/MS and were compared with the MSCs determined by A. F. Swinkels, E. A. J. Fischer, L. Korving, N. E. Kusters, J. A. Wagenaar, A. L. Zomer (unpublished data). We performed phenotypic susceptibility testing by plating E. coli isolates. Additionally, we quantified the read depth of resistance genes present in the resistome to investigate resistance genes besides the indicator organism E. coli. Lastly, we studied whether the antimicrobial residuals had any effect on the composition of the microbiome.

Materials and methods

Housing of the broilers, experimental set-up and sample collection

A total of 158 broilers (commercial Ross 308) were included in this trial and were transported from the hatchery to the animal facility (Utrecht University, Utrecht, The Netherlands) on the day of hatching (Day 0 of their life). The broilers were weighed, tagged and housed in one pen (6 m2) to equilibrate their microbiome. After 4 days, caeca were collected after euthanizing 12 broilers. Subsequently, the remaining broilers were randomly divided into different treatments groups (Day 0 of the experiment) and treatment started. The groups were treated with amoxicillin, doxycycline or enrofloxacin and an untreated control group. Each group was divided into three subgroups containing 12 broilers. The groups stayed in separate stables with hygiene rules to prevent carry-over of antimicrobials and bacteria. The antimicrobial treatment lasted for 4 days and was administered via the drinking water. Doses that were used were 25 mg per kg of bodyweight per day for doxycycline, 10 mg per kg of bodyweight per day for enrofloxacin and 20 mg per kg of bodyweight per day for amoxicillin.21–23 Afterwards, faecal droppings were collected individually from the broilers placed in carton boxes with paper at the bottom and were homogenized to obtain pooled samples. A fraction was stored at −80°C to quantify the presence of antimicrobials. This was repeated at Days 6, 12, 19 and 26 after the start of the treatment. On Day 37, after treatment, broilers were euthanized, and caecal material was collected. The samples were transported to the lab and further processed.

Ethics of experimentation

Broilers were observed daily for the presence of clinical signs, abnormal behaviour and mortality. The study protocol was approved by the Dutch Central Authority for Scientific Procedures on Animals and the Animal Experiments Committee of Utrecht University (Utrecht, The Netherlands) under registration number AVD10800202114909. All procedures were done in full compliance with all legislation. A power calculation was conducted to estimate the total broilers needed for the experiment, based on a simulation performed in R (Supplementary data, available in Zenodo repository; https://zenodo.org/records/11103906).

The size of the pens of the subgroups was set to 2 m2, according to the legal regulations in Appendix III of 2010/63/EU24 to not exceed the kg/m2. Loss of broilers in the first week of the experiment due to health issues was expected, resulting in 12 broilers per subgroup. Solely female broilers were selected to reduce weight increase.25

Phenotypic susceptibility testing

From every subgroup and timepoint, a swab was used to inoculate the pooled samples on three MacConkey plates. Next, 24 single colonies were picked by a pipette tip after overnight culture at 37°C and transferred to a single well in a 96-well plate with 100 µL of LB medium per well. The 72 colonies per treatment and timepoint in one 96-well plate were transferred to square MacConkey plates with epidemiological cut-off value (ECOFF) concentrations of amoxicillin (8 mg/L), doxycycline (4 mg/L), enrofloxacin (0.125 mg/L) and a control plate without antimicrobials via a stamp. Next, non-WT (NWT) colonies were scored after overnight incubation at 37°C. An E. coli colony was scored as NWT if it were able to grow on the selection plate with the respective antimicrobial. We calculated E. coli resistance by comparing growth on control and selection plates.

Shotgun metagenomics

DNA was extracted according to the ‘Ecology from Farm to Fork Of microbial drug Resistance and Transmission’ project (EFFORT) protocol; DNA concentrations were measured with a Qubit.26,27 Illumina sequencing was performed using an Illumina NovaSeq 6000 (USEQ, Utrecht sequencing facility) with a maximum read length of 2 × 150 bp. Libraries were prepared with the Illumina Nextera XT DNA Library Preparation Kit according to the manufacturer’s protocol.28 Each read was trimmed with Trim Galore (v0.6.4_dev) and the quality was assessed with FastQC (v0.11.4). The reads were analysed for taxonomic classification by Kraken2 and the abundance of the DNA sequences was computed with Bracken.29,30 The output was summarized into a biom file using kraken2biom and was analysed with the R programme packages phyloseq (v1.36.0), microViz (v0.9.1) and microbiome (v1.14.0), by which the species composition, and alpha and beta diversity were estimated from a rarefied phyloseq object.31,32 Finally the resistome was investigated by using KMA (v1.4.2) utilizing the ResFinder database with a minimum of 80% gene coverage.33,34 The reads were first normalized for gene length and displayed as sequence depth per Gb of sequencing data (sequence data for this article can be found in the SRA under accession PRJEB73721).

Antimicrobial analysis of the faecal samples

The faeces samples were stored at −80°C until extraction of the antimicrobial compounds. The samples were analysed according to the procedure described by Berendsen et al.14 In short, from the samples, approximately 1 g was weighed into a 50 mL polypropylene (PP) centrifuge tube and internal standard solution was added. The antimicrobial compounds were extracted by 4 mL of McIlvain-EDTA buffer and 1 mL of acetonitrile. Afterwards, 2 mL of lead acetate solution was added to remove excessive proteins. After centrifugation, the extract was transferred to a clean test tube and diluted by 13 mL of 0.2 M EDTA. A Phenomenex (Torrance, CA, USA) Strata-X RP 200 mg/6 mL reversed phase solid phase extraction (SPE) cartridge was conditioned with 5 mL of MeOH and 5 mL of water. The complete extract was applied onto the SPE cartridge and washed with 5 mL of water before being vacuum-dried for 5 min. The antimicrobial compounds were eluted by adding 5 mL of methanol to the cartridges. The eluate was then evaporated at 40°C under nitrogen. The residue was redissolved in 200 µL of methanol and 300 µL of water was added before transferring the extract into an LC-MS/MS sample vial.

The LC system consisted of a Shimadzu UFLC XR (Milford, MA, USA) model Acquity with a Phenomenex Kinetex C18 analytical column of 1.7 µm C18 100 Å, 100 × 2.1 mm, placed in a column oven at 40°C. The gradient profile with a flow rate of 0.3 mL/min is shown in Table 1. The injection volume was 5 μL. Detection was carried out by LC–MS/MS using a Sciex (Framingham, MA, USA) Q-Trap 6500 mass spectrometer in the positive electrospray ionization (ESI) mode. The antimicrobials were fragmented using collision-induced dissociation (N2) and the scheduled Selected Reaction Monitoring (SRM) transitions (20 s window) as described by Berendsen et al.35 Data were processed using MultiQuant software v2.1.1 (Sciex).

Table 1. Gradient profile used in the LC–MS/MS method

Time (min)	Mobile phase A (%)	Mobile phase B (%)	Flow (mL/min)	
0.0	99	1	0.3	
0.5	99	1	0.3	
2.5	75	25	0.3	
5.4	30	70	0.3	
5.5	0	100	0.3	
6.5	0	100	0.3	
6.6	100	0	0.3	
7.5	100	0	0.3	

Statistical analysis

We used R (v4.1.0) and the package lme4 (v1.1-29) for statistical calculations and logistic mixed-effects and linear mixed-effects models. Full models included ‘treatment’ and ‘time’ and the interaction between treatment and time as fixed effects and ‘pen’ as random effect. A logistic regression model was fitted with an interaction term for time after treatment and treatment. This was necessary due to inflated standard errors caused by counts of 0 (WT) or 24 (all NWT). Models without time and/or treatment were compared with the full model. The best model was selected based on the Akaike information criterion (AIC). For the resistome data, we performed a post hoc Tukey test to determine whether the treatment groups at the different timepoints were significantly different from the control group. The microbiome diversity was measured with the Shannon index for alpha diversity and Bray–Curtis distance for beta diversity. All files are in the Zenodo repository; https://zenodo.org/records/11103906.

Results

Antimicrobial residues extracted from the faecal samples before, during and after treatment

The concentrations of the antimicrobial residues extracted from the faecal samples are shown in Figure 1. At timepoint 0 and in the control group we did not measure any antimicrobial compounds. For the unstable antimicrobial amoxicillin we did not find concentrations that reached the MSC. Only at 6 days after the start of the treatment did we measure amoxicillin at 0.05 mg/kg; however, this was below the MSC. On the contrary, the stable antimicrobials doxycycline and enrofloxacin showed relatively high concentrations after treatment. For doxycycline the concentrations were above or within the range of the MSC up to 26 days. Enrofloxacin concentrations were far above the MSC range after treatment and were still in the MSC range at the time of slaughter. These results suggest that doxycycline and enrofloxacin remain effective beyond application.

Figure 1. Antimicrobial extraction in faecal droppings of broilers treated with different antimicrobials. (a) amoxicillin, (b) doxycycline and (c) enrofloxacin, which also contains a close-up view with lower concentrations on the y-axis to make it more visible. In the graphs, the MSCs are also displayed, determined by A. F. Swinkels, E. A. J. Fischer, L. Korving, N. E. Kusters, J. A. Wagenaar, A. L. Zomer (unpublished data). The MSC was determined by concentrations with a 10-fold difference, therefore the established MSC is displayed as a range between the lowest and highest possible concentration in the figures. The ECOFFs are distinguishing bacteria from WT and NWT. The datapoints are presented as the mean and SEM. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Quantifying phenotypic resistance of E. coli isolates over time

The E. coli isolates that were isolated over time in the different treatment groups are shown in Figure 2. We observed that the E. coli isolates from the amoxicillin treatment started with a high proportion of NWT E. coli followed by a declining trend comparable to the control group. For the doxycycline treatment we observed a steep increase directly after treatment and a slight decline at timepoint 12 days; however, the majority of the E. coli isolates remained NWT to doxycycline up until the slaughter age, while in the control group the number of NWT E. coli remained limited. For the enrofloxacin treatment we found, almost exclusively, enrofloxacin NWT E. coli isolates, which is in strong contrast to the control group where we observed only a few E. coli isolates that harboured resistance towards enrofloxacin. The data were best explained (lowest AIC) with a model containing both time after treatment and treatment.

Figure 2. Phenotypic resistance of NWT E. coli colonies. The graphs show the resistance of E. coli to the antimicrobial with which a group of broilers is treated compared with the control group. (a) Amoxicillin treatment compared with the control group, (b) doxycycline compared with the control group and (c) enrofloxacin compared with the control group. The datapoints are presented as the mean and SEM. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Quantifying resistance genes in faecal droppings

We determined the resistome to quantify resistance genes present in the different treatment groups (Figure 3).

Figure 3. Number of resistance genes in the resistome in sequence depth per Gb over time. (a) The amoxicillin treatment is shown compared with the control group, (b) the doxycycline treatment compared with the control group and (c) the enrofloxacin treatment compared with the control group. The datapoints are presented as the mean and SEM. *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Regarding amoxicillin treatment, we observed more resistance genes immediately after the end of treatment (Day 4) and at Day 6 than in the control group (amoxicillin—control Day 4 P < 0.0001, Day 6 P = 0.0023). However, by Day 12 after start of treatment, the quantity of resistance genes in the treatment group approached levels similar to those in the control group. Doxycycline treatment led to a similar trend, with increased resistance genes at Day 4 and Day 6 compared with the control group (doxycycline—control Day 4 P < 0.0001, Day 6 P = 0.0161). However, the difference in resistance genes between the doxycycline-treated group and the control group persisted even at Day 12 (P = 0.0156) and Day 19 (P = 0.0149). Notably, the first decline in resistance genes within the doxycycline group, approaching control group levels, was observed 26 days after treatment initiation. As for enrofloxacin treatment, its impact on resistance genes was closer to that of the control group, especially up to Day 19. On Day 26, a minor increase in resistance genes was observed but it was not a significant difference. It was only after slaughter that a significant difference was observed in the caecal material between the enrofloxacin-treated group and the control group (enrofloxacin—control Day 37 P < 0.0001).

Microbial composition determination after treatment with antimicrobials

We studied the composition of the microbiome to determine if exposure to the residual antimicrobials after treatment had an effect on the microbial composition (Figure 4). The most noticeable change was the E. coli abundance in the microbiota of the treated groups. The enrofloxacin treatment resulted in a strong reduction of E. coli to almost 0% at Day 4; afterwards it returned slowly in the microbiome. Furthermore, the same was observed for the doxycycline treatment as it reduced E. coli to 10% at Day 4 and returned to levels comparable to the control group. On the contrary, the amoxicillin-treated group favoured E. coli at Day 4 and its abundance declined over time. Notably, at Day 12, the control group had almost 90% E. coli abundance.

Figure 4. The relative abundance of the samples taken at the different timepoints after treatment. The 20 most abundant taxa are shown in the legend.

In Figure 5 we determined the alpha diversity of the treated groups at the different timepoints. For the amoxicillin treatment we found a lower diversity compared with the control group directly after terminating the treatment (P = 0.046). The group treated with enrofloxacin showed a significant difference in Shannon diversity with the control only at 12 days after start of the treatment (P = 0.045). In contrast, the group treated with doxycycline had an equal alpha diversity to the control. We also investigated beta diversity and we calculated the distance from the baseline sample to the samples taken at other timepoints, as shown in Figure 6. The group treated with amoxicillin differed from the control group at 4 days after treatment (P = 0.015). Similarly the enrofloxacin treatment group shows a difference 6 days (P = 0.0062) from beginning the application to the control group. For the group treated with doxycycline we did not find a difference with the control group.

Figure 5. Alpha diversity of the samples at the different timepoints after start of the application. The alpha diversity is measured using the Shannon index. (a) Amoxicillin treatment, (b) doxycycline treatment and (c) enrofloxacin treatment. The datapoints are presented as the mean and SEM. *P ≤ 0.05. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 6. Beta diversity distance of the samples at the different timepoints after start of the treatment; higher distance equals a more different microbiome. The beta diversity is measured with the Curtis–Bray distance and is displayed as distance from the baseline. (a) Amoxicillin treatment, (b) doxycycline treatment and (c) enrofloxacin treatment. The datapoints are presented as the mean and SEM. *P ≤ 0.05, **P ≤ 0.001. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Discussion

Our findings show that stable antimicrobial compounds like doxycycline and enrofloxacin remain in the faeces of broilers for the length of a broiler production round (41 days) at or above MSC level, resulting in prolonged selection pressure for AMR bacteria. Moreover, these findings are confirming our hypothesis that both doxycycline and enrofloxacin led to higher levels of NWT than the control group, while amoxicillin did not. These results reveal the importance of considering antimicrobial stability when antimicrobials are priority-classified.

Worldwide antimicrobial stewardship programmes are implemented to create awareness for judiciously selecting antimicrobial compounds, for example in livestock. The results of this research showed that stable antimicrobials, such as doxycycline and enrofloxacin, can retain selective pressure long after cessation of the treatment: doxycycline from 1.5 to 6 mg/L and enrofloxacin from 1.8 to 15 mg/L, within the MSC range. This is in contrast to the antimicrobial amoxicillin where hardly any residues were encountered, which is not surprising due to its instability.15,35 One can argue that amoxicillin could be used instead of stable antimicrobials; however, amoxicillin is still used extensively and has selective properties for β-lactamase genes with possible co-selection for tetracycline resistance.36,37 Nevertheless, persistence of antimicrobials should be taken into consideration in stewardship programmes as it extends the selection for phenotypically NWT E. coli (doxycycline approximately 52% increase, enrofloxacin approximately 100% increase). Moreover, increase of the occurrence of resistance genes after treatment with a stable antimicrobial compound (doxycycline; significance P ≤ 0.05 and P ≤ 0.0001) should also be considered. Therefore, to incorporate this in the antimicrobial classification would be valuable for the global efforts to encourage prudent usage of antimicrobials for reducing AMR, especially for livestock, where coprophagic behaviour is not uncommon.

Investigating the resistance levels phenotypically and genetically in relation to the stable residues gives a better understanding about prolonged selection for AMR bacteria due to recirculation of residues. Extensive research has been conducted into resistance levels in the farm environment38–40 but to our knowledge not in combination with antimicrobial residual concentrations. Peng et al.41 measured the antimicrobial concentrations of amoxicillin, doxycycline and ciprofloxacin (closely related to enrofloxacin) in manure during and after treatment of caged laying hens. In that study, birds were administered 50 mg/kg (among different concentrations) for all three antimicrobials, similar to our study. Peng et al. did find residual concentrations of amoxicillin after treatment; however, up to 2 days after the application they could not detect any concentrations in the group treated with 50 mg/kg. This is a slight difference compared with our study since we did not measure any concentrations of amoxicillin directly after treatment. Amoxicillin was probably degraded rapidly by the β-lactamase enzymes produced by the resistant bacteria or just by the general instability of the compound. This difference could be explained by the β-lactamase genes present in the microbiome. Peng et al. did not study the resistome, which could influence the amoxicillin concentration. Furthermore, in the study of Peng et al., doxycycline and ciprofloxacin were measured with similar concentrations as we observed in our study (doxycycline 6 mg/kg and enrofloxacin 15 mg/kg). After the application time, Peng et al. did not detect any doxycycline after 7 days, which is unlike our findings as we measured residues after 37 days, possibly due to the usage of cages instead of pens. Furthermore, ciprofloxacin was measured with an average concentration of 5.78 mg/kg after terminating the treatment, comparable to our enrofloxacin measurements and the same trend as other studies determined.42,43

The objective of this study was to investigate if stable antimicrobial residues lead to increased AMR levels. The presence of amoxicillin resistance in the baseline measurement suggested existing resistance in the broilers, implying that there were already β-lactamase-producing bacteria present, despite our efforts to select offspring from a untreated parental flock. This might have influenced the results from the amoxicillin treatment group since we observed high amoxicillin resistance directly after the treatment, possibly because of direct selection of resistant bacteria. Another unexpected result was the higher levels of resistance in the resistome data at Day 37 for doxycycline and enrofloxacin. This can be explained as we used caecal material instead of faecal droppings. Caecal samples have been observed to have higher levels of resistance genes.44 This is explaining the increased resistance levels compared with the control group despite the concentrations that are below the MSC at Day 37.

Considering the absence of resistance genes in the resistome after the enrofloxacin treatment, it can be argued that enrofloxacin does not select for resistance genes. Actually, this is described in the literature; resistance to enrofloxacin involves another selection mechanism based on selection or induction of SNPs associated with resistance instead of an increase in species carrying resistance genes.45,46 We sequenced several of the enrofloxacin NWT E. coli strains from the enrofloxacin treatment where we observed SNPs responsible for enrofloxacin resistance (A. F. Swinkels, A. L. Zomer, J. A. Wagenaar, unpublished data).

Our experimental set-up is different from conditions of commercial broilers, due to smaller groups of animals and stricter hygiene protocols. Therefore, parameters would be fluctuating due to the differences in set-up and might influence the results generated from this research. However, we considered the density of the broilers as an essential parameter since this is important for mimicking the commercial setting and enabling coprophagy. To conclude, we believe that persistence is an important factor in development of AMR.

Acknowledgements

We would like to thank the animal caretakers from the animal facility of Utrecht University (Utrecht, The Netherlands) for their assistance during the animal trial. Besides this we would like to thank Robbert van den Beld of WFSR (Wageningen Food Safety Research, The Netherlands) for his guidance and assistance in extracting the antimicrobials. Lastly we would like to thank André Steentjes of Veterinair Centrum Someren for assisting in selection of parental flocks of broilers for our trial.

Funding

This work was funded by the Netherlands Centre for One Health (NCOH), Identifier: RESRISK, Topic: Cmplex systems & Metegenomics #11.

Transparency declarations

None to declare.

Supplementary data

Supplementary materials for this article may be found online at https://zenodo.org/records/11103906.
==== Refs
References

1 Chodak GW . Use of systemic antibiotics for prophylaxis in surgery. Arch Surg 1977; 112 : 326. 10.1001/archsurg.1977.01370030098018 320960
2 Hughes JM . Preserving the lifesaving power of antimicrobial agents. JAMA 2011; 305 : 1027–8. 10.1001/jama.2011.279 21343545
3 WHO . Antimicrobial resistance. 2023. https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance
4 Schäberle TF , HackIM. Overcoming the current deadlock in antibiotic research. Trends Microbiol 2014; 22 : 165–7. 10.1016/j.tim.2013.12.007 24698433
5 FAO . Antimicrobial resistance. https://www.fao.org/antimicrobial-resistance/en/
6 WOAH . Antimicrobial resistance. https://www.woah.org/en/what-we-do/global-initiatives/antimicrobial-resistance/
7 Mulchandani R , WangY, GilbertMet al Global trends in antimicrobial use in food-producing animals: 2020 to 2030. PLOS Glob Public Health 2023; 3 : e0001305. 10.1371/journal.pgph.0001305 36963007
8 Dellit TH , OwensRC, McGowanJEet al Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis 2007; 44 : 159–77. 10.1086/510393 17173212
9 Lloyd DH , PageSW. Antimicrobial stewardship in veterinary medicine. Microbiol Spectr 2018; 6 : 10.1128/microbiolspec.ARBA-0023-2017
10 WHO . Antimicrobial Stewardship Programmes in Health-Care Facilities in low- and Middle-Income Countries: a WHO Practical Toolkit. 2019. https://www.who.int/publications/i/item/9789241515481
11 Pinto Ferreira J , BattagliaD, Dorado GarcíaAet al Achieving antimicrobial stewardship on the global scale: challenges and opportunities. Microorganisms 2022; 10 : 1599. 10.3390/microorganisms10081599 36014017
12 Pierce J , ApisarnthanarakA, SchellackNet al Global antimicrobial stewardship with a focus on low- and middle-income countries: a position statement for the International Society for Infectious Diseases. Intl J Infect Dis 2020; 96 : 621–9. 10.1016/j.ijid.2020.05.126
13 EMA . Categorisation of antibiotics used in animals promotes responsible use to protect public and animal health. 2020. https://www.ema.europa.eu/en/news/categorisation-antibiotics-used-animals-promotes-responsible-use-protect-public-animal-health
14 Berendsen BJA , LahrJ, NibbelingCet al The persistence of a broad range of antibiotics during calve, pig and broiler manure storage. Chemosphere 2018; 204 : 267–76. 10.1016/j.chemosphere.2018.04.042 29660540
15 Berendsen BJA , GerritsenHW, WeghRSet al Comprehensive analysis of ß-lactam antibiotics including penicillins, cephalosporins, and carbapenems in poultry muscle using liquid chromatography coupled to tandem mass spectrometry. Anal Bioanal Chem 2013; 405 : 7859–74. 10.1007/s00216-013-6804-6 23430185
16 Berendsen BJA , ElbersIJW, StolkerAAM. Determination of the stability of antibiotics in matrix and reference solutions using a straightforward procedure applying mass spectrometric detection. Food Addit Contam Part A Chem Anal Control Expo Risk Assess 2011; 28 : 1657–66. 10.1080/19440049.2011.604045 21988199
17 Gullberg E , CaoS, BergOGet al Selection of resistant bacteria at very low antibiotic concentrations. PLoS Pathog 2011; 7 : e1002158. 10.1371/journal.ppat.1002158 21811410
18 Andersson DI , HughesD. Microbiological effects of sublethal levels of antibiotics. Nat Rev Microbiol 2014; 12 : 465–78. 10.1038/nrmicro3270 24861036
19 Boes J , JohansenMV, EriksenLet al False-positive Trichuris suis egg counts in pigs in relation to coprophagia. Parasite 1998; 5 : 91–3. 10.1051/parasite/1998051091 9754304
20 Robertson ID , ThompsonRCA, IrwinPet al The role of dogs in transmission of gastrointestinal parasites in a remote tea-growing community in northeastern India. Am J Trop Med Hyg 2002; 67 : 539–45. 10.4269/ajtmh.2002.67.539 12479559
21 Minister of Agriculture, Nature and Food Quality . Prescription enrofloxacin (The Netherlands). REG NL 10409. 2018.
22 Minister of Agriculture, Nature and Food Quality . Prescription amoxicillin (The Netherlands). REG NL 113695. 2019.
23 Minister of Agriculture, Nature and Food Quality . Prescription doxycycline (The Netherlands). REG NL 8753. 2018.
24 European Parliament and the Council of the European Union. Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes. 2010. https://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2010:276:0033:0079:en:PDF
25 Ross. Broiler Management Handbook. 2018. https://aviagen.com/assets/Tech_Center/Ross_Broiler/Ross-BroilerHandbook2018-EN.pdf
26 Munk P , KnudsenBE, LukjancenkoOet al Abundance and diversity of the faecal resistome in slaughter pigs and broilers in nine European countries. Nat Microbiol 2018; 3 : 898–908. 10.1038/s41564-018-0192-9 30038308
27 Knudsen BE , BergmarkL, MunkPet al Impact of sample type and DNA isolation procedure on genomic inference of microbiome composition. mSystems 2016; 1 : e00095-16. 10.1128/mSystems.00095-16 27822556
28 Illumina . Nextera XT DNA Library Prep Kit. https://emea.support.illumina.com/sequencing/sequencing_kits/nextera_xt_dna_kit/documentation.html
29 Lu J , BreitwieserFP, ThielenPet al Bracken: estimating species abundance in metagenomics data. PeerJ Comput Sci 2017; 3 : e104. 10.7717/peerj-cs.104
30 Lu J , RinconN, WoodDEet al Metagenome analysis using the Kraken software suite. Nat Protoc 2022; 17 : 2815–39. 10.1038/s41596-022-00738-y 36171387
31 McMurdie PJ , HolmesS. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One 2013; 8 : e61217. 10.1371/journal.pone.0061217 23630581
32 Dabdoub S . kraken-biom: enabling interoperative format conversion for Kraken results (Version 1.2). 2016. https://github.com/smdabdoub/kraken-biom
33 Genomic Epidemiology . ResFinder. https://bitbucket.org/genomicepidemiology/resfinder_db
34 Clausen PTLC , AarestrupFM, LundO. Rapid and precise alignment of raw reads against redundant databases with KMA. BMC Bioinformatics 2018; 19 : 307. 10.1186/s12859-018-2336-6 30157759
35 Berendsen BJA , WeghRS, MemelinkJet al The analysis of animal faeces as a tool to monitor antibiotic usage. Talanta 2015; 132 : 258–68. 10.1016/j.talanta.2014.09.022 25476307
36 Bourély C , CazeauG, JarrigeNet al Co-resistance to amoxicillin and tetracycline as an indicator of multidrug resistance in Escherichia coli isolates from animals. Front Microbiol 2019; 10 : 2288. 10.3389/fmicb.2019.02288 31649635
37 EMA, European Surveillance of Veterinary Antimicrobial Consumption . Sales of veterinary antimicrobial agents in 31 European countries in 2022 (EMA/299538/2023). 2023. https://www.ema.europa.eu/en/documents/report/sales-veterinary-antimicrobial-agents-31-european-countries-2022-trends-2010-2022-thirteenth-esvac-report_en.pdf
38 Luiken REC , HeederikDJJ, ScherpenissePet al Determinants for antimicrobial resistance genes in farm dust on 333 poultry and pig farms in nine European countries. Environ Res 2022; 208 : 112715. 10.1016/j.envres.2022.112715 35033551
39 Enciso-Martínez Y , Barrios-VillaE, Sepúlveda-MorenoCOet al Prevalence of antibiotic-resistant E. coli strains in a local farm and packing facilities of honeydew melon in Hermosillo, Sonora, Mexico. Antibiotics (Basel) 2022; 11 : 1789. 10.3390/antibiotics11121789 36551446
40 Byrne N , O’NeillL, DίazJACet al Antimicrobial resistance in Escherichia coli isolated from on-farm and conventional hatching broiler farms in Ireland. Ir Vet J 2022; 75 : 7. 10.1186/s13620-022-00214-9 35459196
41 Peng P , WangY, LiuLet al The excretion and environmental effects of amoxicillin, ciprofloxacin, and doxycycline residues in layer chicken manure. Poult Sci 2016; 95 : 1033–41. 10.3382/ps/pew013 26944981
42 Yévenes K , PokrantE, TrincadoLet al Detection of antimicrobial residues in poultry litter: monitoring a risk through a selective and sensitive HPLC–MS/MS method. Animals (Basel) 2021; 11 : 1399. 10.3390/ani11051399 34069030
43 Pokrant E , TrincadoL, YévenesKet al Determination of five antimicrobial families in droppings of therapeutically treated broiler chicken by high-performance liquid chromatography-tandem mass spectrometry. Poult Sci 2021; 100 : 101313. 10.1016/j.psj.2021.101313 34298383
44 Kang K , HuY, WuSet al Comparative metagenomic analysis of chicken gut microbial community, function, and resistome to evaluate noninvasive and cecal sampling resources. Animals (Basel) 2021; 11 : 1718. 10.3390/ani11061718 34207572
45 Hayer SS , LimS, HongSet al Genetic determinants of resistance to extended-spectrum cephalosporin and fluoroquinolone in Escherichia coli isolated from diseased pigs in the United States. mSphere 2020; 5 : e00990-20 10.1128/mSphere.00990-20
46 Jacoby GA . Mechanisms of resistance to quinolones. Clin Infect Dis 2005; 41 : S120–6. 10.1086/428052 15942878
